{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.5"
    },
    "colab": {
      "name": "Netflix Movie Recommendation System.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "include_colab_link": true
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/bhaveshwadhwani/Netflix_Movie_Recommendation_System/blob/master/Netflix_Movie_Recommendation_System.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Kqp39W9AGHII"
      },
      "source": [
        "<img src='images/netflix-q.jpg'>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "qXxFUPPKGHIJ"
      },
      "source": [
        "<h1>1. Business Problem </h1>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "68x6gvapGHIK"
      },
      "source": [
        "<h2> 1.1 Problem Description </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "aE6iT69QGHIL"
      },
      "source": [
        "<p>\n",
        "Netflix is all about connecting people to the movies they love. To help customers find those movies, they developed world-class movie recommendation system: CinematchSM. Its job is to predict whether someone will enjoy a movie based on how much they liked or disliked other movies. Netflix use those predictions to make personal movie recommendations based on each customer’s unique tastes. And while <b>Cinematch</b> is doing pretty well, it can always be made better.\n",
        "</p>\n",
        "<p>Now there are a lot of interesting alternative approaches to how Cinematch works that netflix haven’t tried. Some are described in the literature, some aren’t. We’re curious whether any of these can beat Cinematch by making better predictions. Because, frankly, if there is a much better approach it could make a big difference to our customers and our business.</p>\n",
        "<p> Credits: https://www.netflixprize.com/rules.html </p>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Ne0jR1z6GHIM"
      },
      "source": [
        "<h2> 1.2 Problem Statement </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "DBuTSu8bGHIN"
      },
      "source": [
        "<p>\n",
        "Netflix provided a lot of anonymous rating data, and a prediction accuracy bar that is 10% better than what Cinematch can do on the same training data set. (Accuracy is a measurement of how closely predicted ratings of movies match subsequent actual ratings.) \n",
        "</p>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "JfA0haeYGHIP"
      },
      "source": [
        "<h2> 1.3 Sources </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "d76zFbnkGHIR"
      },
      "source": [
        "<ul>\n",
        "<li> https://www.netflixprize.com/rules.html</li>\n",
        "<li> https://www.kaggle.com/netflix-inc/netflix-prize-data</li>\n",
        "<li> Netflix blog: https://medium.com/netflix-techblog/netflix-recommendations-beyond-the-5-stars-part-1-55838468f429 (very nice blog)</li>\n",
        "<li>surprise library: http://surpriselib.com/ (we use many models from this library)</li>\n",
        "<li>surprise library doc: http://surprise.readthedocs.io/en/stable/getting_started.html (we use many models from this library)</li>\n",
        "<li>installing surprise: https://github.com/NicolasHug/Surprise#installation </li>\n",
        "<li> Research paper: http://courses.ischool.berkeley.edu/i290-dm/s11/SECURE/a1-koren.pdf (most of our work was inspired by this paper)</li>\n",
        "<li> SVD Decomposition : https://www.youtube.com/watch?v=P5mlg91as1c </li>\n",
        "</ul>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Z1iliA-iGHIS"
      },
      "source": [
        "<h2>1.4 Real world/Business Objectives and constraints  </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "e0mcbCxNGHIT"
      },
      "source": [
        "Objectives:\n",
        "1. Predict the rating that a user would give to a movie that he ahs not yet rated.\n",
        "2. Minimize the difference between predicted and actual rating (RMSE and MAPE)\n",
        "<br>\n",
        "\n",
        "Constraints:\n",
        "1. Some form of interpretability."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "y0xC5pTnGHIU"
      },
      "source": [
        "<h1> 2. Machine Learning Problem </h1>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "PVtbZQwPGHIU"
      },
      "source": [
        "<h2>2.1 Data </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "lrlJdwTHGHIV"
      },
      "source": [
        "<h3> 2.1.1 Data Overview </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "eKyUpY8TGHIW"
      },
      "source": [
        "<p> Get the data from : https://www.kaggle.com/netflix-inc/netflix-prize-data/data </p>\n",
        "<p> Data files : \n",
        "<ul> \n",
        "<li> combined_data_1.txt </li>\n",
        "<li> combined_data_2.txt </li>\n",
        "<li> combined_data_3.txt </li>\n",
        "<li> combined_data_4.txt </li>\n",
        "<li> movie_titles.csv </li>\n",
        "</ul>\n",
        "<pre>  \n",
        "The first line of each file [combined_data_1.txt, combined_data_2.txt, combined_data_3.txt, combined_data_4.txt] contains the movie id followed by a colon. Each subsequent line in the file corresponds to a rating from a customer and its date in the following format:\n",
        "\n",
        "CustomerID,Rating,Date\n",
        "\n",
        "MovieIDs range from 1 to 17770 sequentially.\n",
        "CustomerIDs range from 1 to 2649429, with gaps. There are 480189 users.\n",
        "Ratings are on a five star (integral) scale from 1 to 5.\n",
        "Dates have the format YYYY-MM-DD.\n",
        "</pre>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "pAdsJZkMGHIX"
      },
      "source": [
        "<h3> 2.1.2 Example Data point </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "tCsmW-lrGHIX"
      },
      "source": [
        "<pre>\n",
        "1:\n",
        "1488844,3,2005-09-06\n",
        "822109,5,2005-05-13\n",
        "885013,4,2005-10-19\n",
        "30878,4,2005-12-26\n",
        "823519,3,2004-05-03\n",
        "893988,3,2005-11-17\n",
        "124105,4,2004-08-05\n",
        "1248029,3,2004-04-22\n",
        "1842128,4,2004-05-09\n",
        "2238063,3,2005-05-11\n",
        "1503895,4,2005-05-19\n",
        "2207774,5,2005-06-06\n",
        "2590061,3,2004-08-12\n",
        "2442,3,2004-04-14\n",
        "543865,4,2004-05-28\n",
        "1209119,4,2004-03-23\n",
        "804919,4,2004-06-10\n",
        "1086807,3,2004-12-28\n",
        "1711859,4,2005-05-08\n",
        "372233,5,2005-11-23\n",
        "1080361,3,2005-03-28\n",
        "1245640,3,2005-12-19\n",
        "558634,4,2004-12-14\n",
        "2165002,4,2004-04-06\n",
        "1181550,3,2004-02-01\n",
        "1227322,4,2004-02-06\n",
        "427928,4,2004-02-26\n",
        "814701,5,2005-09-29\n",
        "808731,4,2005-10-31\n",
        "662870,5,2005-08-24\n",
        "337541,5,2005-03-23\n",
        "786312,3,2004-11-16\n",
        "1133214,4,2004-03-07\n",
        "1537427,4,2004-03-29\n",
        "1209954,5,2005-05-09\n",
        "2381599,3,2005-09-12\n",
        "525356,2,2004-07-11\n",
        "1910569,4,2004-04-12\n",
        "2263586,4,2004-08-20\n",
        "2421815,2,2004-02-26\n",
        "1009622,1,2005-01-19\n",
        "1481961,2,2005-05-24\n",
        "401047,4,2005-06-03\n",
        "2179073,3,2004-08-29\n",
        "1434636,3,2004-05-01\n",
        "93986,5,2005-10-06\n",
        "1308744,5,2005-10-29\n",
        "2647871,4,2005-12-30\n",
        "1905581,5,2005-08-16\n",
        "2508819,3,2004-05-18\n",
        "1578279,1,2005-05-19\n",
        "1159695,4,2005-02-15\n",
        "2588432,3,2005-03-31\n",
        "2423091,3,2005-09-12\n",
        "470232,4,2004-04-08\n",
        "2148699,2,2004-06-05\n",
        "1342007,3,2004-07-16\n",
        "466135,4,2004-07-13\n",
        "2472440,3,2005-08-13\n",
        "1283744,3,2004-04-17\n",
        "1927580,4,2004-11-08\n",
        "716874,5,2005-05-06\n",
        "4326,4,2005-10-29\n",
        "</pre>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "qw41PPqqGHIY"
      },
      "source": [
        "<h2>2.2 Mapping the real world problem to a Machine Learning Problem </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "gcl-ilooGHIc"
      },
      "source": [
        "<h3> 2.2.1 Type of Machine Learning Problem </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "LdnjiE2BGHId"
      },
      "source": [
        "<pre>\n",
        "For a given movie and user we need to predict the rating would be given by him/her to the movie. \n",
        "The given problem is a Recommendation problem \n",
        "It can also seen as a Regression problem \n",
        "</pre>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "EBchxZGuGHIk"
      },
      "source": [
        "<h3> 2.2.2 Performance metric </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "sWvLhIk3GHIl"
      },
      "source": [
        "<ul>\n",
        "<li> Mean Absolute Percentage Error: https://en.wikipedia.org/wiki/Mean_absolute_percentage_error </li>\n",
        "<li> Root Mean Square Error: https://en.wikipedia.org/wiki/Root-mean-square_deviation </li>\n",
        "</ul>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "GwDgl7vmGHIl"
      },
      "source": [
        "<h3> 2.2.3 Machine Learning Objective and Constraints </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "L03FzLdeGHIm"
      },
      "source": [
        "1. Minimize RMSE.\n",
        "2. Try to provide some interpretability."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "uRN6G1IuGHIn",
        "colab": {}
      },
      "source": [
        "# this is just to know how much time will it take to run this entire ipython notebook \n",
        "from datetime import datetime\n",
        "# globalstart = datetime.now()\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib\n",
        "matplotlib.use('nbagg')\n",
        "# !pip install scikit-surprise\n",
        "import matplotlib.pyplot as plt\n",
        "plt.rcParams.update({'figure.max_open_warning': 0})\n",
        "\n",
        "import seaborn as sns\n",
        "sns.set_style('whitegrid')\n",
        "import os\n",
        "from scipy import sparse\n",
        "from scipy.sparse import csr_matrix\n",
        "%matplotlib inline\n",
        "from sklearn.decomposition import TruncatedSVD\n",
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "import random"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mxd0uHQVOEF2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RM59l6hBJV1h",
        "colab_type": "code",
        "outputId": "0ec20cb1-347b-4b6b-caf3-6a5e11b7cd97",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3Aietf%3Awg%3Aoauth%3A2.0%3Aoob&scope=email%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdocs.test%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive.photos.readonly%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fpeopleapi.readonly&response_type=code\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Nwow3bg2GHIq"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "rOKh2aUbGHIr"
      },
      "source": [
        "<h1> 3. Exploratory Data Analysis </h1>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "y_bp2_o1GHIs"
      },
      "source": [
        "<h2> 3.1 Preprocessing</h2> "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "LnMF4vI4GHIs"
      },
      "source": [
        "<h3>3.1.1 Converting / Merging whole data to required format: u_i, m_j, r_ij</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "70c17ed1-22c7-4938-ccec-26af46e51fa6",
        "id": "N-pTnvA8GHIt",
        "colab": {}
      },
      "source": [
        "start = datetime.now()\n",
        "if not os.path.isfile('/content/drive/My Drive/Netflix_recommender/data_folder/data.csv'):\n",
        "    # Create a file 'data.csv' before reading it\n",
        "    # Read all the files in netflix and store them in one big file('data.csv')\n",
        "    # We re reading from each of the four files and appendig each rating to a global file 'train.csv'\n",
        "    data = open('/content/drive/My Drive/Netflix_recommender/data_folder/data.csv', mode='w')\n",
        "    \n",
        "    row = list()\n",
        "    files=['/content/drive/My Drive/Netflix_recommender/data_folder/combined_data_1.txt','/content/drive/My Drive/Netflix_recommender/data_folder/combined_data_2.txt', \n",
        "           '/content/drive/My Drive/Netflix_recommender/data_folder/combined_data_3.txt', '/content/drive/My Drive/Netflix_recommender/data_folder/combined_data_4.txt']\n",
        "    for file in files:\n",
        "        print(\"Reading ratings from {}...\".format(file))\n",
        "        with open(file) as f:\n",
        "            for line in f: \n",
        "                del row[:] # you don't have to do this.\n",
        "                line = line.strip()\n",
        "                if line.endswith(':'):\n",
        "                    # All below are ratings for this movie, until another movie appears.\n",
        "                    movie_id = line.replace(':', '')\n",
        "                else:\n",
        "                    row = [x for x in line.split(',')]\n",
        "                    row.insert(0, movie_id)\n",
        "                    data.write(','.join(row))\n",
        "                    data.write('\\n')\n",
        "        print(\"Done.\\n\")\n",
        "    data.close()\n",
        "print('Time taken :', datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Reading ratings from data_folder/combined_data_1.txt...\n",
            "Done.\n",
            "\n",
            "Reading ratings from data_folder/combined_data_2.txt...\n",
            "Done.\n",
            "\n",
            "Reading ratings from data_folder/combined_data_3.txt...\n",
            "Done.\n",
            "\n",
            "Reading ratings from data_folder/combined_data_4.txt...\n",
            "Done.\n",
            "\n",
            "Time taken : 0:05:03.705966\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "HlotH9C7GHIw"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "1c812d97-f4c5-4f0e-f209-8eaa56b0ab78",
        "id": "zbFHJAwSGHIx",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        }
      },
      "source": [
        "print(\"creating the dataframe from data.csv file..\")\n",
        "df = pd.read_csv('/content/drive/My Drive/Netflix_recommender/data_folder/data.csv', sep=',', \n",
        "                       names=['movie', 'user','rating','date'])\n",
        "df.date = pd.to_datetime(df.date)\n",
        "print('Done.\\n')\n",
        "\n",
        "# we are arranging the ratings according to time.\n",
        "print('Sorting the dataframe by date..')\n",
        "df.sort_values(by='date', inplace=True)\n",
        "print('Done..')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "creating the dataframe from data.csv file..\n",
            "Done.\n",
            "\n",
            "Sorting the dataframe by date..\n",
            "Done..\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "7baa43b9-a133-4f1b-bb14-814eb0e07145",
        "id": "tXsJQB-CGHI1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>movie</th>\n",
              "      <th>user</th>\n",
              "      <th>rating</th>\n",
              "      <th>date</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>56431994</th>\n",
              "      <td>10341</td>\n",
              "      <td>510180</td>\n",
              "      <td>4</td>\n",
              "      <td>1999-11-11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9056171</th>\n",
              "      <td>1798</td>\n",
              "      <td>510180</td>\n",
              "      <td>5</td>\n",
              "      <td>1999-11-11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>58698779</th>\n",
              "      <td>10774</td>\n",
              "      <td>510180</td>\n",
              "      <td>3</td>\n",
              "      <td>1999-11-11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>48101611</th>\n",
              "      <td>8651</td>\n",
              "      <td>510180</td>\n",
              "      <td>2</td>\n",
              "      <td>1999-11-11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>81893208</th>\n",
              "      <td>14660</td>\n",
              "      <td>510180</td>\n",
              "      <td>2</td>\n",
              "      <td>1999-11-11</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          movie    user  rating       date\n",
              "56431994  10341  510180       4 1999-11-11\n",
              "9056171    1798  510180       5 1999-11-11\n",
              "58698779  10774  510180       3 1999-11-11\n",
              "48101611   8651  510180       2 1999-11-11\n",
              "81893208  14660  510180       2 1999-11-11"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "f20024d3-5648-4843-a0f1-8292aabc5d4f",
        "id": "KB8urNZWGHI3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "df.describe()['rating']"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    1.004805e+08\n",
              "mean     3.604290e+00\n",
              "std      1.085219e+00\n",
              "min      1.000000e+00\n",
              "25%      3.000000e+00\n",
              "50%      4.000000e+00\n",
              "75%      4.000000e+00\n",
              "max      5.000000e+00\n",
              "Name: rating, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "El0nKTF3GHI8"
      },
      "source": [
        "<h3>3.1.2 Checking for NaN values </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "159f6cf4-3cdf-471b-a4b6-8ce760c56bdb",
        "id": "bBoutCr8GHI9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# just to make sure that all Nan containing rows are deleted..\n",
        "print(\"No of Nan values in our dataframe : \", sum(df.isnull().any()))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "No of Nan values in our dataframe :  0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "lBsgBgJOGHJF"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "p8fPR6NDGHJF"
      },
      "source": [
        "<h3>3.1.3 Removing Duplicates </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "413c7499-8c64-4487-9f00-e5eaa04769d1",
        "id": "f4tN0DQ0GHJG",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "dup_bool = df.duplicated(['movie','user','rating'])\n",
        "dups = sum(dup_bool) # by considering all columns..( including timestamp)\n",
        "print(\"There are {} duplicate rating entries in the data..\".format(dups))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "There are 0 duplicate rating entries in the data..\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "a1ktiJ6KGHJI"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "EQSxmXy8GHJJ"
      },
      "source": [
        " <h3>3.1.4 Basic Statistics (#Ratings, #Users, and #Movies)</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "a4bc67cd-6a7a-41c0-d75c-adc4bb80be98",
        "id": "vQVNTCjzGHJJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "print(\"Total data \")\n",
        "print(\"-\"*50)\n",
        "print(\"\\nTotal no of ratings :\",df.shape[0])\n",
        "print(\"Total No of Users   :\", len(np.unique(df.user)))\n",
        "print(\"Total No of movies  :\", len(np.unique(df.movie)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total data \n",
            "--------------------------------------------------\n",
            "\n",
            "Total no of ratings : 100480507\n",
            "Total No of Users   : 480189\n",
            "Total No of movies  : 17770\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "xEnrVuKDGHJM"
      },
      "source": [
        "<h2>3.2 Spliting data into Train and Test(80:20) </h2>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "hBRN_GB2GHJN",
        "colab": {}
      },
      "source": [
        "if not os.path.isfile('/content/drive/My Drive/Netflix_recommender/data_folder/train.csv'):\n",
        "    # create the dataframe and store it in the disk for offline purposes..\n",
        "    df.iloc[:int(df.shape[0]*0.80)].to_csv(\"train.csv\", index=False)\n",
        "\n",
        "if not os.path.isfile('/content/drive/My Drive/Netflix_recommender/data_folder/test.csv'):\n",
        "    # create the dataframe and store it in the disk for offline purposes..\n",
        "    df.iloc[int(df.shape[0]*0.80):].to_csv(\"test.csv\", index=False)\n",
        "\n",
        "train_df = pd.read_csv(\"/content/drive/My Drive/Netflix_recommender/data_folder/train.csv\", parse_dates=['date'])\n",
        "test_df = pd.read_csv(\"/content/drive/My Drive/Netflix_recommender/data_folder/test.csv\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "FjI2Bwr5GHJV"
      },
      "source": [
        "<h3>3.2.1 Basic Statistics in Train data (#Ratings, #Users, and #Movies)</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "d248bf9c-e7d2-48dd-a538-d12c53f32091",
        "id": "kUEUofOvGHJV",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "# movies = train_df.movie.value_counts()\n",
        "# users = train_df.user.value_counts()\n",
        "print(\"Training data \")\n",
        "print(\"-\"*50)\n",
        "print(\"\\nTotal no of ratings :\",train_df.shape[0])\n",
        "print(\"Total No of Users   :\", len(np.unique(train_df.user)))\n",
        "print(\"Total No of movies  :\", len(np.unique(train_df.movie)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training data \n",
            "--------------------------------------------------\n",
            "\n",
            "Total no of ratings : 80384405\n",
            "Total No of Users   : 405041\n",
            "Total No of movies  : 17424\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "kNzpnzXDGHJX"
      },
      "source": [
        "<h3>3.2.2 Basic Statistics in Test data (#Ratings, #Users, and #Movies)</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "d8b4e924-7c3b-4138-c254-2f33768b7024",
        "id": "KYo7TX2JGHJX",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "print(\"Test data \")\n",
        "print(\"-\"*50)\n",
        "print(\"\\nTotal no of ratings :\",test_df.shape[0])\n",
        "print(\"Total No of Users   :\", len(np.unique(test_df.user)))\n",
        "print(\"Total No of movies  :\", len(np.unique(test_df.movie)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test data \n",
            "--------------------------------------------------\n",
            "\n",
            "Total no of ratings : 20096102\n",
            "Total No of Users   : 349312\n",
            "Total No of movies  : 17757\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "N0tlZ5jIGHJZ"
      },
      "source": [
        " <h2> 3.3 Exploratory Data Analysis on Train data </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "9WIr9e5SGHJZ"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "g768N7pOGHJa",
        "colab": {}
      },
      "source": [
        "# method to make y-axis more readable\n",
        "def human(num, units = 'M'):\n",
        "    units = units.lower()\n",
        "    num = float(num)\n",
        "    if units == 'k':\n",
        "        return str(num/10**3) + \" K\"\n",
        "    elif units == 'm':\n",
        "        return str(num/10**6) + \" M\"\n",
        "    elif units == 'b':\n",
        "        return str(num/10**9) +  \" B\""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "o0lRFBKbGHJb"
      },
      "source": [
        "<h3> 3.3.1 Distribution of ratings </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "abeeef99-a7fb-4ae5-af98-e7cca952c7f3",
        "id": "nqmCKT-bGHJc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        }
      },
      "source": [
        "fig, ax = plt.subplots()\n",
        "plt.title('Distribution of ratings over Training dataset', fontsize=15)\n",
        "sns.countplot(train_df.rating)\n",
        "ax.set_yticklabels([human(item, 'M') for item in ax.get_yticks()])\n",
        "ax.set_ylabel('No. of Ratings(Millions)')\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZcAAAEYCAYAAACQgLsAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxczMSvf/Ab+L4fSATyIbAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Fdo_j64AGHJd"
      },
      "source": [
        "<p style=\"font-size:13px\"> <b>Add new column (week day) to the data set  for analysis.</b></p>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "f19498ad-6c85-4999-d8c6-33b9d300cf3f",
        "id": "_xjyPNEXGHJe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "# It is used to skip the warning ''SettingWithCopyWarning''.. \n",
        "pd.options.mode.chained_assignment = None  # default='warn'\n",
        "\n",
        "train_df['day_of_week'] = train_df.date.dt.weekday_name\n",
        "\n",
        "train_df.tail()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>movie</th>\n",
              "      <th>user</th>\n",
              "      <th>rating</th>\n",
              "      <th>date</th>\n",
              "      <th>day_of_week</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>80384400</th>\n",
              "      <td>12074</td>\n",
              "      <td>2033618</td>\n",
              "      <td>4</td>\n",
              "      <td>2005-08-08</td>\n",
              "      <td>Monday</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>80384401</th>\n",
              "      <td>862</td>\n",
              "      <td>1797061</td>\n",
              "      <td>3</td>\n",
              "      <td>2005-08-08</td>\n",
              "      <td>Monday</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>80384402</th>\n",
              "      <td>10986</td>\n",
              "      <td>1498715</td>\n",
              "      <td>5</td>\n",
              "      <td>2005-08-08</td>\n",
              "      <td>Monday</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>80384403</th>\n",
              "      <td>14861</td>\n",
              "      <td>500016</td>\n",
              "      <td>4</td>\n",
              "      <td>2005-08-08</td>\n",
              "      <td>Monday</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>80384404</th>\n",
              "      <td>5926</td>\n",
              "      <td>1044015</td>\n",
              "      <td>5</td>\n",
              "      <td>2005-08-08</td>\n",
              "      <td>Monday</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          movie     user  rating       date day_of_week\n",
              "80384400  12074  2033618       4 2005-08-08      Monday\n",
              "80384401    862  1797061       3 2005-08-08      Monday\n",
              "80384402  10986  1498715       5 2005-08-08      Monday\n",
              "80384403  14861   500016       4 2005-08-08      Monday\n",
              "80384404   5926  1044015       5 2005-08-08      Monday"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ygJPJ8TGGHJg"
      },
      "source": [
        "<h3> 3.3.2 Number of Ratings per a month </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "8981337b-0e70-4b05-d896-88abc2516e54",
        "id": "Qnco9MwLGHJh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "ax = train_df.resample('m', on='date')['rating'].count().plot()\n",
        "ax.set_title('No of ratings per month (Training data)')\n",
        "plt.xlabel('Month')\n",
        "plt.ylabel('No of ratings(per month)')\n",
        "ax.set_yticklabels([human(item, 'M') for item in ax.get_yticks()])\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEWCAYAAACnlKo3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxCyIGGOMiZgFEWOMMRGzIGKMMSZi/x/3CEEcYw75JgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "xWj3WheyGHJj"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "TKfHd26wGHJk"
      },
      "source": [
        "<h3> 3.3.3 Analysis on the Ratings given by user </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "161d1eaa-b744-4688-83b5-36cf2e6b1681",
        "id": "idlP3rKeGHJq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        }
      },
      "source": [
        "no_of_rated_movies_per_user = train_df.groupby(by='user')['rating'].count().sort_values(ascending=False)\n",
        "\n",
        "no_of_rated_movies_per_user.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "user\n",
              "305344     17112\n",
              "2439493    15896\n",
              "387418     15402\n",
              "1639792     9767\n",
              "1461435     9447\n",
              "Name: rating, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aUdrPjIBTysZ",
        "colab_type": "code",
        "outputId": "a2783438-1c69-4dff-c5fa-adff1dbf7de8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "no_of_rated_movies_per_user.describe()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    405041.000000\n",
              "mean        198.459921\n",
              "std         290.793238\n",
              "min           1.000000\n",
              "25%          34.000000\n",
              "50%          89.000000\n",
              "75%         245.000000\n",
              "max       17112.000000\n",
              "Name: rating, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tz_k92E3T8k9",
        "colab_type": "code",
        "outputId": "6e2f28e3-c782-4af4-81f8-a74c943647e6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        }
      },
      "source": [
        "plt.plot(no_of_rated_movies_per_user.values)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7fd3e660a518>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAD8CAYAAACcjGjIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n9huLmNGJAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hqXHdcYWTOcm",
        "colab_type": "code",
        "outputId": "85712e60-5056-4629-ec3d-bbd9ff6cbffe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        }
      },
      "source": [
        "plt.plot(no_of_rated_movies_per_user.values[:100])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7fd3e66abe48>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 27
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAD8CAYAAACcjGjIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4SAiIt0oHEREpJv/D235KJPrxcFNAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "df9ce16f-a521-4824-82c7-6db68e665ad6",
        "id": "gsw7dn-lGHJ1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "tyfig = plt.figure(figsize=plt.figaspect(.5))\n",
        "\n",
        "ax1 = plt.subplot(121)\n",
        "sns.kdeplot(no_of_rated_movies_per_user, shade=True, ax=ax1)\n",
        "plt.xlabel('No of ratings by user')\n",
        "plt.title(\"PDF\")\n",
        "\n",
        "ax2 = plt.subplot(122)\n",
        "sns.kdeplot(no_of_rated_movies_per_user, shade=True, cumulative=True,ax=ax2)\n",
        "plt.xlabel('No of ratings by user')\n",
        "plt.title('CDF')\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfkAAAEWCAYAAABlpO6zAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3XucXHV9//HXmdlr7vdAsglECB8S\nUIJBUKstVatALbFVrl7ARtQWpC2tv+KlSlH6gN9PsVrBG3e8xIhaUwsGLSJYRTGKaLJ8MUDYbLgk\n2dyzu7M7c87vj3NmM9nM7Jzdnd2dmfN+Ph55ZPbMOd/z/W5y5jPfuxcEASIiIlJ/UhOdARERERkb\nCvIiIiJ1SkFeRESkTinIi4iI1CkFeRERkTqlIC8iIlKnFORFRETqVMNEZ0Bqi5ltAeYDOeAgcB9w\nBfB94JVAPxAAfwC+BXzGOZeJrr0G+AiQKUjyWufc/x2f3ItIITO7GLgKOBHYDzwGXAe8gfBZ7Y1O\nfR64H7jOOfd8dO2ZwANAd0GSP3bO/cW4ZF5iUU1eRuIvnHNTgJcDpwEfjY5f4ZybChwN/CNwIXCv\nmXkF137TOTel4I8CvMgEMLOrgH8H/o3wi/ti4GZgVXTKN6PneRbwl8BRwAYzO7ogmecGPc8K8FVG\nQV5GzDm3jbAmf/Kg4wedcw8C5wKvAv58/HMnIqWY2XTgWuBy59x3ome23zn3X865DxaeGx3fCFwA\n7CD8Ai81QkFeRszMFgHnAL8p9r5zrgP4FfDa8cyXiJT1KqAF+G7cC5xzOeB76HmuKeqTl5H4TzPL\nAnuB/yZs7ruvxLnPETb35Z1vZm8u+Hm5c+65scmmiJQwG9jpnMsO87rBz/MCM9tT8PN7nXNrR507\nqRgFeRmJtzjnflR4wMxKnbsQ+FnBz2udc+8Yq4yJSCxdwBwzaxhmoF8I7Cr4+TnnXFtlsyaVpOZ6\nGTNRc/5K4OGJzouIHObnhLNc3hL3AjNLAX+Bnueaopq8VJyZTQJeAXwG+CVw78TmSEQKOef2mtnH\ngJuirrf7Cae/vgH4UwqmxZlZA7AUuIZwhP2N455hGTHV5KWSPm9m+4EXCafmfBs4yznnT2y2RGQw\n59ynCefIf5Rw1PxWwjUv/jM65QIzO0A49mYdYRP/So2hqS1eEAQTnQcREREZA6rJi4iI1CkFeRER\nkTqlIC8iIlKnFORFRETqVF1MoXvssceC5ubmIc/JZDKUO6cWqVy1oxrK1N3dvXPlypVzJzQTQ4jz\nLEN1/C7HQj2Wqx7LBBNfrrjPcl0E+ebmZpYtWzbkOe3t7WXPqUUqV+2ohjJt2LDh2QnNQBlxnmWo\njt/lWKjHctVjmWDiyxX3WVZzvYiISJ1SkBcREalTCvIiIiJ1KlafvJmdBXwWSAO3OOeuH/R+M3AX\n4WYkXcAFzrkt0XsfAlYDOeBK59z6odI0Mw/4JHBedM0XnHOfG10xJUn6+/vp7Oykt7d3orNymP7+\nftrb28flXi0tLbS1tdHY2Dgu9xOR6lQ2yJtZGrgJ+DOgE3jUzNY55zYVnLYa2O2cO97MLgRuIFz3\neDlwIXASsAD4kZmdEF1TKs1LgUXAic4538zmVaKgkhydnZ1MnTqVY489Fs/zJjo7A3p6emhtbR3z\n+wRBQFdXF52dnSxZsmTM7yci1StOc/3pwGbn3NPOuT5gDbBq0DmrgDuj1/cAr49q5KuANc65jHPu\nGWBzlN5Qaf4NcG1+UxPn3PaRF0+SqLe3l9mzZ1dVgB9Pnucxe/bsqmvJEJHxF6e5fiHh7kR5ncAZ\npc5xzmXNbC8wOzr+yKBrF0avS6V5HGErwF8S7ox0pXPuD0NlMJPJlG0G7e3tHbem0vGkch2pv7+/\nKgNcEAT09PSM2/3GunvAzG4D3gxsd86dXOR9j7BL7hzCrUsvdc79eswyJCJHqMZ58s1Ar3PuNDP7\nK+A24LVDXqB58hOdjYobTbna29vHpVl8uMaruT6vsbHxiN/hhg0bKnmLO4DPE47HKeZswn3IlxJ+\nif8CR1YQEm3wLqBBEOD7xXcGLbVfaLGdREufW+J4kSsqtUFpJuvT25+rTGJVZCzL1dyQqlhLZJwg\nv42wjzyvLTpW7JxOM2sAphMOwBvq2lLHO4HvRK+/C9weI4/D8tute7jsrl9x/z/8MTMmNVU6eZHY\n7rjjDi644IKB4H/ZZZfx6U9/mmnTpk1wzspzzj1kZscOccoq4C7nXAA8YmYzzOxo59zz45PDofX0\n5Xjyxf24F/bz3N4edh7I0HWgjwOZLJl+n95sjp6+HL3ZHJl+n5wf4AcBfgB+4evg8NejD47PVKJ4\nVWbLRGdgjGwZk1RXv2YJ//Lm5RVJK06QfxRYamZLCAPxhcDFg85ZB1wC/Bx4G/CAcy4ws3XA183s\nRsKBd0uBXwLeEGn+J/CnhP/T/wR4cuTFK+4P2w+wfX+G7fszCvIy5oIgKFrbArjrrrs499xzB4L8\nV77ylfHM2lgr1tW3ECgZ5ON0vcHIu3N2HMyy/g/7+HlHN8/s7jus/jq1Oc2MljStjSmaGlI0pT1m\ntXg0pRtoTHmkUh4e4Hng4eF5ASnv0LFUVPNKFamABYQfeuX4vk8qlaJUJc6jeA17OLW+UmcOp+I4\nnDpmzs+RTqWHcUVtGMtynTKbinW1lQ3yUR/7FcB6wulutznnNprZtcCvnHPrgFuBu81sM7CLMGgT\nnbcW2ARkgcudczmAYmlGt7we+JqZ/QNwAHhPRUpaIJMNm1hyJZrFREars7OT1atXc8opp7Bx40Ze\n9rKX8cQTT9DX18eb3vQmrrzySu666y62b9/OJZdcwowZM7j77rt53etexz333EN3dzeXXXYZK1eu\n5De/+Q3z58/n5ptvpqWlhccff5yPfOQjpFIpXv3qV/Pwww/z/e9/f6KLXBFjtazt3u5+blj/BGsf\n3YofBJyyaAaXLDuaBdNbOXp6K3OmNJFKeQRB6abu8dDR8SyLFx8zgTmovHosE4xtuY6bO5l501qG\nPCdu11usPnnn3L3AvYOOfazgdS/hvPZi114HXBcnzej4HuDP4+RrpPqyPqAgnwTf3tDJ2l9tLX/i\nMJx/2iLeurKt7HnPPvssN9xwAytWrGDPnj00NzfT1NTEpZdeyhNPPMG73vUu7rjjDu68805mzZpV\n9Pobb7yRT37yk/zd3/0d69evZ9WqVXz4wx/mE5/4BKeeeiqf+tSnKlq2CovT1Tfmfr9tL5fe/ii7\nu/v4q1MX8qaT5jO1pZHBj78+DqQeVePAuzGXiYJ8Vk+1jKEFCxawYsUKAO677z7WrFmD7/vs2LGD\np556ihNPPHHI69va2gZqqyeddBLbtm1j3759HDx4kFNPPRWAN7/5zTz44INjWo5RWAdcYWZrCAfc\n7R3v/vjHO/fwjlt+wZTmBv79ghXMndJMgAK6JEcyg3x/vibvT3BOZKy9dWVbrFr3WJg0aRIAW7du\n5bbbbuOrX/0q8+fP5+qrryaTyZS9vqnp0HiRdDod65rxZGbfAM4E5phZJ/BxoBHAOfdFwpa6cwjX\nx+gG3j2e+dvX28/ffPXXTGlp4JOrTqK1qXFCm+JFJkIig3xfLuyTz+b0yMvYO3jwIK2trUyZMoWd\nO3fy0EMPcfrppwMwefJkDh48WLS5vphp06YxefJkfvvb33LKKadw771H9HiNG+fcRWXeD4DLxyk7\nR/jXdZt4fm8PN56/gtYmLe8ryZTIIH+oJq8gL2PvxBNPZPny5bzlLW9hwYIFvPzlLx947/zzz+c9\n73kP8+bN4+67746V3nXXXcdHP/pRUqkUr3jFK5gyZcpYZb1mPbZ1D9/+dSeXvOoYjpreUrE53yK1\nJpFBvi+X75NXc72Mjba2tsNGvF9//fVFF8N55zvfyTvf+c6Bnx944AEAZs2addj1q1evHnh9/PHH\n81//9V8AfPnLX+bkk49YbC7xPv/AZqa3NvLG5UcpwEuiJTLI52vyaq6XWvSTn/yEL33pS+RyORYs\nWMD1119f/qIE2fTcPn7U/iKrX3MsqWKT1kUSJJlBPponr9H1UovOOecczjnnnInORtX6xi87aG1M\nc+YJ2sBSJM4udHVnoLk+p+b6elVqhbmkSGr5gyDgh5te5JUvmUVjOpEfbyKHSeRTMNBcr5p8XWpp\naaGrqyvRga6rq4uWlqFXzKpHv9u2lxf29fKKJbM0XU6ExDbXK8jXs7a2Njo7O9mxY8dEZ+Uw/f39\nNDaOz1SulpYW2tomZn2AiXT/xhdJpzxOOrr6N/gRGQ+JDPJa1ra+NTY2smTJkonOxhHqdVvganL/\nphc4ddEMWhrTWtVOhKQ21+cH3qlPXqRu7O3u58kXD3Dq4hkK8CKRhAZ51eRF6s3G5/cCsGjWpAnO\niUj1SGSQ71OfvEjd2bhtHwBtMxTkRfISGeQ18E6k/vz+ub0cNa2FKc2JHGokUlRCg3zYJ59L6BQr\nkXr0+217OeGoKXquRQokNMhrWVuRenIwk+XpnQc5bq426xEplOggr/3kRepD+/P7CAJYrEF3IodJ\nXJAPguDQwDvV5EXqQvsL+wFYOL21zJkiyZK4IN9XMDdeA+9E6kPnrm6aGlJMnzQ+KwqK1IrEBfl8\nUz1onrxIvejc08NR05K3Vr9IOckL8v0K8iL1pnN3D0dPb9FKdyKDJC7IFzbXK8iL1Idtu3uYN7V5\norMhUnUSF+Qz/bmB1+qTF6l9vf05dh7IMFdBXuQIyQvyWQ28E6kn2/b0ADB7ioK8yGCJC/J9hwV5\nzZMXqXWdu8MgP1Mj60WOEGuRZzM7C/gskAZucc5dP+j9ZuAuYCXQBVzgnNsSvfchYDWQA650zq0f\nKk0zuwP4E2BvlPylzrnHRl7EwxXW5H3V5EVqXufubgBmT1ZNXmSwskHezNLATcCfAZ3Ao2a2zjm3\nqeC01cBu59zxZnYhcANwgZktBy4ETgIWAD8ysxOia4ZK84POuXsqUL4j5NetBzXXi9SDbbt7aEh5\nTG9tRE+0yOHiNNefDmx2zj3tnOsD1gCrBp2zCrgzen0P8Hoz86Lja5xzGefcM8DmKL04aY6JPs2T\nF6krnbt7mD+tBc+b6JyIVJ84zfULga0FP3cCZ5Q6xzmXNbO9wOzo+CODrl0YvR4qzevM7GPA/wBX\nO+cyQ2Uwk8nQ3t4+ZCF6e3tpb2/nqS0HBo7tP3Cw7HXVLl+uelOP5arHMlWDbXt6OEpz5EWKqsaN\nlz8EvAA0AV8G/hm4dqgLmpubWbZs2ZCJtre3s2zZMp7o7QS2k/KgqaW17HXVLl+uelOP5aqGMm3Y\nsGFC7z8WXtjby8kLp010NkSqUpwgvw1YVPBzW3Ss2DmdZtYATCccgDfUtUWPO+eej45lzOx24J9i\n5DG2fHN9S2Nau9CJ1IG9Pf1Ma9HIepFi4vTJPwosNbMlZtZEOJBu3aBz1gGXRK/fBjzgnAui4xea\nWbOZLQGWAr8cKk0zOzr62wPeAvx+NAUcLD+6vrkhRU4xXqSm9ed8DmSyTGmuxkZJkYlXNsg757LA\nFcB6oB1Y65zbaGbXmtm50Wm3ArPNbDNwFXB1dO1GYC2wCfgBcLlzLlcqzSitr5nZ74DfAXOAT1am\nqKH82vVhTV6deCK1bG9PPwCTFeRFior1ZDjn7gXuHXTsYwWve4HzSlx7HXBdnDSj46+Lk6eRyq9d\n39KY1hQ6kRq3pzsM8pOa0hOcE5HqlLivv/m165vSKfXJi4xCjEWyFhNOrZ0RnXN19OW+Yvb29AEK\n8iKlJG5Z20zWp7khRTrlqbleZIQKFsk6G1gOXBQtflXoo4RdcacSjru5udL5yNfkWxXkRYpKZJBv\nakiR8rQYjsgoxFnQKgDyc9umA89VOhP5ID+lSaPrRYpJXnN91qcpnSKV8sgFCvIiIxRnkaxrgPvN\n7APAZOAN5RKNs7AVHFpYyG0Jt7jY2/UiB3bVfvdbX18fHR3PTnQ2KqoeywRjW67W/ul0beurSFoJ\nDPI5mhpSpD2P/mztfyiIVLGLgDucc582s1cBd5vZyc65kg9enIWt4NDCQvdtdXheF8cdu6guVrzr\n6HiWxYuPmehsVFQ9lgnGtlwL5k5m3rSWIc+Ju7BV4prr+6I+edXkRUYlziJZqwmn0OKc+znQQjgt\ntmL2RAvheGjhepFiEhfkM1mfxnRYk9cUOpERi7NIVgfwegAzW0YY5HdUMhN7uvuZ1tpAoP3nRIpK\nZJBvakjhaeCdyIjFXCTrH4HLzOy3wDeAS6OVMCsmX5NXo5xIcYnrk+/L98mnPHwFeZERi7FI1ibg\nj8YyD3u7+8IgP5Y3EalhyazJq7lepC7s6elnakvi6ioisSUuyPfnfBpSXjjwTkFepKbt6e5nioK8\nSEmJC/I5H1KepxXvRGpczg/Y19vPlCYFeZFSEhfkgyAglfLCFe80WkekZu3v7ScItAOdyFASF+T9\nIMDzwtq8Bt6J1K78kraTtW69SEmJC/I5P1BzvUgd2DOwl7yCvEgpiQvyQQApDwV5kRq3pztc27tF\nffIiJSUuyIfN9R4pT8vaitSy7r4cAC0NifsYE4ktcU9HLghUkxepA5lsGOQb0lq3XqSUxAV5P5pC\nl/LAD8LR9iJSe3r7w83smtKJ+xgTiS1xT0dQMLoetH69SK3K9Ic1+UY114uUlLinIxcEeITN9YCW\nthWpUZlsWJNvTCXuY0wktsQ9HX5waMU7UE1epFblm+sb1VwvUlLino4gCKI+edXkRWpZJpsb2IdC\nRIpLXJDP+eqTF6kHmaxPc0MK7TMrUlrigrwfQCrlkW/hU5AXqU29/TmaG9MEivIiJSUwyIcD71Lq\nkxepaZmsT5NG1osMKdZ6kGZ2FvBZIA3c4py7ftD7zcBdwEqgC7jAObcleu9DwGogB1zpnFsfM83P\nAX/tnJsy4tIV4efXrh/ok/crmbyIjJNM1qc5nVI9XmQIZb8Gm1kauAk4G1gOXGRmywedthrY7Zw7\nHvgMcEN07XLgQuAk4CzgZjNLl0vTzE4DZo6ybEX5BWvXg2ryIrUqbK5Xn7zIUOK0dZ0ObHbOPe2c\n6wPWAKsGnbMKuDN6fQ/wejPzouNrnHMZ59wzwOYovZJpRl8A/h/wf0ZXtOIK164Hja4XqVWZrE+T\navIiQ4rTXL8Q2FrwcydwRqlznHNZM9sLzI6OPzLo2oXR61JpXgGsc849b2ZxykAmk6G9vX3Ic3p7\ne2lvbyfn+xw8sJ9U334Ann5mC307a3eryny56k09lqseyzSRMv059cmLlFFVezSa2QLgPODM4VzX\n3NzMsmXLhjynvb2dZcuWEfAM06ZNY97UFmAPbYsWs+zoaSPO80TLl6ve1GO5qqFMGzZsmND7V1Jv\n1mdSY+1+QRcZD3G+Bm8DFhX83BYdK3qOmTUA0wkH4JW6ttTxU4Hjgc1mtgWYZGab4xUlHj8ISAEp\nTaETqWlhTV4L4YgMJU5N/lFgqZktIQzEFwIXDzpnHXAJ8HPgbcADzrnAzNYBXzezG4EFwFLgl4BX\nLE3n3EbgqHyiZnYgGsxXEUEQEATgeRSMrleQF6lFfVlfS9qKlFH2CXHOZQn7ydcD7cBa59xGM7vW\nzM6NTrsVmB3Vuq8Cro6u3QisBTYBPwAud87lSqVZ2aIdKR/Pw8Vw8qPrNYVOpBb1qk9epKxYffLO\nuXuBewcd+1jB617CvvRi114HXBcnzSLnVHSOfL5p3qNgdH1ONXmRWpQfXS8ipSXqCfGDMKCnPK14\nJ1LrMmquFykrUU9IFOPxtOKdSM3r7c/RmNbAO5GhJCrI56Io73kMbFCj5nqR2pPzA7J+oD55kTIS\n9YQcaq7Xincitaw/+nKu5nqRoSXqCQmilnmvYO36bE7N9SK1JjMQ5NVcLzKURAX5XOHAO9XkRWpW\n30CQ14p3IkOpqmVtx1q+ud7DGxhdryAvMjLltouOzjkfuIZwr7jfOucGL6Q1Ivnm+gbV5EWGlKia\n/ECQLxhdryl0IsMXZwtqM1sKfAj4I+fcScDfV+r+aq4XiSdZQf6wPvnwtfrkRUYkzhbUlwE3Oed2\nAzjntlfq5n3Rc6uBdyJDS2RzfWGfvGryIiMSZwvqEwDM7H8Jm/Svcc79oBI3H2iuT6kmLzKURAZ5\nz1OfvMg4aCDclOpMwp0mHzKzlzrn9pS6IJPJ0N7eXjbhAz0ZAPbs6qIjs7sima0GfX19dHQ8O9HZ\nqKh6LBOMbbla+6fTta2vImklK8gXNtdrdL3IaMTZgroT+IVzrh94xsyeJAz6j5ZKtLm5mWXLlpW9\n+SMdvwJgwVHzaZs5aXg5r2IdHc+yePExE52NiqrHMsHYlmvB3MnMm9Yy5DkbNmyIlVaiOrQGmusL\nRtfnp9WJyLAMbEFtZk2E20WvG3TOfxLW4jGzOYTN909X4uZ90ZfzxlSiPsJEhi1RT4hfuKytdqET\nGbGYW1CvB7rMbBPwY+CDzrmuStw/k42CfIP65EWGkqzm+sPWrtd+8iKjEWML6gC4KvpTUfmBd9pq\nVmRoiXpC8t3v4dr14WvV5EVqT9/AYjiJ+ggTGbZEPSGHVrxDo+tFalifFsMRiSVRQT4/J96LdqHz\n0Dx5kVqkxXBE4knUE5IfSJ8fkJtKeQryIjWoLxfQlE7hearJiwwlUUG+cIMaCEfYK8iL1J6+XEBz\nQ4pAU2BFhpSoID/QXB/9nE556pMXqUF9uYDmxkR9fImMSKKekkOj6w/9rSAvUnv6cgFNDYn6+BIZ\nkUQ9JQNNe1E/XirlkdU8eZGa05cN++TVWi8ytEQF+Xxzfb4mn/Y8fNXkRWpO2FyfnuhsiFS9RAX5\nfDzPD7xrSHtksqrJi9Sa/Oh6fUUXGVqignxQsKwtQHNDmt7+3ATmSERGoj8X0KSFcETKirV2vZmd\nBXwWSAO3OOeuH/R+M3AXsBLoAi5wzm2J3vsQsBrIAVc659YPlaaZ3QqcRjgI/kngUufcgdEVM5Tf\ncS4VRfmmhhQ9CvIiNScXBFrSViSGsk+JmaWBm4CzgeXARWa2fNBpq4Hdzrnjgc8AN0TXLifcgvIk\n4CzgZjNLl0nzH5xzpzjnXgZ0EO50VREDzfVRkG9uSNHTpyAvUmty/qEv6yJSWpyvwqcDm51zTzvn\n+oA1wKpB56wC7oxe3wO83sy86Pga51zGOfcMsDlKr2Sazrl9ANH1rVC5bjffP7y5vqUxrZq8SA3y\ng2BgAK2IlBanuX4hsLXg507gjFLnOOeyZrYXmB0df2TQtQuj1yXTNLPbgXOATcA/lstgJpOhvb19\nyHN6e3t5dkcHALt2bid1MEW2r5f9vX7Za6tZb29vTee/lHosVz2WaaLkgkPbRYtIaVW5n7xz7t1R\nk/5/ABcAtw91fnNzM8uWLRsyzfb2dha2zQJeZN7c+bTNmsTMp/vZ2XOg7LXVrL29vabzX0o9lqsa\nyrRhw4YJvX+lhDV5BXmRcuI0128DFhX83BYdK3qOmTUA0wkH4JW6tmyazrkcYTP+W2PkMZbCXegg\nHF2v5nqR2uOrJi8SS5ya/KPAUjNbQhiILwQuHnTOOuAS4OfA24AHnHOBma0Dvm5mNwILgKXALwlH\nzh+RZtQPf5xzbnP0+lzgidEWMi8IDl8Mp7khRUZBXqTm5Hz1yYvEUbYm75zLEo5wXw+0A2udcxvN\n7FozOzc67VZgtpltBq4Cro6u3QisJexb/wFwuXMuVypNwuB/p5n9DvgdcDRwbaUKe8To+sYUvVlf\nq96J1Bg/CJelFpGhxeqTd87dC9w76NjHCl73AueVuPY64LqYafrAH8XJ00jkjqjJh8tiZrI+rU1a\nIlOkVuQCTaETiSNRq0kcueJdWPzuvuxEZUlERsAPAvXJi8SQqCDvlwzy6pcXqSXhYjgTnQuR6peo\nIJ+L9qJJFax4B2j9epEaoyl0IvEkKsj7g9auz/fJqyYvUltyvqbQicSRqCB/RJ98Y1h8zZUXqS2q\nyYvEk6ggn2+uz380DPTJZzTwTqSWhFPoJjoXItUvUY/JwMC71OHN9QcU5EVqSi4ISKsmL1JWooJ8\ncESfvEbXi9QiX/PkRWJJVJAfWLs+WuCuSUFepOYEQRAF+YnOiUj1S1SQz69em18Os6UxbK4/qMVw\nRGpG/su6lrUVKS9hQT6qyUc/52vyvarJi9SMrH94t5uIlJbIIJ//cEh5Ho1pT831IjVk8HMsIqUl\nLMiHfxe28rU0pOnWPHmRmqHmepH4Yu1CVy8OrV1/6MOhqSFFj2ryIsNmZmcBnwXSwC3OuetLnPdW\n4B7gFc65X432vgNBXjFepKxk1eT9w1e8g3DVux4NvBMZFjNLAzcBZwPLgYvMbHmR86YCfwf8olL3\nzqlPXiS2ZAX5geb6Qx8OzQ1pevr9CcqRSM06HdjsnHvaOdcHrAFWFTnvE8ANQG+lbpwP8loMR6S8\nhDbXQ/SS5oaU1q4XGb6FwNaCnzuBMwpPMLOXA4ucc/9tZh+Mk2gmk6G9vX3Ic3YcDFve9u7dQ0fH\nweHkuer19fXR0fHsRGejouqxTDC25Wrtn07Xtr6KpJWsIO8HeER98lGUb1afvEjFmVkKuBG4dDjX\nNTc3s2zZsiHPmbKrG+hg1qyZLF48e8R5rEYdHc+yePExE52NiqrHMsHYlmvB3MnMm9Yy5DkbNmyI\nlVbimusH9+M1N6a1n7zI8G0DFhX83BYdy5sKnAw8aGZbgFcC68zstNHe+NAUutGmJFL/klWTDwIG\nd+M1p9VcLzICjwJLzWwJYXC/ELg4/6Zzbi8wJ/+zmT0I/FNlR9cryouUk6iafC4ISKc8CA4da25M\nqSYvMkzOuSxwBbAeaAfWOuc2mtm1ZnbuWN5bU+hE4ktUTT4IOLIm36DmepGRcM7dC9w76NjHSpx7\nZqXum9OKdyKxJasm7wekPI+goCrf3JCit98f2IZWRKpbNqcV70TiSlSQ94PgiLm1zQ0pAiCT1Vx5\nkVqgtetF4ktUkC/aXJ/fbjajVe9EaoF2oROJL1FBPt9cXyi/p7x2ohOpDQPLU09wPkRqQayBd+U2\nojCzZuAuYCXQBVzgnNsSvfchYDWQA650zq0fKk0z+xpwGtAP/BJ4n3Ouf3TFDPlBcEQ/XmsU5A+o\nJi9SE7LahU4ktrI1+ZgbUayARmbxAAAcqUlEQVQGdjvnjgc+Q7hWNdF5FwInAWcBN5tZukyaXwNO\nBF4KtALvGVUJC4SL4Rx+rKUx/BWouV6kNviaQicSW5ya/MBGFABmlt+IYlPBOauAa6LX9wCfNzMv\nOr7GOZcBnjGzzVF6lEozmpZDdPyXhCtpVYTvB4dtMwuHmusPqrlepCZoCp1IfHGCfNmNKArPcc5l\nzWwvMDs6/sigaxdGr8ttbtEIvJNwm8ohxdnUore3l1179hH4ObZu7RhYUGP3/rAn4A/PbGVebme5\nW1Wd3t7esmWvRfVYrnos00TQwDuR+Kp5MZybgYeccw+XOzHOphbt7e1Mm5aiccdO2hYtHtiFrnV/\nL/xsJ5NnzGHZssWVyPe4am9vL1v2WlSP5aqGMsXd1KKa+eqTF4ktTpAvtxFF4TmdZtYATCccgDfU\ntSXTNLOPA3OB98XIX2x+UHp0vQbeidSGrPrkRWKLE+SH3Igisg64BPg58DbgAedcYGbrgK+b2Y3A\nAmAp4Yh5r1SaZvYe4E3A651zFV2hJhxdz2Fr1yvIi9QWX831IrGVHV0fcyOKW4HZ0cC6q4Cro2s3\nAmsJB+n9ALjcOZcrlWaU1heB+cDPzewxMyu6FvZI+AGkBs2ubUynSKc8ja4XqRH5mvzg1StF5Eix\n+uTLbUThnOsFzitx7XXAdXHSjI6P2TiBcHT9kcdbGlMK8iI1Ir+srZeopbxERiZRj4mf32p2kNbG\nNAf6FORFakFONXmR2BIX5AfPk4ewX747o3nyIrVAA+9E4ktUkM/54QfD4E1lWxrTHFRNXqQm5Afe\nFWuVE5HDJSrIB0Wm0EEU5FWTF6kJWgxHJL5EBfli8+QBWhpSdKsmL1ITDg28U5AXKSdRQT5XZD95\nCAfeaatZkdqQzUU1+QnOh0gtSNRzUrK5vklBXqRW+NqgRiS2RAX5sLn+yONqrhepHTkNvBOJLVFB\nPucHRTe1aGlM058L6M9VdBVdERkDGngnEl+igny4rO2R8uvXa668SPXTFDqR+BIV5IMSi+G05jep\nUZO9SNXL1+RVkRcpL1FBPucXX9b2UE1eQV6k2uXH1hT7wi4ih0tUkPcDKPax0NIY/hq03axI9cv6\nxQfQisiREhXkyzXXa9U7kern+9FU2MHrU4vIERIV5HNBQKpIiZvzffKZ/nHOkYgMV9YPSHsQKMqL\nlJWoIO/7kCrSYJ+vye/vVXO9SLXL+cUXtRKRIyUryAdB0RG5+T55BXmR6ucHxde7EJEjJS7Il1oM\nB+CgBt6JVL18c72IlJewIF98lazmhhQpD/b1qk9epNr5JVauFJEjNUx0BsaT7wdFp9B5nsf01kZ2\n7O8b9zyJ1CozOwv4LJAGbnHOXT/o/auA9wBZYAfw1865Z0d7X02hE4kvYTX50jWAmZOa2L6/d5xz\nJFKbzCwN3AScDSwHLjKz5YNO+w1wmnPuZcA9wP+txL3zU+g0tl6kvGTV5ANK1gBmTm5i+/7M+GZI\npHadDmx2zj0NYGZrgFXApvwJzrkfF5z/CPCOStx4oE9eUV6krEQF+Zwf4BVtsIeZkxp5avuBcc6R\nSM1aCGwt+LkTOGOI81cD95VLNJPJ0N7ePuQ5e/buxfNg69aOgb3l60VfXx8dHaPu0agq9VgmGNty\ntfZPp2tbZbqPExXkgxKL4UDYXL+np59MNkdzQ3p8MyZSx8zsHcBpwJ+UO7e5uZlly5YNec7kDd00\n7NnFokWL664y39HxLIsXHzPR2aioeiwTjG25FsydzLxpLUOes2HDhlhpJaxPvvja9RAGeYAdarIX\niWMbsKjg57bo2GHM7A3AR4BznXMVebiyWgxHJLZE1eRzQekPh5mTGwF4cV+GtpmTxjNbIrXoUWCp\nmS0hDO4XAhcXnmBmpwJfAs5yzm2v1I3DKXTqkheJI1aQjzFVphm4C1gJdAEXOOe2RO99iLA/Lgdc\n6ZxbP1SaZnYF8PfAccBc59zOUZZxQDBUkI9q8i/s7QFmVuqWInXJOZeNntX1hM/wbc65jWZ2LfAr\n59w64P8BU4BvmRlAh3Pu3NHeOxx4p5q8SBxlg3zBVJk/Ixxc86iZrXPObSo4bTWw2zl3vJldCNwA\nXBBNqbkQOAlYAPzIzE6IrimV5v8C3wcerEQBC/kBRZe1hUNB/vm9mkYnEodz7l7g3kHHPlbw+g1j\ncV9/iC/rInK4OH3yA1NlnHN9QH6qTKFVwJ3R63uA15uZFx1f45zLOOeeATZH6ZVM0zn3m3wrQKXl\n/OJbzQJMaWkgnfJ4QUFepKplc1oMRySuOM31cabKDJwTNePtBWZHxx8ZdO3C6PVwpt8MKc60m97e\nXrLZHD0HD5ac9jCtKcVTz3eVTaua9Pb21lR+46rHctVjmSZCOLZmonMhUhvqYuBdnGk37e3teKkU\nU6dOYfHitqLnzJm2n26/oWxa1aS9vb2m8htXPZarGsoUd9pNNfP9gLSivEgscZrr40yVGTjHzBqA\n6YQD8EpdG2v6TaXlyqx5HS5tqyl0ItVMU+hE4otTky87VQZYB1wC/Bx4G/CAcy4ws3XA183sRsKB\nd0uBXxJOVy+XZsWF+8mX/nCYMamRTc/vG+tsiMgo+EFAg2K8SCxla/LOuSyQnyrTDqzNT5Uxs/x0\nmFuB2Wa2GbgKuDq6diOwlnA96x8AlzvncqXSBDCzK82sk7B2/7iZ3VKpwgZDrF0PMH9aCwcyWXYd\n1G50ItVKA+9E4ovVJx9jqkwvcF6Ja68DrouTZnT8c8Dn4uRruHJlavILZoTLCD614wCzJs8aiyyI\nyChpCp1IfAlb1jYYssALprcCsFkb1YhULe0nLxJfYoJ8EAQEQyyGAzBnSjONaY8nX9w/fhkTkWHx\nNfBOJLbEBHk/Wug6NUQVIJXyOHp6q7acFaliqsmLxJeYIJ/fzKLUfvJ5R09v4Zmug2OfIREZkZwf\nDPllXUQOSUyQz9fky7XyLZjRynO7e8hkc2OfKREZNj8ISCvGi8SSoCAfRvlyfXkLZrSSC6Cjq3s8\nsiUiw5QdYg8KETlcXSxrG0cQtyY//dA0uqXzp45xrkRkuHz1ySeHn6O5bzcpP3voQ7xKHDcjTcPB\n58ck7e29Kbqi/+QtLS20tbXR2Ng4orQSE+QPNdeXr8kDtD+/n7NOPnqssyUiw6SBd8nR3LebebNm\nMH3GrKprvenry9DU1DwmaTc3pmhMpwiCgK6uLjo7O1myZMmI0kpgc/3Q57U0ppk/rVnL24pUKd8P\nSJUZQCv1IeVnqzLAjxfP85g9eza9vSPfAj0xQT7f0hOnBnDMrMm4FzRXXqQaZf2AVGI+uRKuzCql\nSTDa8ifmUfEHXpX/hS2aNYmtu7rp7suOZZZEZARyZVauFJFDEvOsxB1dD3DM7EkEwJMvalEckWrj\na568VKGvf/Uuenp6Bn6+8vL3s3/fxHf7JibIxx1dD7B41iQANj63dwxzJCIjoYF3MlGCIMD3/aLv\nff1rdx/Wd/65m77I1GnTxitrJSVvdH2Mc+dObaa1Mc3GbQryItXE9+O3yEl9+d5j2/jub7ZVNM2/\nPHUhq1YsHPKc57Zt44q/eS8nv/RltG/ayEknv5TNm/9Ab08Pb3jjm3j/317BN772VXZs38773vNu\nZsyYwZdvvYM3n/1n3P31tfR0d/OBy9/PilNP5fHHHmPuvPnc+Nn/oKWlhY2//x3XXvMxUimPM175\nan7204dZ+53vVbSMiajJ/2jTi/xwcziQLk4zX8rzWDSrlXYNvhOpKtkoyCvGy3jq6HiW8y64kG99\ndx3/8E8f5KvfWMvd3/gmv97wK/7wpOOit7+DufPm8aVbbufLt95xxPVbO57l/Asu4lvfXcfUaVN5\n4Ec/BOBfP/ZRPvIvH+cba79DaoxGkyaiJv/dx7Zx/+/3APE/HJbOm8r9m16g60CG2VPGZi6kiAzP\nwNiaCc6HjL9VK8rXusfK0Ucv4KUvOwWAH65fz3e+/S2y2X66urp4+qmnWHqCDXn9goULsROXAbBs\n2XKee24b+/ft42D3QV52ygoAzj7nz/npQz+peN4T8azMn9pCf74GEHN+7Z+cMJf+XMA9GzrHMmsi\nMgxZNdfLBGhtDRdJ29bZyd133c4Xv3IrX/3GWl7z2j8m05cpe31TY9PA61Q6TS43fnujJCLIHzX9\nUE087oCdRbMmYfOn8rVfdBBU2XKKIkmVU3O9TKCDBw/Q2trKlClT2dXVxc9++tOB9yZNmkz3wfg7\nmE6dNo3Jkybzu8cfB2D9D+6reH4hIc3186e1DLwezofD65fN4+YHn+KRp3fxquNmj0HORGQ4NPBO\nJtIJdiJ24jLeuurNzJ03j1NWnDrw3l+99Tw+8LfvY87cuUX75Yv5l2s+wSev/TiplMfLV76CKVOn\nVDzPiQjyRxUE+eF8OJyxZDZ3/mwLX31ki4K8SBU4NPBOrWsyPhYsXHjYiPd//cS/AUeuXX/hxW/n\nwovfPvDz9+8LB9fNnDnzsOvfdcm7B14fd/xxfPOe7wJw+61fYdnykyue/2QE+ekFNflhXNfUkOK1\nS+dy/6YX2XWwj1mTm8pfJCJjJj/wLq2avNSBnz70ELff9hVy2RxHLVjAv157XcXvkYggf3hz/fA+\nHF534jx+sPEF7vnVVt77J8dVOmsiMgwDNfkJzodIJbzxrLN541lnj+k9EjHwrqUxzdTmsKjDnYq4\naNYkTjxqKp97YDO/7tg9BrkTGbldB/t4Ye/Id6iqNfk+eS8Rn1yC5yV+4PNoy5+YR2X2pEYg/hS6\nQpf/6fFMbWngHbf8QrvTSVW5au1jvOu2X0x0NsbNwBQ61eUTwU81sHfPrsQG+vx+8i0tLeVPLiER\nzfUAsyY1sGV3ZkQfDXOmNPPRP1/Oh7/7O/7xW4/xn3/7RzSkE/P9SKpUT1+Onz3VRV/Wp3N3N20z\nJ010lsacptAlS6ZpJtt37Wbnzp2HNiCpEtlcjoZ0ekzSbkinSEfzvVtaWmhraxt5WpXKVLWbPSks\n6kg3tpg1uYlLXnUsn3vgD/z1nb/ijCWzWP2aJbQ0js0/skgp9/3ueR7buodXHz+Hvmy4WcbDT+7k\nojMWT3DOxl4+yGvgXUKk0mRa5kx0Lorq6HiWxYtHHnyHctzcycybNvLae6FYQd7MzgI+C6SBW5xz\n1w96vxm4C1gJdAEXOOe2RO99CFgN5IArnXPrh0rTzJYAa4DZwAbgnc65vtEVM6zJA6OqArzyJbN4\nZufRPLplNw89uYMfu+1c9YYTmDm5iWVHT/xuQ1J/+rI+TQ0pft2xm5+4Hax+7RL+5Xu/Z+eBPn7x\nzC4a0x6Tmxp48Mnt4x7kR/O5MFKqyYsMT9kgb2Zp4Cbgz4BO4FEzW+ec21Rw2mpgt3PueDO7ELgB\nuMDMlgMXAicBC4AfmdkJ0TWl0rwB+Ixzbo2ZfTFK+wujLejsyWGf/Gi2qPQ8j4vPOIaLzziGXzzd\nxU0PbubiW8L+0NcunUNzQ5re/hx/eepCuvtzBEHAK46dxYv7emlpTHNK2wye3XWQloY0bTNb2ban\nh8nNDcyc1MT2/b1MbWmkuSHF9v0ZZk1qojHtsaenn+mtjaQ8j4N9WaY0NRAAmWyO1sY0OT+gtz9H\nc0OKIAinGDWkU2RzPgHQmE4dqv2kPLI5n5TnkYpep1Menufh+wGex8Dr/EY+QRAMzEgodTz/OggC\ngiDcBCjfh+Z5pV8Pdc+8/PHBaRd7PbBQSsoj5wd40evCMvdlfRpSHp4HmaxPUzqF50FPf46WhrBV\nprs/x6TGNH4Q0N2fY0pTA/2+T09fjumtjfT05+jpyzFzUhP7e7P05XxmT25ix4EMQQBzpjSxdXcP\nTQ0p5kxp4g8vHmB6ayMH+3x++oedHD2jhclNDfzsqZ2ceNQ0/CDgx09s59XHz6Fzdzf3b3qRt6xY\nyH2/f57/fvx5PvC64/nyQ0+zrzfL9x7bxs4DfUxpbuCxrXs4fcksWhvT/PzpLnJ+MNDEN9ZG87kw\nmvv6gYK8yHDEqcmfDmx2zj0NYGZrgFVA4cO8Crgmen0P8Hkz86Lja5xzGeAZM9scpUexNM2sHXgd\ncHF0zp1RuqMP8lFNvjGdqshe1K86bjbHz5vC7p5+ntq+n+899hxTWxrJ+QH/+K3flr3e8w51MaW8\nQ1vhFnud/0ALgjBQ+1Fga0h50UCkZ2hMe/Tnwgub0in6cv6RrxtS9GV9PC/8PfRlfVJe2P+TD375\nQNiY9vDw6MuFgTAgoD8X0NSQwvcDsn5Ac0P4BSL/uj/n4wfQ3HDons0NKTJZHy+6f29/+MWiIeWR\nyYav00Xu2ZjywHvmiHs2RffMFblnJnv4PQvLXPi78Lzwd9efC79kpD3vsD3K/UG/53T0pQEOf13q\n363w3/ZIW0r+n/j0D58EoLUxzX8//jyeBy+ZM5lP3f8ksyc38caT5nP/xhf546VzWHb0NL700NOc\nungGzekUP3lyB0/vOMDS+VNLpl9hI/5ccM6NunO1qULPcbVpqMNy1WOZoHbKFSfILwS2FvzcCZxR\n6hznXNbM9hI2ty8EHhl0bX4boWJpzgb2OOeyRc4vqbu7e+eGDRueHeqcFbO8Sd8+f0Ej7CLYs8sv\nl2YcRwFHNcGyNviLtrkQzlbwYeZho/LyH/opjwAgits+4AWAx2HHg0HH8/+NqmvUSTJ5HPp3GHhd\n8O9W9HUQnuvlv7MFAV7Kww8gFf2/8P0gnOmS8giiRoj87CEPpudfB+87aUEqTKjXf+P5C1LQSwD+\nH5+/gH2dT/Zs6CQ7ONODHFOh38VoPhd2lko0zrP87fOOSuOlJgV7OkaU8Wp2zBSgzspVj2WCsS3X\n1j34WwO/3EL4sZ7luhh4t3LlyrkTnQcRGT09yyKVFWce2DZgUcHPbdGxoueYWQMwnXCgTalrSx3v\nAmZEaZS6l4hMvNF8LojIOIkT5B8FlprZEjNrIhxIt27QOeuAS6LXbwMeiPrd1gEXmllzNGp+KfDL\nUmlG1/w4SoMoze8hItVmNJ8LIjJOygb5qH/8CmA90A6sdc5tNLNrzezc6LRbgdnRwLqrgKujazcC\nawkH4/wAuNw5lyuVZpTWPwNXRWnNjtIWkSoyms8FERk/XlKXCxQREal3WptVRESkTinIi4iI1Km6\nmEJXTrnlN6uNmW0B9hMuBZx1zp1mZrOAbwLHEq6ocr5zbne06NBngXOAbuBS59yvo3QuAT4aJftJ\n59yd41gMzOw24M3AdufcydGxipXDzFYCdwCtwL3A343HwK4S5boGuAzYEZ32YefcvdF7VbG0cz2o\ntWcZ9DzHLcdEPM9JeJbrviZfsPzm2cBy4KJoud1q96fOuRXOudOin68G/sc5txT4Hw4NYjqbcNbC\nUuC9RKsDRg/fxwkXKDkd+LiZzRzH/EP4wJ416Fgly/EFwocxf93ge42VO0rc6zPRv9mKgg+FwqWd\nzwJuNrN0mf+X+aWdjwd2E36oJF4NP8ug57lan+c7Stynbp7lug/yFCy/GX2Dyi+/WWtWES7zS/T3\nWwqO3+WcC5xzjxCuM3A08Cbgh865Xc653cAPGb8gCIBz7iFg16DDFSlH9N4059wj0bf9uwrSGlMl\nylXKwNLOzrlngPzSzkX/X0Y1oNcRLgMLh/+Okq5enmXQ81wVz3MSnuUkBPliy2+WXSp3ggXA/Wa2\nwczeGx2b75x7Pnr9AjA/el2qfNVa7kqVY2H0evDxiXSFmT1uZrcV1E6GW64RLe2cENX6f7ocPc+1\n9zzXzbOchCBfi17jnHs5YfPP5Wb2x4VvRt90a37uY72UI/IF4DhgBfA88OmJzY5UET3PtaWunuUk\nBPk4y29WFefctujv7cB3CZuDXoyatIj+3h6dPtylgydapcqxLXo9+PiEcM69GC305ANf4dBui1ra\nuXKq9f/0kPQ8AzX0PNfbs5yEIB9n+c2qYWaTzWxq/jXwRuD3HL5EaOFyv+uAd5mZZ2avBPZGzWfr\ngTea2cyouemN0bGJVpFyRO/tM7NXRn1f72ICl0DOf9BF/pLw3wy0tHMl1dSzDHqea/F5rrdnue6n\n0EVbXOaX30wDtxUsoVuN5gPfNTMI/32+7pz7gZk9Cqw1s9XAs8D50fn3Ek5T2Uw4VeXdAM65XWb2\nCcL/gADXOufiDjCpCDP7BnAmMMfMOglH1V5P5crxtxyacnNf9GfMlSjXmWa2grC5cgvwvij/G80s\nv7Rzlmhp5yidUv8v/xlYY2afBH6DlnYGavJZBj3PVf08J+FZ1rK2IiIidSoJzfUiIiKJpCAvIiJS\npxTkRURE6pSCvIiISJ1SkBcREalTCvJVxMwCM/t0wc//ZOGOSKNNt9nMfmRmj5nZBSNM40wze3XB\nz+83s3eNNm9D3O8OM3tb+TNFqo+e5cPup2d5AinIV5cM8FdmNqfC6Z4KEO2o9M1SJxWszFTMmcDA\nB4Nz7ovOubsqlsMqV+Z3IzKYnuUqlbRnOVGFrQFZ4MvAPwAfKXzDzI4FbgPmEO5z/G7nXMegc2ZF\n57yEcAGK9xJuGvFVYK6ZPQa81Tn3VME1DwKPAa8BvmFmTxLu9dxEuCzj2wkXp3g/kDOzdwAfAF4P\nHHDOfSpK4xfAnwIzgNXOuYfNbBLh4hYnAw5YAFzOoUUhTiNccOI259xnivw+3mBmVwPTgKucc983\ns4cI93F+LMr/TwkXpfhtQZkuBU5zzl0R/fx94FPAw8Xua2bHEW4VOTf6vV3mnHvCzO4Aegk/WP8X\nuKpIHkWK0bN8OD3LE0Q1+epzE/B2M5s+6Ph/AHc6514GfA34XJFr/xX4TXTOhwm3etwOvAd4OPr2\n/1SR65qcc6c55z4N/BR4pXPuVMItE/+Pc24L8EUO7bH8cJE0GpxzpwN/T7hqFIQrWO12zi0H/gVY\nGR1fASx0zp3snHspcHuJ38WxhOtG/znwRTNrIXywLwUwsxOAlsIPhTJK3ffLwAeccyuBfwJuLrim\nDXi1cy4xHwpSMXqWDzkWPcsTQkG+yjjn9hHupXzloLdeBXw9en034bf1wV4TvYdz7gFgtplNi3Hb\nwma/NmC9mf0O+CBwUsysfyf6ewPhA53Pz5ooP78HHo+OPw28xMz+w8zOAvaVSHOtc853zv0huuZE\n4FvAm82sEfhrwtpFXEfc18ymEDZdfiuqHX0JKFy7+lv5pStFhkPP8mH0LE8QBfnq9O/AamDyON3v\nYMHr/wA+H307fh/QEjONTPR3jjLdQM653cApwIOETYe3lDh18JrLgXOuG/ghsIpwneyvFbkuy+H/\nt1uGuG+KcM/nFQV/lhVcW/i7ERkuPcshPcsTREG+CkUbNqwl/HDI+xnh7kYQ9q0Va2Z7OHoPMzsT\n2BnVJoZjOoe2Q7yk4Ph+YOow0/pfog0rzGw58NLo9Rwg5Zz7NmGf4ctLXH+emaWifraXEPYFQvhA\nfw54NHrYB9sCrIiuXUS0VWSx+0a/n2fM7LzoHM/MThlmOUWK0rM8QM/yBNHAu+r1aeCKgp8/ANxu\nZh8kGqxT5JprgNvM7HHCQSeXFDmnnGsIm7t2Aw8AS6Lj/wXcY2arorzEcTNwp5ltAp4ANgJ7gYVR\nWfJfMj9U4voOwq0cpwHvd871AjjnNpjZPkr3//0v8AzhblHtwK+j46Xu+3bgC2b2UaCRsFkybt+g\nSDl6lvUsTxjtQidjxszSQKNzrjf6Bv8jwJxzfaNMdwFhM92Jzjl/9DkVkaHoWa5dqsnLWJoE/Dga\nWOMBf1uBD4V3AdcRTsPRh4LI+NCzXKNUkxcREalTGngnIiJSpxTkRURE6pSCvIiISJ1SkBcREalT\nCvIiIiJ16v8DVNbwziTrvw4AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 576x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "0217ea22-3546-4f56-befc-f9a78c21b651",
        "id": "AVN_18IHGHJ5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "no_of_rated_movies_per_user.describe()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    405041.000000\n",
              "mean        198.459921\n",
              "std         290.793238\n",
              "min           1.000000\n",
              "25%          34.000000\n",
              "50%          89.000000\n",
              "75%         245.000000\n",
              "max       17112.000000\n",
              "Name: rating, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "D5c1dpEBGHJ8"
      },
      "source": [
        "> _There,  is something interesting going on with the quantiles.._"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "ftxPLUPcGHJ9",
        "colab": {}
      },
      "source": [
        "quantiles = no_of_rated_movies_per_user.quantile(np.arange(0,1.01,0.01), interpolation='higher')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9KOcgUldwWcm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "6a7ef012-2612-40b9-8516-43637e338b1f",
        "id": "1EK2PIQMGHMv",
        "colab": {}
      },
      "source": [
        "plt.title(\"Quantiles and their Values\")\n",
        "quantiles.plot()\n",
        "# quantiles with 0.05 difference\n",
        "plt.scatter(x=quantiles.index[::5], y=quantiles.values[::5], c='orange', label=\"quantiles with 0.05 intervals\")\n",
        "# quantiles with 0.25 difference\n",
        "plt.scatter(x=quantiles.index[::25], y=quantiles.values[::25], c='m', label = \"quantiles with 0.25 intervals\")\n",
        "plt.ylabel('No of ratings by user')\n",
        "plt.xlabel('Value at the quantile')\n",
        "plt.legend(loc='best')\n",
        "\n",
        "# annotate the 25th, 50th, 75th and 100th percentile values....\n",
        "for x,y in zip(quantiles.index[::25], quantiles[::25]):\n",
        "    plt.annotate(s=\"({} , {})\".format(x,y), xy=(x,y), xytext=(x-0.05, y+500)\n",
        "                ,fontweight='bold')\n",
        "\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "9fb24bb8-f60d-4361-c834-1e2ff0cfd469",
        "id": "7PBxPq52GHMz",
        "colab": {}
      },
      "source": [
        "quantiles[::5]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.00        1\n",
              "0.05        7\n",
              "0.10       15\n",
              "0.15       21\n",
              "0.20       27\n",
              "0.25       34\n",
              "0.30       41\n",
              "0.35       50\n",
              "0.40       60\n",
              "0.45       73\n",
              "0.50       89\n",
              "0.55      109\n",
              "0.60      133\n",
              "0.65      163\n",
              "0.70      199\n",
              "0.75      245\n",
              "0.80      307\n",
              "0.85      392\n",
              "0.90      520\n",
              "0.95      749\n",
              "1.00    17112\n",
              "Name: rating, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "aPsoDhzwGHM3"
      },
      "source": [
        "__how many ratings at the last 5% of all ratings__??"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "c7f1eab1-d5ad-4381-d9f2-0a0e11a38ab6",
        "id": "YQXyKqLxGHM3",
        "colab": {}
      },
      "source": [
        "print('\\n No of ratings at last 5 percentile : {}\\n'.format(sum(no_of_rated_movies_per_user>= 749)) )"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            " No of ratings at last 5 percentile : 20305\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Bw4YrJhSGHM6"
      },
      "source": [
        "<h3> 3.3.4 Analysis of ratings of a movie given by a user </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "f9e2f900-fc75-48a6-98c5-378f9711552e",
        "id": "Dk-VW2CnGHM7",
        "colab": {}
      },
      "source": [
        "no_of_ratings_per_movie = train_df.groupby(by='movie')['rating'].count().sort_values(ascending=False)\n",
        "\n",
        "fig = plt.figure(figsize=plt.figaspect(.5))\n",
        "ax = plt.gca()\n",
        "plt.plot(no_of_ratings_per_movie.values)\n",
        "plt.title('# RATINGS per Movie')\n",
        "plt.xlabel('Movie')\n",
        "plt.ylabel('No of Users who rated a movie')\n",
        "ax.set_xticklabels([])\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"960\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "GjNmHPGCGHM_"
      },
      "source": [
        "- __It is very skewed.. just like nunmber of ratings given per user.__\n",
        "    \n",
        "    \n",
        "    - There are some movies (which are very popular) which are rated by huge number of users.\n",
        "    \n",
        "    - But most of the movies(like 90%) got some hundereds of ratings."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "88gf03axGHNA"
      },
      "source": [
        "<h3> 3.3.5 Number of ratings on each day of the week</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "f2b2b524-96cb-4edc-ce04-006380c7cd3f",
        "id": "LEMc4h97GHNB",
        "colab": {}
      },
      "source": [
        "fig, ax = plt.subplots()\n",
        "sns.countplot(x='day_of_week', data=train_df, ax=ax)\n",
        "plt.title('No of ratings on each day...')\n",
        "plt.ylabel('Total no of ratings')\n",
        "plt.xlabel('')\n",
        "ax.set_yticklabels([human(item, 'M') for item in ax.get_yticks()])\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "7038fd53-fe53-4eed-f34b-40f13a718bbf",
        "id": "1Mfqv6UbGHND",
        "colab": {}
      },
      "source": [
        "start = datetime.now()\n",
        "fig = plt.figure(figsize=plt.figaspect(.45))\n",
        "sns.boxplot(y='rating', x='day_of_week', data=train_df)\n",
        "plt.show()\n",
        "print(datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"1066\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "0:01:10.003761\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "e9bd9666-b045-41bf-d5c9-6fff7c8d7c9f",
        "id": "AQPi83aoGHNG",
        "colab": {}
      },
      "source": [
        "avg_week_df = train_df.groupby(by=['day_of_week'])['rating'].mean()\n",
        "print(\" AVerage ratings\")\n",
        "print(\"-\"*30)\n",
        "print(avg_week_df)\n",
        "print(\"\\n\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " AVerage ratings\n",
            "------------------------------\n",
            "day_of_week\n",
            "Friday       3.585274\n",
            "Monday       3.577250\n",
            "Saturday     3.591791\n",
            "Sunday       3.594144\n",
            "Thursday     3.582463\n",
            "Tuesday      3.574438\n",
            "Wednesday    3.583751\n",
            "Name: rating, dtype: float64\n",
            "\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "5o1LK_m-GHNJ"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": true,
        "colab_type": "text",
        "id": "p8ppntjZGHNK"
      },
      "source": [
        "<h3> 3.3.6 Creating sparse matrix from data frame </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "NaR7yrqPGHNK"
      },
      "source": [
        "<table>\n",
        "<tr>\n",
        "<td>\n",
        "<img src='images/data_c.jpg' width='250px' align=left/>\n",
        "</td>\n",
        "<td>\n",
        "<img src='images/arrow.jpg' width='60px' align=left/>\n",
        "</td>\n",
        "<td>\n",
        "<img src='images/data_sparse_c.jpg' width='400px' align=left/>\n",
        "</td>\n",
        "</tr>\n",
        "</table>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ioKS3o7VGHNL"
      },
      "source": [
        "<h4> 3.3.6.1 Creating sparse matrix from train data frame </h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "cf2444aa-18be-4e29-95a3-693b1c22e027",
        "id": "SfWUsCnLGHNL",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "\n",
        "start = datetime.now()\n",
        "if os.path.isfile('/content/drive/My Drive/Netflix_recommender/data_folder/train_sparse_matrix.npz'):\n",
        "    print(\"It is present in your pwd, getting it from disk....\")\n",
        "    # just get it from the disk instead of computing it\n",
        "    train_sparse_matrix = sparse.load_npz('/content/drive/My Drive/Netflix_recommender/data_folder/train_sparse_matrix.npz')\n",
        "    print(\"DONE..\")\n",
        "else: \n",
        "    print(\"We are creating sparse_matrix from the dataframe..\")\n",
        "    # create sparse_matrix and store it for after usage.\n",
        "    # csr_matrix(data_values, (row_index, col_index), shape_of_matrix)\n",
        "    # It should be in such a way that, MATRIX[row, col] = data\n",
        "    train_sparse_matrix = sparse.csr_matrix((train_df.rating.values, (train_df.user.values,\n",
        "                                               train_df.movie.values)),)\n",
        "    \n",
        "    print('Done. It\\'s shape is : (user, movie) : ',train_sparse_matrix.shape)\n",
        "    print('Saving it into disk for furthur usage..')\n",
        "    # save it into disk\n",
        "    sparse.save_npz(\"train_sparse_matrix.npz\", train_sparse_matrix)\n",
        "    print('Done..\\n')\n",
        "\n",
        "print(datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "It is present in your pwd, getting it from disk....\n",
            "DONE..\n",
            "0:00:04.955848\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "XFbTLnY3GHNO"
      },
      "source": [
        "<p><b>The Sparsity of Train Sparse Matrix</b></p>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "99164627-26ef-4692-b31d-265f7bafce06",
        "id": "0Zy4dRnnGHNO",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "us,mv = train_sparse_matrix.shape\n",
        "elem = train_sparse_matrix.count_nonzero()\n",
        "\n",
        "print(\"Sparsity Of Train matrix : {} % \".format(  (1-(elem/(us*mv))) * 100) )"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Sparsity Of Train matrix : 99.8292709259195 % \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "hOlZ-zQAGHNR"
      },
      "source": [
        "<h4> 3.3.6.2 Creating sparse matrix from test data frame </h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "80f3aabe-a508-45d4-a5db-93a50e22dadb",
        "id": "5VMK9kIMGHNR",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "start = datetime.now()\n",
        "if os.path.isfile('/content/drive/My Drive/Netflix_recommender/data_folder/test_sparse_matrix.npz'):\n",
        "    print(\"It is present in your pwd, getting it from disk....\")\n",
        "    # just get it from the disk instead of computing it\n",
        "    test_sparse_matrix = sparse.load_npz('/content/drive/My Drive/Netflix_recommender/data_folder/test_sparse_matrix.npz')\n",
        "    print(\"DONE..\")\n",
        "else: \n",
        "    print(\"We are creating sparse_matrix from the dataframe..\")\n",
        "    # create sparse_matrix and store it for after usage.\n",
        "    # csr_matrix(data_values, (row_index, col_index), shape_of_matrix)\n",
        "    # It should be in such a way that, MATRIX[row, col] = data\n",
        "    test_sparse_matrix = sparse.csr_matrix((test_df.rating.values, (test_df.user.values,\n",
        "                                               test_df.movie.values)))\n",
        "    \n",
        "    print('Done. It\\'s shape is : (user, movie) : ',test_sparse_matrix.shape)\n",
        "    print('Saving it into disk for furthur usage..')\n",
        "    # save it into disk\n",
        "    sparse.save_npz(\"test_sparse_matrix.npz\", test_sparse_matrix)\n",
        "    print('Done..\\n')\n",
        "    \n",
        "print(datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "It is present in your pwd, getting it from disk....\n",
            "DONE..\n",
            "0:00:02.357853\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "t3xebt2ZGHNT"
      },
      "source": [
        "<p><b>The Sparsity of Test data Matrix</b></p>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "d49234ce-e450-43f4-aac2-4185ae6daf80",
        "id": "6EKxr7vwGHNY",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "us,mv = test_sparse_matrix.shape\n",
        "elem = test_sparse_matrix.count_nonzero()\n",
        "\n",
        "print(\"Sparsity Of Test matrix : {} % \".format(  (1-(elem/(us*mv))) * 100) )"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Sparsity Of Test matrix : 99.95731772988694 % \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "XiTGR4RmGHNe"
      },
      "source": [
        "<h3>3.3.7 Finding Global average of all movie ratings, Average rating per user, and Average rating per movie</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "wHOby1-ZGHNe",
        "colab": {}
      },
      "source": [
        "# get the user averages in dictionary (key: user_id/movie_id, value: avg rating)\n",
        "\n",
        "def get_average_ratings(sparse_matrix, of_users):\n",
        "    \n",
        "    # average ratings of user/axes\n",
        "    ax = 1 if of_users else 0 # 1 - User axes,0 - Movie axes\n",
        "\n",
        "    # \".A1\" is for converting Column_Matrix to 1-D numpy array \n",
        "    sum_of_ratings = sparse_matrix.sum(axis=ax).A1\n",
        "    # Boolean matrix of ratings ( whether a user rated that movie or not)\n",
        "    is_rated = sparse_matrix!=0\n",
        "    # no of ratings that each user OR movie..\n",
        "    no_of_ratings = is_rated.sum(axis=ax).A1\n",
        "    \n",
        "    # max_user  and max_movie ids in sparse matrix \n",
        "    u,m = sparse_matrix.shape\n",
        "    # creae a dictonary of users and their average ratigns..\n",
        "    average_ratings = { i : sum_of_ratings[i]/no_of_ratings[i]\n",
        "                                 for i in range(u if of_users else m) \n",
        "                                    if no_of_ratings[i] !=0}\n",
        "\n",
        "    # return that dictionary of average ratings\n",
        "    return average_ratings"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "lKKAvkjWGHNh"
      },
      "source": [
        "<h4> 3.3.7.1 finding global average of all movie ratings </h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "85b83613-6952-46c0-d21a-0b396c442f1d",
        "id": "E_2UJPQ0GHNi",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "train_averages = dict()\n",
        "# get the global average of ratings in our train set.\n",
        "train_global_average = train_sparse_matrix.sum()/train_sparse_matrix.count_nonzero()\n",
        "train_averages['global'] = train_global_average\n",
        "train_averages"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'global': 3.582890686321557}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "WxLw0l4cGHNr"
      },
      "source": [
        "<h4> 3.3.7.2 finding average rating per user</h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "77ba4505-2a69-4e94-dde9-3d4090837044",
        "id": "FDxKEdnRGHNs",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "train_averages['user'] = get_average_ratings(train_sparse_matrix, of_users=True)\n",
        "print('\\nAverage rating of user 10 :',train_averages['user'][10])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Average rating of user 10 : 3.3781094527363185\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "UUK-RMqTGHNz"
      },
      "source": [
        "<h4> 3.3.7.3 finding average rating per movie</h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "10f11d17-ecbd-4155-8c66-164f02e89cc9",
        "id": "6SOIqs_9GHN0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "train_averages['movie'] =  get_average_ratings(train_sparse_matrix, of_users=False)\n",
        "print('\\n AVerage rating of movie 15 :',train_averages['movie'][15])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            " AVerage rating of movie 15 : 3.3038461538461537\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "JxskKIsNGHN3"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "wyRb1XHUGHN4"
      },
      "source": [
        "<h4> 3.3.7.4 PDF's & CDF's of Avg.Ratings of Users & Movies (In Train Data)</h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "c20f89fb-1639-4d71-d5d6-1bdc331b060e",
        "id": "Brg-5Lf8GHN5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 311
        }
      },
      "source": [
        "start = datetime.now()\n",
        "# draw pdfs for average rating per user and average\n",
        "fig, (ax1, ax2) = plt.subplots(nrows=1, ncols=2, figsize=plt.figaspect(.5))\n",
        "fig.suptitle('Avg Ratings per User and per Movie', fontsize=15)\n",
        "\n",
        "ax1.set_title('Users-Avg-Ratings')\n",
        "# get the list of average user ratings from the averages dictionary..\n",
        "user_averages = [rat for rat in train_averages['user'].values()]\n",
        "sns.distplot(user_averages, ax=ax1, hist=False, \n",
        "             kde_kws=dict(cumulative=True), label='Cdf')\n",
        "sns.distplot(user_averages, ax=ax1, hist=False,label='Pdf')\n",
        "\n",
        "ax2.set_title('Movies-Avg-Rating')\n",
        "# get the list of movie_average_ratings from the dictionary..\n",
        "movie_averages = [rat for rat in train_averages['movie'].values()]\n",
        "sns.distplot(movie_averages, ax=ax2, hist=False, \n",
        "             kde_kws=dict(cumulative=True), label='Cdf')\n",
        "sns.distplot(movie_averages, ax=ax2, hist=False, label='Pdf')\n",
        "\n",
        "plt.show()\n",
        "print(datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAecAAAEVCAYAAADaVy1GAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n0zRN8zD/DzK4xFoJKDILAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 576x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "0:00:41.239658\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "fDkc5fbDGHN9"
      },
      "source": [
        "<h3> 3.3.8 Cold Start problem </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "OV7pWM-JGHN_"
      },
      "source": [
        "<h4> 3.3.8.1 Cold Start problem with Users</h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "dfe1438c-6717-4cb3-bba3-dff0f65e0e49",
        "id": "0B336WmIGHOA",
        "colab": {}
      },
      "source": [
        "total_users = len(np.unique(df.user))\n",
        "users_train = len(train_averages['user'])\n",
        "new_users = total_users - users_train\n",
        "\n",
        "print('\\nTotal number of Users  :', total_users)\n",
        "print('\\nNumber of Users in Train data :', users_train)\n",
        "print(\"\\nNo of Users that didn't appear in train data: {}({} %) \\n \".format(new_users,\n",
        "                                                                        np.round((new_users/total_users)*100, 2)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Total number of Users  : 480189\n",
            "\n",
            "Number of Users in Train data : 405041\n",
            "\n",
            "No of Users that didn't appear in train data: 75148(15.65 %) \n",
            " \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Ce-YrA2tGHOC"
      },
      "source": [
        "> We might have to handle __new users__ ( ___75148___ ) who didn't appear in train data."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "QW49HowmGHOD"
      },
      "source": [
        "<h4> 3.3.8.2 Cold Start problem with Movies</h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "e7ec447d-f40a-4d62-d2c7-14f06806ad9e",
        "id": "PdYapFIaGHOE",
        "colab": {}
      },
      "source": [
        "total_movies = len(np.unique(df.movie))\n",
        "movies_train = len(train_averages['movie'])\n",
        "new_movies = total_movies - movies_train\n",
        "\n",
        "print('\\nTotal number of Movies  :', total_movies)\n",
        "print('\\nNumber of Users in Train data :', movies_train)\n",
        "print(\"\\nNo of Movies that didn't appear in train data: {}({} %) \\n \".format(new_movies,\n",
        "                                                                        np.round((new_movies/total_movies)*100, 2)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Total number of Movies  : 17770\n",
            "\n",
            "Number of Users in Train data : 17424\n",
            "\n",
            "No of Movies that didn't appear in train data: 346(1.95 %) \n",
            " \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "vfLw-_WgGHOG"
      },
      "source": [
        "> We might have to handle __346 movies__ (small comparatively) in test data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "6egkG4CgGHOH"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Q7WCBgwSGHON"
      },
      "source": [
        "<h2> 3.4 Computing Similarity matrices </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "moMgv9DqGHON"
      },
      "source": [
        "<h3> 3.4.1 Computing User-User Similarity matrix </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": true,
        "colab_type": "text",
        "id": "bq_qbJCkGHOO"
      },
      "source": [
        "1. Calculating User User Similarity_Matrix is __not very easy__(_unless you have huge Computing Power and lots of time_) because of number of. usersbeing lare.\n",
        "\n",
        "    * You can try if you want to. Your system could crash or the program stops with **Memory Error**\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Eefgxw-ZGHOP"
      },
      "source": [
        "<h4> 3.4.1.1 Trying with all dimensions (17k dimensions per user) </h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "ChJI9dHJGHOQ",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "\n",
        "\n",
        "def compute_user_similarity(sparse_matrix, compute_for_few=False, top = 100, verbose=False, verb_for_n_rows = 20,\n",
        "                            draw_time_taken=True):\n",
        "    no_of_users, _ = sparse_matrix.shape\n",
        "    # get the indices of  non zero rows(users) from our sparse matrix\n",
        "    row_ind, col_ind = sparse_matrix.nonzero()\n",
        "    row_ind = sorted(set(row_ind)) # we don't have to\n",
        "    time_taken = list() #  time taken for finding similar users for an user..\n",
        "    \n",
        "    # we create rows, cols, and data lists.., which can be used to create sparse matrices\n",
        "    rows, cols, data = list(), list(), list()\n",
        "    if verbose: print(\"Computing top\",top,\"similarities for each user..\")\n",
        "    \n",
        "    start = datetime.now()\n",
        "    temp = 0\n",
        "    \n",
        "    for row in row_ind[:top] if compute_for_few else row_ind:\n",
        "        temp = temp+1\n",
        "        prev = datetime.now()\n",
        "        \n",
        "        # get the similarity row for this user with all other users\n",
        "        sim = cosine_similarity(sparse_matrix.getrow(row), sparse_matrix).ravel()\n",
        "        # We will get only the top ''top'' most similar users and ignore rest of them..\n",
        "        top_sim_ind = sim.argsort()[-top:]\n",
        "        top_sim_val = sim[top_sim_ind]\n",
        "        \n",
        "        # add them to our rows, cols and data\n",
        "        rows.extend([row]*top)\n",
        "        cols.extend(top_sim_ind)\n",
        "        data.extend(top_sim_val)\n",
        "        time_taken.append(datetime.now().timestamp() - prev.timestamp())\n",
        "        if verbose:\n",
        "            if temp%verb_for_n_rows == 0:\n",
        "                print(\"computing done for {} users [  time elapsed : {}  ]\"\n",
        "                      .format(temp, datetime.now()-start))\n",
        "            \n",
        "        \n",
        "    # lets create sparse matrix out of these and return it\n",
        "    if verbose: print('Creating Sparse matrix from the computed similarities')\n",
        "    #return rows, cols, data\n",
        "    \n",
        "    if draw_time_taken:\n",
        "        plt.plot(time_taken, label = 'time taken for each user')\n",
        "        plt.plot(np.cumsum(time_taken), label='Total time')\n",
        "        plt.legend(loc='best')\n",
        "        plt.xlabel('User')\n",
        "        plt.ylabel('Time (seconds)')\n",
        "        plt.show()\n",
        "        \n",
        "    return sparse.csr_matrix((data, (rows, cols)), shape=(no_of_users, no_of_users)), time_taken      "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "id": "RNLuKop0GHOT",
        "colab": {}
      },
      "source": [
        "start = datetime.now()\n",
        "u_u_sim_sparse, _ = compute_user_similarity(train_sparse_matrix, compute_for_few=True, top = 100,\n",
        "                                                     verbose=True)\n",
        "print(\"-\"*100)\n",
        "print(\"Time taken :\",datetime.now()-start)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "QsRzdbEUGHOU"
      },
      "source": [
        "<h4> 3.4.1.2 Trying with reduced dimensions (Using TruncatedSVD for dimensionality reduction of user vector)</h4>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "WIDu_I28GHOV"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "6h5654MXGHOV"
      },
      "source": [
        "* We have  **405,041 users** in out training set and computing similarities between them..( **17K dimensional vector..**) is time consuming..\n",
        "\n",
        "\n",
        "- From above plot, It took roughly __8.88 sec__ for computing simlilar users for __one user__\n",
        "    \n",
        "    \n",
        "- We have __405,041 users__ with us in training set.\n",
        "\n",
        "\n",
        "- ${ 405041 \\times 8.88 = 3596764.08  \\sec } =  59946.068 \\min = 999.101133333 \\text{ hours}\n",
        "= 41.629213889 \\text{ days}...$\n",
        "\n",
        "    - Even if we run on 4 cores parallelly (a typical system now a days), It will still take almost __10 and 1/2__ days.\n",
        "    \n",
        " IDEA:  Instead, we will try to reduce the dimentsions using SVD, so that __it might__ speed up the process..."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "bf53b3a5-828f-457c-d354-4d52b6559767",
        "id": "ntGRZKe8GHOW",
        "colab": {}
      },
      "source": [
        "from datetime import datetime\n",
        "from sklearn.decomposition import TruncatedSVD\n",
        "\n",
        "start = datetime.now()\n",
        "\n",
        "# initilaize the algorithm with some parameters..\n",
        "# All of them are default except n_components. n_itr is for Randomized SVD solver.\n",
        "netflix_svd = TruncatedSVD(n_components=500, algorithm='randomized', random_state=15)\n",
        "trunc_svd = netflix_svd.fit_transform(train_sparse_matrix)\n",
        "\n",
        "print(datetime.now()-start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0:29:07.069783\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "T5F0rmM1GHOX"
      },
      "source": [
        "Here,\n",
        "\n",
        "\n",
        "- $\\sum \\longleftarrow$ (netflix\\_svd.**singular\\_values\\_** )\n",
        "\n",
        "\n",
        "- $\\bigvee^T \\longleftarrow$ (netflix\\_svd.**components_**)\n",
        "\n",
        "\n",
        "- $\\bigcup$ is not returned. instead **Projection_of_X** onto the new vectorspace is returned. \n",
        "\n",
        "\n",
        "- It uses **randomized svd** internally, which returns **All 3 of them saperately**. Use that instead.. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "xw5xEQ6MGHOX",
        "colab": {}
      },
      "source": [
        "expl_var = np.cumsum(netflix_svd.explained_variance_ratio_)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "237de9aa-848a-47b8-ffcf-e7ebacce3905",
        "id": "KUps11F8GHOZ",
        "colab": {}
      },
      "source": [
        "fig, (ax1, ax2) = plt.subplots(nrows=1, ncols=2, figsize=plt.figaspect(.5))\n",
        "\n",
        "ax1.set_ylabel(\"Variance Explained\", fontsize=15)\n",
        "ax1.set_xlabel(\"# Latent Facors\", fontsize=15)\n",
        "ax1.plot(expl_var)\n",
        "# annote some (latentfactors, expl_var) to make it clear\n",
        "ind = [1, 2,4,8,20, 60, 100, 200, 300, 400, 500]\n",
        "ax1.scatter(x = [i-1 for i in ind], y = expl_var[[i-1 for i in ind]], c='#ff3300')\n",
        "for i in ind:\n",
        "    ax1.annotate(s =\"({}, {})\".format(i,  np.round(expl_var[i-1], 2)), xy=(i-1, expl_var[i-1]),\n",
        "                xytext = ( i+20, expl_var[i-1] - 0.01), fontweight='bold')\n",
        "\n",
        "change_in_expl_var = [expl_var[i+1] - expl_var[i] for i in range(len(expl_var)-1)]\n",
        "ax2.plot(change_in_expl_var)\n",
        "\n",
        "\n",
        "\n",
        "ax2.set_ylabel(\"Gain in Var_Expl with One Additional LF\", fontsize=10)\n",
        "ax2.yaxis.set_label_position(\"right\")\n",
        "ax2.set_xlabel(\"# Latent Facors\", fontsize=20)\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAA8AAAAHgCAYAAABq5QSEAAAgAElEQVR4XuxdB3hUxdp+zyahCQgovQXJng1VpCqgVBVR8IJgoahYAPltiPUi13sVewNsXEHFLooIFgSvha5U6WRnlxAgVJFek+yZ/5mTbNzElLPnnC0n883z3OdiMjNn3vf7Zr55M00BJWKAGCAGiAFigBggBogBYoAYIAaIAWJAAgYUCTASRGKAGCAGiAFigBggBogBYoAYIAaIAWIAJIDJCYgBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggASwFGYmkMQAMUAMEAPEADFADBADxAAxQAwQAySAyQeIAWKAGCAGiAFigBggBogBYoAYIAakYIAEsBRmJpDEADFADBADxAAxQAwQA8QAMUAMEAMkgMkHiAFigBggBogBYoAYIAaIAWKAGCAGpGCABLAUZiaQxAAxQAwQA8QAMUAMEAPEADFADBADJIDJB4gBYoAYIAaIAWKAGCAGiAFigBggBqRggARwGTbzihUrzktISLhS07QMl8t1pgxDJWjEADFADESEAU3TKrhcruRAILCgU6dOf0bkIw6qlOKKg4xFTSUGiIG4ZIDiSuzNQgI49jaIWAtWr149BMDHEfsAVUwMEAPEgDwMDG3fvv0n8sAtGinFFdk9gPATA8SAjQxQXLGRzHCqIgEcDlsOy7ty5crOLpdrWXJyMipWrGi69YFAAIwxqKqKhIQE0/U4qaBsmGXDK3xRNsyE19wIdPr0aWRkZEDTtC4dO3Zcbq4WY6VUVR0D4CEAdRVF2cw5v58xtqS40m63+zpFUZ4C0BTANgDjGWNfBfOrqvpvADcqitKQc54FYA3nfLzP51sRzNOqVavqZ8+enQKgf97Pvs7KyronIyPjSFHftRJXZPPBIH+Em+YOxkYA5+YiHw/Px6MZV5zrVZFtOQngyPIb09pXr17dVkx4mjVrhkqVKpluixjY1q1bhzZt2kglgGXCTDYu+3/Ykc3GduE9deoUtm7dKsbPdu3bt19reiAtpaCqqjcA+BDAGM75MgCjFEW5Iycnp3l6evrOwsVTUlIucblcQhxPACBE7wAAT3LOuwYFrtvtFruADgBI55xXdLlcYwEM1jQtxe/3/yHqVFX1e855A875SPHfiqK8rShKBmOsX1FNthJX7LJJpGwQqXoJN80dIuVb8VIv+Xh4Ph6tuBIv/hGP7SABHI9WsalNViYqoU2QcWCTDbNseIV/y4aZ8JobWKM1UVFVdYWiKGu9Xu9dISu4QnnPYYw9Vrj1qqrOBFCVMXZV8Hdut3u+oiiHGWM3FYU2JSWlqsvlOgqgN2PsJ7fb3UxRlC2c84uDojklJeVil8v1q6IoqV6v11u4HitxRTYfDF0BlukPqjLjJh8PTwiaG5Xjp5RZe0crrsQPU/HXEhLA8WcT21pkZaJCAliuVW+zg7htzhqDimTDTHjNOVk0JirNmzcvl5OTc0qszoZuYfZ4PJM5520YY92KEMBiVfhVxtirIYJ5rKIo93u93saF84tvZGdn36soyuMAUhhjBz0ez22c81cYY9VC86uqeoRzPtbn871HAtic38geQ2X8I6OsmAl3+DskoxFXrI9cZbsGEsBl2L4kgM0bl8SCee6cUpJs7BRLmWunXfaNxkTF4/HU45zvdrlcXdLS0vLPGauq+k8AtzDGPEUI4CzO+a0+ny//Yi6x5VlRlPcYY+WD+d1u9zWKonwGQJyD2QvgH4yxVeL3ov68OtTQ+t1uN8ur59niBLC4EyLcozXCJhs3bkSrVq2kOU4TFAeEu+wfMyFbU982Gq1EXBF360T6aI3R9siYjwRwGbY6CWDzxrVr8my+BdEtKRteGf9iLZuN7cIbTQGsaVpnv9//a4h4Ha8oynDGWGoxAvgWn8/3afB3Ho9nKOf8HcZYheDPWrdufc7JkyfFpVrnK4pyp6IoPQOBQKdt27YdKE5gq6rqAyDqea44ARzdEYq+RgwQA8RAmWQgondLlEnGbAJFAtgmIuOxGhLA5q1i1+TZfAuiW1I2vCSAy/5qjF0+HQ0BHI0t0MERJU/cvssYe9bKFmhaATY+RtPKtzwr/mRreWxtZcWfVoCNj5+RykkCOFLMxkG9JIDNG8GuybP5FkS3pGx4SQCTADbaw6IhgEVbxCVYnPM1Pp9PPIWkJ1VVtwCYW8IlWFUYY31D8n8P4Ehxl2Dl1ekH8BFj7N/BS7AAdGKMrRS/d7vdnRRF+Y0uwTLqIaXnk3F8lXGMlRUz4aYzwKWPgvGXgwRw/NnEthaRADZPpWwTFtnwyhiwZbOxXXijKID1Z5AURRmtaZrYBj1SbFlOSEhosXXr1h2qqn4AYHdQDKempnbWNG2xoijjOedzFUW5lnM+MfgMktj6fObMmfGapn2dlJS0Nzs7+zyXyyXE9TBN09r5/f7NeYJYiOZ6mqaNEv/tcrneBrCDnkEyHz8Kl7TLF+1rUXRqkhG3jJhljKfBHmTW3tGKK9Hp6c78CglgZ9rNUKtJABuiqchMZgc181+MbUnZ8MoYsGWzsV14ozlRUVVVCNSHAYgzu5s0TRM3MS8W/urxeBZyzsX7vLcGRwuPxzNIiF4AFwDYxjkf7/P5ZovfJycnVyhXrpy4IKsTgPMB/Mk5X6UoysTgJVgiX4sWLWpkZWVNURSlv/hvzvnX2dnZd2dkZBwpalSyElfssklsR8vwv0645Xkah2wtj62tzCOiGVfCH7HkKEECuAzb2cpEJZQWGQd02TDLhtdK4HLqkCGbje3CSxOVgh5vJa7YZROn9UHCLY8oIlvLY2sr8wiKK7EfxUkAx94GEWuBlYkKCWB6BzhijhknFcs2USG85hyPJiokgM15zl+lZOt7QeQy4pYRsxUhaLVvxbq8WXtTXIm15QASwLG3QcRaQALYPLVmBzXzX4xtSdnwyhiwZbOxXXhpokIC2OrobJcvWm1HtMvLiFtGzDLGU6t/5KG4Eu3R6O/fIwEcextErAUkgM1TK1sQkw2vjAFbKhsHAgisWYgda35D43YXI6FddyDB3M3XNFEhAWw+kuSWlKrvhZAlI24ZMZOP0y3QVsfIWJQnARwL1qP0TRLA5omWLYjJhlfGgC2NjX+eDbx0H3Ag868BoFYD4MHJQM+BYQ8KJIDtE8AL0/bj+1VpmHhjFyQlJYZtC6cWkKbvFTKQjLhlxCxjPKUVYKeOxn+1mwSw821YLAISwOaNK1sQkw2vjAFbChsL8fvwIHGXcaHOnxfqXphlWAQfPnwYvXr10uuZPHkyKlWq1K59+/ZrzY8qZaOklbhy5auL4N1/AnPGXII2jWqUDUIMoJCi7xXBg4y4ZcQsYzwlAWxg4IvzLCSA49xAVppnZaIS+l0ZB3TZMMuGV8aAXeZtHAgA/ZILrvwWGEAVoHYD4OvthrZDv/TSS5g2bRqGDBmCa665RtTUbtiwYYc0TdtezLh8J2NsevB3KSkpV7hcricAXAQgR1EU8bbveK/XuzqYp127dknHjx9/HMBwAA0AHOCcz3K5XBO8Xu/xcMf/5s2b18nJyXkWwNUAzs17Gmmqz+ebYqQuj8dzO+d8tKIozTnn4q8IGZzzl30+33vB8lbiSo+XfsH2g6cw885O6NRUvMwkRyrzfa8YM8qIW0bMMsZTEsDOH7tJADvfhsUisDJRIQFMt0CX4a6hQ5NtolLm8a5eCIzugQBc+K16K8yvdQnaH9mKa/cvKujKU38B2ncv0b2zs7Nx6aWXQqwCf/TRR3C5XEUJ4BWhlSiK8ozX6/1a/ExV1asAfANAHDzeDaB83lu8pwFcwhhbn5dPvNN7EwANgC/vPd8kAEsYY6KR4ueGUvPmzStnZ2evVRTFDeA05zwz79+i/DOMsfElVaSq6msA7hZ5FEXZyTk/BKCewMEYu8MOAdzr5YXY9sdJfHpHR1ySUtMQrrKQqcz3PRLA+QyQrekZJCNjFh2tMcJSZPOQAI4svzGtnQSwefplC2Ky4SUBbO5CKPM9KvIl/V/OxJeff485dbtjb4VccdXymB/frry/4McnfgL0EZqz+LRo0SKMHDkSNWvWxA8//ICtW7f+TQAzxoqNn263e6OiKC0B/FavXr1L9+7dW5FzvgFAMoBvGWP9UlNT22maFlwNvocx9npKSko/l8uli2jO+SCfz/elUeZUVR0H4CVRVFGUNl6vd4Oqqi8DeECsQCcmJjbcsmXLvqLqS0lJucTlci3PE9yDGGNfBfN5PJ4qoavRkyZN6t+1a9e5zZo1E9vCjTZPz9f75YXw/3ESH9/eAV3ctcIq6+TMMo6vMo6xsmIm3HQJlhPHZxLATrSawTaTADZIVBHZZJuwyIZXxoBdFm186GQWvlm/B1+uzcSGzKP5Pblq9gn0278Yd+74Csmn9xbs4QZWgF9++WW8/fbb6NmzJ8S/ixLAAA4CqAjAzzl/0+fzie3Pmqqq9QEEb+D6J2NMbEkWq8JvA7hTrM4yxqq43e5HFUWZKH6XkJBQb+vWraKhLlVVTwKoAGAaY2yk0VFMVdX/CY0JgDHGPKJcampqZ03TluUJ6qE+n0+sOP8tud3uVxVFEX8p2AVArE5fBkAQOlfTtPF+v/+YKNS9e/fEypUrr/33v//dyowAvvyVRfAdOIGPbu+AriSAjZrWsfnK4phTmjFkxCxjPA36gVl70wpwaT0p8r8nARx5jmP2BRLA5qk3O6iZ/2JsS8qGV8aAXVZsnB3Q8HPaAcxak4lf0g4gR8u98CrRpaD7kXUYmDEPPQ+uRAUtu1CnMn4G+N5778WCBQswfPhwPPDAA0UJYCFWDwBIAXBOnsB8wefzPZKSknKxy+US533FVuI7vF7vO+LfHo/nKc65OO+LxMTEuoFA4N+c81Hivxlj4krkQJ5QFiK0Aed8gc/n62N0ZFBVNU18BsBSxtilolxKSkpTl8vlz6vjMcbYc0XVp6rqPABi27ZIYpt2OoDmAkJwxTqIoXHjxo8/88wzMCOAr3h1Edj+E/jgtg64TKUVYKO2dWq+sjLmhMO/jJhljKckgMPpFfGZlwRwfNrFllaRADZPo2xBTDa8MgZsp9vYf+AEvli9C1+u3Y2DJ87md+6W9ati4EUN0L9NPZy/8ru8W6B1SRoyAIR3C/SIESOwfPlyjB49GqNGjcoXwLfddps3Ozu7ydatWzeJylu0aFEjOzt7SZ5YPJ2YmFgtKyurXd524sICeCLnXD+Hm52dXScpKelJAPoKbyEBLFaPxSryfMZYUJSWOpipquoVC82hArhp06YpCQkJ4myxaMujXq/3+WIE8A8ALs/73aWMsaUej2cC51y0UZyBbsI5P59z/lurVq0WPPbYY33NCOA+kxYjbd9xfDCiPS7z1C4VU1nJ4PS+Z9YOMuKWEbOM8ZQEsNlRIX7KkQCOH1vY3hISwOYplS2IyYZXxoDtRBufPJuD7zbuxeerdmH1jsP5Hfr8yuUwsG0DXNe2ATx1qhTs6EW9A1y7ITBukuEnkO655x797O/NN9+MsWPH5gvgop5BUlVVnLsV52+FyKzPORdxNbgFejxj7Bnxu5K2QItyXq93jx1boDnnPp/PJ4Sw4S3QqqqKW55vFWU0Tavg9/vPpqamXqlp2vw8AdwjEAgkK4ryXpMmTc4+/fTT5c0I4L6TF2PL3uOYcWt7dE8lAWw+QjmjpBPHHKvMyohZxnhKAthqT4l9eRLAsbdBxFpAAtg8tbIFMdnwyhiwnWJj8frO2p1HdNH77YY9OJml7wyGSwF6eGrh+g4N0TO1FpIS9JuZi06BAAJrFmLHmt/QuN3FSGjX3dDTR8HKgk8g9e7dGy+++GK+AB46dGjDQCBw2u/3ixVTJCcnVytXrtxSsRgM4GSVKlWqr1mzJjvkEqyV9erV65KZmVnJ5XKJS7AaK4ryjdfr7R96CRbn/D7xVJHH47mWcz5H1M05v87n881OTU3trmnaL0EhmpaWtrAo0KGXYHHO2/p8vnUhZ3tzsrOzG2zfvn2/qqripmf9tmfGWKr4f4/HM5Rz/lGeiO/q9XqXud3u8XlnlLmmaY0URektBHBycjLMboG+evISbN57DO/e0g49m9UxP0A7rKRT+p7dtMqIW0bMMsZTEsB2jxbRr48EcPQ5j9oXSQCbp1q2ICYbXhkDdrzbWGxr/mrtbsxcvQtiu3MwJZ9XCYPbN8Sgdg1Qu6q4G8pYsoL3l19+0bc/161bF/PmzcsXwEOGDOkPQLztKy6I2pH3bFHlvBY9wRjTtwznPYP0rdDtVp9BMiqAjT6DpKrqv/MwCAGszwHy3iMWt0C3VxTllKZp6YqiCFGvcM7f9fl8twdZtxJXrpmyBJv2HMM7N7dDr+YkgI15snNzWemDTkUtI2YZ4ykJYKf20L/aTQLY+TYsFoGViUpopTIO6LJhlg2vjAE7Hm2saRxL/Afx6Yqd+HHr/vwLrSokudC3VV3c0L4hOjapIbYWhz1SW8GblZWlvwN85MgRfPbZZ2JbsK4Thw0bpmiadi+AznnndE8BEGdvpzDGZoY2Mm8L8b8AtBXPEAEQF2OJLdGrgvnyhKe4GOvmvPr+UBRlFoDHg08Pqao6AMBscTmVpmmq3+8Pbq/+GyfNmjUTl2uJW6evBnAugG2Kokz1er2Tg5mLEsDid61ataqelZX1LOf8WgDVOOd+l8s13ev1vh68oEvksxJX+r22BBt3H8P04W3Ru0XdsG3q1AJWfNGpmGUcY2XFTLjpGSQnjlPhzyqciFLSNluZqJAADmDduvAHNae6mowTNNkwxxNesdr7xepMfLpyJ3YeEhoyN13YsJoueq+5sC6qVkiy1J2s4hVbn6dPn45hw4ahb9++ugAu6gywpUYaKKyq6isAxpZ0iZWBamzLYiWuXPv6UqzPPIppw9vichLAttkkXiuy2gfjFVdJ7ZIRMwng8OeK9AxS7Hs3CeDY2yBiLbAyUSEBTAI4Yo4ZJxXLNlGJNV5xtvfX9D/xyYqdWLB5H7IDubc0V6mQqF9mdWPHhkitU9U277CK9/Dhw+jVq5e++jxp0iRUqlQpJgLY4/Gs0TStXP369S9auHChWEmOabISV/7xxlKs23UU/x3WFle2pBXgmBoyCh+32gej0ETbPyEjZhLAJIBt70hRqJAEcBRIjtUnrExUSACTAI6V30bru7JNVGKF98ipLP3N3k9W7kT6HyfzzdumYTUM6dQI/VrXQ8VyCbab3S689Jf6gqaxElcGvLkMv+88gqlDL0KfVvVst3m8VmiXL8YrvuLaJSNuGTGTACYB7LSxSbSXBLATrWawzVYmKiSASQAbdDPHZpNtohJNvGK1d82Ow/pq77cb9yIrRz9Di3PKJeDai+pjSMdGaFlfHFGNXLILLwlg+wTwdW8tw5odR/DmkIvQtzUJ4Mh5f3zUbFcfjA80xlohI2YSwCSAjfWO+MpFAji+7GFra0gAm6dTtiAmG14ZA3Y0bHz8TDa++n03Pv5tJ7z7j+d3wOZ1q2LoxY1wbZv6qFw+0XzHDKOkXXhJANsngAe9tVx/z/mNm9rg6gvrh2FNZ2e1yxedxoKMuGXELGM8DfZFs/amuBL70YwEcOxtELEWkAA2T63ZQc38F2NbUja8MgbsSNrYf+A43l++A7PXZua/2ytuchbbm4de3BgXNjjX1E3OVnqFXXhpomKfAL5+6nKszDiM129qg2tIAFtxb0eUtasPOgJsXiNlxCxjPCUB7KReWXRbSQA734bFIiABbN64sgUx2fDKGLDttnFA4/rTRR/8moFl/j/zO1tKrcoY1qkRBrRtgHMrWrvJ2XwPBuzCSwLYPgF8w3+XY8X2w5hy44Xo36aBFfM6qqxdvugo0OLtrIBcR4lkjCtWhaDTfLpwe836OMWV2FueBHDsbRCxFpAANk+t2UHN/BdjW1I2vDJOVOyy8eGTWfhs1S589NsO7D5yWndclwL0blYbt3RORuem50V9tbeo3mMXXpqo2CeAb3z7V/yWfgiTb7gQ115EAji2o37kv25XH4x8S+37goyYZYynVoU/xRX7+pzZmkgAm2XOAeVIAJs3kmxBTDa8MgZsqzbetPso3l+ega/X78HZvEutqldKwo0dG2Fop0ZoUL2S+Q4XgZJW8QabRBMV+wTwkGm/Yfm2P/Hq9a0xoG3DCFg9Pqu0yxfjE13xrZIRt4yYZYynJICdNhr9vb0kgJ1vw2IRkAA2b1zZgphseGUM2GZsLG5vnr95ny58xa3OwdSyflXcckky+l1YDxWS7H/CyHzP/aukGbxFfZcEsH0CeOi037Bs2594ZXBrDGxHAtgOP4/nOuzqg/GMsXDbZMQsYzwlAeykXll0W0kAO9+GJIAjYEPZgphseGUM2OHY+OCJs/pNzh+t2IE/jp/Ve1hSgoK+reri5kuS0bZRtbjY5lxS1w8Hb0n1kAC2TwAPn/4blvj/xMuDW+M6EsARiFzxVaVdfTC+UJXcGhkxyxhPSQA7qVeSAHa+tcJEQCvAYRIWkl22ICYbXhkDthEbp+07hneXbsecdXvy3+6tVaU8hnZqjJs6NUStKhXMd6oolzSC10iTSADbKIDfWYElvoN4cVArDG7fyAj9ZSKPXb7oNDJkxC0jZhnjKQlgp41Gf28vrQA734bFIiABbN64sgUx2fBKFbADAeD3JdAO7IbvyAmkDBqBhHLl8juHpnEsZAfwztLtBW5zbtOwGm7r2gR9WtRBuUSX+c4Uo5J2+TQJYPsE8C3vrsAidhAvXNcK13cgARyjrhG1z9rVB6PWYBs+JCNmqeJpIR8xa2+KKzZ0NotVkAC2SGA8FycBbN46Zgc181+MbUnZ8EoTsH+eDbx0H3AgM9/BeK0GUB6cjFNd++PLtbvx3rLtSP/jpP57cZvzVS3r6sK3XePqsXVKi1+3y6dpomKfAL713ZVYyP7A8wNb4oaOjS1a2DnF7fJF5yDObamMuGXELKutreCmuBL70YwEcOxtELEWkAA2T61sQUw2vFYCl3mvinJJIX4fHgSAF/jw3vLn4YMG1+AT9yAczckNAVXKJ+LGjg31Z4yKus358OHD6NWrl5530aJFqFKlSpTBhP85u3za4kQlQVVVHwCx3NmSMZYWPpL4KmElrtz23kr87P0Dzw1oiRs7kQCOL8va3xq7+qD9LYtcjTJiliKeFuMyZu1tMa5EzoElqpkEcBk2tpWJSigtZju4k6mVDbNseMt8wBbbns5ChN8AACAASURBVPslF1j5XV/VjXcaXYt5tboix5Wod89GNSpiRJcmGNy+ISqXz/1ZUemll17CtGnTcPPNN2P8+PF/y7Jw4UKMGjUq/+cbNmxA+fLl8/976dKleP3117F161YkJCTgoosuwv33349WrVrl58nOzsZbb72FuXPnYv/+/ahRowauvPJK3HfffahcuXLYw8m+ffswYcIEbNq0CcePH0ejRo1w44036hiMpC+++AKfffYZ/H4/OOfif9vOnj37tM/ne0+Ud7vdlymK8iiADgDOFz/jnN/l8/mmhtbv8Xju45xPAvAZY+wmI9+O5zxW4sodM1bhx7QDeOYfLTDk4uR4hmlr22QcX8v8GGuzILLV4WJQGfl4Gz22GU0kgI0yFbl8JIAjx23Ma7YyUQltvIwDm2yYZcNb5idnqxcCo3tAg4Ifa3bEtEYDsap6i/xu3enwRty2cy56P/kMEjr0KHGsEsL00ksvhVgFnjVrVgHRKgoePHgQ/fv3x59//lmkABYrxnfddZe+HbJ27drIysrS66pQoQJmzpyJ1NRUvdy4cePw7bffwuVyoXHjxsjMzIT4dvv27fHhhx/qPzeaTp48iQEDBmDHjh36d+rUqYOMjAy9+OjRozF27NgSq3rqqafw0Ucf6XlEWSHmMzMzDwUCga8YY3eIn3s8nvs55y8CSAegFieAmzZtWishIWEvAC0QCNTftm3bAaM44jGflbhyx/ur8OPWA5h4bQsMu4QEcDza1842UVwxLojs5D0WdcloayvzCBLAsfDSgt8kARx7G0SsBVYmKiSAA1i3bh3atAnvr3oRM2aEK5YxeJVlzGe++xRz/jsDbzcegPRzct9bTdKy0W//El34tjy+LdejJn4C9Cl5UVII2JEjR6JmzZoQK7mFk/jd8uXLcdlll+Gnn37Sfx26AtyvXz8wxvS+9PHHH+PMmTO6YN69ezd69OiBqVOn6qu01113nV5WrNoOGzYMP//8sy6cRZoyZYq+Gmw0vfvuu3j++ef1p5pmz56N5s2b47nnnsN7772HxMREiBVrgaeo9Pvvv+srxUJwi+926dJFX7kG0G7o0KE+r9d7XPxHamrqeSdOnDhVuXLl2pqmbRc/K2oFWPxcVdVfAVysKMoYr9f7llEc8ZjPSlwZ+cFq/LBlP566tjmGX9IkHuFFpE1leawpiTAZccuIWfgA4Q5vrkgCOCJDbViVkgAOiy5nZbYyUQlFKuPAJhtm2fCW1YB99FS2/nbvjEUMf5zJPftbJfsEhu7+HiN2foPaWYcKDmJTfwHady9xYHv55Zfx9ttvo2fPnvoW5dAkVmYnTpyIRx99FCdOnNC3OYcKYLGVWQhjkR544IH8bdJC5H7++ef66uzatWv1+idNEruEgSVLlqBWrVrQNE0XzWfPnsX1118PsSprNI0YMUIX5XXr1tVFudiaJr5z0025Yl9s6RbCvKj0zDPP4P3339fLitXpVatW6SvAiqLMPHDgwEi/338stFxqampyaQLY4/FM5pzfC+BTxtgQozjiMZ+VuDLqw9VYsHk/nuzfHDd3JgEcj/a1s00UV2gF2E5/ise6zPo4CeDYW5MEcOxtELEWWJmokACmFeCIOWacVGw2cMVJ8ws0Y/eR0/r7vZ+t3ImTWQH9d3WzDuH2jNm4YfcCVAmcLtRsBajdAPh6O1DKuaV7770XCxYswPDhw/H444/n1+Pz+fRV2w4dOmD69Om6+C0sgMUuihtuuEEvI4Ty4MGD9X8LsRsU02JV+bXXXtO3Q4u0ZcuW/LNU3bp1gzjL27VrV7zzzjuGqe/Tpw+2b98utinjq6++0uvbuXMnLr/8cr0Osd1arFwXle68804sXrxY/5UQ6PXr10d6erp+DhjAt4yxAsrZiABWVfUBAC8DWM0YE2eGHZusxJW7PlqN7zftx3/6NcctXUgAO9YJDDa8LI2xBiHTSqgku+aC/mDWx0kAG+1RkctHAjhy3BqqWVXVMQAeEvNVRVE2c87vZ4wtKa5wcnJytXLlyj0NYCAA8U6J2Ho3jjE2r3AZKxOV0LrMdnBDBMRpJtkwy4ZXuF1ZwLx17zG8vTgd36zfgxwtd8U3tU4VjLzsAvQ7tBJJj4pboEX66yZoDgX6wP/CLKCnGEZKTsHV1MJnZ8U25gMHDuCbb77RtxMLEVtYAAe3E4svhArgV199Vd/6LNKyZcv0rcZFCWCxeixWkcUZZCGyjSaxXVqc+Q0VwOI88BVXXKFXUZIAvu222/Q2iSS2bIvt008//bR+/lkkl8vVJC0tLfdAce5W6FJXgFVVFeeGp3HOfT6fTz8v7NRkJa6M+XgN5m3chyeuaYYRXS9wKgVht7ssjDVhgy4jY2y4uMnW4W0FDpffeMtv1t4kgGNvSRLAMbSBqqpiaeRDAGM452LGNUpRlDtycnKap6en7yzctObNm5fLyckR+Q4oivKMy+XKDAQC4oDfccbY+sL5rUxUQusy28FjSK3lT8uGWTa8wkGcilmsRP667U9MXZyOxeyPfF/v3PQ8Xfh2U2vqZ1/1VNQ7wLUbQBk32ZD4FVXcc889+OGHH/52A7QQl+I8bbly5fRPiQurxP9EqlSpEh588EH07t07fwu0uHhKiGiRStoCLVZfxWVZdmyBFhdYibPE4WyBfuyxx/RzwyJt3LgROTk5+m3Q4kyxSC6Xq0daWtrCIPEGBbC+Asw5X+Pz+dpbHqBiWIGVuPJ/H6/Bdxv3YcLVqbj90qYxRBHdTzt1rLHKkoy4ZcTs5HgaKx8vSQCLFwbq16+/fOHChTlW20fli2eABHAMvUNV1RWKoqz1er25N73kXpYibluZwxh7rHDT3G73aEVRHqpSpUrqmjVrcmeaJSQrE5XQamUc0GXDLBteJwZsTeP4cet+vLFwG9bvOpIrxhSgb6u6GHVZU7RqcG7Ro4F4Eun3JdAO7IbvyAmkDBqBhDzRWtoYIn4ffAJJiNk33ngjv4gQwCUlISRvvfVW/aytuASrdevW+PTTT0u9BEs8sySeKvrxxx/xf//3f/onxOqyWL1dsWJF/jNGH3zwATp16lRkE0IvwRIrty1btkTwbK8Q7eJir/PPP1+/6Tl42/P8+fP1ur7++ms89JDYlAN88sknaNasGZ599ln9zLJYStc0rZHf788MftiIAA45AzyTMXajEd7jNY+VuHLPJ2vxzYa9ePzqVNxBAjheTWxbuyiu0Blg25wpTisy6+MlCWBVVQOBQKBu8MUAVVXF7ZM3MMZ2xykNjmwWCeAYmS1vNfcUgMGMsa+CzcibKLVhjHUr3DRVVedxzg8piiLKXQtALP98whgTSxO5B/9CUnCioqqqviJjNokOLlZCxJud4bxzZvZ78VBONsyy4RU+5hTMOQEN8zbtw1uL0sH2n9C7R4UkFwa3bYDbuiajUQ1jfdssXnFj8pgxY/TngMRqanFJbH9+88039V+Lrc/Bd4DFiq4oL1Z0Cz+DJARm8BkkITq/++67/GeQdu3apa++tmvXTr+UStzKvHLlSl1UizRjxgx07NixyOaIZ5AGDRqU/wyS+K7YAi2SOPsr3iAWKbTN4uyxSGIVe+jQofrN1BUrVkS9evXyzwBzzt/1+Xy3i3xut3ugoigvABAPKDfOa4gYk48pirLC6/UODTYueAs0gLsZY3/9FSEeBrsw22BFAN/76Vp8vX4vxvdNxZ2X0QpwmNQ7LrtZceA4oCENlhFzMJ7K9HJG0ORm7V2KABZP5tUJEcDHOecX+nw+8eQeJZsYIAFsE5HhVuPxeOpxzne7XK4uaWlpy0MmSv8EcAtj7G/LK6qqpgFI5px/nJCQ8KamaW4AYjI1mTH2ZOE2BCcq4baN8hMDxEDsGcgOcPyy4zTmpJ3E/pO5f9+qlKigT0olXOOuhHMrRGd1QYhQIWDFLc9iFTU5uej3W8VKa3DrsBCnwa3Rot3r16/XL6MSF1OJP6K53W79ZuemTf8SQeI7c+bM0W+BPnToEKpWraoLXJFPCFGRxI3M4vywqFvcTn3eeecVayjx1rA4VyzEuJhsCBHcq1cvXHXVVfllQtssxHgwCayi7Jo1ayDEtBD/SUlJL82fP//R4B8b3W73rYqivFdUAxRFWeT1evXrtfPeAd4jVo+zs7MbbN++fX/svct8C6wI4Ps/+x1z1u3BP6/yYGS3FPONcFhJs5Nkh8H8W3NlxC0jZmF4wh3e2WcSwLEf3UgAx8gGQQGsaVpnv98v3ojUk9vtHq8oynDGWGrhprndbqYoSgXGmLg+U58R590u+hBjrG7h/LQCbN64ZlfLzH8xtiVlwxsM2PG4s+FUVg4+W5WJ6Uu2Y//xs7pj1KiUhFs7J2P4xY1QtWKSKWexYmMhNsUtzLfccgseeeQRU9+3o5A4hytWg8WTSnfcIe6VKj5ZwRtaq5ioiC3c4h3g9u3brw0Xh9vtvldRlMkAHL/9WWC3QwA/2seD0d1JAIfrS07LL6MokhEzCeB1+pN94eyQLG0LtKZpdfx+v37Jh6qqYlfRhV6vV39vnpI9DJAAtofHsGsxuQV6kdihxxjrHfygqqpiOWNeYmJi+S1btmSFNsTKRCW0HhkHdNkwy4Y3HgP20dPZ+GB5Bt5dth2HT+Ue8a9TtYJ+sdWNHRuiUjmx09Z8smJjsZoqVk/F5VpiS3SVKlXMN8RCyYEDB+pblMVqsjjLW5oAtmNLnsXbOhNUVfWJLdKc85Y+n0/c8eDoZCWuPDDzd8z+fQ8e6ePBXSSAHe0HRhpvZcwxUn885pERczzG02j5hll7l7YCDGAdgOAlWG0BbAaQ+xfxvMQYK/oMULTAO/w7JIBjaEBxCVberaDiKSQ9qaoqDqLNLeoSLFVVnwEwhDEm3o/QRH6Px3Mf5/wRxli9wlCsTFRC6zLbwWNIreVPy4ZZNrzxFLD/OH5WF70f/roDJ87mxrvG51XCXd2aYkDb+iifaM9WZ9lsbBdeiwLY8lgUbxVYiSsPfr4Os9buxsNXqhjTQ5zgkSPZ5YtOY0tG3DJijqd4Gu0+YtbeJcUVj8fzlBEcXq93gpF8lKdoBkgAx9Azgs8gKYoyWtM0sQ16pKIodyYkJLTYunXrDlVVPwCwOyiGmzZt2jAhIUEI5Bkul+u1QCDgVhTlXc75FJ/PJ94GLpCsTFRIAAdgx+pRDN0rrE+bHcTD+kicZY415gPHz+C/i9Lx8YodOJOt/z0LntpVMKZHU1zdqi4SE1y2MhZrvLaCMVCZXXhJANsXVx76Yh2+WLMbD17hxt09Hf0csgEP/CuLXb4Y1kfjILOMuGXETALY3i3QcdB1pWgCCeAYm1lVVbH6+zCAuoqibNI0bazP51usT4Y9noWc8wzGWO61pwBSUlIucblcrwJoI8QxgHdKuwVaPONh9RZomcSgjIO5jEE7VpgPHDuDqXnC92xOrvC9sGE13N0jBb1Sa8El3jaKQIoV3ghAMVSlXXhJANsngB+etR6fr87EuMvduKcXCWBDjuzgTHb1QSdRICNmGedMQZ80a28rccXj8bTmnK9mjJVzUt+It7ZGZqYVbyglbQ+tAJs3vNlBzfwXY1tSNryxCNhC+L61aBs+WbETQeF7UaNquL+3isvc5+vnayOZZLOxXXitTFQiac9Y1W0lrjwyaz1mrs7E2N5u3NebBHCsbBit79rVB6PVXju+IyPmWMRTO2xlRx1m7W0lrqiqeqF4bZAxZu82MTsIcVAdkZ1xOYiIsthUKxOVUD7MdnAncyobZtnwRjNg7zsqVny34ZOVO5GVt+LbNk/4XhoF4Wv1L9VO7cd2+bSViYpTuSup3VbiymNfbsCnq3bh/l4puP/yv730Vxbp0jHZ5YtOI0hG3DJiJh+P7hboPAG8ljFmzwUhThtYbGovCWCbiIzHaqxMVEgA0xngePRpO9sU6YmKEL5vLfTrE/6g8G3fuDru6+1G15TIr/gW5irSeO20jR112YWXBHBBa1iJK/+cvQGfrNyF+3qmYOwVJIDt8PN4rsOuPhjPGGUfZ2X9A6tV3FbiCglge0YEEsD28BiXtViZqJAAJgEcl05tY6MiNTnbc+Q03lq4DTOF8A3knvHtkFxd3+rcuel5Ed/qXBxFkcJro0lsrcouvFYmKuECyrsT4qG8OyE2c87vZ4wtKa4et9t9naIo4sbQpgC2ARjPGPtK5G/Xrl3SsWPHJgLo63K5LuCcHwXwo6Ioj3q93j3BOlVVzRAXj4d+g3P+vM/ne7So71qJK49/tREfrdiJe3o0xbgr//bUfbh0OSa/Xb7oGMB5DZURt4yYhbkJt33vANerV69SSX29UqVKrV0u1zJaAbY2IpIAtsZfXJe2MlEhAUwCOK6d24bG2R2wxRnfN37x49OVfwnfjk1q4P5eblwSQ+Fr9S/VNlAdkyrssm+0BHDwVQAAYzjnywCMUhTljpycnObp6ek7C5OYdyGiEMfiKQwhegcAeJJz3tXn86244IILzk1MTJylKMo0AOsDgUB1l8s1iXOe6PP52ocKYEVR3klISBD5gunEli1bTtgtgCfM2YgPf9uJu3s0xYMkgGPSL6L5Ubv6YDTbbPVbMmImAWzvFmhVVcVfznkJvii0GycBbK23kgC2xl9clyYBbN48sgUx2fDaGbAPnczSz/i+vzwj/3KrTkL49lZ14RsvSTYb24U3igJ4haIoa71e710h4nQrgDnFvAs/E0BVxthVwfxut3u+oiiHGWM3FeV3qqp2ALAyJyencVBUixVgRVEmeb3eSUZ81Upc+dfcjfjg150Y0/0CPNynmZHPlYk8dvmi08iQEbeMmO2Mp7L4eElxRVXVXkZ4YIz9ZCQf5SmaARLAZdgzrExUQmmRcUCXDbNseO0I2EdPZ+OdJel4Z+l2nMwK6F1GXG714BUedE45P+5GFtlsbBfeaAjg5s2bl8vJyTkFYHBwC7NwII/HM5lz3oYx1q2wQ6mqKlaFX2WMiWfx9KSq6lhFUe73er0FtjSHCOTeiqL8oGlaNb/ffyyvjNgCXR6AeFJjF+f8i6SkpBe3bNmSVZQTB+OKqqphP6/35Ddb8P5vOzH6siZ46Eq5zgBv3LgRrVq1QkKCPPfWiD4oG24ZMQfjqWy2toJbxBXGmKiiXfv27dfG3YRBggaRAC7DRiYBbN64dk2ezbcguiVlw2tFAJ88m4MZyzPw9uJ0CBEsUsv6VTHuCg+6qzVjdsa3NI+RzcZ24Y2GAPZ4PPU457tdLleXtLS05SGC9p8AbmGM/U0tqqqaxTm/1efzfRIicIcoivIeY0wI2gIpOTm5QlJS0lIAaT6fb1ioaNY0bW1CQsJhznlHAM8CmMsYu6MkAVyavxX1+/fWHcO3vlMYkHoOhrWqYqYKKkMMEAPEQFligARwjKxJAjhGxEfjsySAzbNs1+TZfAuiW1I2vGYE8JnsAD76bYd+wdWfJ3MXx9y1KmPcFSqubFEnboVv0JNks7FdeKMpgDVN6+z3+38NEbTjFUUZzhj7241ReQL4Fp/P92kwv8fjGco5f4cxViF0BBEXYh0/fvwLAI00TeseXP0tapTJu1hrlsvlOj8tLe3PwnmsrABP/G4L3lu+E3d2TcajV8l1CRatjsmx8k0rwLTLwcjsjVaAjbAU2TwkgCPLb0xrJwFsnn67Js/mWxDdkrLhDUcAiyeMPl+9C6//7Me+Y2d0wySfVwljL1dxTet6SHA5YxiVzcZ24Y2GAI7kFug88fs5gAtcLlfPokRt6Gijqmp9AJmc84vFZVrFCeBmzZqFvQV64rebMX1pBu68tAnGX908uoNcDL9mly/GEIKpT8uIW0bM4cRTU44Ux4XM2jsacSWOaYuLpjlj5hYXVDmvESSAzdvM7KBm/ouxLSkbXiMBO6BxzPl9Nyb9xLDr0GndQPWrVcS9vVIwsG0DJCW4Ymu0ML8um43twhutiYqqqis452t8Pt+YoGlVVd2Stx35scLmVlVVXIJVhTHWNyT/9wCOBC/BChG/bk3Tevj9/j9Kcxu3232NoijfhF6UFVrGSlx5+rstmLZkO27vmowJ17QorSll5vd2+aLTCJERt4yYjcRTp/mu0faatXe04opRHDLmIwFchq1uZaISSovZDu5kamXDLBvekgI25xy/eA/ghflepO07rrtxzSrlcU/PFNzQoSHKJzpzK59sNrYLb7QmKsFnkBRFGa1pmtgGPVJRlDsTEhJabN26dYeqqh8A2B28ETo1NbWzpmmLFUUZzzmfqyjKtZzzicFnkLp37564e/fuLxVFaQvgmsTExP0hY/IhcclV3lNKF3POf3G5XOKd4A6c81c556t9Pt+1RY3hVuLKs/O24L+Lt+P2LsmY0I8EsJNjpJG229UHjXwrXvLIiJkEsL3PIMWLL5f1dpAALsMWtjJRIQFM7wCX2a4RCAC/L4F2YDd8R04gZdAIJJQTF+ACa3cexnPfp2Hl9kP6f1etkIgxPVJwyyXJqFjOmcI3aEfZJmZ24Y2WABZ2UlVVrP4+DKCuoiibNE0b6/P5FovfeTyehZzzDMbYrUGbejyeQUL0iu3NALZxzsf7fL7Z4vepqanJmqZtL6ofu1yuHmlpaQtTUlLaulyuN0X2vJugdwD47MSJEy/s2bNH3Er9t2Qlrjw3bwumLt6OEZ0b44n+LcvsEFMYmF2+6DTCZMQtI2YSwPYJYLfbLY6rGEo+n+96QxkpU5EMkAAuw45hZaJCApgEcJnsGj/PBl66DziQmQ+P12qAbaOn4IUjDfHDltxFsvKJLtzaJRljuqXg3EpJZYIK2SZmduGNpgB2gqNZiSvPf78Vby1Kx62dG+PfJICdYG5LbbSrD1pqRJQLy4iZBLB9AlhV1Q+NuixjbLjRvJTv7wyQAC7DXmFlokICmARwmesaQvw+PAgAz4e2t/x5ePWCoZhVrxc0JQHiPqvB7Rri/svdqHtuxWIpOHz4MHr1yn2rftGiRahSJf6fdImHidnx48fRrVvuk7Y//fQTqlevHjE3swsvCeCCJrISV16YvxVvLkzHzZc0wpPXtoqY7eOtYrt8Md5wldYeGXHLiJkEsH0CuLQ+Rb+3jwESwPZxGXc1WZmokAAmARx3Dm2lQWLbc7/k/JXfo4nn4M3kwZjRsB/OJuQ+mXrF0XV4+F93I6XuuaV+6aWXXsK0adNw8803Y/z48Xr+tLQ0PPfcc/D5fDh69CjOOeccNGrUCDfccAMGDRLC+6/0zTff4N1338W2bdtQoUIFXHzxxXjggQeQnJycn+nEiROYPHkyFixYgEOHDqFOnTro378/7rrrLiQlhb8qLb71xBNPwOv14uzZs2jatCluv/12XHPNNSXiXbFihY6zqPTee++hc+fOyMzMzP+DQFH57r77btxzzz36r5555hm8//77GDlyJMaNG1cq12Yz2DURJQFsnwB+aUEaXv9lG4Z3aoSnBpAANuvbTilnVx90Cl4SguELQSfZtqi2mvVxiiuxtzwJ4NjbIGItIAFsnlqzg5r5L8a2ZJnHu3ohMLoHzriS8H7Dfngj+XocS6qsk97x8CY84p+BdkfTgKm/AO27l2iM7OxsXHrppRCrwLNmzUKrVrkT+R9//BEPPfQQ6tatq4vajIwMnDx5Uv/dK6+8gquvvlr/98yZM/Gvf/1L/3eDBg1w5MgRCLFbo0YNzJkzB7Vr14awx/Dhw7FmzRpd7Ip8O3bsgKZpumB9+eWXw3KY/fv349prr9XbXLlyZVSrVk0XrSI99dRTuP764o8SBQWwaEfz5gWfrnn88cfRunVrHDhwAELkhqZjx45h+/bcI6j/+c9/cOONN+r/Fu+hij8IiNXfJUuWmBLzRsDb5dM0USnItpW48soPaZjy8zYM69QIE0kAG3FjR+exqw86iQQZMQv7EO42SEgwfk+I0biiquo/AIgA3QhA7mUleYkx1tFJfSPe2koCON4sYmN7rExUQpsh48AmG+ayjlf7/lN88/p/8ULKLdhdsZbu3p4TGXjENwM9/lyN/IFw4idAn5tK7IViy7NYvaxZsyaWLl2anzcrK0sXc4qSW9vOnTtx+eWX6/8eNmwYJkyYAJFHiGcheq+88kpMmTIFQpxeddVVulgO5hOrvvfee69edurUqejRowc+/PBDTJwo7jsCvvzyS7RsafwSISFyP/roI1SsWBHff/+9LtJF/eI7QgwLIVou7yKwwuCDArh+/fr4+eefDY9QTz75JD7++GOce+65+OWXX/QV8WDq2rUr/vjjD7z99tv5W6INV2wwo10+bXSiYrBZjs9mJa68+oMXk3/2Y0jHhnhmYGvHc2EUgF2+aPR78ZJPRtwyYhb+RrjtF8Cqqoq/Kj+nKMpHnPMRAD7gnKcoinIR53yqz+d7NF76uhPbQQLYiVYz2GYrExUSwLQF2qCbxX22VRmHMHHmCqw/rOltrXPmIB7c9gEG7F2IBOT+LD8ZWAEWq69CuPXs2RNvvfVWgeJiEnDTTTchJydHX7EVK7sivfrqq+jbt6++ojtkyBD9Z6Ke4Pbj2267DcuWLdO3QAtRKlZWv/jiC30l+ffff4fL5dKF8mWXXaaXHTt2LEaPHm2Y+yuuuEJvj1itFivQ4i/VYhv2gw8+qNfx6aefom1b8VrO31PoCrBoj0hNmjTRt0/36dOnyDJC4Hfv3h2nT5/W2ynaG5rENm4hpkeNGqVv/Y5EsmtCRgK4oHWsxJVJ//Ni0k9+3NShIZ69jgRwJPw+nuq0qw/GE6bS2iIjZsEJ4Y6IAE5TFOUpr9f7saqqxznnF/p8vnRVVZ/OewM+96/klEwxQALYFG3OKGRlohKKUMaBTTbMZRFvxsGT+pNG8zfv0925UuAMxmz/ArfvnIOK2tlCnVgBajcAvt4OlLKNKbhyKrYoC6EamoTwbdHir/dNExMT8cgjj+Sfof3uu+/yBd+MGTNwySWX6MXF1umvv/5aX4UV0ovwIwAAIABJREFUW4SFuBSry+Lcr1hxDk4wgluQxbliscJqNAnhK1afxcqrEO9CAP/666+49dbcF3VCt2gXrjMogOvVq6df9iW2NYu6RBJnioOCPrTcG2+8oa9uCzxC6IrV8tAkVrLFinZwFdwojnDy2eXTJIALsm4lrkz+keHVH324oX0DPD/ownDM6ei8dvmi00iQEbeMmIPxad06OgNstI8aiSuqqp5KSEholvcO/AEAlzPG1jdr1swdCAR+ZYydb/R7lO/vDJAALsNeYWWiQgKYVoCd2jWOnMrClJ/8+PC3DGQHuH6z8w0dGmFsRR9qTQheRvXXTdAcSu4W6BdmAT0Hlgp7xIgRWL58eZErm8HCYuX3f//7n35BlhCbr7/+ur7V99tvv82/+ClUAIuVWLEiW758eWzYsKFIARwqrsV5WnGu1mgqSgALDAKLSMEV6qLqExdwie3ZDRs21H+9Z88eDB48GAcPHtQv+RI4Q5MQx2LLtvi9OOv79NPij9UFk/ie2NrdpUsX/TKwSCS7JqJGJiqRaH+81mklrrz2E8PL//Ph+vYN8AIJ4Hg1sW3tsqsP2tagKFQkI2YSwOELfyNxRVXV7ZqmDfT7/b+rqroKwNuMsWlut7u3oiifM8ZqRMGly+wnSACXWdMCViYqJIBJADuta2TlaPjg1wy89rMfR09n683vptbEP/s2g6dO3jNFRb0DXLsBlHGTDYlfUae4zfiHH34ocAN0cVwNGDAAW7ZswUUXXYTPPvuswBZocZN0v3799KLFbYEWglj8VT2WW6CLwhZcBRdnnjdt2lQgy+eff66fdxZnocWKt7htunCiFWCn9a6/2mslrrzxsw8v/sAwuF19vDi4jXNJCLPlJIrC2x4aJr1xlZ1sLY+trQh/IwLY7Xa/oyjKDsbYk6qqjlEU5UXO+WIAnQB8zRjL3cJFyRQDJIBN0eaMQlYmKiSASQA7w8sBzjnmb9qH5+anYcefp/Rmp9apogvfy9SCW2/1X4onkX5fAu3AbviOnEDKoBFIKOYCqKI4CD6B1Lt3b4itvsEktjB36tRJv8VZJLFV+LrrrtNXTz0ej77FOfQSLHEu97XXXtPP9oqztCIgBrdVh16C9eabb+pPDImng8QTQiLNnj1b32ot/v+xxx7Tfybe1RW3RReVQi/BmjdvHsR2ZnFrs1i9Db0ES5xLFj8TGMT3RBI3U4szvxdemLtldd++fTouscIbPLMc/KawhTjrnJ6erq8Ci1XeolLwDPCdd96Zfw7Zbn+zayJqZKJid9vjuT4rceXNX3x4YQHDdW3r4+XrSQDHs53taJtdfdCOtkSrDhkx54ZVueZMQX8yi9tIXOnevXtiZmZmgt/v189sud1ucYFIV0VR/JqmvRH8ebR8u6x9hwRwWbNoCB4rE5VQWsx2cCdTKxtmp+LdkHkET36zBat3HNbdrWaV8hh3uYrB7RsiQex9LiGZxSxuNBYXO4mblBcuXJj/BSFeV61apYtLceOxEIFi27JIjz76aP524+KeQRLPAs2dOzesZ5CMCuDCzyCJm5l3796tty30GSTRzq+++gqhNz4HfybaV6tWLf15J/GOsEji3WOxyh1MQoSPGTNG/09x63SHDh2KtEDwFuj//ve/+mVZkUhm7Vu4LUYmKpFof7zWaSWuvLXQj+fnezHwonp45YaL4hWi7e2yyxdtb1iEK5QRt4yYhRsR7vBWvimuRHjwMVA9CWADJDk1i5WJSihmGQc22TA7De+B42fw4nwvZq3NBOdAhSQXRl56AUZ1a4pzyica6rJmMYeu4orV0WbNmunfE2d6xSrvrl279FVfIYLFyq84B/uPf4in/P5KIp84+7pt2zb93K+4DGvcuHH6imowiXPEkyZN0m+FFu/3CvEp6hGrp2LrcfCbzz77rH7JlNiWXalSpWKxi2+JS6vS0tJ0ASu2JovLtoLbsEXBogSwuCxLbN8Wl3OJVV9xEZaqqhCrt507dy7wvaFDh4qjF/pt0+KN5KKS2DItVpBLe37JkBFLyGTWviSAS2beSlyZutCP5+Z78Y829TDpRhLAVn083svb1QfjHafs8yUSwJE5Ayx4TUlJqZqQkNCec16Lc+4K9TWfz/eJk/pGvLWVBHC8WcTG9liZqMg+oMsWuJ2CV5zznbF8u37J1YmzuaurYjL9yFWpqHtuxbB6jxXML774IqZPn66v6grRGKsU3MZc0iVWwbZZwWsnPiHYxR8LIrn92c4JGf2lvqD1rcSVtxf58cz3JIDt7E/xXFe8jDnR5EhGzHaOt9G0lR3fMmtvI3FFVdW+AITIrQpAnO/66/ZOgDPGxM8pmWSABLBJ4pxQzMpEhQSwXOdZzA7i0eoH4mzpL94DeOrbrdh+8KT+2dYNzsUT/ZqjXWNzFyFawSxWZMW5XHHRk9gGLVZFo50EJ2LluGXLlroYLy1ZwVta3UZ/f/z4cf02bJHEdmmxrTpSyS68RiYqkcIQj/VaiSvTFm/D0/PS0P/CuphyU9HvTscjZqttsssXrbYj2uVlxC0jZhLAkVkBVlXVC+B/iqI85vV6j0e7/5b175EALsMWtjJRIQFMAjheuob/wAk89e0WLGJ/6E06v3J5PNLHg+vaNoCrlHO+JWGQbaJCeM15NAnggrxZiSvTl2zDxO/ScE3rOnh9SDtzBnFgKdn6XtBEMuKWETMJ4MgIYI/Hc1LTtFY+ny/dgcNe3DeZBHDcm8h8A61MVEgAkwA273n2lBRPGU35yYf3l2cgR+NISlBwW9cmuLtHCqpUyD0DayXJNlEhvOa8hQSwfQL43aXpePLbrbi6VR28MZQEsDmPdE4p2cYcEoLhC0HneHPRLTXr40biitvtnutyuT70er1FX6jhdPJi3H4SwDE2QCQ/TwLYPLtmBzXzX4xtyXjCG9A4Pl+9Cy8t8OLPk1k6Mb2b1cL4q5ujyfnn2EZUPGG2DVQJFRFecywbmaiYq9mZpazElfeWpuM/325F35Z18OYwEsDO9ADjrZZtzCEBTALYaO8wElfcbvcIRVGe4JxPd7lcGznn2aH1M8bmGf0e5fs7AySAQzhRVfWfFpxEHEh/1kJ524tamaiENoaCWILttom3CuPFxmt2HMK/5m7G5j3HdIqa1jwH/+rXAt2Kes/XIonxgtkiDMPFCa9hqgpkNDJRMVezM0tZiSvvL9uOJ77Zgqta1sZbw9o7kwATrZat7wUpkhG3jJhJ+Icv/I3EFVVVtRKGG6E5yv7k1MR4a7QICeCCAlg4m7hlrTAvoTevBX9X+Gdx54xWJiokgGkLtNFBxI58B0+cxfPfp+GLNZl6dVUqJGJsbxXDL2mMpIQCN//b8Tm9DtkmKoTXnOsYmaiYq9mZpazElQ+Wb8e/vt6CK1vUxn+HkwB2pgcYb7VsY46McUXmP3ZYsbfBuFKawA0Y742UszADJIBDGPF4PLcXJohz3gnAHQAyOeezFUXZKfJwzhspijIQQAMA0xVFWeH1et+JJxezMlEhAUwCOBq+LLY7f7JyJ16cn4ZjZ3KfNbqhfUM83MeD8yqXj2gTZJucEV5z7mRwomKucgeWshJXPvx1OybM3YIrmtfG2zeTAHag+cNqsmxjjhVBFBaxcZhZRltbsTfFldg7MQngEmygqmpHAIsAvFKvXr0nFi5cmDtDz0vdu3dP3Lt377855w9yzrv5fL4VsTfpXy2wMlEhAUwCONK+/PvOw5gwdxM27c7d7tyiXlU8eW1LtGscuadxZPZr2SYoduGliUrBkcBKXPn4twyMn7NZP9M//ZYOkR5i4qZ+u3wxbgAZbIiMuGXEbEUIGnSluM1m1t5G44rH4+kiNAaAZnk7VLdqmvai3+//NW5JcUjDSACXYCiPx7OAc16XMda6JHuqqrpBUZS9Xq/3yniyu5WJisxCQcbB3OwgbsbfD53Mwgvz0/DZql16cbHd+aErPRjaqTESLDxrFG5book53LZFIj/hNceq0YmKudqdV8pKXPnktwz8c85m9EqtiXduFX9fliPJ1veCVpURt4yYZZwzWfVxI3HF7XbfpCjKhwDmAliWdzyzM4D+AIYxxmbKMYJGBiUJ4BJ4VVX1iKIo33m93qEl0e/xeD7mnF/NGKsWGTOZq9XKRIUEMK0Am/O64ktpGtdF7wsL0nDkVO5lhoPaNcCjV6Xqb/tGO8k2USG85jzMyETFXM3OLGUlrny2Ygce/WoTenpq4t0RJICd6QHGWy3bmENCMPzLoIx7U3zmNOvjRuKKqqpbOOfv+ny+l0LRu93uhxRFGcEYax6frDijVSSASxbARznnPp/PV+JhJbfbvVpRFDdj7Nx4MruViQoJYBLAdvryhswjmDBnE9ZnHtWrTa1TBU/9oyU6JNew8zNh1WU2cIX1kTjKTHjNGcPIRMVczc4sZSWuzFy5A4/M3oTuak3MuI0EsDM9wHirZRtzSACTADbaO4zEFVVVzwQCgZbbtm3zh9bbtGnTlISEhE2MsQpGv0f5/s4ACeCSBbB4Y0tsa36QMfZqUVk9Hs/9nPNXAMxnjPWNJyezMlEhAUwC2A5fPn4mW3/P94PfdoBzoEr5RDxwhYrhFzdGYoRudzbabtkmZ4TXqGcUzGdkomKuZmeWshJXPl+1Ew9/uRHd1PPx/m3ifkk5kmx9L2hVGXHLiJmEf/jC30hcUVV1G4BnGWPTQ0dKVVXvBPAIYyxFjhE0MihJAJfAq9vtbuNyuZZxzsVfWTYDmMU536koingCqbHYwSnu7lEU5YymaV18Pt+6yJjJXK1WJiokgEkAm/O63FKcc8zftA///mYz9h87q//sH23q4Z99m6FW1fj4o6VsExXCa86jjUxUzNXszFJW4soXq3fioVkbcan7fHx4OwlgZ3qA8VbLNuaQEAxfCBr3pvjMadbHjcQVVVX/D8DLAKZxzpfnaY+uAMSLNeMYY2/GJyvOaBUJ4FLslJqa2jkQCMxQFEX8paWo94C3KYpyq9frFQfU4ypZmaiQACYBbNaZdx85jSfmbsKPWw/oVSSfVwlPD2iFLinnm60yIuXMBq6INCYKlRJecyQbmaiYq9mZpazElS/X7MK4Lzaga8p5+OiOi51JgIlWy9b3ghTJiFtGzCT8wxf+RuOKx+MZzDkfl3cLtKB6K+f8RZ/P96WJoYiKhDBAAtiYO7jcbncvRVHEX17qiZvYOOd7OOdL/X7/jwA0Y9VEN5eViQoJYBLA4XprTkDDjOUZeOV/DKeyAkhKUHBXt6YY0yMFFZJKe8893K9Zzy/bRIXwmvMZoxMVc7U7r5SVuPLV2l0Y+/kGdGl6Hj6+kwSw86wfXotlG3NICIYvBMPzqPjLbdbHKa7E3pYkgGNvg4i1wMpEhQQwCeBwHFNccvXY7I3YvCf3Td8OydXxzIBWcNeuEk41Uc1rNnBFtZE2fozwmiOTJioFebMSV4IC+JILauDTkZeYM4gDS8nW94ImkhG3jJhJ+Icv/CmuxH4gJwEchg26d++emJmZWV3TtKz09PTc62zjOFmZqJAAJgFsxLXFJVcv/8Dwwa8Z0DhQtUKifs73+vYN4Yrim75G2lo4j2wTFcJrxksAmqjYJ4Dn/p6J+2aux8UX1MBnJIDNOaSDSsk25pAQDF8IOsidi2yqWR8vLq6oqirOjjVnjB1UVfWPQkcvC7SBMVbL6fzFsv0kgA2wL25c45yPUhSlNYAERVHe93q9t4miHo9nkKZp13POH/P7/eLGtrhJJIDNm8LsoGb+i7EtaQavfsnV15ux79gZvfHikqvxVzdHzSrRf9PXDHtmMJv5TryUIbzmLEEC2D4B/PW6TNz72Xp0alIdM0d1NmcQB5aSre8FTSQjbhkxk/APX/gXF1c8Hs/tgUDgI7/ff1ZV1TvyLr4qctTzer3vOHA4jJsmkwAu2RQJqqqKg+b9AOQAEAI3VVGUGSECuDXnfB3nfILP53s6biwLgASweWvIFsTCwXvg2BlMmLsJCzbv1wlufF4lTPxHS1zqrmme8BiUDAdzDJpn+ycJrzlKSQDbJ4C/Xb8bd3+6Dh2Tq+Pz0SSAzXmkc0rJNuaQEAxfCDrHm4tuqVkfp7gSe8uTAC7BBiFv/M5LTEy8Y8uWLftUVdVCBbAorqpqOoA9jDFxSVbcJBLA5k1hdlAz/8XYljSCVzxt9MXqTDz13RYcP5ODRJeCUd0uwD093XF5yVVpjBrBXFodTvo94TVnLZqo2CeA523YjTGfrEP7xtUx6y4SwOY80jmlZBtzSACTADbaO43EFVVVszRNq+/3+8VW6PzUokWLGtnZ2fsYY+WMfo/y/Z0BEsAleIWqquJd31oVKlRwb9iw4WSe2C1KAP9PXFHOGGsQT05GAti8NWQL3KXh3XXolH7J1VL/QZ3U1g3OxfPXtUazulXNkxzjkqVhjnHzbP884TVHqZGJirmanVnKSlyZt2EPxnzyO9o2qobZY7o4kwATrZat7wUpkhG3jJhJ+Icv/I3EFbHgFggE6mzbti33Tcm85PF46nHO0xljFUwMR1QkjwESwCUL4FMAFjDGBgSzFbMC/CmAAfHmjFYmKqG0yDigy4a5OLwBjeP95Rl4cYEXp7MDKJ/owrgrVNzWpQkSE1yOHkjJxo42X6mNt8u+RiYqpTamDGWwElfmb9yD0R//josaVcNXJIDLkFcUDcWuPugkomTETALYXgGsquqYPJ9/TVGUxzjnJ4J9gHOe4HK5unHO3YyxC53UN+KtrSSASxbARxRFWeH1eq8sRQAvAyCcMa5uZLMyUSEBLNEt0IEAAmsWYsea39C43cVIaNcdSEiA/8BxPDxrA9buPKK7Q8cmNfRV3ybnnxNv45ip9sg2USG8ptyEboEuRJuVuLJg016M+mgt2jQ8F3P+L65ODJlzDoOlZOt7QVpkxC0jZhLAtgvgXXl9qD6AvQC0kKEmC0CGy+WakJaWttzgEETZimCABHDJAngJgBZJSUkpmzdvPiSyFl4BTklJaeByuRiAhYyxvvHkZVYmKiSAJRHAP88GXroPOJCZb/LsWo3w34FTMGV7OWQFNFQun4hHr0rFkI6N4v5po3D6n2wTFcIbjnf8lZdWgAvyZiWu/G/zXtz54Vpc2OBczL2bBLA5j3ROKdnGHBKC4QtB53hz0S016+NG4oqqqkvKly/ff+PGjYedzlM8tp8EcAlWcbvdIxRFEdeMz8vJyRkq3v4NFcD16tWrdM4558xWFOVyzvn1Pp9P3BgdN8nKRIUEsAQCWIjfhwcVeGZuU5UL8FDz+7G1ygW6C/Tw1MTTA1qhXrWKcePXdjXEbOCy6/vRrofwmmPcyETFXM3OLGUlrvy4eS/u+HAtWtWvim/uudSZBJhotWx9L0iRjLhlxEzCP3zhT3HFxEBqcxESwCUTqqiqOkuc7wUg9oEuBtAfQBqADQB6ATgPwEzG2E0228ZydVYmKiSAy7gADgSAfsn5K7/ZSgLeTB6M15rciBxXIqpnHcMT+2fh2vdnQklMLNEXDx8+jF69RFcAFi1ahCpVqlj23WhUEOuJivj+FVdcgb179+Kbb75B06ZNIwo71ngjCq6Iyu3CSxOVguRaiSs/bdmH2z9Ygxb1quK7e0kAR7tPRPt7dvXBaLfbyvdkxEwC2D4B7PF4Xihfvvx/xMW74t8l+aLX633Yiq/KXpYEcOke4FJV9VEADwCoUSj7cQCvMMaeKrRHv/Rao5DDykSFBHAZF8CrFwKje+hmZuc0wrgWY7Gxqlv/7z4HlmHi1jdxfvZRYOovQPvuJXrrSy+9hGnTpuHmm2/G+PHj9byrV6/Gxx9/jI0bN+LgwYMoX748LrjgAtx+++3o3bt3gfqWLl2K119/HVu3bkVCQgIuuugi3H///WjVqlV+vuzsbLz11luYO3cu9u/fjxo1auDKK6/Efffdh8qVK4fdm/744w+Idv/88884ffo0GjVqhBtvvFHHUFLKzMzMF/uF802cOBGDBw/Wf/ztt99ixowZ2LVrF06ePIlq1aqhWbNmGDlyJDp06JBf9P3338czzzyDq6++Gq+88krYOMIpINvEzC68JIDtE8C/pO3DiBlr0KJuVXx3HwngcPqvE/Pa1QedhF1GzCSA7RPAYttzVlZWv4yMjCPi3yX4PmeMXeakvhFvbSUBbNAi7dq1Szp69Gg7l8uVrChKQiAQyMzJyVmRkZFxxmAVUc9GAtg85WU+iM3/FIHHh2Fa4wF4pekwZLmScG72cTyZNhX99y9C/sAw8ROgT/GbG4QwvfTSSyFWgWfNmpUvWl977TVd1ApxWq9ePWzfvh0ir0ivvvoq+vbNPS4vVozvuusuCL5r166NrKwsva4KFSpg5syZSE1N1fONGzdOF5UulwuNGzeGEKKivvbt2+PDDz/Uf240CUE6cOBAZGRkoFy5cqhbty527NihFx89ejTGjh1bbFWhAvjCCwtewDhq1Kh8cfz000/r7a1ZsybE+8np6enIycnR/xAwb948NGiQ+2Lan3/+ia5du+rtX7x4Mc47T2woiUwq8z5diDa78JIAtk8AL0zbj1tnrEazOlXw/f3yzN3s8sXIjAyRq1VG3DJiJgFsnwCOXG+kmgszQAK4DPsECWDzxi3rQSz9l5/x4My1WFutmU5Sj4Or8NyW11A7S7/r7a9UygqwELBiVVMIPbGSG0zz58/Xt0J37twZiqJg8+bNGDRoEDRNQ8+ePfXVXJH69esHxhjatGmjrxifOXMG/fv3x+7du9GjRw9MnToVmzZtwnXXXafnnzBhAoYNG6av3ArhLNKUKVP01WCj6d1338Xzzz+vt+vZZ5/Vv/fiiy/ivffeQ2JiIhYuXKjjKSqFCmCv11vsJ8+ePauL3WD64osv8PjjjxfZ3htuuAHr1q3DE088gSFDhhiFEXa+su7ThQmxCy8J4ILMWokri737cfN7q5FapwrmkwAOuw87rYBdfdBJuGXELOxDuNvoO9iMJoorRpmKXD4SwJHjNuY1W5mohDZexoGtrGLWxLu+v2bg+flpOJOtoXLOKfyLTcPgPf/7a9VXN74C1G4AfL1dfxKpuPTyyy/j7bffLiBqi8orVkHF1t/jx4+jT58+mDx5sr6V+bLLcleBHnjgAYgVVJGEyP3888/1VeC1a9fq9U+aNEn/3ZIlS1CrVi1dSAvRLITm9ddfj6eeEqcQjKURI0Zg+fLlSE5O1rcfi3rWr1+Pm27KXekWW6OFMC8qhQrg6tWr698X26eFcBXbn0NXojds2ACxLVpssQ6ugAtR/P3336N+ffG6QW4SecQq9jXXXAPBZ6RSWfXp4viyCy9NVAoybCWuLGYHcPO7q6DWrowfxnaLlKvHXb12+WLcASulQTLilhGzcAPCbY8Adrvdnxvt5z6f73qjeSnf3xkgAVyKVzRv3rxcIBC4nnMuZup1Afy1rBNSVlEUHvpecDw4m5WJSmj7ZRzYyiLmXYdO4aFZ6/Fbeu4qb9caGp6fewfqn/2jwE3QuvgV6YVZQM+BJbryvffeiwULFmD48OH5K5xFFQiugIpV1+nTp+vbfsWqp1j9DIrA4PlZIXaDK8RiVVlspxbboUXasmVL/l9Zu3Xrhn379ul1vfOOuKzdWBICXAjStm3b4sEHH9QFsFhxvvzyy/UKxHZrsapdVAoKYLFCLLYr79y5U38nVqQ77rgDDz30UH4xIbL/n73rAI+q2ML/7CaAFEUEFARFyM4GkCYgKlL1WbBifyqKFVSUYkXsvWBv2BXE3lBsKF1AOlKzs6EpoCBIbyl33ndusrxNSDZ3797dvTcz831+anbaf86ZOfPfmTlDZDuS6N4y7VZH3wGm32jn+YknnsDRRx+NL75IXiD5ymjTsTTuFF5NgEtKORG/8mt4Ay5/ezYC9Wvi5yGaAFubsbyby6kx6CUJqIiZ9KNxO0OAOeejIvYupWSMMfoav4MxNpf+LqU8BgBFGv1GCNHHS2PDbX3VBDiGRpo3b96gsLBwPIBg0ZZYzEQX0q2ff0iBJSSyUInunooTW2XCTLuvn8z+Ew+PXYqdeYU4INOPu89ojss7HQE28av93gHGoY2BW5+vkPySjUR2U2PdnaW7wXS8l+7A3nXXXftI4fz5883AU5SiA0jRHWE6+kxp2rRpJmksiwDT7jHtItMdZCLVVhMdl6b7v9EEmIgtRWSmFIsAExmivJxzM++WLVtw2WWXITc319yxnj17tnmvODpRELDXX38dI0eOxGGHHYaPPvrIvBcdSZGPA7QjTR8TkpUqk01bkZFTeDUBLintRPzK9Nx/cOlbs5BVrwZ+uTV2cD0rOvZKHqds0St4I/1UEbeKmEnfGrczBDh6jHPOH2OMHVazZs1+c+fONYOodO/ePWPdunUjpJQbw+EwBejVyaYEKiJ1NqutHMWCweBoKSWdi5zJGHtWSikMw9hRHrrc3NzlbkKeyEIlGoeKE1tlwbxpx17c9eUi/Lx0vanSY5vUwdMXtsaRh9T4v4oLC1E4dxJWz/0NR7Y/Dv723WMee462jZtvvhnjxo0rEQE68jsRb9rNJTJLd2sffPBB8x5wJEUfgabAU0SiKcU6Ak2BoihYlhNHoCmYFt0BjucIdFnjm+4T071iSpH+lc63Y8cOtG/f3vwzHfWmI9+RFNkBbtmyJb788sukTSGVxaatCsgpvJoAl5R4In5lRu4/+O9bs9C0bg1MuE0TYKu27NV8To1BL+FXETPpR+NOCgH+hzF2YqhUwBHOOUUH/VUIUddLY8NtfdUEOIZGOOcbGWM7d+7cmb1mzZrdblNeRf1JZKGiCbD3n0GaGNqA2z9biI079iLTz3DbKUFc26Up/L79h71d5xV5AomeNnrllVf2mQ1Fcx46dKgZCZmCYdEuLgXEKp0iQbBat25t7oxWFASLnlmip4p++eUX3HTTTWZ1dESadm9nzpy57xkj2m3t1KlTmUMkOgg+TSP7AAAgAElEQVQWRWs+99xzzaBY9CQREXUK7FW3bl188MEH5j+UKKgXJWqXdnrp2DWlbdu2mfd/w+EwqlevjlmzZiEzM9MM6NW7d2/zb5RIDrSzTKn0cfHIHWCKjE2738lKdnWcrP4ku16n8GoC7BwBnrl8Iy5+cyaaHFIdk24veoZNheSULXpNViriVhEz2aXGnRQCvIUxdkUoFPomeuwHg8FzpJTvCSEO9tqc4Kb+agIcQxvBYHCnYRjfh8Phosc9PZY0AbavMC9P5rvzCvH4D8swckbR0z4UcOb5i9uhRcMDyxWIXbwTJ040d27pKSGKnhxJdOQ38q4t7djS0d9IovuzEbJMZJPK045uos8gWSXAVp9BijzlRP2OfICN/I1IPR1jpnd+I3eAaTd8wIABJsxgMGgSYQqQRUe/I88sEcH+8MMPEf2EUiQK9H333Wcep05WsqvjZPUn2fU6hVcT4JKaSsSvzF65ERe+PhNH1DkAU+7omWwTcE39TtmiawBZ7IiKuFXErAlwcp5BCgaDL0gpL2OMPVJYWPgbydnv9x8npRzGGPswFAoNtDgUdbYyJKAJcAyz4Jz/xhgzQqHQ/ltXHjCnRBYq0fBUnNC9innx2q0Y+PF8LP9np6nCqzo3wZ2nZaNaZuzr6Xbx0k4v3cGlu7Bff/01mjcvelYpmjyWHioUAZmeMYokiuxMhDgS4Kpdu3YYNGgQaFc4kujNXwqMRW1s2LABFIGZ7vJSPnprmNLPP/9sElDaoaW7tNGku3QfqA6KuDx+/HgzSjMRVbqPfOWVV+7LWhYBpieZKGIz3V+mI9zU1lFHHWXuPEfeNqYKaPebIlhTOyQj6i8dtaZAWfTvSKJ3gEl+FBwssvOcrKnFro6T1Z9k1+sU3lQSYM75jQAokloDxtgSKeUgIcTU8mQVCATOZ4xRCPRmAOgKzjAhxFeUn96u37Zt2yMAevl8vqZSyq10iIExdlcoFFoXqbNVq1YH792790UAZxf/7Zu8vLybV61ataWsdhPxK/NWbcJ5I37D4bWrYdpdJyXbBFxTv1O26BpAFjuiIm4VMZM5aNzO7wAT3+Wc3wmAiG7kbUaKWvqCEOJJErvFoaizlSEBTYBjmEXx4oLCz/aItQhxq2UlslCJxqTixOY1zIWGxOtTluPZcQIFhkT9WlUx/MI26MrLfs+2tM0mgpfe0KUgVBQQi4JcpSvRfd733nsvZhArt9k1HdWmY9jJPv6s4gIlEZuOtpNUEWDOOYVEpwigN0opp9F1ccbYtQUFBS1WrFjxR+lxlZWVdbzP5yNyfC8AIr29ATwkpTwxHA7PbNq06UEZGRmfM8beBPB7YWHhwT6f73kpZUY4HO4QqY9z/oOUspGU0gx9zhh7gzG2SghR5ltgifiVBX/8i3NfnYHDDqqG34ZqApyuuTJV7To1BlPVXyfaURGziv4lYit29R2vX2nZsmWdwsJClpOTs8kJO9V1VBzZWGkZBYPBhoZhDGSM9ZNSDqdNJsbYGnryqCzBRH9Vd4PgElmouI0opFqedie1VPeT2luzeReGfPI7Zq0qet7otJaH4fHzWuHgGiWjEcfqWyJ4N2/ejJNOOsncxaRj0HQ8OB3pvPPOA+0Uf/XVV+Zd3opSIpgrqtvK79Q+3V1et26deUe4WTPaxEteSjfe5CEru2an8Ma7ULGLk3NOwRbnhUKhGyJ1cM6XAfhaCDG0dL2cc/o4e6AQ4vTIb4FA4EfG2GYhRNGj1qUS57wjgFkFBQVHEqkOBALNGWNLpZTHEWmm7FlZWcf5fL4ZjLHs0sFX6PdE/MriNZtx5svTzQ90s4adbFdUnivnlC16DbiKuFXETHapcSdlB9hrQ95T/dU7wDHUxTk3ih9IJTmVSXqjitMzSBWvulNoHoksVKK7qeLE5hXM3y/6C3d+sRDb9xSgRhU/7j+7JS5s38gko/Ekr+CNB1NFeVXDrPFWZBFl/54KAkzvzRcUFNCD0hdGjjBTb4rvgLUVQuz3aC7nnHaFnxNC7IucxjkfzBgbFAqFjiwLTSAQOJkxNs4wjNq5ubnbgsHg1VLKZ4UQtaPzc863SCkHh8Phd0vXE/Er9BRYJMibVckuW7cVZ74yA3VqZGL23WrtAC9atAitWrXa9465VZl5OR/NOarhVhFzhACrputEcJNfEUJQFe07dOgwLzLOOeezLfANM7sQ4lgvzw/p7nt8q+R09zbF7XPOf7VqiMXG2CXFXYzZnCbA9rXhdrKwJ78QD41dig9nFp2MbNu4Nl64pG3J543igO92vHFAsZxVNcwar2XTKJExFQSYThtJKdf6fL7OOTk506MWQ3cDuFIIQW/Rl0ic8zwpZd9wOPxh5IdAIHApY+xdIUTV0vmbNGlSLTMzk3xaTjgcvpx+55zfXVxH0cPWxSkQCIjieh4vXU/Er9iR5pptBRj400bUrMLw/jmH2qlCl9ES0BLQEqhMEihBgIPBIMV0iKSqUsp+xHUBzCj+43H0bVRKOSIcDtP9YJ1sSkATYJuC80IxTYDta8nNZEGs344BH86DWL8DtNHbv1szDPkPR6bfZxuwm/HaBlVBQdUwa7z2LCmVBNgwjBNyc3MjCx0EAgGK9tlHCEHvPpZIxQT4ynA4/FHkh2AweJmU8m0hRLXozBQQa/v27Z8BOMIwjO60+0u/EwEui2BzzsMAqJ4nyiPAdnaAc9dvw6kvTketahlYcK9aR6D17ljsQIz2Rqf7SukdYH3KwYpVlrcDHF2Wc/4GgI1CCJqn9yXO+aMADhVCXGulLZ2nbAloAlyJLUMTYPvKdSNZkFLio1l/4qGxS7An30DdmlXx3MVt0CVgLdBVLGm4Ea997VkrqRpmjdeaXZTOlQoCnMwj0MXk91MATX0+X8/oICqJHIGmiO/xHoFesWE7ej47xbyuseSh0+wpxIOlVBt7ERWpiFtFzKRvjdv5O8B0FcXv93dctmwZfZDcl5o3bx4oLCycI4Q4yIPToWu6rAmwa1ThfEc0AbYvU7dN5lt35+PuLxfhu0V/maAouvMzF7ZBvVr7nXS0BdpteG2BiLOQapg13jgNpDh7KggwNUVBsKSUc8PhMD2FZCbO+VIAY2IEwaolhOgVlf8HAFsiQbCiyG/AMIweubm59ITGvhQJggWgkxBiFv0QCAQ6McboCUDHg2Ct3rgd3YZPQbVMH3Ie3he7y55iPFRKtbGnCXD878J6yJzL7Kq28aQQ4PUA7hBCvB8tdM45vdf4tBCivtftJp391wS45GLgUvrfzMzMb5YuXbqD7lPFo5zou1jxlEtWXk2A7UvWTZP5vD8245aP5mPN5t3I8DHcfmoQ13VpCp/PueHrJrz2tRZfSdUwa7zx2UckdwoJsPkMEmOsv2EYdAz6esbYdX6/v+WyZctWc85HAlgbIcPZ2dknGIYxhTE2TEo5hjF2jpTykcgzSN27d89Yu3btF4yxYwCcmZGRQYupSPp36dKlecUkm0gzvXhAd83g8/no2N3qZDyD9OemHejy9GRU8TOIR/fxdnuK8VAp1caeJsCaAHtoeCbUVbtj24pfKb4Cc4+U8nX6KEkdpYj9Pp/vOsMwHguHw3QUWiebEnBuBW2zA24qFon67PP5mufk5IioKNAVddOMEi2EcNUlF02AK1Jb+b/bndTst7h/ScN823cFho8Lgd75PaJOdbz433ZmwCunkxvwOo2povpUw6zxVmQRZf9uZaFir+b9S3HOaff3DgANGGOLDcOgSMxTKGcwGJwkpaT3eftGSgaDwQuI9NLxZgDLpZTDwuHwl/R7dnZ2E8MwVpbVN5/P1yMnJ2cS/UbvS+bl5b3IGDu7eIH1TX5+/oBVq1ZtKatsIn7lr807cfyTk0Df7lY8foZTYnN9PaqNvYhCVMStImbSt8bt/A4wybU4sOFAAM2Lx9UyKeULbttwc/0kXEYHNQGOEkowGKSFhMzIyHhuyZIl/0b+36piQ6HQvVbzRvIVL3huL17wLJFSDhJCTC2rnkAg0Jcic5b+LS8v74BVq1btKf33RBYq0XWpOLGlG/O/O/Mw+JMFmCyKTiye1aYhHu19NA6slhmviVnKn268ljrpcCbVMGu89gwolQTYXg9TWyoRv7J+6y50enyi2eGVj/eK+7m21CJ1rjXVxp4mwHoH2LnR4+6a7I5tB/wKbbgVuls67u6dJsBp1A/n3DzyBuBGKeU0AP0YY9cWFBS0WLFiRdH7NlGpmAC/kJGRUeJJjKVLl/5dFoxEFiqaABdiwYL0ODE68jxg9Dys27rHvCv34NktcVGHxkldLNqdxNM4fBJuWjXMGq89k3FgoWKvYZeWSsSvbNy+Gx0enWAiW/5YL/gdvMbhUnGZ3VJt7GkCnJ61QzrHgLbx5OwAl9ZpMBgMGoZxNWPsCiFEg3Tq3OttawKcRg1S0BPG2LxQKHRDpBuc82UAvi4r6EkxAX5eCGHpDGwiCxVNgFNPgCnK87vTVuGx75ehwJBoWrcGXr38GGQfdmDSrVRF56UaZo3X3jDSBLik3BLxK5t37EG7R8abFYpHTkeVDPtPt9nTZnpKqTb2NAHWBDg9Iy31rdod2/H4lYYNG1avUaPGRYyxawAcD2COlPKLcDj8dOoRV54WNQFOky7tPHtRTIDfYoytlVLS8YcFhmHcm5ubO78sGIksVDQBTi0B3rYnH3d+vhA/LC7azD+jdQM8cV4r1ErSkefS9mJ3Ek/T8HGkWdUwa7z2zCaehYq9FrxVKhG/snXXXrR56BcTcM7Dp6FapqvCZiRNEaqNPU2ANQFO2mByWcV2x7YVv5KVlUUBr+it3wspACI9DACgR3nXJF0mGtd3RxNgCyrinB8upTyTMRYAUAtAWXKjIFhmFE0rKRgMNpRSrvX5fJ1zcnKmR8pwzunB6yuFECWOOdPvNBj8fn+WlHIRgAMZYwOllL38fn+b0u+EUf7IQoVzHvd7jaUJ8KJFi9CqlToPnNOklirMy/7ahps+WoDVm3Yh089w9+nZ6HPcEUk98lwWAU4VXivjIxV5UqnjVOCpqA2NtyIJlf07LVSEEPRj+w4dOsyzV0vlKZUIAd6xOw9HP/izKYwlD56KGlUzKo9gYiCxu0j2unBUxK0iZrJTjdu5I9Cc8yEAaLe3jpTyY7/f/0FOTs5cznm+YRhtcnNz6Xk8nRKUgCbAFQiwmJA+ACD6U3VEbrK4eNxRoCME2DCME3Jzc+nJCzMVhz3vI4TItqBbH+ecFmRThBC3lM4fWahYqEdnSZMEflm5C2/P24Y8A6hb3Yfbjq+NQJ0qaeqNblZLQEsghgQ0AY76sNq8efO4P6zu3puP5vePM0X8+/2n4KADkhPUz21WrMlBfOTAbfqLpz9a1+roOhHiH2sHmHNeAOBJIcR90YGuNAGOZyRWnFcT4BgyKn5i4lPG2J/05pbP5ztPSnkyvakIoBmAiwCcCOBZAD8IIYouN1lIdo5Al1Ut5/xNAI2EEKeXR4D1DrAFhZTKkuzdsr35hXhg7DJ8OmeN2XKPYD0Mv6AValdPD/lNNt74NZD8Eqph1njt2ZTeAS4pt0R2gPPyC8Dv/cmscP69/8HBNdIz39mzBPulNClShxRpXauj6yQSYCK+V9Kz7AA+pGC5QogcTYDtz8FlldQEOIY8A4HARMbYCT6fL5iTk7OKc05PEFHktX27wYFA4HbG2KM+n6979FFmK2qiIFhSyrnhcJjefjQT55yONowpKwhWGXUyzvksxtiiUCh0dXkE2M6X+ui6VJzQk4l53ZbduGH0PPz+5xbzPcxbTwnihm7N4EtjRNRk4rUyFtKRRzXMGq89K7NyV8tezd4slQgBLigoQNY9RQR4zj0no27Nqt4UQpy9Vm3sRcSjIm4VMSdCBOMcSq7LblffVvwK5/wkxtjVUsrejDF6/7eNYRhdok+Nuk4gHuqQJsAxlMU530wfqoUQPSlbMBh8R0pJ93Pp4lLk+HOEtK4SQvSKR/eRZ5AYY/0Nw6Bj0Nczxq7z+/0tly1btppzPpIuvkfIMOf8fp/P91thYWGY7gD7fL5bpJR9AHQWQszSBDge6cfOa3dSq6gHv63YhJtGz8OmnXmoXT0TL17SDl15vYqKJf33ZOFNescTaEA1zBqvPWOxslCxV7M3SyVCgMkGA8N+hAFg1t0nof6B1bwphDh7rdrY0wRYB8GKc4h4NrvdsR2PX2nSpEntKlWqXM4Yu4pIMIDpUsrPw+Hwi54VnAs6rglwbAK8G8BXQohLKVsgEHiVMdYvLy/vkFWrVm2JFOWc0xGFU4QQdePVKeecdn/vANCAMbbYMIzB4XB4SjHhniSlJGLdt7j954qPYR8GYCtjbH5hYeED5X0NSmShEo3D7gCPVxZuyu80Znri6J3iJ44KDYkWDQ7E633ao3Gd6q6A7TReV4CqoBOqYdZ47VllPAsVey14q1QifoVsMHjPjyiQwIyhPdHgoAO8Bd5mb1Ube5oAawJsc6h4rpjdsW3Xr2RlZbXz+XwUIOsSO5zDcwJOYoc1AY5NgJcDWB3ZAS4OUPUQY6xrKBSaFikaDAaJqLa1+j5vEvVZoupEFiqaADv3DNLuvELc9eVCjFmwzhTruW0b4vHzWuOAKu55AsTuJJ4qW05GO6ph1njtWZHdhYq91txfKhG/QjbY/N4fzaB/v97ZA40OdscHwGRLXbWxpwmwJsDJHlNuqd/u2E7Ur2RlZVXNzc3d6xY5eLEfmgDH0FogEBjDGOsohGhI2TjnFPCKdmfH79ix45x169btCgQC5zHGPmOMzQyFQie4yQgSWahoAuwMAf7z313oN2oulv61DX4fw7BezXFV5yYpfeLIik3ancSt1O3WPKph1njtWWKiCxV7rbq3VCJ+hWyw5f0/YU+BxJTbe+CIQzQBdq+mE++ZanMOSUxFzBp3/B88tF9JfH5JtAZNgGNIkHNOD1C/wRg7ORQKTaCsxYGxugHIB0DHoM1jz1LKs8Ph8HeJKsTJ8oksVDQBTpwAT8/diBs/nIctu/JRt2YVvHzpMTiu6SFOqtixulR02qph1njtDRe9UCkpt0T8Ctlgqwd+wq58iQm3dkPTejXtKcVjpVQbexH1qIhbRcyaAGsC7LEp2eyuJsAxtNakSZNqGRkZzXw+33ohxEbK2rRp04MyMjKeoZOsAGoDoIBUjwkhRrnNABJZqGgCnBgBHjVjFR74dinovm+bRgdhRJ/2cPN9NxWdtmqYNV57M7QmwM4S4DYP/oQdeRK/DOmKrPq17CnFY6VUG3uaAMdPiDxm0vt1V9t4fM8/ab+SfovXBDj9OkhaDzQBti9au5N5fqGBB79dgg9++8NsnO77PnF+a1TLdM9937KkYhevfQmnv6RqmDVeezanFyrOEuB2D43Dtr0GfhrUFcHDNAG2Z5XeKKXanKN3QjXxtzoytV+xKqnk5dMEOHmyTXvNmgDbV4Edx715Zx5uHD0PM1ZsAmPA7acWve/L6H9cnuzgdTmkCrunGmaNt0KTKDODXqg4S4DbPzwOW/YY+P6WLmjR8EB7SvFYKdXGXkQ9KuJWEbMm/vETf6t+JTs7u4lhGPQSTLPCwsJbly9fviErK+sUxtif4XB4mcemQld11/0rc1eJy1ud0QTYvr7idWLh9dtx7cg5WL1pF2pU8eOFS9rh5BaH2u9AikvGizfF3UtKc6ph1njtmZHVhYq92r1XKhG/QjbY8ZFx+He3gbE3n4ijDz/IewKw0WPVxp4mwPETIhtm5aoi2sadPwLNOe/CGPvRMIxZjLETpJTNw+HwCs75XVLK9uFw+EJXGYHHOqMJcJTCOOfj7OqPMSZDodCpdssno1wiC5Xo/qg4scWDeWJoA27+cD527C1Ao4MPwNtXdvTc0b548CbDVtNRp2qYNV57VqYJsLM7wJ0eHYeNuwyMuakz2jSmMBqVP6k29jQB1gS48o/qIoR2x7YVvxIMBqcbhvFVOBx+mnO+XUrZppgAHwvgCyFEY1XknAycmgCXJMBGAkKWQghXXfTUBNi+Nq1OaqN+W437xyyGIYFjj6qD1y47BofUrGq/4TSVtIo3Td1LSrOqYdZ47ZmRlYWKvZq9WSoRv0I2eMJjP2P9zkJ8ccMJaH/kwd4UQpy9Vm3saQKsCXCcQ8Sz2e2ObSt+hXO+gzHWKhQKrYwmwMXHonOEENU8KzgXdFwT4CglZGVlNUtEJ7m5ucsTKe902UQWKtF9sTvAncaTyvoqwmwYEk/+lIPXJ68wu3VB+0Z4rHcrVMnwpbKbjrVVEV7HGnJRRaph1njtGZ+VhYq9mr1ZKhG/QjZ44uM/468dhfis//Ho2KSON4UQZ69VG3uaAGsCHOcQ8Wx2u2Pbil/hnK8xDOPC3NzcGaV2gOkVmmeEEAlxFs8K3aGOawLskCDdWE0iCxVNgMt/BmlPfiFu++x3jF34lymmIf/huLlnlieCXZVnp3YncTfavdU+qYZZ47VqGSXzWVmo2KvZm6US8Stkg12f+Blrtxfi4+uPc+276E5rRrWxpwmwJsBOjyG31md3bFvxK5zz4QA6GoZxgc/nWy6lbAugHmNsNIDRQoj73SoXL/RLE2AvaMlmHxNZqGgCXDYBpkjP14+ag9mrNiPTz/Dk+a1x3jGNbGrIPcXsTuLuQRB/T1TDrPHGbyNUwspCxV7N3iyViF8hG+z+5C/4c1sBPry2E07IqutNIcTZa9XGnibAmgDHOUQ8m93u2LbiV1q0aFGloKBgFIDzAdDxwjwAmQA+bdiwYZ9JkyYVeFZwLui4JsAWlNC+ffvMbdu2nQ3gRMZYQyoipVzn8/mm+f3+b5YuXUpG6bqUyEJFE+D9CfDqTTtx1buzsWLjTtSqloHXL29faRZwdidx1xl9HB1SDbPGG4dxRGW1slCxV7M3SyXiV8gGez71C1ZvLcCoa45Fl0A9bwohzl6rNvY0AdYEOM4h4tnsdsd2PH4lOzubG4ZxjJTSxxibJ4TI8azAXNRxTYArUEZ2dnZ3KeX7Ukra5istL8kYWwPgqlAoNMFFejW7kshCRRPgkgR4wZ9bcM17s7FpZx4Or30A3r2qI/ihtdymctv9sTuJ227QBQVVw6zx2jO6eBYq9lrwVqlE/ArZ4MlP/4KVWwrw3lUd0T1Y31vgbfZWtbGnCbAmwDaHiueK2R3b2q+kX9WaAMfQAee8I4ApACis71wAH0kpVxERZowdAeC/ADoA2AugqxBidvpV+v8eJLJQUZYAFxYC86fC2LAW4S07kHXBVZi2aiv6fzAXu/IKcfThB+KdKzui/oGVK/ie3UncTfYeb19Uw6zxxmshRfn1QqWk3BLxK2SDpw4fj9zN+Xinbwf0zPbOW+n2rKeolGpjTxNgTYATGS9eKmt3bFv0Kz7OeR8AJ0kp6zPGSkRZFUKc4iVZua2vmgDH0EggEPiRMfYfALcIIV4pKyvn/EYALwP4SQhxupsUnMhCRUkCPOFLYPhAYANt6helbwPnYEiTa5EvGboE6mLE5e1Ro2qGm9TsSF/sTuKONJ6mSlTDrPHaMzSLCxV7lXuwVCJ+hWzw9GfGQ/ybjzf6tMcpLQ/zoATi77JqY08TYE2A4x8l3ixhd2xb8SvBYPAFKeV1AH4AQFFXZbSUhBA3e1Nq7ui1JsAx9MA53wogJISgR6fLTZzzWQCCQoiD3KHWol4kslBRjgAT+b3jghLzy8hGZ+D+YD9I5sOZhxXimQFnoGpGxU89b968GSeddJIpwsmTJ6NWLfcflbY7iTtp79SHU045BX/99Re+/fZbNGuW3Aj/bsDspPwqqkvjrUhCZf9uZaFir2ZvlkrEr5ANnvHseORsyseIy4/BaUc38KYQ4uy1amNPE2BNgOMcIp7NbndsW/ErnPN/pJRXhcPhsZ4VkIs7rglwbGK7hTH2XSgUuiyWDoPB4Ggp5RlCiNpu0nUiCxWlCDAdez6ryb6dX/rE9lzTS/Fi00tNMVzx51jcv+V7+L9ZAfgrJsDDhw/Hm2++iSuuuALDhg0z69i7dy/uvfdeLFy4EKtWraIgamjTpg0+/fTT/Uzm119/xcsvv4xly5bB7/ejXbt2GDRoEFq1arUvb35+Pl577TWMGTMG69evR506dXDqqadi4MCBqFmzZtxm+Pfff5v9W7x4MbZv344jjjgCl1xyiYnBapo0aRL69eu3LzthrVqVbg8AwWCw3Gp69+6NJ554wvz9/fffx2OPPYYzzjgDzz77rNWmbeWz67hsNeaCQhqvPSVYWajYq9mbpRLxK2SDZz03Hks35uOVS4/BGa01AfamFVjrtWpzDklFRcwad/wfPKz4Fc75Xz6fr1tOTo6wNuJ0rngkoAlwDGlxzn8GUF8I0SaWUDnnvwPYIISg49KuSYksVKJBVPoJfc4koH8PE7IBZu76jmp8pvn/g5d/gFtWflwU/WzERKBD95j6JWLapUsX0C7w559/vo+0btu2DR07dkT9+vWxe/duk2SWRYBpx/iGG24wneihhx6KvLw8s65q1arhk08+QXZ2ttn+rbfeirFjx8Ln8+HII4/EmjVrQG136NABo0aNMv9uNe3cuRNEQlevXm22c9hhh5kknVL//v0xePDgCqvauHEjzj77bGzatGlf3mgCfNFFF5WoY8+ePQiFQubfrr/+ehMPJSp/4oknmv2fMmUKDjnkkArbtpuh0tt1KcFovPYsxcpCxV7N3iyViF8hGzzn+QlY/E8eXvxvO5zdxnxUodIn1cZeRKEq4lYRM+lb425rblhYTVb8SjAYvIMC8AohbrFar85nXQKaAMeQVVZW1nE+n28ygEeFEA+XPn9PRYPB4L1SynsNw+iWm5s7w7rok58zkYVKdO8q/cT240fAPZeiED7c2eJmfN7wP2DSwEM5r6HPWrp6UZwe+RA4jeKelZ+IwBKhq1evHmgnN3ohQOSOCHCfPn0wa9asMgnwWWedBW4zANEAACAASURBVCEE2rZti9GjR4OIIhHLtWvXokePHhgxYoS5S3v++fQsHMxd28svvxwTJkwwiTOlF1980dwNtpreeecdPPnkkxTZDV9++SVatGhh7si+++67yMjIAO3sEp5YiTBPnz4dXbt2xfjx482s0QS4dNm3334bTz31lFn/uHHjcPjhh+/LcvHFF2PBggW4//77cemlRbvwyUiV3q5LCU3jtWdFVhYq9mr2ZqlE/ArZYO8XJmDhhjw8f3FbnNvu/+Pem9Kw1mvVxl6036O5nPxZPOTAmlTdmUvrWh1dkwXa1bcVvxIIBD4rjkP0D4DFUsr8aKsPh8MldxbcOSRc2ytNgGOoJhAIXOrz+brSJXQp5XLG2BdSytVUhDF2JAC6NNqUMfamYRgULbpECofDH6ZT84ksVKL7bXeApxN7XG3PmYT8G07GkJZD8O1h3eA3CvHskmdxznr69hGVLOwAP/PMM3jjjTfQs2dP84hyWak8AkxHmYlAUhoyZMi+48REcumoNO3Ozps3z6z/+eefN/NNnTrVJNWGYZiLDDpqTbutDz9M32uspauuusokrw0aNDDJKy1UqJ3//reI7NORbiLm5SXacX7kkUdw1113YceOHebxbUrlEWDaqT755JNBx67PPPNMkMyiE9VFdZb1mzVE1nJVersuJQaN15pdlM5lZaFir2ZvlkrEr5ANnv/iBCxYn4dnLmyD89vT64KVP6k29iIaVRG3iphJ3xp3fMTfil/hnI+KNTsKIShCtE42JaAJcAzBcc6N4l3faDlForCV9bdIbfSbFEJYPw9hU4GxiiWyUImut7JPbHv35uOWQU/ip4PaIdPIx4uLn8bpG6ZHiYABhzYCvllZ4R3gW265BT/99JO5y3vPPfeUqZ7yCDB9KafdT0pEAi+88ELzv4nsRsg07Sq/9NJL5nFoSkuXLt33Zb1bt24mqaQjxLTDajWddtppWLlypXlP96uvvjLr++OPP/Cf/xSd6KfjybTDW1YKh8PmbjQd737rrbdM8lsRAf76669x5513mtVRe7TjHJ1o55l2oI8++mh88cUXVmHEna+y23VpgWi8cZuIWcDKQsVezd4slYhfIRu86OWJmPvXXjx1fmtc1LGxN4UQZ69VG3sR8aiIW0XMpG+N23kCHOc0o7PHKQFNgGMILBgMPlLWsWerMg6FQvdazZuMfIksVKL7U5kntj35hbjhg7mYGPoHVYx8vLbwcZy0kYJ6FyUJVnT/96nPgZ7nVaimyG5qrLuz5RHg+fPnm4GnKEUT4Oeee848+kxp2rRp5hHnsggw7R7TLjLdQSYyajXRcWm68xtNgOk+MEVkphSLANPx7A0bNphRm+mYNJHzigjwOeecg5ycHBx//PF477339uvmZ599Zn48aNKkifkxIVmpMtt1WTLTeO1ZkibAJeWWiF8hG7zklYmYvW4vHj+vFf577BH2lOKxUqqNvYh6VMStImbSt8adHAKclZXVwufzccaYLCwsDOfm5i712PTn2u5qAuxa1STesUQWKioQ4F15Bbhu5BxMy92Eapk+vNl6D7q8e0uJd4DloY3Abn3BEvklmd18883mndboCNClNWnlCDQFniISTSnWEWgKFEXBspw4Ak3Br+gucTxHoIk00z3eKlWqmH2l4830D6Xq1avjtttuw2WX/T+IOhH4q6++2vydSDqR9dIpsgPcsmVL805yspJqDlvjtWdJmgA7S4AvfXUiZq7di0fOPRqXH0c3iSp/Um3saQKs1r1nTYDj13dFfiU7O7u9YRh0lI+e/4hwNTqButAwjGtyc3PnVf6ZM7kINQFOrnzTWrsmwOWLn8hv33dnY9bKf1Gjih/v9O2ITk0Poc+YwPypMDasRXjLDmRdcBX8xeTOijIjTyDRHddXXnmlzCJWgmC1bt0aH330UYVBsOiZJSLbv/zyC2666SazPdqFpd3bmTNn7nvGaOTIkejUqVOZ/YkOgkWRq+noMT1FRE8SEbmlwF5169bFBx98YP5D6ccffzT/Het5I/p96NCh6Nu37752ifwSCeacm7vGZaXIHeBevXqBdr+TlVRblGq89iypooWKvVq9WyoRv0I2ePlrEzFjzV48dE5LXHF8E+8KIo6eqzb2NAGOnxDFYU6uzKpt3LkdYM45PfdBRxFz6VVOxtgyKSXxteb0OAnFHgLQSQiR40pj8EinNAGOoahgMHhOKBQaU5Euu3fvnrFu3bqHhRBDK8qbyt8TWahE97OyTWy78wpx1Xuz8NuKf1GrWgbev/pYHHPEwSVUYxfzxIkTzZ1bCihF0ZOjU+ROLR1TpmBVtGtKu66UiFjSTi6RTSpPO7qJPoNklQBbfQYp+nhz5Amj0vYc6wg0HXum48+UKOr0ueeeW+ZwiESBvu+++0rsHjs9duzq2Ol+pKo+jdeepDUBLim3RPwK2eAVIyZh2p97cP9ZLXBV56PsKcVjpVQbe5oAawLssSFqu7t2x3Ysv8I5/0RKWS0cDtMiKRJ3KNJHFggEvmaM7RFCFAWN0cmWBDQBjiG24iBYb+7YsWPwunXrdpWVNRgMBg3DGM0Ya5fuoFel+5fIQqWyEmC683vt+3Pwa+5G1KyagVHXHIt2pcgvYbc7qdG7vXSsd8uWLaBgT82b0we7ohRrt5SiLzdqVBQRlSI70+5xJMBVu3btMGjQINCucCTRMWMKjEVt0B3cgw8+2Hz6iPLVrFnTzPbzzz9jwIABZvRouksbIdtl2TEFz6Kj1osWLTIjOTdu3Ni8j3zllVfuy54oAb7jjjswZswYk9gT3szMzP26Qk9FkfzoSabIzrOtmc1CIbs6tlC1K7NovPbUoglwSbkl4lfIBvu+PglT/9iDe85ojmu70EZG5U+qjb2IRlXErSLmRNZMXh/9dvVdAQHeAOAMIcTssuTDOT8WwFghRH2vyy+d/dcEOIb0Oed0vIADEIyxy0Oh0Jzo7MFg8AYp5dN03ZEx9mUoFKJnkVyTElmoRIOwO8BdI4jijhD57TdqLiaLf1C9ih8jrz4WHZrUKbObiWB++umnzfutFBCLngZKV3r88cfNIFOxgli5baFCR7UfffRRJPv4s4oOOxGbTpcNJ9KuU3g1AXaWAF/zxiRMWr0HQ0/PRr9uzRJRsWfKOmWLngFc3FEVcauIWUV/mujaqQICvKewsDCwfPnyP8sa882aNWvs9/vDQohqXpsT3NRfTYBjaKNRo0YH1KhR4zkpJb0Bky+lfCgcDj+WlZVV1+fzvQOgF4CdUsqB4XD4XTcplvqiCfD/NZJXYKD/B3MxIWcDDsj0472riu/8lqO0RJzY5s2bcdJJJ5m7mHQMulatWmkxjfPOO88MSEVPDdFd3lgpEbxOgaM+0N3ldevWYezYsWjWLLmLYzdgdkp2VurReK1Iaf88mgA7S4Cve3MyJqzajTtOC+LG7ln2lOKxUqqNvUTJgcfUW6K7Wtfx3YX1sq4TIf6x/EogEKBNtzuFEF+VJZ9AIHA+gMfD4TBt0OlkUwKaAFsQXFZW1lk+n4+isR0CgI4kUOQOOnowo3hneKWFalKeRRPgIpEXFBoY8OF8/Ljkb1TN8OHdvh1xQlZd1xPCVBqMik5bNcwar70RpQmwswS431uT8cvK3bjtFI4BPQP2lOKxUqqNPU2A9R1gjw1R2921O7Zj+ZVgMPiwlPJyKWWvcDi8LLpzxc8ijWWMjU73U6u2heaSgpoAW1REMBg8Sko5H4C5nccY+zgUCl2eyDvBFpu2nU0TYEBKibu+WIRP5vyJKn4f3rqyA7ryehXK1O6kVmHFLs2gGt5Evty6VIUVdks1HTuFVxNgZwnwDW9PxrgVuzH4ZI6BJ2sCXOHA9XAGp8agl0SgImYV/WmiH3li+RU6fVq9evWJANrToxsUBZrak1K2AHAqgHm7du3qvmbNmt1eGhtu66smwBY0kpWV1c7n840GQKHJ/wZAoXsLGGMPh0KhRwEYFqpJeRZNgIEnfsjBiMnL4WPAq5e1x2lHF0Vdriip5sRUw6uiw1ZNx07h1QTYWQJ80ztT8OPyXbjlpACG/EeNE3xO2WJFfsttv6uIW0XMKvrTZBJgqjsrK6sqY+w2xth/AUS+FIallB/n5+cPX7Vq1R63jXev9UcT4Ao0xjmnKEYPAKjCGHtj586dg6tXr94dAN35pa3E36SUfcLh8Aq3KV91Avz65OV4/IeiZ9KeOr81LurY2LKKVHNiquFV0WGrpmOn8GoC7CwBvvndKfg+dxcG9MjCbacGLc/JXs7olC16TQYq4lYRs4r+NNkE2Gtj3Yv91QQ4htaCweAkKWUXAJsMw7gmNzf320h2znldxtg7UsozAexgjA0KhUIUGMs1SWUC/MnsP3DnF4tMXdiJNqqaE1MNr4oOWzUdO4VXE2BnCfDA96ZgbHgXbujeDHeeRoeqKn9yyha9JikVcauIWUV/qgmw12aj/furCXAMHRa/A/xjfn7+VStXrlxfVtZAINDf5/M9Q49W63eA3TEgxi3524z4bEigX7emGHr6/9/itdpD1ZyYanhVdNiq6dgpvJoAO0uAB78/Bd+IXbi+a1Pc3Sv+udnqHO6mfE7ZopswWemLirhVxKyiP9UE2MoM4O48mgDH0E8gELg5HA6/VJEKg8FgUEr5gRCiY0V5U/m7ijvAC/7cgkvemIE9+QYu6tAIT57f2nyOKN6kmhNTDa+KDls1HTuFVxNgZwnwrSOn4uvQTlxz4lG490yK6VL5k1O26DVJqYhbRcwq+lNNgL02G+kd4KRprHv37hmTJk0qSFoDNipWjQD/+e8u9H51GjbuyEP3YD28dUUHZPh9NiQHqObEVMOrosNWTcdO4dUE2FkCfPuoqfgyZyf6ntAED5zd0tb87LVCTtmixu1+CWhd63eArVip9itWpJTcPPFvjSW3P7p2ByWgEgHeuisf54+YjtwNO9C8wYH4rP/xqFk1w7Y0VXNiquHVBNhve2x4paBTNq0XKs4S4Ds/mIrPl+3EFccfiYfOOdor5pRQP52yxYQ6kYbCKuJWEbOK/lTvAKdhQnG4SU2AowTKOR8H4CchxDOl5cw5v5He4gqFQvQ2V4nEOX8RwMVCiEMd1k9C1alCgPMKDFzxzkz8tuJfHHZgNXx9U2ccdlC1hGSnmhNTDa+KDls1HTuFN5UEmPwMgNsBNGCMLZFSDhJCTC1vMgsEAufTc3wAmgFYDmCYEOKrSP5AIHCez+frJ6Wk9yQPkVK2C4fDC6LrKw722K1UG58IIS4pq91E/Arp5O7Rv+KTpTtwWacj8GjvVgnN014p7JQtegVvouTAazij+6t1rXeArdivRb/i45z3AXCSlLI+Y6zEkUYhxClW2tJ5ypaAJsAlCbDBGHsvFApdXQbJjfUbPYl0hQ6ClfphJqXE7Z8vxOdz16BGFT8+638CWjQ8MOGOqObEVMOrCbDeAbY6SVhcqFitrtx8nPOLAYwCcKOUchqAfoyxawsKClqsWLHij9IFs7Kyjvf5fESO7wVApLc3gIeklCeGw+GZlL948XQUgHUA3oxBgEVGRsZ9kTb27Nmze8WKFVuTQYDv+fBXfLRkB/57bGM8fl7rhOXmhQpUnF9VnGNVxaxxL0DbtvERfyt+JRgMviClvA7ADwD+AiCj5zshxM1emP/c2kdNgDUBrtA2XeW8CwuB+VOBjX8BdRvgnd2N8dB3OfD7GN6+sgO6B+tXiMdKBldhttLhBPOohldFh62ajp3Ca2WhkuDwM4tzzmcyxuaFQqEbIvVxzpcB+FoIMbSMj7KfADhQCHF65LdAIPAjY2yzEOK/0fmzs7ObGIaxsjwCbBjGgnA4PMgKjkR3gO/7+FeMXrQDF7ZvhKcvbGOlSc/nccoWvSYIFXGriFlFfxoZi3b1bcWvcM7/kVJeFQ6Hx3pt7Huhv5oAawJcoZ3aHeAVVhxvhglfAsMHAhvWmCWnHdwGVxzzEAqZ34wmSlFFnUquwewUoArqUQ2vig5bNR07hdfKQiXRYdqiRYsqBQUFuwBcGH2EuXgHoK0QovQRZSLMtCv8nBDiuSjCPLj4Tfoj4yHAUkqKRkXrAXru7wfG2IOhUGh7WbgSJcAPfDINoxZux3nHHI5nL2qbqOg8Ud4pW/QE2KhOqohbRcwq+tMUEeC/fD5ft5ycHOG1se+F/moCrAlwhXbqigmdyO8dF+w7AfJntUNx9rHPYnOVg3DeXxPwzBWdwU46r0IsVjO4ArPVzjqQTzW8Kjps1XTsFN5UEOBgMNhQSrnW5/N1zsnJmR5FaO8GcKUQIlh6mHPO86SUfcPh8IeR3wKBwKWMsXeFEFWtEmDO+XVSypUZGRl/FxYWUlSqxwHkCiH+E4sAc85RvXr1uGYf0smjX/yG93/fjnPaNMCzF6mzA7xo0SK0atUKfn/lv34QTQ5Uw002rhrmiD/VuK2PbfIrQpi8tn2HDh3mlTWRBoPBO6SUjYQQt8Q10erMliSgCbAmwBUailMLyQobKi8DHXs+q8m+nd9dvqo4r+Nw5NQ6Cq23Cnw6dyiq1asPfLMSsLC42Lx5M0466SSztcmTJ6NWrVr7tZx2zLaFZa/g888/j9deew3XX389br31VnuVeKyUajrWeO0ZaCoJsGEYJ+Tm5s6IIrTDGGN9hBDZ5RDgK8Ph8EeR34LB4GVSyreFECWiAMY6Al263uzs7PaGYcwxDKN9bm7ufguzyA6wPWkCY8M78e6C7ejcuBqGHFfbbjW6nJaAloCWQGWQQAkCHAgEPo0GxRijQFf/AFgspcyP/i0cDl9UGQSQLgyaAGsCXKHtpX3hPGcS0L+H2U+KAHDz0Xdg7GFdUXfvZnwzazAa7t1YhGHERKBD9wrxDB8+HG+++SauuOIKDBs2rET+zz77DB9//DGWL18OwzDQuHFjXH311Tj//PP35fv222/xzjvvmHmqVauG4447DkOGDEGTJk0qbLt0hvz8fJN4jhkzBuvXr0edOnVw6qmnYuDAgahZs6al+v7++2+cffbZ2Lq1KGYNYevatet+ZSdNmoR+/frt+/vChQtRtWrRRtGGDRvQo0cPVKlSBePHjzf7UdlT2u06xQLWeO0JPBUEOJ1HoMuQCuOc7wVAxJvuGZdIEQJsdwf4ya9+w9vzt6PX0Yfhpf+qcwRa745Z3x2zN1LdUUrvAOtTDlYssbwdYM45BUK0lIQQFCFaJ5sS0AQ4SnCccwPAp4yxIaXlKaWki6fl/UZ3sC7QUaBtWmFFxX78CLjnUjPX+43OxP3Z/ZFhFODDecNw7JYl/y/9yIfAaSViv+xXMxHOLl26gHaBP//8c/NIWiQ9/PDD+OCDD8z/bdCggUkGd+zYYRLDRx991Pz7J598gvvuKwqW2qhRI2zZssXMQ4Tx66+/xqGHxvcSFu22jh07Fj6fD0ceeSTWrFkD6mOHDh0watQo8++xEpH0vn37YuZMM+irmcoiwBs3bjRJ8qZNm/bliybA5LQvu+wyzJ8/H0OHDjXrrOxJE8LKrWGn9JsKAkyaoCBYUsq54XCYnkIyE+d8KYAxMYJg1RJC9IrKT9FCt8QTBKu0FTRv3vzowsLCRVLKbuFweErp3xO9A/zEF9Px5rxtOK3lYRjRh15nqvzJKVv0mqRUxK0iZrJLjdv5KNBeG+9e668mwFEaKybAJcKMx6NQTYDjkVYceYt3gBfXamYefc7zZeLe0Bu45s9vSlZiYQeYjjzTMd969erh119/3VeeiN8ll1xiEs4XX3wRPXv2xIIFRaHtd+/ebe7G5uXlmeSZSC/t0lI+2rU9/fTTsXPnTlx++eW49156kcRaWrx48b6dZSpH5SdMmIAbbigKAkv1Uzux0htvvIFnnnnG7MMPP9Dat2wCTJinT59u7gzTDi+l0gSY6nn77bfRunVr0E54ZU+qOWyN155Fp5AAm88gMcb6G4ZBx6CvZ4xd5/f7Wy5btmw153wkgLURMpydnX2CYRhTGGPDpJRjGGPnSCkfiX4GqWXLlnXy8vKOYIw1BPAdgEuklKHMzMy/ly5d+ndWVlYzn893GWPse8bYRsMwWgB4BsBuIURHWtc6TYCf+nI6Xp+7DSc3PxRvXdnBnlI8Vkq1sRdRj4q4VcSsCXBynkHinI8rKCi4sPSTdMFgsJaU8gv9DnBijkAT4JIEmHZ5EyHAjRNTh7OlE/lSH92TtE/ohYXYfk42zmx6B1ZXb4j/bJiBNxY+aoYsLUoMOLSRpTvARPKINBLBpaPHkfTYY4/h/fffN3d+s7OzMXv2bPN4MBFQ2qUlAjx37lxcemnRTjTVc+aZZ5r/TUekp02bZh6B/umnnywrkdqnu7eUpk6divr165vHrol07927FxdddBFoV7q8tGTJElx88cU48cQTcdVVV5lHuimV3gGmneRHHnkEd911l7lb/fLLL5v5ShNg2sG+++67kZGRYeKPN8CNZeAuyZh2u06xHDReewJPFQGm3nHOaff3DjqEwhhbbBjG4MgubDAYnCSlXCWE2Hc8IxgMXkCkF0BTAMullMPC4fCXEaSBQKAvBcUqA/mDQogHmjVr1tjv99OxFwp+RXcu/mSMfZeRkfHgkiVL/i1LYon4FbLBZ76ajlfnbEPP7Pp4py9x7MqfVBt7EY2qiFtFzKRvjdv5HWDalCssLDxs+fLlG6JnyaysrHo+n2+dECKz8s+eyUOoCXDyZJv2mhNZqER3Pt0Tm5QSA57/Ft+t9+Pw3Rvw3cxbULtgR3EXi034qc+BnhVHgb7llltMktqnTx/cc889+2Bed911mDKl6LQf3eul4810x5fapiPQI0aMwHfffWfe9aX03nvv4fjjjzf/+/bbb8c333xjHpmme15WEx2lpiPVlJYuXbovOmi3bt1A93qJ2NKObFmJdqV79+5tElpqOxwOl0mA6e90f7ljx4546623TPJbHgGmHXHaKaZEx7IDgYBVKJ7Ml267TrXQNF57Ek8lAbbXw9SWSsSvkA0+9/UMvDx7K7rxenj/6mNT2/k0taba2NMEOP4dwTSZpmPNaht3jgBnZWXRSRw6kbgYQFfDMPZ9jMzIyPAXFhae5vP5BoRCoRLP3TmmTEUq0gS4Eis6kYWKmwjw6JmrMeyrxchgEp+Gn8Ixq6b+v3uHNgZufd4S+aVCtFNKR4H79++PwYMH76snsotLfxg9ejTatWtn3vWle8KU6NgwHYmOREiOJsC33XYbKDAW7RjTrqrVVB4BpmPKdLSajlsTaS0r0Y4u3Vcmgty5c2fzDnBZO8B075cCXFH/6Nj3Sy+9VC4Bph1u+jBA6aOPPsIxxxxjFYon86nmsDVee2aqCXBJuSXiV8gGXxgzAy/O2oougboYdU0ne0rxWCnVxp4mwJoAe2yI2u6u3bEdy6+UcR0zwtXohCr9NwUpvEUI8abtjuuCUadItTAqnQQSWai4hQCv+GcHer04FXvyDdzdKxvXd24CzJ8KbPwLqNsAaNfF0tNHETw333wzxo0bt18EaAr89OWXRScHaReX3mqkI9FPPvmk+beRI0eaR4MjR6ApkvRZZ51l/ubEEWjafaYAWlaPQBNRpWPKBxxwgNkHmoTp2DQlIuInn3wynn32WQSDQbPftDtNiQJs0T+U6IgzkXcKfkXlo3eAiTBTlNfKnOw6Lq/KROO1pzlNgJ0lwC99MwPPz9yKE5odgg+vO86eUjxWSrWxpwmwJsAeG6K2u2t3bMfyKxSbQUrJ/H6/kFIeL6UsfuoEdCoxr3bt2n/PnTu3xJNItgEoXFDvAFdi5XudABcUGjh/xAz8/ucWdM46BKOu7gSfLzGTjTyBRATxlVde2ad9OkZMR5kpffjhh+Y93AceeACffvopGGOgJ4Qo0nMkCNYpp5xi7qbSTu1pp50Gmsyij1UT8aQ0YMAAEOkuK0UHwaLnmGgH95dffsFNN91kZqf6qR3aVb7jDroWCDz11FNmkCpqa9asWeVaL71z/Oqrr5oEOFaKRHymSTxyB5jIP5HrGjVqVOLRod6dJbuO2qtG4BReTYCdJcCvfDsDz/62FZ2OqoNP+hVdI6nsySlb9JqcVMStImayS43buSPQUeOc3g7bLxCh1+YBt/Y3MTbhVlS6X6YEvE6AXxwfxrM/C9SqloGfBnVFw9pFu52JpIkTJ5rHnynYFZHaSKJdUYoCTaSUdlXpDnBubq55B5ju0FKQLErlPYN08MEHm2/5Rp5BskKAqT4rzyBFH2+mnehOnfY/NljeEejSsop1BDoSBfroo4/GF198kYiYPVFWNYet8dozS02AnSXAr439DcNnbEHHJgfjs/4n2FOKx0qpNvYi6lERt4qYNQGOf8e/PL/COe9Vq1atn2mHl/471lQnhPjeY1Ohq7qrCbCr1OFsZ7xMgBet2Yrer05DgSHx/MVtcW67wx0RTvRTRrTj2bx58331bt261YzuTE8Rbdu2zYzKTE8T0W4r7YpGEu0Wv/POO2aQLDpuTMGwiMhSFGhKVM+xxxYFd4ns4pbXeSLeFA2a+kJ3dYlIU+TpQYMGmZGnKaWKAEfeAb7zzjvNY92VPam2UNF47Vm0JsDOEuA3vvsNT07fgnZH1MZXN3a2pxSPlVJt7GkCHD8h8phJ79ddbePO7ABHR34uvgtcnmlItz296jUb1gTYaxqLo79eJcB5BQbOfGkqxPodOKNVA7x8aTvzGLJT6emnnzaDS1FALHoaqKyUyGROAbNuvPHGmEGsnMLiVD3//POPGe2a7gvTBwA67l3ZUyI69qJsNF57WtME2FkC/Nb3v+HxaVvQptFBGDPgRHtK8Vgp1caeJsCaAHtsiNrurt2xHcOvRB97/v/OS9k91MejbWuuKJqYTpVUAl4lwJGjz4fUqIKfh3RDnRpFAZycSps3bwbdkY3c7a1Vq9Z+Vdud1KiiJ554woyiTE8JNW7sqqehyxUhvUdMO9H0HBQFxlIhzEU5rwAAIABJREFUJaJjL8pH47WnNU2AnSXA7/wwE4/+uhlHH34gxt7cxZ5SPFZKtbGnCbAmwB4bora7a3dsW/ErWVlZVXNzc4uim+rkuAQ0AXZcpO6p0IsEOHfDdvR64VfkFRp44ZK2OKetM0ef49WK3Ukt3nbckl81vCR31TBrvPZGm5WFir2avVkqEb9CNvjejzPx8NTNaN7gQPwwUBNgb1qBtV6rNueo6FdU/tiRiL6t+BXO+R4As6SUkxhjk3ft2jV9zZo1u62NPp2rIgloAlyRhIp/b9q06UGZmZn0MGpdwzD+CIfDMy0WTVu2RBYq0Z1OlRMzDImLXp+BOas3o0ewHt7p29HRo8/xKCJVmOPpUzLzqoY3EceVTD0ks27VdOwUXisLlWTqzW11J+JXSCcjf5qJB6dsRvDQWvhpcFe3wUtKf5yyxaR0LomVqohbRcwq+tNEib8Vv8I57yKl7MYY684Yo+eQ6Ej0HMbYJMMwJofD4Z+TOHwrfdWaAFeg4qysrHp+v/95KeWFAMzz+Iyx90OhkBklKBgM3iClfJAx1jsUCk1zk8UkslBJBwH+4LfVuOfrxahexW8efT7cgajPdvWhmhNTDa+KDls1HTuF18pCxe4848VyifgV0skH42bi/smbkVW/Jn4Z0s2LIoi7z07ZYtwNp7mAirhVxKyiP00FAY4evu3bt8/cvn07PQFyLYDLAPh0EKzEJjhNgGPIr2XLlnXy8/Npp7cZgEVSymmMsf6MsfciBJhzTmd0V0spXw6Hw4MSU4ezpRNZqKSaAG/csRc9h0/Ctj0FuP+sFriq81HOCiPO2lRzYqrhVdFhq6Zjp/BqAlxy8kzEr5BOPvplFu6Z+C+OqlsDE2/rHufM7M3sTtmi19CriFtFzCr601QR4GbNmmX5/X6aKOlrIf27JmNsqpRyshDiGa/NCW7qrybAMbQRCASeY4wNBPCgEOJBykphyaMJcPHfFtKVQiFEOzcpN5GFSqoJ8J2fL8Qnc/5EiwYH4tubT4Tfl17TVM2JqYZXRYetmo6dwqsJsLME+JPxs3D3hH9xRJ3qmHJHDze5zKT1xSlbTFoHk1SxirhVxKyiP00FAeacrwFQHcBkOvZcWFg4JTc3dwEAmaQhq1S16WUZLhc153wFgL1CiH2PxZZDgMcCOFYIUd9NkLxCgBf8uQXnvlJ0evzz/sejQ5P0P8GjmhNTDa+KDls1HTuFVxNgZwnwZxNm4a7x/5pXXKbd1dNNLjNpfXHKFpPWwSRVrCJuFTGr6E9TRIAXAwjQvV8AkygY1s6dO6etW7duV5KGrFLVagIcQ93FEdjGCCEujmQriwAHAoHPGGNnCSGqucl6vECAKfBV71en4fc1W3HeMYfj2YvaukKEqjkx1fCq6LBV07FTeDUBdpYAfzFxNu74ZRMaHFQNM4ae5Ir5PtmdcMoWk91Pp+tXEbeKmFX0p6kgwNRGdnb2IYWFhd18Pl83CohFoYcAzGOMTQyFQvc4PWZVqk8T4NgEeD0AIYTY91ZDOTvA8xljdUKh0JFuMh4vEOAxC9Zi4McLULNqBibc1g31a7njG4JqTkw1vCo6bNV07BReTYCdJcBfTZqN237ehPq1qmLWsJPd5DKT1henbDFpHUxSxSriVhGziv40VQQ40k7xSzREgs/TQbCcmbA0AY4hx0AgMIYxdpqUsm04HF5GWUsT4KysrON8Pt90AKOFEH2cUYsztbidAOcVGDj52cn4499duP3UIG7qkeUMcAdqUc2JqYZXRYetmo6dwqsJsLMEeMzk2RgybhPq1qyCOff8x4HZ2v1VOGWL7kdasocq4lYRs4r+NBUEOCsr6yyfz0dPIHWXUrYGsBXAr3QfWEo5UQjxu9fmBDf1VxPg2AS4Kx0zALDCMIxrc3Nzp3DOCyNBsIrJ7wcAjvT5fMfl5OTMdZNy3U6AR85YhfvGLEHdmlUx5Y7uqF4lwzXiU82JqYZXRYetmo6dwqsJsLME+NspczDop404uHom5t93imvm/GR2xClbTGYfk1G3irhVxKyiP00FAeacbwIwtfjd30nhcJgIrw6A5dBkpQlwBYLknA8G8DQ9/1v89eWg4n/nAahHxaWUg8Lh8EsO6cSxatxMgHfuLUC3pyeBnj96+JyW6HN8E8dwO1GRak5MNbwqOmzVdOwUXk2AnSXA30+dg5t/3Iha1TKw6IFTnZiuXV+HU7boeqClOqgibhUxq+hPU0GAi3mHJrxJmvg0AbYgWM75iQDuKn6Di0KSUyICPNXn8z2ak5MzyUI1Kc/iZgL88oQwho8TOPKQ6vhlSDdk+n0pl0+sBlVzYqrhVdFhq6Zjp/BqAuwsAf7x1zm46YeNqFHFjyUPneaqeT9ZnXHKFpPVv2TVqyJuFTGr6E9TRICTNTR1vcVfF7QgrEvA36JFi3oFBQX+WrVqbZg7d26+9aKpz+lWArxlVx66PDkR2/cW4IVL2uKctoenXjgVtKiaE1MNr4oOWzUdO4VXE2BnCfC4aXNww/cbUS3Th5yHT3fd3J+MDjlli8noWzLrVBG3iphV9KeaACdz5khN3XoHODVyTksrbiXAz/0s8ML4MLIPq4Xvb+kCn899ZqiaE1MNr4oOWzUdO4VXE2BnCfAv0+ei33f/oIrfB/GoJsBpWRykqFGnxmCKuutIMypiVtGfagLsyHBJayXuYx5pFUfJxrOyslr4fL5zAXxXXrS1QCDQljHWizH2RSgUCrmo+3AjAd6xtwCdn5iArbvz8cqlx+CM1g3cJLJ9fVHNiamGV0WHrZqOncKrCbCzBHjCjLm4buw/8PsYlj/Wy5Xzv9OdcsoWne5XsutTEbeKmFX0p5oAJ3v2SH79mgDHkDHn/HUAVzHGmoRCoXVlZQ0Ggw2llKsBvC6EGJB8lVlvwY0EeMTk5Xjihxw0rVcDPw/uZi6C3JhUc2Kq4VXRYaumY6fwagLsLAGe9NtcXPPtP2alKx/vBcbc6QOc9EtO2aKTfUpFXSriVhGziv5UE+BUzCDJbaPye54E5Mc5px3dLUKITrGq4ZzPBHCgEKJ5vM1xzm8EcDuABoyxJRRRWggxtaJ6OOeXAPgIwBghBO1S75fcRoD35BfixCcnmpGfh1/YBhe0b1QRzLT9rpoTUw2vig5bNR07hVcTYGcJ8NSZ89D3mw1mpbQD7NaPoE46H6ds0ck+paIuFXGriFlFf5osAhwIBD61OjbD4fBFVvPqfPtLQBPgGFYRDAZ3Sim/FUIQ2Sw3cc4/BtBLCHFgPEbGOb8YwCgAN0oppwHoxxi7tqCgoMWKFSv+KK+u5s2bH1lYWEj5VwD41ysE+P3pq3D/N0vQ6OADMPG27q6L/Bwtb9WcmGp4VXTYqunYKbyaADtLgKfNmocrxhQRYPHI6aiS4a4XAOLx4VbzOmWLVttzSz4VcauIWUV/miwCzDknTmApCSH6WMqoM5UpAU2AYxPbLQB+FUKcWQEB/hZANxsEeCZjbF4oFLohUj/nfBmAr4UQQ8tp0885n8wYe1dK2QVAbS8Q4LwCA92fnoh1W/fg4XOPRp/jjnT1kFTNiamGV0WHrZqOncKrCbCzBHjG7Hm4/OsiApzz8Gmolul3tS9wonNO2aITfUllHSriVhGziv40WQQ4leNT9bY0AY5NgGmXtWVBQcGRK1as2FpW1iZNmtSuUqUK3QEWQoiOVg2qRYsWVQoKCnYBuFAI8VWkXDAYfEFK2VYI0a2sujjnDwJoLYTozTl/zwoB5pyjevXI88VWe/j/fDShL1q0CK1atYLfb2+x8s3v6zD404WoV7MqJt/WFVVdvuhxAnP8kk5fCdXwRhx2onadPo3F37JqOnYKLxFgIQQJvH2HDh3mxS/5ylUikas1pJOZc+bj0q/Wm0JZ8uCpqFE1o3IJqAw0mhS1tb128JpxaF2ro+tEiH88H1azs7MPKSgo4ABkRkZGOCcnZ5PXxoUb+6sJcGwCTEGtXpRSjsvMzOy7dOnSv6OzN2vWrL7f738fwCmMsVtDodDzVpVcHDxrrc/n65yTkzM9Uo5zfjeAK4UQwdJ1BYPBzlLKTwAQQd5olQBb7VMy8w0dvwni33xc0rImLmxRM5lN6bq1BLQEtASSIQFNgIGEXhcgcjB77nxc8mURAV74wCk4sFpmMnTlqjo1KVKHFGldq6PrZBPgFi1a1CwoKHgJwGUAIl8KCwGMNgzjltzc3G2umug81hlNgGMorH379pnbt2+fBOB4ADsBfMsYW24YhmSMZQE4C0ANADMzMjK6LV26NM+q/iME2DCME3Jzc2dEygUCgWGMsT5CiOzouoLBYC0p5UK6LyyE+IF+s0qA070DvHDNVvR+bQaq+Bmm3tEddWtWtSqmtOVzavcobQDibFg1vBHHpXeA4zQUD2V3yqb1DnBJpSe6Azx33nxc9EURAV5w339Qu3oVD1mVva5qUqQOKdK6VkfXySbAxfGFOkgpB1apUsXkCXl5ecRHnqPrkxXFJ7I3W6lTShPgCnRd/AXmZQCXAygdrcOgLzGMsQGhUGh7PGYT7xHo4veG5wOgrz+RFOmPYRhGMDc3d3l0HxJZqETXk+iEPuTTBfhy3lr0bnc4nru4bTxiSlveRDGnreM2G1YNbyKOy6aI015MNR07hTeeo2ppV3IKOpCIXyGdzJ8/Hxd/uQGFhsSMoT3R4KADUtDr9DbhlC2mF0X8rauIW0XMKvrTyGiwq28rfqU4EO9ppV+GCQQCXX0+3w+hUIg24HSyKQFNgC0KLisrq5HP5+sBoHFxkT8Nw5iYm5u7xmIV+2Wj55OklHPD4TA9hWQmzvnS4qeNSgTBatKkSbUDDjiAdp33pcLCwkcA1PL7/QMZY6L0DnQiC5VS7WDBggVo2zb+L3v05NEJj09AXqGBr2/qjLaNa9sVV0rL2Z3UUtpJBxtTDa+KDls1HTuF18pCxcGh6PqqEvErEZ3c9NNm/L1tL8bc1BltPOITElGMU7aYSB/SUVZF3CpiVtGfpoIAc87/9Pv9py9btmxx9PjNzs5uZRjGj0KIw9MxritLm5oAp1GTkWeQGGP9DcOg4w3XM8au8/v9LZctW7aacz4SwNryIkJbPQLdvHnzhINg2SXAr0zMxdM/hcxFDi12vJJUc2Kq4VXRYaumY6fwagJcctZ2ggA/NGM3fl+zFW/0aY9TWh7mFbdgu59O2aLtDqSpoIq4VcSsoj9NBQEOBoP0Qsy5BQUFfZYvX26Gzo+KPURPtL6apqFdKZrVBDjNauSc0+7vHQAaMMYWG4YxOBwOT6FuBYPBSVLKVUKIvmV10+0EWEqJns9MxsqNOzH8wja4oH2jNEvbevOqOTHV8KrosFXTsVN4NQF2ngC/vkRi3NL1ePicluhzfBPrE7NHczpli16DryJuFTGr6E9TQYA557OJCgCgQAmritukCZPiDeVEzwdCiGO9Nj+ku7+aAFvQAOf8RABdiaQCKC+CkxRC9LNQXcqyJPKlPrqTdif0+X9sRu9Xp+OATD/m3HOyp567sIs5Zcp1uCHV8KrosFXTsVN4NQF2ngB//UcVjJr5B27q0Qy3n1oi3qPDM5s7qnPKFt2BxnovVMStImYV/WkqCHAwGHzY6mgLhUL3Ws2r8xVJQBPgGJaQlZV1oM/nozd6uxdniyUvIsD2HslNkjWmmwDfN2YxRs5YjXPbNsTzl7RLEsrkVKuaE1MNr4oOWzUdO4VXE2DnCfCMrbUwfFwY5x/TCM9c1CY5k7iLanXKFl0EyVJXVMStImYV/WkqCLClQaYz2ZaAJsAxRBcIBF6l+7kAVkopR/h8PiGl3FFeESHEeNuaSELBdBLgvAIDnR77BZt35eP9q49FN14vCQiTV6VqTkw1vCo6bNV07BReTYCdJ8CioC6GfrUYPbPr452+HZM3kbukZqds0SVwLHdDRdwqYlbRn6aCAGdnZ3fPycmhp1j3S8Fg8JpQKPS25cGoM+4nAU2AYxgF5/wv2iX3+Xwtc3JyNnnNftJJgH9euh7XjZyDerWqYsZdPZHhL/2ClLulqZoTUw2vig5bNR07hVcTYOcJ8CrUx62fLUSXQF2MuqaTu52BA71zyhYd6EpKq1ARt4qYVfSnqSDAnPO9UsrnDz/88GGTJk0qoDazs7MPMQzjLTqZKoQ4OKUDupI1pglwbAK8G8D3Qojzvaj3tBDgwkJg/lTcNHkLvvsnE9d2boJ7zmrpOfGp5sRUw6uiw1ZNx07h1QTYeQK8xncoBn7yO45rWgcfX3+85/xDvB12yhbjbTfd+VXErSJmFf1pKghwdnb2CYZhjAKwFcClPp/vSMMw3gMQ9vv9fei1mHSPcS+3rwlwbAK8EMBmIUQ3Lyo55QR4wpfA8IHYvmkT2ncdjTx/FYwVj+LoAbcBPc/zlAhVc2Kq4VXRYaumY6fwagLsPAFen9kAN344H+2PPBhf3HCCp3yDnc46ZYt22k5nGRVxq4hZRX+aCgJMbbRo0aJmQUHBGwB6A6CjlA8KIZ4AYKRzbFeGtjUBjqHFQCBA7/K+aBhG+9zc3CVeU3hKCTCR3zsuACAx5tBuGNjqdjTd+SfGz7ixKNLaU59bIsGbN2/GSSedZIp68uTJqFWrVlrE7iYn9sILL+DVV19F//79MXjw4KTIw014kwKwjEpVw6zx2rMsTYCdJ8D/VmuI60bNQ5tGB2HMAHpkoXIn1cZeouTAy9agdd0Wfr+r4sEm1Zzs6tuqX+GcU5TAjwBUK36JZtSuXbsGrlmzhk6o6pSABDQBrkB4nPNnAFzIGLs7Ly/v55UrV65PQN4pLZoyAkzHns9qAmxYY+Lr33oofqzfGQNWfozbln9QFGz80EbANyuBCibG4cOH480338QVV1yBYcOGmfXNmTMHo0ePxqJFi7Bx40ZUrVoVTZs2xTXXXIOTTz65hEx//fVXvPzyy1i2bJk5Cbdr1w6DBg1Cq1at4pJ9ZFKjNj/99FP88ccfJhnv3r07hgwZgnr1Ygf1Gjp0KObOnYt//vkH9B5y3bp1zbIDBgxA7dq19/Xlhx9+wMiRI7FixQrQhFinTh0ce+yxuOWWW9C4cWMz36ZNm9CtWzdkZmZi/PjxZh6nk91J3Ol+pLI+1TBrvPasy+pCxV7t3iuViF+J2OC2GofjqvfmonmDA/HDwC7eE0KcPVZt7GkCvABt22oiGOcw8WR2u2Pbil8JBAK3McYeYYy9u3PnziE1atQISCk/kFJWY4xdLoSY5UmhuaTTmgDHUATnnB6bppRhbm0WJTp2EPnv6NL0DFJ5bwSnRd2JLFSiO1zhAJ8zCejfwyyy018Nx3Qdjb3+qvjut5vRcsfK/1c1YiLQIfKi1P4iyc/PR5cuXUC7wJ9//vk+0vrSSy+ZpLZmzZpo2LAhVq5cCcpL6bnnnkOvXr3M/6Yd4xtuuAHU30MPPRR5eXlmXdWqVcMnn3xCwQMs64HquOuuu/DNN9+YZZo0aYK///4be/bsMf/7q6++QvXq1cutr1OnTmZ/DzroILMP69atM/OeeOKJePvtosB9v/32G/r27buPIBNJFkLAMAxkZWXhu+++21d/v379MGnSJBCxpjJOpwp17HSDLqhPNcwarz2js7JQsVezN0sl4lciNrj7wMa4/O3ZCNSviZ+HePKGUVzKU23sRYSjIm4VMZO+Ne74PnhY8Suc8/WGYVybm5v7bWRMtWjRokp+fv6TjLEbhBC0K6yTTQloAhxDcJxz2tIsi+yWWUoIUbRl55KUyEIlGkKFE9uPHwH3XGoW+a5+Z9zUeiiO2PUXJk+/ruRD0498CJz233KlQwT2+uuvN3dXaSc3kn788Udz9/WEE04AYwxLlizBBRdcYBLFnj174rXXXjOznnXWWSaBpC+vtGNMZPXss8/G2rVr0aNHD4wYMcKyZojsUt2E/eqrr8add96J/7F3HeBRFO/73Uuj9x6KkNxeCBI6Ik2agqggWGgi4t8fioiCCBYQpSiKIMUCAiqKihRRmhTpCAoiIC25vQBBmoAQpJPkbv/Pt8kdl5Byt7tX9mbmefLkbm/K9843M9+8OzPfJCUl4eGHH1YIK32n53mFmzdvKivVztC7d29lRZhI8549e5THtK2ZtjdT2Lp1KypUqIDRo0crZJ3i7d69W8FL4fvvv8dbb72FhIQELFq0yGMcnkYsUMeeZmSgeKxh5njVNU5PJirqcjZmKi12xdkGM0rVQI/ZO3BH2SLYNDzz5WkoB9b6nlOXLOJmETPpm+PWnwDHxMRUOHz48NncxkaLxdLOarVuCOVx09fYOAH2dQ0HMH8tExV3sQsc2NxWgAfdOQIrK7XGsymL8XoyOatzCwWsAE+ePBmzZs3KRmpzqz4ioE2aNMHly5fRqVMnhUSeOXMGrVu3VqLTFmVaMaXw5ptvKluYaRWYCKWnZ1N++uknheQ6ySdtpabQsWNHpKSkoEWLFvjiiy/y1S6tXG/ZskXZwkwknAKtcM+ZQx7sgR07dqBfv363rQDTFulRo0bhgQcecOV/6NAhdOvWDeHh4fjjjz/yXX1W0+QK1LGaTIM8DWuYOV51DZIT4Oz1psWuONugUPYOdJ/5O6JLFca219qpU4yBUrHW95yqYRE3i5hJ3xy3/gQ4vyGuTp06ZQ4ePHjBQMNg0InKCXDQqUQ/gbRMVNylKHBgyzoDnHbuDBre8y2uhBfBjztfRoNLUlY2np0BpnOva9asQd++fRUCmFegFVD6nVZHiUzStuK9e/eiR48eSpLx48fjscceUz5PnTrVtUJMq8oFnd11lkmrxbS9msIvv/yC6tWrK5+dK7k1a9YErUznF1555RUsX+7auaKsYBMppq3RzkB4aas1TbKdoV69ehg7dmy2LdsXL14EbaumsGLFCpjNZv0aCqPGq8B2rWsNBz4zjledDjgBzl5vWuyKsw1GVKiFLp9sR4XiUdg5MrsfB3VaCu5UrPU9pzZYxM0iZtI3x60fARZF8RKAWpIk/Ut1K4riyrCwsGcSExNP0/eaNWtWjIiIOCVJEjvexnwwxHMC7INKDZYstUxU3DF4NLBtWILtE8aid6N3UO5mKnZufRImZfd4VhPzwAt0//79sX379ny9HdPZYNoKnJGRoRBHSkOBthX37NlT+exOgInEOrc+b9u2TXFG5UmgbdVEnim4E+BevXopK8nkhIscWBUU6KxycnIyRowYoWzP7tq1KyZOnKgks1qtivyXLl3C3LlzFcL7xhtvKC8BiKivW7dOWbmmQHjr1Mm8T3n+/Plo2LBhQUV79btHOvYqx+CPzBpmjlddm+QEWH8CXKRyDO6fvg1likZi95v3qlOMgVKx1vc4AeZOsAzUPTWJqrZv52dXRFF02O32Ss7tz6IoXpZluZ7NZjviRoBPS5JE1yLxoLIGOAH2sOIsFovF4XDQshvdy5Nrvdlstu88zM4v0fxKgAG8M2MpZh8LxyOn1mHyoUzyiIrVgGFTPboCafDgwVi7dm02D9DOiqJtz0RIiczSNuAxY8Yo54CdwX0LNF0VRFcGUdBjC7Q74fRmC7S7ksnT8zvvvKM8opVjWkEeNmyYspprsVhczraIbJOnaArujsDcV4BpVVkURV3bkNpBXFch/JwZa5g5XnUNjBNg/QlwyapmdJiyFcWjwrF/TEd1ijFQKtb6HifAnAAbqHtqElVt39aBAPMVYE2ay4PIacwzpJJbLJa2six/AsCSDzAixOQFOqi2I/ibAHf4cDOSz17BJ82i8EChc0C5ykCDVgVefeSsV+cVSHS10SefUJVnBvLmTN6PiSySM6zp06cr24lzBqcTLHIURaQ1LydY5NyKzuTSmdr33qP7xG8P7k6wyOsylU9XK1EaIuPO1Wci3nSOlwIR2nvvvRf79u3D9evXXVuWSX7a3r1x40YlHnmQjo+PV84pk2dnwkTEn643cl+x/vnnnxETE6OkcZ4BpjPMdAa4aNGiuvYztYO4rkL4OTPWMHO86hqYPwmwKIrPAxhO9z0KgnBQluUhkiRtzUtys9n8iCAI4wDQQHEYwEhJkn50xjebzd1NJtOzsiw3AlBWluUGNpttr3t+sbGxUSaTaRKAXoIgFAaw3m63P5+cnJx5r12OoMWuONtguRoi7pm0BYUiTEgad786xRgoFWt9z6kaFnGziJn0zXHrugXakxVgToA12gC+ApxPBcbGxjY0mUzbs1Z8lwCIB3CnLMs0WaglCEJ7ACUBfC0Iwgmr1fqmRn3omlzLRMVdEE8GtlMXr6P5exsQZhKULW0lC0d4jYUIIq3cVq5cWSGGzvDZZ5/hww8/VL7S9UaVKlVy/UZbhZ1kmbxIU3ryDp3fNUieEGDC7Mk1SCdOnED79tQMgAkTJqB79+5YsmSJQpjpCiTCQmSaVnAp1K5dW/ndZDLhhx9+ULY8UyhRooTiBZq2S1PI6e3Z6QX6zjvvVNLpHTzRsd5lBjo/1jBzvOpanL8IsCiK5MRgHoDnZVneBuBZQRCeycjIiD9y5MjfOaWPjY2922QyETkmu0OktxuAsbIst7TZbDsoviiKfenIGAC6h212bgTYYrHMkGX5IVmWn5Jl+bzJZJoMoIwkSUSa7b4gwJVqxqHFxE2KvTj8buY1dqEcWOt7Tl2yiJtFzKRvjltXAmx3OByVkpOTz2WN45cEQahntVqVe0X5GWB9rAUnwPnUo9lsXiQIQndBEB6wWq2rRVH8EsCTzpVe8sKWkZExS5blFg6Ho0leb8z1UZX3ufiTAP+45wSGLvgL9aqVwtJBLbwXNmull7wkE1kkL8xEFik47wHOLdPo6Ghs2HDLEzxdJ0SEmFZMabWUvDcPGTJEIZQUaPW2adOmyrlb9zPEOfOmwZzOFR84cEC5luj48eOK86o2bdooK71OZ1q5EeD9+/crnqnp2iTCQmS3atWquOeeexSCTsTYGWg783fffafcbUyrxkTu6comWh158qM9AAAgAElEQVQuXbq0Kx6loxcEBV2/pKriGTVerBlsjldd7/AjAd4hCMJuq9U60CmpKIqJAH6SJOn1nNKLoriA3p1JkuRaQjWbzasFQUiVJCnbfXNxcXF3OByOozkJcK1atUqGh4fTJKuvJEmUHx3JqCLL8nGTydQ5KSlpjS8IcDVzPJq+mzluH53Q2XXdmzoNBX8q1voeJ8B8C3Tw90p9JFTbtwvaAg3gvNs1rOS8hjw+O7KkJu5GLymDatepPjXqv1w4Ac6nrkVRPC3L8kmbzdaYouUkwPTsjjvuKBQZGZkCYJUkSZkemYIk+JMAv7p4HxbsOo5nW9fC650ziaua8MEHHyienck5FBFUvQORUnJERWdoaSsynSfOLagd1PSWl/K7cOGCQp5J1vXr1ytbpfUOwYRXb2x55ccaZo5XXcvyBwGOj4+PzMjIIFfwj7lvYbZYLNNkWa4vSdI9OaUXRZFWhadIkpTprj7TRg0VBGGI1Wqt4R4/LwJMd0nKsrw+KiqqzP79+1Pd8vkri3i/lbNcLXbF2QZrWeqgwfj1StbS+PsRGR7avlxY63vONsMibhYxk745bv1WgC0Wy/95Yq2sVuvnnsTjcXKvAU6A82kZoijepK1lkiQp7oVFUfwMwDNXrlwpfurUKde9NWazeaEgCC0kSYoOpoamZaLijsOTga3NBxuRcv4avniqMdrFVVRdDampqcqWYrriyHk+VnVmuSScN2+e4oyKzgg77/YNdgJMq8mffvqpsipMdxz7IniiY1+UG8g8WcPM8aprbf4gwFmrridNJlOLpKQkOnbjJLR0RqKfJEm3+aAQRTGNti27O180m829BUH4UpKkKE8IcF7xRVFcS4uzkiRlXqjuFpx2hV4iFilSxKtKpTZIO2RiLfGoNz5zBXjf6A4oGpX7i0ivMg/iyE7cdevW9fgu+iCG47FoLOJmETM1CI7bu75NdoVuBgHQqHHjxrs97lS5RLRYLI9dvnx5pTsv0ZIfK2k5Ac5H06IokhOQfZIkKYeURFEkN76vhYWF1UtMTDzgTGqxWJYBaG+1WvX1TKSxFfqLAP/z3w00m7AeJgHY+9Z9KFHI+/O/GqHqnpyTBd2rNOgy5DoOOpXoKpBe+vUnAXY4HM2Tk5N/c1aE2WweKQgCbU+Oy1k5WQS4n81mm+9mi/rIsvy5JEmZ96dlhbxWgPMhwL8IgnDYarVmutN3C067okVZGQ4ZPX44o2TxVdcKKBYZ2ivAWuqKp+U1wGsgpGtAMwGme4Npp5DzmqSQri0dwXECnE9lWiyWTbIsV3a+fRdF8WEA5AzrY0mSXqSkZrOZnGHtk2X5qM1mq6ujbjRn5S8CvHTvSbz0/V7UjS6J5YNbapY7GDLQa/IcDFg8kYE1vFQnrGHmeD3pCbfH8QcBNuIWaC0rwOTMz/LWOqWyd7zeFuWKZVuwVqeoIE7FV8e8Wx0LYlUWKBrXNTu6ds4jaFeLt7s79FwBznlPcIGNlEdQaoAT4HwagiiKdAj1HYfDUTc5OflQ1iSF9ixUEwRhBzkKAdABQCm6ukKSpExXxUES/EWAX1+yH/N3/o1nWtbEqAfJUbbxAycLxtdhQQi4jguqIWP/rpd+/UGAqaZFUSSb8qfNZqOrkJQgiuIhAEvzcYJV3LlDKSv+KgAXvXWCJQjCE1ardSHlUbt27cp2u/2EL51g1a9fH3Gj1yDdLuO319uhckm6fSl0g15t0Wg1xCJuFjFTu+S49TsD7G0/5wTY2xrLjM8JcD71lnUuq5vJZNqSlJS0P2uCUU+WZfIOHZuVVJZlea7NZvufm4c2ddrQOZW/CHD7yZtw+NxVzH6yMe6NV3/+V2f4mrJjbTBnDS+LBps1HeuF148EWLkGSRCE5xwOB22DHiAIwv/CwsLqJCYmHhNF8WsAJ51kOC4urrnD4dgiCMJIWZaXCoLQVZbl8e7XINFNBWlpadUFQagCYCWAnrIsWyMiIv45dOjQP9QPsq5BehDAUw6H40LWncBlfXkNEhHgumN+wbU0O7YMb4vqZb07S6xpcA9AYr3aYgBE11Qki7hZxMyiPXV2DLX61tOucAKsbpjiBFhdvdFb8jvtdnvp8PBwm3MioTIrnyXzBwH+73o66o0hfylQ7v8tUzTSZ3j8mbHaQc2fMupZFmt4WTTYrOlYL7x6TlQK6rOiKNLq7wgAlQVBOOBwOIbabLYtWUSVjuSkSJL0lDMfi8XyKJFeupcewGFZlkfabDY6pqMEs9n8FDnFyqXcMZIkvU3Ps24y+ABAbwC0FLvebrc/f/jwYdrhdFvQYlfcddJg3DpcupGB9cPuQUz5YgVVjaF/16stGq0SWMTNImYW7SknwEYbjW6XlxNg4+swTwRaJirumeY3oG9L/hd95uxAtTKFsXVEu5CpTdaMGGt4WTTYrOlYL7z+JMBGGEC12BV3ndw1YQP+vZKG1UNaIa5SCSNAVy2jXm1RtQABSsgibhYxs2hPOQEO0KCiY7GcAOtYmcGWlZaJiqcE+JONyfhgjRUPJlTGx70bBlsVqJaHNSPGGl4WDTZrOtYLLyfA2YdRLXbFXSct3t+Efy7dwIrBLXFndEnVY7UREurVFo2A1dO5g9GweCov17V3Z2E9rddgjadW3x7YlTCLxdLs5s2bB1NSUi7mh18UxaSMjIz7jhw5QnfD8+BhDXAC7FZRoijSnYuIioqasX///lTndw/rUpYkaYKHcf0STctExVMj9uy8XVhz8AxGdq6N/7WmXXihEdQOakZFzxpeToDDjNpUPZZbrzbtwUTFY5lCIaIWu+KukzaTN+P4hetY8nxzNKxeOhSqJk8MerVFo1USi7hZxMyiPXX2RbX69sSuiKJ4QxCE2lar9ajR+r4R5OUEODsBdgCQTSZT7aSkJEkUReW7h87CiAAH1axSy0TFUwLc7N31ylv8BQOa4a5aZY3Q5j2SUe2g5lHmQRiJNbwsGmzWdKwXXk8mKkHYpX0mkha74q6Te6dsxZF/r2Lhs3ejac0yPpM3GDLWqy0GAxZvZGARN4uYWbSn/iDAZrN5l8lkGmG1Wjd40+94XM9qgBNgt3qyWCz/l/V1odVqvSyK4jOCIBAB9ihYrdbPPYrop0haJiqeEOAzl27grnfXwyQA+9/uiKJR4X5C5vtiWDNirOFl0WCzpmO98HICnH281WJX3HXSefo2WM9cxrfP3IUWseV8P6gHsAS92mIAIagqmkXcLGJm0Z76iQDfazKZ3gMw0m63/3nt2rWr7h3x1KlT11R1TJ5IqQFOgEO4IWiZqHhCgNce/AcD5v0JS8XiWDO0dUjVJGtGjDW8LBps1nSsF15OgH1DgLt8sh0HT13C3P5N0MZSIaTsR04werVFo1USi7hZxMyiPfUHAc7aheos6rbFuGDbdWq08YkT4Hw0ZrFYJsqyfFGSpHeNpliS19cEeNIaKz7emIzHG1fFxEfrGbGK8pSZNSPGGl4WDTZrOtYLLyfAviHAj8z8HXuPXwyp++PzMih6tUWjGVkWcbOImUV76icC3D6/Pi9J0nqjjQnBJC8nwPloQxTFNEEQllmt1keDSWmeyuJrAtz/y53YaD2HcQ/fib7NangqliHisWbEWMPLosFmTcd64eUE2DcEuOfsHfgjJRUz+jTE/XUrG8IuqBVSr7aotvxApWMRN4uYWbSn/iDAgeq3rJTLCXD+BPioIAh/cgJsx969e1G/fnb39vd8sBHHzl/D/P81w90xoeMAi8XBnEWjzRpmjledWecE2DcEuO8Xf2D74fOY1rM+utaPVqccg6Rire9pJQcGUWuuYnJd82uQPGm/3tiV2NjYKFmWqwmCEOmed3Jy8iFPyuJxcq8BToDzaRlms3mKIAhPhIeH1zx06NAVozUiX64Ap2U4EPfmKjhkYOcb7VGhRCGjVU++8rJmxFjDy19yBJXDep+MHXq1aW8mKj4BEmSZarEr7jrp/9Wf2CKdw+TH6uGRRlWDDKW+4ujVFvWVyve5sYibRcws2lOtL3k8sSuiKJJ3wDkAHsqtt/IzwNrGME6A86m/2NjYEiaTaasgCKmCIAxOSkrar626/Ztay0TFXdLcBvTks5fR4cMtKBoZhgNjOkIQQqspsWbEWMPLosFmTcd64fVkouLfkT2wpWmxK+46efab3ViXeBbvda+Lnk2rBxaUj0vXqy36WEzds2cRN4uYWbSnfiLA8wDEOhyOl00m0y8AHpNluaIgCK/LsjzMZrOt0L3TMpRhaLEWnRUniuJaAEUANM+6D/g4gL8B3MhZFF2XZLVaO+osgqbstExUCiLATg/QdaNLYvnglprkDMbErBkx1vCyaLBZ07FeeDkBzj5Ca7Er7joZ9N1erD74T0j6kMhp0/Rqi8FoK/OTiUXcLGJm0Z76iQCflmX5YZvNtkMUxUthYWGNEhMTbaIoPgzgZUmSQuv6FT8PcJwA51PhOVyQF6QaOdi2I2iZqBREgD/bfBgTViWhS70qmN6rQUF1Y7jfWTNirOFl0WCzpmO98HIC7BsC/NKCv7Bi32mMfjAeT7esaTgb4Y3AerVFb8oMhrgs4mYRM4v21E8E+LLJZKqblJSUYrFYjgHobbVat1kslpqyLB+UJIkW6HhQWQOcAOdTcbGxsTHe1GtycvJhb+L7Oq4vCfCri/dhwa7jeKm9GUPvFX0Nxe/5s2bEWMPLosFmTcd64eUE2DcEePji/Viy5yTe6ByHAa29MrV+twdaC9SrLWqVw9/pWcTNImYW7ak/CLDZbN4ly/IbycnJa81m81JBEC6kp6e/FhER8RKAnpIk1fJ3nw6l8jgBDiVt5sDiSwL8+MzfsDPlQsh68GTNiLGGl0WDzZqO9cLLCbBvCPDrPx7Awl0nMLyjBYPaxoawJQb0aotGqyQWcbOImUV76g8CLIrik4IghFut1i/i4uIaORyO1QDKAEiXZflpm832ndHGhGCSlxPgYNKGzrL4jADb7Wgybg3O3ZCxrF0hJLRvA4SFlldZ1owYa3hZNNis6VgvvJwA+4YAj152CN/u+BtDOpgxpEPo7SJyrzW92qLOUwSfZ8cibhYxs2hP/UGAc3bQ+Pj4YhkZGfHp6enHjh49esbnHTjEC+AE2AsFU+OTZbm4w+HItd6sVuspL7LzeVSfEODNS3Fj8jDE1f1YkX/P5l4oXbok8Mo0oF13n2PyVwGsGTHW8LJosFnTsV54OQH2DQEetzIJc7enYFDbGAzvGOevoT0g5ejVFgMivIZCWcTNImYW7akvCbDZbJ4aHh4+JzEx8YCG7seTFlADnAAXUEExMTEVwsPDx8qy3BVAhXyikxOs8GBqcboT4NSjCHvtcaQUroQ2LWajsP0GDm18FAKymtHExSFDglkzYqzhZdFgs6ZjvfByAuwbAvzeaitmbz2KZ1vXwuudaweT6dRdFr3aou6C+ThDFnGziJlFe+pLAiyKYhIAM4A/ZVmeI8vy98nJyZd83F2Zy54T4HxUHh8fXykjI2MngGgAtN2A9vmWB/AHADp8XjbreiSKkyFJUqtgakG6EuDdf6LhW49AOHsC20snoHejdxFz9TjW/zYwC7IAVKwKLDta4Hbo1NRUtG/fXkm3efNmFC9ePJiqTZFFLyN27NgxdOrUCZUrV8aaNWsQERERdFj1xBuU4PIQSi8dGwUzx6tOU5wA+4YAT/rFhhmbDuPpFjUx+qF4dcoxSCrW+p5WcmAQteYqJtd1fYSF2JG4/NqjWn0XZFcsFksLh8PxtCAIjwEwAVhCZNhms20xcv8IJtk5Ac5HG6IofgRgEIAxkiSNEUXxSwBPOq87slgs7WRZ/kQQhDN2u71jcnLyzWBSrp4EOHnBHFg+fE6Bt7ByB4yoMwSt//0TX+99KzvkmRuBxm3yrYZJkyZh9uzZePLJJzFy5EhX3FWrVuHrr7/GkSNHQINDmTJl0LRpU7z44ouoVq2a11W7f/9+TJ06FXv27FEIbe3atTF48GC0aNEi37xWrFiBuXPn4ujRo7hx4wZKly6tpB0wYACaNGlyW9orV67g4YcfxvHjdE008Pbbb6NXr16ueIMGDcK6deswZswY9OzZ02sc/kigdhD3h2y+KoM1zByvupZU0ERFXa7GTaXFrri3wWkbDmP6ehv6Nquh3AUcyoG1vscJ8F7Ur8+JYCj3aa1t3FO7UqVKlSJFixbtIQhCfwAtAdBtM58LgvB1sB27NJq+OQHOnwAn05sXSZLojgY5JwGmpDExMdXCw8Npu8KHVqv1zWBqAFomKu44yHgfm/Uean0+Snk8tWYvTI3pg14nVmNCUuZZYFcY/x3Q6Rb5y1kf6enpaNWqFWgVePHixahbt64S5ffff8dTTz0FWZZRrlw55U+SJDgcDsTGxmLlypVeVW1iYqJCNp0ENjIyEmfOnFHeTM6cOROtW+d9f/g777wDIsHFihVDoUKFFEKekZGBqKgo/Pzzz6hatWo2WV599VX89NNPrmc5CfDq1avx0ksvwWKxYNmyZV7h8FdkFidorGHmeNX1Jk8nKupyN14qLXbFvQ1+uukIJv8ioVfTapjQPcF4FeGFxKz1Pa3kwIuqDbqoXNec+HvSKNXYFbqa1WQyPQ2Atl4WkyQp0pOyeJzca4AT4PwJ8HUAqyRJUrw7iaI4B0D/8PDwwocOHUpzJhVFcYUsy6LNZgsqV5ZaJio5CbD7CvDw+JewqMq9GHZ4HgYfXZC9BgtYAaYtz7SSWr58efz666+utJ9++immTZumfN+6dSsqVKiA0aNHY8GCBShSpAh2794NQfC8uT733HPYuHEjoqOjFdJJRLZ3797466+/SI9Yvnx5npq/efMmwsPDsXdv5lvcJUuWYNSoTPI/ffp0dOzY0ZWWCPHQoUNx//33g1awKeQkwDTQNW7cWFmFpvgxMcF35yWLRps1zByvummAmomKupKMkUqLXXFvg7N/TcF7q5LwSMOqmPx4PWOAVykla32PE2C+Aqyyqxgumdq+7a1dSUhIKHrz5s0edP0RgOYArJIkhbbzBB+3Bs8ZhY8FCcbsRVE8R3zMjQBPAjBUEIRYq9V61Cmz2WxeKAjCg5IkFQkmHFomKjkJ8F7XGeCT6NNgHLaVrY8PD0xG9382ZkX17Azw5MmTMWvWLLRr1w4zZsxwFbNjxw7069fvthXgUqVKKeTzgQce8LhqabW2UaNGyupvjx49MHbsWCUtrfxOmTJF+bxlyxZUrFgxzzyJ/NL2bJPJpGyFppVrWgEmkkukmsLp06fRpUsXZXv2hx9+6CLGOQkwxaV4VqtVkYVkCragdhAPNhzeyMMaZo7Xm9ZxK663ExV1pRgnlRa74t4Gv9x+DONXJqJr/SqY1rOBcSpAhaSs9T1OgDkBVtFNDJlEbd/21K6YzebWWdufH6UKEgRhEW2Btlqt2wxZYUEkNCfA+SjDYrH8ST9brdZG9N9sNvcXBIFWgQdLkvQpPatatWrhIkWK0Fbpm5IkkWOsoAlaJiruIFwdPMsLdNu7Z+Jo0Wh8v+s1NLtIXto99wJN53nJGVTfvn1dq6rOsuj5a6+9ppz/dYZ69eoppDEuzvNrMs6dO4eWLemoBDBw4EAMGTJE+bxo0SJXmbSyTKu7eQVahX7mmWdcP9N5ZFr9dZ4Bpq3ZRNgPHDigrBCTcyunY6/cCLBzRfrpp58GbZkOtqB2EA82HN7Iwxpmjteb1nErrqcTFXW5Gy+VFrvi3ga/3Xkco5ceROe6lfBpH8XEhmxgre85FckibhYxk745bu+2fudnV2JjY6sKgtBPEISn6KQlgB1EesPDw78/dOjQlZAdKP0MjBPgfCrcYrF8IMvyS3a7verhw4fPxsXFlXU4HCkA6LqjDwVBOAGgryzLd9ECoyRJ5DAraIKWiYo7CPeBTdi0FHFrTEgzRWDrr/+HajfOABWrAcOmenQFUv/+/bF9+3YQIaStw85Aq6P026VLlxQHVER433jjDYUs03ZpciJF25g9CWfPnlXOGVPIiwAvXLgQRK7zCk7MdN53zpw5inOuSpUqYf78+ahSpYoi44QJEzB+/Hg89thjOHHiRL4E+JVXXlG2XT/++OMYN26cJzD8GodF48UaZo5XXZfiBDh7vWmxK+5tcMGuk3jjx/24N74iZj/ZWJ1yDJKKtb7HCTBfATZI19Qsptq+nZ9dEUUxA8B5APNkWf7cZrMlahaUZ3BbDXACnE+jMJvNtEQ40mQyfWq1WpW9vhaL5QlqkADoPhs5a/kzKS0trXlKSsrFYGpjWiYqeRHg1OsZaDx+nbLmK7W+hIgKlYEGrQq8+siZH3lhXrt27W0eoIcNG6Y4nnJ3FPXLL7/ghRdeUJK6O8wqqI7dt0C7E05vtkC7D2rXr19XtlRTePbZZ/Hyyy8rK9U//vijcj6ZAjnvongUyOFWnTp18P3337tE5SvABWnN/7+rNVz+l1SfEjledfXICbBvCPCSPacwfPE+tLGUx9z+TdUpxyCpWOt7nABzAmyQrqlZTLV9Oz+7Yjabu0dHRy/btGkTEWEefFQDnACrqFiLxVJTluUHBEEoLcuyFB4e/qO7UywVWfokiS8IsO3cVXSauhVlikZi95v3ei238wqkDh064JNPPnGlJ2K5adMm5U5gIsi05ZjO6xJppeB0HvXRRx/h448zPU/TqnFeIacTLDq/S06w9u3bl80J1ogRI5RnCQkJmDhxopLdt99+q5zZJS/UtE2azv0SQafg3LrtJMB5lU8r2EuXLnX97DwDnNv2aK8r0QcJ1A7iPhDFb1myhpnjVde0OAHOXm9a7Ip7G1y+7x8MWbAXLWPL4ZtnaBNV6AbW+p5TkyziZhEz6Zvj1m8LdOiOhMGFjBPg4NKHrtJomai4C+I+sP1+NBV95uxAbIViWPfyPV7LS56ZiZxWrlxZIbzO8MMPPyhbnimUKFFC8QKdnExHq6GQUzq/S8FTApyUlKQ4m3Jeg0RndGlrdM5rkIjQ7ty5U7lveN68eUoZtApN8UkGiv/3338rz8kz9HfffZfr1un8tkA7V5CpHmmV22w2e11vvk7AovFiDTPHq64XcQLsGwK8+uBZDPpuN5rWLIOFz96tTjkGScVa3+MEmK8AG6RrahZTbd/mdkVz1WvOgBNgtyq0WCzHZFn+xuFwfJucnHxIc+0GOANfEOCVB87gxfl7cFfNMligYtKSlpamnM+9ePGicndu7dq3vLjTGVkimOR1mUgjnblt27atsu24dOnSSm3SmVsiqgVdZURxaWV36tSp2LNnj/J2Mj4+XtlS7XSQRXFyI8Cvv/66cu0SeXmmdFQ2rQSTU6y8HGflR4DpHDM5//JE5kA1GbWDeKDk1aNc1jBzvOpaDZ+o+IYAr0s6h2fn/YmG1UthyfMt1CnHIKlY63ucAHMCbJCuqVlMtX2b2xXNVa85A06A3apQFEVH1rleOtO5x2QyfRMWFkZe1/7RXNMByMAXBHje73/j7eWHNHnu/OCDDxTHUuT0irYSexO6deuGxMRExRlVgwa+uzpD7aCWGxYi3XSeOVi3P5PMeuL1Rp+BjMsaZo5XXWvjExXfEOAttvPoP/cP1I0uieWDM732h2pgre9xAswJcKj25Zy41PZtblcC30I4Ac5OgMl18BMA6L4tWnIkJ1dEitfJsjzv6tWrP506derWHT2B11++EviCAE9bn4zpG5LxRLPqGP9wXVU1kJqaqnhMFgTBde7Xk4wuX76sbFUmx1ZjxozxJInqOGoHtZwFHjt2DJ06dVK2fK9evVpxkBWMQS+8wYgtL5lYw8zxqmudfKLiGwK8/cgF9P18J+IqFcfqIa3VKccgqVjre5wAcwJskK6pWUy1fZvbFc1VrzkDToBzqcJGjRpFXLlypXPWFUf0n+7fITJM5HdJ1hbpX5yrxZq14KMMfEGARy87hG93/I0X25vx8r2ijyQPfLZqB7XAS65OAtbwUi2xhpnjVdc3+ETFNwT4j2MX0XPW74gpXxTrh7VRpxyDpGKt73ECzAmwQbqmZjHV9m1P7EpCQkLRmzdv0jbJ9rIsVwBgchdYkqRamgEwnAEnwAUoPzY2tkRYWNhjsiz3AUCvqakBEhk+A+A7OjNss9n2BmMb8gUBfmH+Xqw68A/Gdq2DJ+++Ixhh6yKT2kFNl8IDkAlreDkBDgtAK/NvkXq1aU8mKv5FFtjStNgVd53sPfEfHpnxG2qULYLNw9sGFpSPS9erLfpYTN2zZxE3i5hZtKdaX/J4YldEUZwP4B7agWoymU7nXHSzWq3TdO+0DGXICbAXyrZYLFUcDkcfQRBom7Rz/y+R4URBEOZZrdb3vcjO51G1TFTchXMf0J/4/A/8duQ8pvWsj671o32OIVAFsGbEWMPLosFmTcd64fVkohKocSoQ5WqxK+46OXj6Mrp8vA1VShbC9tfbBwKK38rUqy36TWCdCmIRN4uYWbSnfiLAFwVBeMBqtW7TqUvybNxqgBNglc0hNjY23mQy9QXwDICy9GZGkqSgWlbRMlHJiwA/9PF2HDp9CXP7N0EbC+3ICM3AmhFjDS+LBps1HeuFlxPg7GO8FrvirhPrmavoPH0ryhePwh8jO4SmIclCpVdbNFolsYibRcws2lM/EeCjsix3ttlsiUbr+0aQlxNgFVqKi4srK8vy47Is00pwMwBUj0wQ4HsmbcaJ1OtY8nxzNKyeeTVRKAbWjBhreFk02KzpWC+8nAD7hgAfPX8NHT7cglJFIrB39H2haEZcmPRqi0arJBZxs4iZRXvqDwJssViecDgcXa9evdrPSA54jTJOcQLsoabuuOOOQpGRkV0FQegjy3JHAOFZxPdfAAsAzJMkaaeH2fklmpY39XmtADcYtw6XbmRg/bB7EFO+mF9wBKIQ1owYa3hZNNis6VgvvJwA+4YAn7+WjqbvrIdJAKzj70dEWDb/LoEY9n1Wpl5t0WcC+ihjFnGziG01bMYAACAASURBVJlFe+oPAiyK4h4AMcQ1ZFlOEQQh3b2rSpLU0Eddl4lsOQHOX82C2Wxun3XmtxsAYnxUZzdlWV5OB9OrVq26atOmTRnB2Fr0JsAJCfUgjl4DWYayZY22roVqYM2IsYaXRYPNmo71wssJsG8IsCCYEDd6NdIyHNg6oi2qlSkSquaEOY/zWsmBkRuCXuOO0eqA466PsDDPT0F6YldEUXwrv3YgSZJv7wM1WiP0Ul5OgHOpsNjY2IYmk4m8PvcEUMm5xRnAr7TSm5GRsejIkSP/eVnXfo+uNwGuZamDBuPXKziSxnVCoQjPO7vfwWsskLXBnDW8nACHbt/Ve/LtyURF43BjqORa7ErOcabtpE04+u9VzP9fM9wdQ640QjOwOL6yOMayipnj9v7aK25XAj/WcwLspgOz2TyStjgDsGQ9pvqx0lVH4eHh3yQmJh4LvMo8l0DLRMW9FKfxLldDxD2TtiAq3KRsWQvlwNqEhTW8LBps1nSsF14+Uck+0muxKzl18sScHfg1+V9MeqweHm1UNWRNil5t0WgVxCJuFjGzaE+1vmjldiXwoxknwG46EEXRkfWVzvV+n3Wu94/Aq0mdBFomKrkR4IgKtdDlk+2oUDwKO7nXTnVKCdJULBpt1jBzvOo6H5+o+I4Av7p4HxbsOo6hHUS81MGsTkEGSMVa39NKDgyg0jxF5Lr2biuwkXWthfjnZVdEUbwAQJQk6V9RFFNz3v3rXl+SJJUxev0FUn5OgLMT4AV0rjc6Onp1sJ7r9aax6E2Ar5eoBroHOLZCMax7+R5vRDFcXNaMGGt4tRguwzXmLIFZ07FeeDkB9h0B/nCtFdM3JKNvsxoY9/CdRu1aBcqtV1sssKAgi8AibhYxs2hPtb7kyYcA93M4HN8nJyffFEWxX35dWpKkr4KsyxtKHE6ADaUu74TVmwCfiaiM57/bg4bVS2HJ8y28E8ZgsVkzYqzhZdFgs6ZjvfByAuw7Ajx7yxG883Miutavgmk9GxjMSngurl5t0fMSgyMmi7hZxMyiPfUVAQ6OnsuGFJwAh7Ce9SbAyfZyeHXJAbSxlMfc/k1DuObAnNdOFo02a5g5XnVDFifAviPAC/74G6/+sB/t4irgi6eaqFOQAVKx1ve0kgMDqDRPEbmu+RZoT9ovtyue1JJv43AC7Nv6DWjuehPgP6+UxLurrCH/tp7Ft5ksGm3WMHO86oZjPlHxHQFetf80Bn67G41rlMbigc3VKcgAqVjre5wAe+8V2ADNOF8ReRv3jvhzuxL4Fs8JcOB14DMJ9CbAG/8tio83Hg7581qcAIf+FTlcx6GvY70mZHyi4jsCvC35X/SZswNixWJYOzR0/Uro1RZ9NlnwUcYs4mYRM4v2VOtLHm5XfDToeJEtJ8BeVJbRoupNgJeeiMTXv/2NQW1jMLxjnNGqwyt5WTNirOFl0WCzpmO98PKJiu8I8L4TF9Hl422oVKIQfn+jvVdjtJEi69UWjYSZxTGWVcwct/cr/tyuBH404wQ4wDoQRfF5AMMBVBYE4aAsy0MkSdqam1hms7k7gDcEQYgVBCFClmUbgMmSJM3LLb6uBHj3n5i38zx+Ou7AG3UjMaBnOyAsdFeRWJuwsIaXRYPNmo71wssnKr4jwCn/XkWbSZtQNDIMB8d2CrA19l3xerVF30nom5xZxM0iZhbtKV8B9s2Y4c9cOQH2Z23nKEsUxR501zCA52VZ3gbgWUEQnsnIyIg/cuTI3zlFi4uLa+NwOErLspwky3JaWFjYg7IsTzaZTA8kJSWtyRlfNwK8bjHs7w/CwOrPYX35pnjv0HT0zDgEvDINaEecPPQCa0aMNbwsGmzWdKwXXk6AfUeAL1xNQ8NxvygFJL9zP8LDTKFnTMCeU0Wt5MDIjUCvccdodcBx638GOCEhoejNmzdfA9BeluUKALINkJIk1TJaOwkmeTkBDqA2RFHcIQjCbqvVOtAphiiKiQB+kiTpdU9EE0VxtyAIK61W65s+IcAblkAe8ahyF/fjjd7HH6Xr4NN9E9D57PbM4iYu9pgEp6amon37zG1umzdvRvHixT2BGJA4eQ3m9Py+++7D6dOnsXz5csTExAREPr0LZdF4sYaZ41XXa/xJgL3ZEURozGbzI4IgjANAA9FhACMlSfrRDakgiuJbAAYAKC0Iwg673T4oOTn5oJvNSQFQw712ZFl+32az0cTrtqDlxWrONphud8A8cpVSxp4370XpopHqlBTkqVjre5wAe78lNsibcIHi8TauPwEWRXE+gHtkWZ5nMplOKxNxt2C1WqcVqBgeIc8a4AQ4QI0jPj4+MiMj4xqAx9wnLBaLZZosy/UlSSrIIwhNbNoJgrDM4XA8bLPZMl+juwXnREUURRQpUsR7pHY7TA/HAGdPgBpKx2Yfw1rsDnyzeyRaXvgLMj2tGA3Hj4c92g794YcfYs6cOejbty9ef/12fk+keOBA17sA7NmzB1FRUV7LfeDAAUybNg179+5VrjOqXbs2Bg0ahObN8/cyunLlSnz99df4+++/cfXqVRQrVgx33nknBgwYgMaNG7vkmDdvHiZMmIDOnTtj0qRJXssXjAmonvbv34+6desiLIS3trvXPWuYOV51PY8IsCRJlLhR48aNd6vLpeBU3u4Iio2NvdtkMtFxGXr5SaS3G4Cxsiy3tNlsO6hEi8XyqizLI2VZfio8PFyy2+2jALQWBMFitVovUxxRFFMEQfg8LCxstpuUVw4dOnQlN6n1JMCUf+03V+N6uh1bhrdF9bIq7FTBVRvwGJwceEcOAq4wDQJwXbOja2omavXtyYtVURQvCoLwgNVqpR2iPOhcA5wA61yhnmZnsViqyLJ80mQytUhKSspaTlUmI28A6CdJkiW3vGrVqlUyPDz8JABihnZBEJ63Wq1f5DdR8VSmnPGKWXfB8uFzrsfNWs7FP4XKYfmOl1D3Mi02ZAbryzNxxXKLIOZWXkZGhkJCL1++jPHjx6NWrew7N/777z+8+uqruHTpkiv53LlzERnp3YpASkoK3n77baSlpSkrzBEREbhw4QJMJhOGDx+OevXq5VkdRH63b9+OUqVKQZZlnDp1ShncKA8iuuXLl1fSkqzPP/+8kufHH3+MkiVLqq1ino7XAK8B49SArwmwVzuCRFFcAKCEJEn3O6vQbDavFgQhVZKkXgDoJekpQRCmWq3W9ylObGxslMlkOiPL8qs2m+0zepZFgCnOVE9UoTcBbvHeBpy8eB0/DLwbjWqU8UQEw8VRO0k2HNAcArOIm0XMpHaO2zvi7yEBPirLcmebzUY7Q3nQuQY4Ada5Qj3NzkmAHQ5H8+Tk5N/cJjAjBUHoK0lSXm6WTTExMbVMJlMxQRBoP/GbJpPp4aSkpE05y9a6Aiys+R6m0U+4sq3bZgEuhxfFhu0DUOvaKddzx9hvIHfsmS905+oukUj6nDM899xz+O2339CqVSts2LBB+VnNCjAR002bNiE6Oho//fSTsoL8xBNPYN++fbRdEEuXLs1Tzps3b7pWnGkwnz59OmbPzlwUmTp1qrL12Rl69eqFv/76C6NHj0bPnvlj97RNBDIea6uDToPN0qo3azrWC68/VoDV7AgSRZH8REyRJGmKc+wQRXGoIAhDrFZrDbPZXEsQhMMOh6NhcnLyHjcbs1QQhIuSJPWjZ0SAs16o0tvG47IsL4qIiPjg0KFDabmNSVrsSm466Tl7B/5IScXUHvXwUELlQA6DPitbr7boMwF9lDGLuFnEzKI9dXYZtfr2xK5YLJYnHA5H16tXr/Y7deoU7RjlQcca4ARYx8r0Jis1E57c8hdFcY4gCNWsVmvHnL9reVOv5LVrE/BcW+UjHTyIab8UDiEMO7f0RYW01FvFzdwING6TL/zJkydj1qxZaNeuHWbMmJEtLm0pplXh1157DVeuXFFWVSkQafVmCzStMjdq1Ag3btxAjx49MHbsWCWfmTNnYsqUzDnili1bULFixTxlpTJJluvXr+Pw4cPKW02SYdWqVQqpdgaKQ3I/+OCDIGxGDyy+vWUNM8errpd68qZeXc63UqnZESSKYhptbbbZbN85czKbzb0FQfhSkqSouLi45g6HY5sgCNFWq9X1xlIUxVmCINRw2gwizQ6HY3dYWFiqLMtNAUwAsFSSpGdyw+W0K1oxO9NP23ERW/6+gb51i+HhuGJ6Zcvz4TXAa4DXgBFqIM+dRaIo0otL8u8gyLJMR1XS3QFJktTQCACDVUZOgAOoGXKCJcvynzabja5CUoIoioeyJh8eOcEym82fm0ymGKvVehsD1UyA7XbgoTsgnz2Jm6YIxLVbosh4YONjKGa/Tn0SqFgVWHa0wDPAL774ItasWaOc/x01io6hZQabzYZHHnkETZo0Uc4HE/lVS4DPnTuHli1bKvnSWeIhQ4YonxctWuQqc8GCBahfv36eWqct0P3793f9XqZMGWUlmORzD19++SXee+895YzwDz/8EMBWpE/RrJEjqjXWMHO86vqKPwmwNzuCsghwP5vNRo5SlGCxWPrIsvy5JEmFnAQ4LCysSmJiIjlQcdqY2bIsV7PZbLneO5TlWGuxyWQql5SUdD5nrem9AjxprYQZm4/gibuqY0yXeHVKCvJUaleJghxWgeKxiJtFzE57ytKOKmfjV6tvT1aAsxwY5tnPJEkaU2An5BHyrAFOgAPYOJxOTwRBeM7hcNA26AGCIPwvLCysTmJi4jFRFL8GcNLpEVoUxddlWd5F3j4FQaDtap0B0NmugZIkzclrokJOoFQ5waIMs7xAn48ojsb3ZC40HF7XBWFOZ3QeeoEmUknkkrY6Dx061CVqly5dcPbsWcWjMm2P/uijj1QTYMqHtlBTyIsAL1y4MN9zwE7Bzpw5g3fffRerV69GpUqVMH/+fFSpUsUlt5NU33HHHQqxN3pgjRw5DTY5SqMXIiw4/mJNx3rh9QcBVrMjSK8t0DnHLlEUaavLCVmWmzmdabnH0fJiNTedfLvjGEb+eADt4yrg86eyv2g0+rjqPklmaaxhGbde447R2j7H7d08wh92xWhtyN/ycgLs7xrPUV7WtRcjAFQWBOGAw+EYarPZtlA0i8WyibY9SJL0VNb38bIs093BVQFcFwSB7gOeJkkSOUO5LWiZqLhnZl+3GClT3kL7uhNRyH4DSRsfBSpWA4ZN9fgKpMGDB2Pt2rV48sknMXLkSFf2FosF4eHhLmdX6enpoD8KRNpfeeUV9OnTxyMtuW+BfvzxxzFuHN0O4t0WaHfDTYT9mWcydwE+++yzePnll11yOFeA69SpgyVLMlfGjRxYNF6sYeZ41fVQf01UvN0RlOUEq7gkSfQiVAmiKNKdQnS+190J1hSr1TqRfs8i2mfdnWDlrBWz2fygIAjLMzIyauR2H70Wu5JbG9yQdAZPz92FO6NLYMXgzBeYoRZY63vudpQ14s917R0RNHpfV6tvf9kVo9evL+XnBNiXtRvgvLVMVNxFpw6+bOMODF2XirIRDvzZtTjQoFWB257d8yAvyuRQqkOHDvjkk09cPxEBzi/QdUlPPfVUtpVhq9WaZxJaYd64caNyXnfZsmXK+d3evXsr54npOihaaaYwYsQI5VlCQgImTlTmhvj222/RrVs3hXgT5k8//dS1Gp1z67bzDDBdheQ8XxxgdWsqXu0grqnQACdmDTPHq67B+Wui4u2OoKwtzlsEQaBrjsixVVdZluklac5rkOg4Tf+wsDCb3W6nWwbaOK9ByrpKqZksyxtNJtN/AJrIsjyFdhrZbLauudWYFruSWxv8I+UCHpv5G+4oWwSbhmf6nAi1wFrf4wSYnZ1FLOuasKvt23nZFVEUL9C7TEmS/hVFkZztZLv7131slCQpNN3m+8kAcALsp4oORDFaJio5CfDC9Tvx+oYLqFamMLaOaOc1HCKlRE4rV66seGnOK+S1Bdr9eX4EOCkpSXGARY6wSpcurVxhRFujaYsrOcNq3bq1UjQR2p07d6Jp06aKMysKRMYpfvXq1ZVVaLoPmAKtUH/33XfZtk5TGfRmm7xAe7pC7XWl+TGB2kHcjyLqXhRrmDledU3IXwSYpPNmR1DWmPUokV4AdK/cYbrz12azuW9JoauQ3qJNLABKA9gRFhY2KDEx8QClj42NbWgymT4FQLcO0NV6xwB8f+XKlYl5eR3VYldya4OJpy/h/mlbUa5YJHaNuledkoI8FWt9j2VSxHXNV4A9GY7yIcD9HA7H98nJyTdFUVQ89ecVJEn6ypOyeJzca4AT4BBuGVomKjkJ8FdrdmDsllTEVSqO1UMySaQ3ge7lpfO5Fy9eVK4nonPJuYW8CLBzxdV9FTev8mlll64tomuUyBjFx8fjhRdecDnIonS5EWBabd69e7dCmEneYsWKoXHjxvjf//6XzXHW+fPnFSyCIChXOpUrV86bqgjKuCwabdYwc7zqup4/CbA6Cf2bSotdya0Nnki9hpbvb0RUuAnW8a4rjf0Lyselsdb3OAHmK8A+7lJBk73avs3tSuBVyAlw4HXgMwm0TFRyEuDPVv6OidsvomH1UljyfAtVMn/wwQeKp2dyiEVXHnkTaGtyYmKi4oyqQYMG3iRVFTe/Qe3rr7/GO++8g1DZ/kwVpHYQV1W5QZKINcwcr7qGxycq2etNi13JrQ3+dy0d9cauVQqRxt+PyHCTOkUFcSrW+h4nwJwAB3F31FU0tX3bW7tStWrVwoUKFYpwFz45OfmSrmAYy4wT4BBWuJaJSk4CPG3pb5i+8z+0jC2Hb565S1Wtpaamon379srKKW2DLl68uEf5XL58WdmqTI6txozxj9f3vAY1en7ffffh1KlTWLFiBWJi6Io24we1g7iRkbOGmeNV11q9naioK8U4qbTYldzaYIbdgdiR5LsL2PPmvShdlC44CK3AWt/jBJgT4NDqwXmjUdu3PbErCQkJRW/cuEE3vTwOoGxOKSRJCmOlnn2BkxNgX9RqkOSpZaKSkwBP/GE7Ptt9CffFV8SsJxsHCULfiaF2UPOdRL7NmTW8VJusYeZ41fUhTyYq6nI2ZiotdiWvNlj7zdW4nm7H1hFtUa1MEWNWTD5Ss9b3OAHmBDjkOnEegNT2bU/siiiK5DG2rSAIo2VZ/lqW5UEAogVBeFYQhNesVuu3rNSzL3ByAuyLWg2SPLVMVHIS4LELtuGrfZfRrUE0pvSoHyQIfSeG2kHNdxL5NmfW8HICHPovjvVq055MVHzbO4Mrdy12JS+dNHlnHc5dvomVL7ZEnSolgwuwDtLo1RZ1EMWvWbCIm0XMLNpTrS95PLErdNe7yWR6MikpaZMoipfsdnvDw4cPJ4ui2BdAL/cr8PzasUOkME6AQ0SRucHQMlHJSYDf+PZXLDh0BX3uqo53utUN4VrLhMaaEWMNL9cxJ8CeDmKeTFQ8zSsU4mmxK3mNM+0mbcKRf69iwYBmuKvWbTv9DF9tLI6vLI6xrGLmuL1f8ffEroiieCUsLKxOYmLiMVEUTwDoLknSTovFUlOW5f2SJBUz/OAYQACcAAew8n1dtJaJSk4C/PLXW7HUehXPtKyJUQ/G+1r0gOfP2oSFNbwsGmzWdKwXXk8mKgEfsPwogBa7kpdOun78K/468R8+79cY7WtX9CMa/xSlV1v0j7T6lcIibhYxs2hPnb1Erb49sSuiKO5zOByDk5OTN4uiSJ4C90mS9IrZbH5REIQRkiRV1a+3spcTJ8AhrHMtE5WcBPiFL7dgVfI1DG4Xi2H3WUK41vgKMN2ZzEJQa7iMWjccrzrNeTJRUZezMVNpsSt5tcE+c37HtuTzmNazPrrWjzZmxeQjNWt9Tys5MHID4Lrm9wB70n49sSuiKA6VZdlus9mmWyyWtrIsrwRAE7RwQRBetlqt0zwpi8fJvQY4AQ7hlqFlopKTAA+Ysxnrj17H8I4WDGobG8K1xgkwJ8Ch2bxZm5jphdeTiUpotpjcUWmxK3np5Plv/8TP+//BG53jMKB1aHjWz2lD9+71fpuk0duVXn3QSPXAIma+Aux931ZjV2rVqlU9LCyssSAIhyVJ+stI/SIYZeUEOBi1opNMWiYqOY13v5mb8OvxGxj1QG0806qWThIGbzasGTHW8LJosFnTsV541UxUgndk0y6ZFruSl05mbj6M91Yl4d74ipgdgrcM6NUWtWvPvzmwiJtFzCzaU627HLhd8e9YlFtpnAAHXgc+k0DLRCUnAe71yUbsPHUT4x++E080q+EzmYMlY9aMGGt4WTTYrOlYL7x8opJ9VNZiV/LSyd7jF/HwJ9tQsnCEchewyRRaUxO92mKw2EdP5WARN4uYWbSnviTAVatWLVy4cOH2NpttBZUjiuIEAFHOMgVBsN+8efPNlJSUG572RR7v9hoILSvDNZytBrRMVHIS4Eemb8DeM2mY/Fg9PNIo9M/ds2bEWMPLosFmTcd64eUE2PcEOMPuQL0xa3E1zY6fX2yF+ColQsqa69UWjVYpLOJmETOL9tSXBNhsNtM9vw9KkvRQFgG+DOCgLMvX6bsgCHEAJkqSNMVoY0IwycsJcDBpQ2dZ9CTAD01Zj0P/puOT3g3xQEJlnSUNvuxYM2Ks4WXRYLOmY73wcgLsewJMJfT7Yic2S+cw+sF4PN2yZvAZBQ0S6dUWNYgQkKQs4mYRM4v21JcEWBTFLQCmSJL0o5MAy7Jcz2azHaHvFovlCVmWB0mSdHdAOnaIFMoJcIgoMjcYehLg+yatw+HUjJC9piJn/bFmxFjDy6LBZk3HeuHlBNg/BHjGpsN4f3VongPWqy0abbrCIm4WMbNoT31MgP9xOBztk5OTD2YR4HMmk6lJUlJSCn2Pi4sTHQ7HH5IklTTamBBM8nICHEza0FkWPQlwm/fX4filDHz7zF1oEVtOZ0mDLzvWjBhreFk02KzpWC+8nAD7hwDvO3ERXT7ehuJR4dgz+l6Eh5mCzzColEivtqiy+IAlYxE3i5hZtKc+JsDXBUGob7Varbl1XlEUaQv0XkmSCgWsc4dAwZwAh4AS84KgJwFu/u4vOHPVjh8GNkejGqVDuNYyobFmxFjDy3Uc+nc969WmOQH2DwG2O2Q0Gv8LLl5Lx/cDmqFZrbIhY2f0aotGqxAWcbOImUV76mMCbJNl+TWbzfZDbn3eYrE8Lsvyu5Ikhf6dpD4c9DgB9mHlBjprPQlw43FrkXrDgZUvtkSdKqG/64I1I8YaXhYNNms61gsvJ8D+IcBUyvBFf2HRnyfQo3E1vP9oQqBNqG7l69UWdRPITxmxiJtFzCzaU18SYIvFMk2W5Q5paWmNcnp6Jg/RRYoU2SUIwjqr1fqSn7pySBbDCXBIqjUTlJ4EOOHtNbiaLmP9sHsQU75YCNcaXwEOCwv91UEWDTZrEzO98HIC7D8CvOPIefSY9TuKRYXjj5EdUDgyNMYivdqi0Qwvi7hZxMyiPfUlAa5Zs2bFiIiIvYIgpDkcjo9NJpMEQKbjv7IsvwAgPD09vcHRo0fPGG1MCCZ5OQEOJm3oLItuBDgtDXFvrUW6LGBb58KIbnEPEOIkiTUjxhpeFg02azrWCy8nwP4jwA6HjDaTNuHvC9cwtUd9PNwgWmerGJjs9GqLgZFefaks4mYRM4v21JcEmPK2WCw1HQ7HDEEQ7qWbj7LKk2VZ/gXA806P0Op7J0/JCXAItwFdCPCGJciYNASxCTOUmvpzc2+ULV0CeGUa0K57yNYea0aMNbwsGmzWdKwXXk6A/UeAqaRp62yYsk5Cy9hy+OaZu0LCxujVFo1WGSziZhEzi/bU1wTYmX+dOnXKpKenK2d9IyIikg8ePHjBaONAsMrLCXCwakYHuTQT4A1LgBGP4qopEnXaZZ7FP7ThERRxpGVKN3GxxyQ4NTUV7du3V5Jt3rwZxYsX1wGh77LQYsSOHTuGTp06oXLlylizZg0NWr4TVKecteDVSQS/Z8MaZo5XXRPjBNi/BPj4hWtoNXEjBAHYNbIDyhaLUqe4IErFWt/TSg6CSHVei8J1XR+sHKPSQvy5XfG6a+megBNg3as0eDLURIDtduChO4CzJ3AhogQa3vOdAuzIuodgUo4iCEDFqsCyox5th540aRJmz56NJ598EiNHjlTy2rVrF7799lvs378f//77L6KiolCrVi383//9Hzp06KCqIimvqVOnYs+ePYon59q1a2Pw4MFo0aJFvvktWbIEK1euhM1mA5H1UqVKoUaNGnj99ddRp06dbGmPHz+OTz75BL/++isuXryIEiVK4M4778TkyZNdxH7QoEFYt24dxowZg549e6rC4s9ELBpt1jBzvOp6FJ+o+JcAU2mtJm7A8QvXQ8YbNGt9jxPgvahfnxNBdSOusVKp7dvcrgRez5wAB14HPpNAEwHetQl4rq0i26mocmjeai4iHemQNnTLLu/MjUDjNvliSE9PR6tWrRRiuXjxYtStW1eJ/9FHH+Hjjz9GsWLFUKVKFRw9ehQUl8KUKVPQuXNnr+omMTFRIZs3btxA6dKlERkZiTNnzihvI2fOnInWrVvnmV/fvn2xc+dOlC9fXpGHZKFQuHBhLF++HNWqVVO+03Mqg4gv/Va9enVFZiLFRHgrVaqkxFu9ejVeeuklOseBZcuWeYUjEJHVDuKBkFWvMlnDzPGqazl8ouJ/Atzvi53YLJ3DhO510atpdXWKC6JUrPU9ToA5AQ6i7udTUdT2bW5XfKoWjzLnBNijajJmJE0EePV8YFTvTNJXuAratpiF4hlXsX9Tj+yVMf47oFOvfCuItjwPGDBAIZe0auoMRBJpK3Tz5s0hCAIOHjyIRx99FA6HA+3atcOMGZnnjj0Nzz33HDZu3Ijo6GiFdBYqVAi9e/fGX3/9BVEUFSKbV5g7d65C0mNiYpQo9H3ChAnKZ1oFfuqpp5TPtDpNGO666y6FvNPqLwUi3eHh4cofBRrcGjdurKxCfrRtbAAAIABJREFU//zzz658PcXi73hqB3F/y6lneaxh5njVtR4+Ucleb1rsiqdtcMzyg/hyWwr63FUd73TLfGFq5OApbiNjzE12FnGziJl0z3F7t+LP7UrgRztOgAOvA59JoGWiArcVYKloddx396cok/Yfdm/pk11eD1aAaWvwrFmzCiS1siyjSZMmuHz5snKGdtq0aR7XTUZGBho1aqQQ0R49emDs2LFKWlr5pdVkClu2bEHFihU9ypO2UhMZpzB69Gj06dMH//33n0J8Sc777rsPSUlJytZts9msrPbm3GbdpUsXWK1WRRaSKZgDi8aLNcwcr7oeyCcq/ifA3/x+DKN+OqAUvHRQC9SrVkqd8oIkFWt9z1ntLOJmETMnwN6v+HO7EvjBmRPgwOvAZxJoIsCuM8AncaB4LTx41zRUuvEvfv81cyXUmzPAL774ouIMirYZjxo1Kk+8ixYtUn6n1eA5c+agZcuWHtfNuXPnXPEHDhyIIUOGKGmdedLnBQsWKOdyPAnDhw9XVpFLliyprOCWK1cO+/btw2OPPeZKXrVqVVy9elXZ2k3brOfPn4969eq5fneuSD/99NN49dVXPSk2YHFYNNqsYeZ41XUvPlHxPwE+e/kGHp/5G1LOX1PI748Dm8NkMu50hbW+xwmw94RI3egUPKl4G+crwMHTGj2TxLgWxTN8TMfSRICp5rK8QO8uYUH3ppNQ7do/2Lr9mVtXknnoBbp///7Yvn07iBAOHTo0V53Q2eC33noLtJL72muvgdJ4E86ePatsYaaQFwFeuHBhNoKaW/5UPjmuori0hZpWrmnVl8Lu3bvRq1fmdm/atv3FF1/gypUrisMuOhPcrVs3vPfee65sX3nlFWXb9eOPP45x48Z5A8fvcVk0Xqxh5njVdStOgP1PgKlEIsFtP9iEq2l2jOhkwcB7YpSXo0YMrPU9ToA5ATZiP1Ujs9q+ze2KmtrWN40xrYm+dRCyuWkmwFkk+LcZH6NXzDDEXD2O9b8NBCpWA4ZN9fgKJPLCvHbt2mweoJ2VTtuJyWszbVWm87NEPp1bj71RjPsWaHfC6c0WaCKztHK8detWlC1bViHr3bt3d7n0P3nypLKNmwKdaR42bJjymbY37927F02bNsW8efNcYvMVYG806P+4ag2X/yXVp0SOV1098olKYAgwlTpry2G8+3OSIkC3BtF4/5EERIab1CkygKlY63ucAHMCHMDu5tei1fZtblf8qqZcC+MEOPA68JkEuhBgurf30Gn0+3o34orasfr+EkCDVh5dfeQE5rwCiVZK6fogZ0hLS1McTK1YsUJxhjV9+nRlZTVncHqLpud0pjavkNMJFl2rRE6waOuyuxOsESNGKM8SEhIwceJEJTvyFk2kls71xsbGKg64aFt1zqsMOnbsiJSUFOW87+eff65sgab7jWkFmLZHjx8/3iWe8wzw22+/7Vo59pmyNWasdhDXWGxAk7OGmeNV19z4RCVwBJhekH7+61G8tyoJGQ4ZIzvXxv9a11KnyACmYq3vcQLMCXAAu5tfi1bbt7ld8auaOAEOfHX7VwK9CPAvB0/jf/N2IyG6JJYN9vxcrhMteWYmclq5cmVs2rTJVQmfffYZPvzwQ+U7OadyXiFE38ljtJMse0qAibzSaqzzGqSIiAjQ1uic1yA5rzxyX7F1enemsskTdNGiRRVPzkWKFEGbNm1A9/pSoJVsOtNMEzO6GokI8IULF5R4tI3b6UX6+vXrilMuGhyJ4JOjrGAOagfxYMZUkGysYeZ4C2oRuf/OJyqBI8DOkuduO4q3lx9CtTKFsWFYG0SEGWsVmLW+xwkwJ8DqRlvjpVLbt7ldCbyu+Qpw4HXgMwn0IsA/7zuJ57/bi8Y1SmPxwNtXaAsCQCu9dD6XVkl/+ukn1K5dW0niTmxz5kFXGW3YsEF5TKuqtLW4oKuMKC6t7NKW6j179ijkMz4+Hi+88EI2h1q5EWDns9yw5Dzbu379emWFWJIkZeWaVolffvnlbFcdkdMvIsqeyFxQ/fnjd7WDuD9k81UZrGHmeNW1JD5RCTwBvp5mR4v3N+DC1TS8fK+IF9sH9wvFnC2Ntb7HCTAnwOpGW+OlUtu3uV0JvK45AQ68DnwmgV4E+MfdxzF04T7cXasM5g+4W5W8H3zwgeLZmZxbkZMrbwIR0MTERMXLcoMGDbxJqjqu2kHNWSCR7l9++QVG2P5MMmvFq7qiA5iQNcwcr7rGxicqgSfAJMGS3Sfw8sK/QH6wZvRphE53VlKn0ACkYq3vcQLMCXAAullAilTbt7ldCYi6shXKCXDgdeAzCfQiwAv+OIZXfziAe8Ry+OrpTI/I3ga6KojOypIXT9oGTSunngS6E5i2KpNjK3KQ5a+gdlAj+Y4dO6bcY0xbvlevXo3IyEh/ia26HC14VRca4ISsYeZ41TU4PlEJDgJMx07eXHoA3/z+t0KCp/aoj671o9Up1c+pWOt7nABzAuznLhaw4tT2bW5XAqYyV8GcAAdeBz6TQC8C/M1vKRi19CA61K6AOf2a+EzeYMpY7aAWTBi8kYU1vFQ3rGHmeL3pEbfi8olKcBBgkiLd7sDwRX/hp72nEBEm4PHG1fDcPTGoVqaIOuX6KRVrfY8TYE6A/dS1Al6M2r7N7UrAVQdOgAOvA59JoBcB/uLXIxi7IhGd61bCp30a+UzeYMpY7aAWTBi8kYU1vJwAh3nTPAwZV682zScqwUOAlX7rkPF/X/2BTdZzimB0LdLw+yzo3+IOhAepcyy92qLROiKLuFnEzKI91fqSh9uVwI9mnAAHXgc+k0AvAvzZ5mRMWGXFw/WrYGpP/5zB9VmleJgxa0aMNbwsGmzWdKwXXj5RCS4CTNLQduhfk//FjE2Hsf3weUXAWuWKYsi9Ih5KqKwctQmmoFdbDCZMnsjCIm4WMbNoTzkB9mQECO44wWUlgruuDCedXgT4o/USJv9iw+ONqmLiY/UMVw9qBGbNiLGGl0WDzZqO9cLLCXDwEWCnRESEv9v5NyatsSL1WrryuMkdpfHBo/VwR7miaoZ+n6TRqy36RDgfZsoibhYxs2hPOQH24cDhp6w5AfZTRQeiGL0I8IdrkzB9w2H0aVoN73RPCAQUv5fJmhFjDS+LBps1HeuFlxPg4CXATsmu3MzAF78eVVaEr6fbUTQyDO1rV8QDCZXR1lJB2SYdyKBXWwwkBjVls4ibRcws2lNOgNWMCMGVhhPg4NKHrtLoRYDf+/kQZm45iv7Na+CtLnfqKmOwZsaaEWMNL4sGmzUd64WXE+DgJ8BOCU9evI6h3+/FzpQLLqErFI/Co42qom1cBTSoViog54T1aovBai/zkotF3CxiZtGecgJstNHodnk5ATa+DvNEoBcBHrf8ID7floIBrWrijQfiQ7jGbkFjzYixhpdFg82ajvXCywmwcQiw0q8dMn47fB6bpbP4YfdJXLia5gJQpmiksiJ8Z3QJ3BtfEVVL+8d7tF5t0WjGl0XcLGJm0Z5yAmy00YgTYONrzAsEehHgN3/cj3k7/sYLbWPwSsc4LyQwblTWjBhreFk02KzpWC+8nAAbiwC7S3szw47VB/7BhqSz2Cydw8Wsc8LOOOWLR6Fh9VKIq1QCd9UsA3PF4ihXLFJ3J1p6tUWjWVQWcbOImUV7ygmw0UYjToCNrzEvEOhCgO12jJj9Cxam2PFyfDhe7NMBCONXqHihBkNEZdFos4aZ41XXFTkBNi4Bdpc8w+5QPEf/eSwVO45cwK5jF+CQb28TxaLCEVO+KGpXLoG4SsUVUlytdBFULlUIESqvWWKt72klB+p6anCk4rqujzAG5oha2zi3K4Hvr3wLdOB14DMJNBPgDUuASS/h5XKPYUmV9njd9gWevb4TeGUa0K67z+QOhoxZM2Ks4WXxjTVrOtYLL5+ohAYBzmlX/ruejuSzl7H72EUcOPUfdqWk4tR/1yHnQoopLd2sVLF4IYUIly0ahYolohBTvhjKFotEdKnCoO3V9FeiUARMpuxTK73aYjDYRm9kYBE3i5hZtKecAHszEgRnXE6Ag1MvukiliQAT+R3xKN24iBfuHIEVlVrjLetn6H98RaZsExeHNAlmzYixhpdFg82ajvXCywlwaBLg3IzsjXQ7jl+4BunMFST9cwmJpy/hyL9XcSL1OtIyHB7ZZeK+pYpEolSRCJRR/keiVOFwpF1JhXhHNEoUjgStMit/hcJRPCoCpYtGoFyxKBSKCK3dVXr1QY8qPkgisYiZRXvKCXCQdDgNYnACrKHygj2pagJstwMP3QGcPaFAfDbhDayp0BzjEz/BEydX0btwoGJVYNlRj7ZDp6amon379kpemzdvRvHixYO96qDWiB07dgydOnVC5cqVsWbNGkRERAQ9VlaNl1odG0KhuQjJ8arTHCfA7BDgvFqIwyHj/NU0nLp4Haf/u658JlJ87PxVnL+SpqwcX7yajss3M9Q1sqxUhSPCUCQyTCHChSPDQN/pLyrClPk561nO3wu5xS0cacpMnxW/UHhmOueziDBB9zPOeYFmbcxh1ZZy3HtRv753W7+5XdE0VOqSmBNgXaoxODNRTYB3bQKea+sC1b/+W9hYrgkmHpyKx0+vuwV25kagcZsCwU+aNAmzZ8/Gk08+iZEjRyrxb968iTfffBP79u1DSkoKZFlGvXr1sHDhwgLzyyvC/v37MXXqVOzZs0chsLVr18bgwYPRokWLfPNcsmQJVq5cCZvNBiLrpUqVQkJCgkLau3bt6jrPMnnyZKxbtw5nzpxBRkYGypYti2bNmuGFF15AdHS0q4xBgwYp8caMGYOePXuqxuPPhHyiElorL7m1HdZ0rBdePlHhBNjTsZicbpGjrdRraUi9mo6L9P9aOs5fuQHp2ElEFSuNa+l2XL6RAbq7+MqNDOUzeapOs3u2wuypLHnFCzMJIBJM55kjw0zK/4hwwfWZ7kxWnt0WJ/OZK02OdM40t9KbECbIOHn8b4gxtRAVGX5b+sjwLDlcZWaWEW7K/C/QvnODBb3GHYPBVr1oYDScOeVVq29uVwKveeONLoGvM8NIoJoAr54PjOp9i9DVfRUrK7bCnL1j0eHfnbfwj/8O6NQr3/pIT09Hq1atFGK5ePFi1K1bV4l/6dIlNGnSBBUqVMD169dx+fJlTQQ4MTFRIZs3btxA6dKlERkZqRBVcsYwc+ZMtG7dOk85+/bti507d6J8+fIoVqwYjh49qsSNiorC8uXLUaNGDeV7t27dFBxlypTBlStXQKu9FGrWrInVq1e78qfPL730EiwWC5YtW2aI9qJ2EDcEuDyEZA0zx6uutfKJCifA6lrOrVQF9T16AXzpRoZCmG+kO3A93Y7raXbQlmznZ/pP3289y4ynfE/LipfjO+VFv19Ly8jV4ZdWXL5OT9vJw4lomwRkEneT63+4QpQzn9HnMCLNWfHy+k7xlT8lfibRzvzv/J6Zz63fczzPikdpsqXLKp/yESDj6OFkxFlERIaHZ8ZzlZe9XGc+zrxMQubvhNto5L+gNu7rthKo/NXi5nYlUBq7VS4nwIHXgc8kUE2Ac6wAHy9UAb+VTkC3fzYiQrbfkteDFWDa8jxgwACFXP7666+utDRonD9/XiHATgKqZQX4ueeew8aNG5WVWCKdhQoVQu/evfHXX39BFEWFyOYV5s6dq5D0mJgYJcpXX32Fd999V/n86quv4umnn1Y+06o1kWJnGD58uIvg/v777wrxpkADW+PGjZU3oj///LMrX58pWoeM1Q7iOhQdsCxYw8zxqmtqfKLCCbC6luM5Adaaf0HpiWCn22UXYabzzOl2+qPnDmX1OV155vZd+Z2ey8rvt9JQ/Mx4mWncvmelScvIyodWxC9dRmShIq68lfgZt6ehsnm4VQMKORYEmExQiDqR4kzCTKQdWb9lkndyukZx3Qm1M31uz1zxiZi7SLdbXln5O8tyxncvi545XxwQcSfif/rUSdSoXg0RYWGKTNllI/kzZXcSfXcZXfFz4nDK58zP+Tv9z5Kf6siJw98vDtTaVW5XAt/bOQEOvA58JoFqAuw6A3xScYJ1e/D8DDBtG541axbatWuHGTNm5IpVKwGm7ciNGjVSVn979OiBsWPHKuXQyu+UKVOUz1u2bEHFihU9quuDBw+ie/dML9e0TfuJJ55wpfv+++/xww8/4L///nOtAMfGxmLFihXZ3th26dIFVqtVkYVkCvagdhAPdlz5yccaZo5XXWv150RFFMXnAQwHUFkQhIOyLA+RJGlrXpKbzeZHBEEYB4De3h0GMFKSpB/d4guiKL4FYACA0oIg7LDb7YOSk5MPOuPUrVu39M2bN6cD6JL1bFlaWtrglJSUi7mVq9quAHybpJfnBNW12OBJ5c2Y4yTpRJAziGA7HLA7Mol05n/Z9T3DIYOutcr8nxmX/lM8+stw+05x7PQ9K27O73Y5K42SPjNeZh7u/2/lf+t5ply3ynTm48DV6zcQHhGZSz6ZsjnxBI+mQkcS54sCItkuAp7jBYGTgOdG6DPJvvNlg3NF/taKv2uV3iSgSEQY+jevgev/HOZngA3YhDgBNqDSPBVZy0QFLi/QVJo7Cc5qMh56gX7xxRcVZ1BEckeNGuUTAnzu3Dm0bNlSyXvgwIEYMmSI8nnRokWuMhcsWKAMUJ6EESNGYOnSpcp2aFrBdSfORKiJWDtDfHy88j0nuXauSNPqMa0iB3vwZqIS7Fg8lY81zByvpy0jezx/EWBRFOlN2TwAz8uyvI38DwqC8ExGRkb8kSNH/s4pfWxs7N0mk4nI8ZsAiPR2AzBWluWWNpttB8W3WCyvyrI8Upblp8LDwyW73U6DcGtBECxWq/UyxRFFcZUsy1VlWSaSTC/yZgmCkCJJ0kOcAKtrMzlTsdb3nPhZxO0NZnKuRoTaId8i3O7P3Am2k6jnfOaK7yTtsgx6RvGy5asQfSgkX/mfFU8pn+JnvQhw5k9p8yzL9VtWfjK93HLgfGoqihcvCTrN7iqfXiw4ZcqlDPfyHQ5kvSCAInte5evTK/XJ5anmNfBQ9E1OgPWpTr/mwgmwX6vbv4VpIsAkatY9wE5v0Ir0FasBw6Z6fAVS//79sX37dhAhHDp0qE8I8NmzZ5UtzPkR4P9v78zj5CjKPv483bObELNCECFZQEIyXb27IQiCnArIIYfI4YEcIiByiPoKAoKgwqu8comASnhFUDnkEJXD9xUEUQQ5ggSQI7tdPQmLCeHlECJHjp3pet7PM+leeyczu3PtzOz0U/9Adqqrq75V3dW/qqeeh51rsYn1aIl3ktlxFeedMmUKnHrqqXDYYYetFdSdJzg+/3vuuefC/PnzYbvttgM2o44Hfz/ttNPyZteHHHIIfPe7vEHT2qmSSbu1W1J+7ZLWZmlv+WMjnrOBAng+Ij7hed4Xo/srpfoB4Hat9TcKa6+UugUA3q213jf6zXGcuxHxDa01O2fg3d9liHiZ53kXcp50Oj3JsqyXiegM3/d/4jhOLyIuJKIdItGcTqd3sCzrEUTs8diMpSDVMq8kbQwmWQhy25PY30lsc6P7ei3BHhP9edEciudIUK8R02tE+YjfeeEhWLNQMPz3UQR7lCdaYLBtC/bu2xCW+AtFAFc3vTb1KhHATcU/vjev5UNluGZBAMGC++GFBY/CZtvsAPY2u5UV+ii6nr0w33PPPSM8QBe2up4m0HHBWYkJNDu14p3jBx98EDbYYAOYN29e3tPzaK7t77vvPjjpJLZYBLjmmmuGd6H537IDPL5jux6lJ+1DRdpb3ahphADu6+vrzOVyKwDg03ETZtd1LyeirbTWuxYRwLwrfKnWes05jzW7uacg4sme523mOM4sRFxkjPlAJpN5MiaS70DE5Vrro1zX/TwR/UBrvV68fKXUciI6xff9n4sArm7cxK9K2rOXZOEvfV1ZOKDan67mllBtfzdiXmkumda/uwjg1u+jqmtYFwFc4ypuFAJpzz33hCuuuKJoW0YTwD/60Y/gxz/+cf66IpsRw+UVOsFiZ1XsBIvDLMWdYLF5M/+NwxxddNFF+evZWzQ76hoYGOAdkvyZ5enTp8NTT/07thuHalq0aBF85CMfAcuywBiTP99700035cvgOu61117D9YnOAPMuMe8it3qq9iXe6u0arX5Ja7O0t7rR2ogPFdd1u4noRcuydh4YGHg4JmjPAoCjtNZuYe2VUkNs2uz7/o0xcXs4Iv5caz2pp6dnJ2PMQ4i4sed5y2JlsonzZp7n7a2UOissQ8XLdxxHh+WcX0oA83uVLWUqSTwGOVwdRwOIW8xUUsZEzCvtTk5/S18np6/5XVRtf/O8orXmIrbZdtttn5iI77WJXmcRwBO9B0epfysIYPbMzOJ0xowZcP/994+obSQYWYCyh2UOXcTCk9MNN9yQP1dbrgBm8crOpqIwSB0dHcCm0YVhkCKxzWbL11/Px+0Ajj322GEP1ewJms/+skMOfkHtu++++Ti/bOrMcYz5g2/TTTfNe7B+7bXX8tdznTmOMF/HicM6sVMufjGycyzHcVp+lCVNHEUTV3yRo+U7qcYKJq2P69XeRgpgY8xOmUzmkZigPRsRj9Ra95QQwEf5vr9mFW7Nmd8jiOgarfXkSADbtt3d39//UpRHKfVTItrU9/19WAAXE9hKKZ8NW7TWF5QSwDUOR7lcCAgBISAERAA3bQyIAG4a+vG/cSsI4KGhofz53OXLl8Ptt98Ovb29ww3nOLmlEpsXb7LJJnDeeeflhepYoYy4HN7Zveyyy+DJJ5/Mi092UMXiNXKQxXmKCeDob8XqctBBB8GFF14IS5YsgQsuuADYQzSLXxbILOp33HHHvBl0JNy5DHb6xc6/yqnz+I+C8u5QL7FQ3t1aI1fS2iztrW7cNUIAT0QTaNkBLn88VbtLVP4dWjNnEtudxDbz6JN2V7bzLTvAzX9niQBufh+MWw1aQQBz4y6++GK4+uqrgR1inXnmmRW19+CDD4b+/v68qfHWW29d0bW1ZK5FLLDovvfee/NOsiaC+XM0eSVpNzSJba5lTNfyLDXr2nq1txECmBkppeYT0QLf99c4Fljzt4UAcMcoTrC6tNb7xfLfBQB8vjfuBOtSz/Py5z1Cof1KoRMsANhea/0Y53EcZ3tEfFScYNVv5NZrLNavRo0pKYntTmKbkzifRk9Qtf3dqHmlMU/6xLyLCOCJ2W9l1bpVBPAbb7wBe+yxRz5OLptBd3V1lVX/t956K+9hmR1bsXfmRqZqX2rsHXqfffbJ7w7ffffdebPuiZCqbe9EaFupOiatzdLe6kZroz5UojBIiHiiMYbNoI9HxONs257T39//glLqOgB4MRLDoYnzA4jIYY7YsdWBRHRekTBI7EH6GNu2/SAI2OR5t8IwSADQbYw5gQlZlnUVALwgYZCqGy/Frkras1erOKgf+caXJH0tTrDKGXWNmlfKqUtS84gAbuOebxUBPBERJ20SS1p7k7hinbQ+rld7G/mhopTi3d+vA8AMRHzWGMOemB/g8eq67v1ExPF5j47eqa7rfopFLwDMAoBFHPPX9/3fxt65HArpHI4pDADTAGC+bdtf6u/vfzbKM2fOnPWHhoZ+iIgH8N+I6M5sNvvlwcHB5cXe3bXMK/Xqk4k2p0i7kyOKpK+T09e1fEc0cl6ZaO/LRtVXBHCjSDfhPrV8qMSrm8QXetLanLT21jJxNeFRrsstk9bH9WqvfKiMHH61zCv16pO6PBANLETanRxRJH2dnL6u5TtC5pUGvoBL3EoEcPP7YNxqUMuHigjgYEQYpHHrpBYpOImTdtLaLO2t7mGTDxURwNWNnH9flbRnL2p5EtudxDbXIgRrfbaafX21/S3zSrN7DkAEcPP7YNxqIAK4erTVvtSqv2Nzr0xae5M4YSetj+vVXvlQEQFc69u5XmOx1no0+voktjuJbU7ifFrrIo/MK41+G619PxHAze+DcauBCODq0SZtEktae5M4YSetj+vVXvlQEQFc/Uyy5sp6jcVa69Ho65PY7iS2Wcb4U7DVVpWZfsu80ui3kQjg5hNvYA1EAFcPO2mTWNLam8QJO2l9XK/2yoeKCODqZxIRwBJer9bRMzGur9f7dmK09t+1rLbdMq80v6dlB7j5fTBuNRABXD3aal9q1d+xuVcmrb0igO3mDrgG3L1eY1o+VEQA1zpc6zUWa61Ho69PYruT2OYkzqfRs1Rtf8u80ui3kewAN594A2sgArh62NW+1Kq/Y3OvTFp7kzhhJ62P69Ve+VARAVzr27leY7HWejT6+iS2O4ltTuJ8KgK40W+T+t9PdoDrz7RlSnzsscd2sizroZkzZ8I666xTdb34ha61BqUU2Hb77xxFL/MktVn6uP3HddL6uF7tXblyJQwODoIxZuftttvu4apfpG1yYS3zSr36ZKKhlHbLt8NEG7OV1lfGeGVjXOaVSkdY/fOLAK4/05Yp8fHHHz8cAH7ZMhWSiggBISAEJi6BI7bddtsbJ27161NzmVfqw1FKEQJCQAgAgMwrTRoGIoCbBL4Rt50/f/57bNve2xgzaFnWqkbcU+4hBISAEGgnAsaYyZZlzQyC4A/bb7/9P9upbdW0ReaVaqjJNUJACAiBfxOQeaX5o0EEcPP7QGogBISAEBACQkAICAEhIASEgBAQAg0gIAK4AZDlFkJACAgBISAEhIAQEAJCQAgIASHQfAIigJvfB1IDISAEhIAQEAJCQAgIASEgBISAEGgAARHADYAstxACQkAICAEhIASEgBAQAkJACAiB5hMQAdz8Pmj5GiilTgKA0wFgBiI+R0Qna60fbPmKF1TQcZxdEJHbsQ23BQAO1lrfHsuGSqlzAOB4AJiGiPODIPhSJpN5Lsozd+7caatXr/4hABwQ/u3OoaGhrwwODi5vNR5KqW8AwCcAoAcAVgLAw4h4hud5XlTXdDo9ybL6fx8yAAAgAElEQVSs7wPAYYjIsbLuC4LgpEwmszTKM2vWrPelUqkrEHF3IuJybkylUqctXLhwqNXa7LruF4noiwAwM6wb9913tNZ38b/brb2F/MM+/x4RXe77/snt1mal1LkAwM9oPL2stZ4e/qGtnuFWe75qrU+7zCXMIWnzSdT3Mq/kSSRqXino+7adX2LtlHmm1pf9BLheBPAE6KRmVlEp9RkAuB4ATiKihwDgBET8Qi6X61u8ePE/mlm3Su+tlNoXEXc2xjyBiL8pFMCu655BRGcT0dGpVEoHQfBNAGDR7Hqe9xbfTyl1FxFtQkQskgERr0LEQa31xyutz3jndxznbgC4mYj+ZllWCgD+CxHnTpo0qe/pp59+h+/vuu6VRPRxbjMR/dOyrEsAYH2tNS8SBABgO47zFCK+aow5FRHfg4jXAsBvtdZfGe82VFp+Op3+uGVZgWVZGb7WGHMUL94YY7bmhYx2a2+cj1LqgwDwKwB4k4j+HAngdmpzKIA/lUql9ozaPjQ0FGQymVfD8dxWz3Cl47+V87fTXBLOBYmaT6KxJfNKsuaVmChs+/mlQADLPNPKE0od6iYCuA4Q27kIpdR8RHzC8zzeVcsnpVQ/ANyuteYdxgmZlFJUIIB552gZIl7med6F3Khwt/BlIjrD9/2fOI7Ti4gLiWgH3/fnh3l2sCzrEUTsie+stiKUdDr9XsuyXiGiXX3ff2DWrFnrplIpFg5Haq1vCQVENxEtsSxrv4GBgT/wogEA/A8ibup53rKw/w8FgF8YYzbMZDJvtmJbC4Th67zzn81mf92u7e3r65uazWaf4IUqy7K+aYx5igVwu/VxKIAP0lpvVWTctf0z3OrP2mj1a9e5JHwnJm4+ifpa5pX2nVeiPk7K/FIggGWemcgTThl1FwFcBqSkZunr6+vM5XIrAODTWuvbIg6u615ORFtprXedqGwKBbDjOLMQcZEx5gOZTObJqF2O49yBiMu11ke5rvt5IvqB1nq9AoG1nIhO8X3/563MY/bs2Wnbtn3btuf29/c/67oumzTfN2nSpPWfeeaZN2Iv/7+HCxznOI7zHUQ8UGv9/uj30AycReXunuf9uYXbbCulPg0A1/IOsG3b09u1vUqpa4nodd/3T3Fd9/5IALdbH7MA5sUMIvoXAKwGgPlEdJbv+4uT8Ay38LM2atXaeS4pJoCTNBZlXmnfeSX2TZCI+SUugGWemaizTfn1FgFcPqvE5XRdl3cDX7Qsa+eBgYGHYy+HswDgKK21O1GhFArgnp6enYwxDyHixtFOZ/hhwybOm3met7dS6iw2FfZ9X8Xb7TiORsSfa63Pb2Ee6LruHUQ0TWv9Ya6n4ziHh/WeVCDo7wGA57XWJyilruLztFrrjxbkWR2yuKnV2tzT0zPXGPMIAEwGgLcB4HCt9e/btb1KqUPZdD+bzX5wcHBwVVwAt1ub2SKBiKbYts1HFDbi3W4i6rEsaw5b9Lf5M9xqj1rZ9WnnuaSYAE7AfBL1vcwrbTqPxr73EjO/xNos80zZb/eJm1EE8MTtu3GvefTRYozZKZPJsKDIJ8dxzkZENptl50oTMpUSwLZtd/f3978UexH+lIg29X1/HxbAxYS/UsoHgGu01he0Kgyl1BUA8DFjzIciB1ejiKN7eTfc87wTWQBHCwAFApgdYH1Oa31zq7WZd5tWr17NjrvWI6JPAsAXjDG7IuJWJQT/hG3v7NmzN7Vt+3EA+KjWmnfu+Vz38A5wu/ZxNOa23HLLd61atWoRAFxkWdajLIDb9Rluteeskvq081wymgBu97Eo80p7zivRs530+UXmmUre8hMvrwjgiddnDatxO5utJckEWin1IwA4CBF38Tzv+WgAtZt5bKkHQyn1RwBYhIi3tJsJtFLqIADg4wnssCxKNgDwmUQDAHsDwB/b2MydfRLwAkbGGHNxux9jaNjLv843aue5pJgAToIJtMwr7TuvxDYAEj+/xFjIPFPneaHZxYkAbnYPtPj92XEJES3wfZ9DIeWTUmohANzRpk6wLvU87yJuZ/jRxk6jRjjBAoDttdaPcR7HcbZHxEdb1AkWOwVi8Xuwbdu79ff38071cIocJCHiZz3PY+/B0NvbOyMIgqWFTrBs294k2hkPvbnyudqJ4gTrPgBYksvlvspOsNqpva7rdlmWtVm8X4Mg4LPoA7ZtX7h69eol7dbmeFtDR3W8A3yV1vq7oSO7dnqGW3yGKL967TqXFBPAHCCgjceizCtrvoPadl6JLZInen6JOMg8U/57fiLlFAE8kXqrCXWNQlcg4onhucrjEfE427bn9Pf3v9CEKlV9y9CTYZoLQER2dPU1DhcTBMHrHNIpDIPEnq2PYWdRQRCwyfNuhWGQAKDbGHMCl2NZFp+RfaFFwyDNQ8TDjTEHdnZ2Dsf+ffPNN/+1dOlSjucbhUHaHwCONsa8HsYEfk+RMEgvG2NOtyxrffYAHTrJarkwSEqp7wHAXZZlLQGALmMMe6w+k4j28X3/3jAkUNu0t9jDEDeBbrc+Vkp93xjzO9u2/8ELMIjIocp2DR27vdBuz3DVL7sWvLCd5pJwgZQ9rydmPomGlOM4Mq8kcF6JieLhIzbtNr9EbZR5pgUnkHGokgjgcYDabkUqpXj39+sAMAMRnzXGsMfjByZaO3t6enYzxhTzWnyt1vrocNX+HI51DADT2MOsbdtfYo/JUVvnzJmz/tDQ0A8R8QD+GxHdmc1mvzw4OLi81XiEZt5rVYuIjvF9n0UszJw5c3JnZ+fF7CgKANYBgPuCIDhp0aJFLCDzadasWXyedh4A7A4ALJxvNMaclslk2AtvSyXHca5BxD14rAIAewp+moguZPHbju0tRwC3Ux8rpfjM+S4AsAEAcAgvPvf7rUwmw1YpnHh3qm2e4ZZ6uOpQmXaZSxhF0uaTmDjg4xUyryRsXok6vHCBtZ3ml9gYl3mmDu/7Vi9CBHCr95DUTwgIASEgBISAEBACQkAICAEhIATqQkAEcF0wSiFCQAgIASEgBISAEBACQkAICAEh0OoERAC3eg9J/YSAEBACQkAICAEhIASEgBAQAkKgLgREANcFoxQiBISAEBACQkAICAEhIASEgBAQAq1OQARwq/eQ1E8ICAEhIASEgBAQAkJACAgBISAE6kJABHBdMEohQkAICAEhIASEgBAQAkJACAgBIdDqBEQAt3oPSf2EgBAQAkJACAgBISAEhIAQEAJCoC4ERADXBaMUIgSEgBAQAkJACAgBISAEhIAQEAKtTkAEcKv3kNRPCAgBISAEhIAQEAJCQAgIASEgBOpCQARwXTBKIULg3wSUUj8HgM8ZY6ZlMpk3AcBSSr2BiPd6nvepSlgppQYBYDPLsjYfGBjg/5dUQCBiNAqYF7TWMwWcEBACQkAICAEhIASEgBAQASxjQAjUmYDjOM8gImqtt+Ci0+n0HMuynkXEMz3Pu7CS2423AHZd934i2rXZAruWekSMEPE3RPR2Eb6vaa1Pq4S75BUCQkAIlEMger8DQJBKpdZbuHBh9A6ylVLLAWCqMWanTCbzSDnlFcvT09Mz0xjzfPjbtVrro6stS66rnkBPT89uxpg/l1sCER3j+/4vys0v+YSAEGgcARHAjWMtd0oAgS233PJdq1at+hcRXev7/rHcZNd1P09E1yDi7p7nlT158rUigMceNOPNaOwaSA4hIASSSsBxnBMQ8b+JaIHv+9tGHFzX3ZaI/gYAq1Kp1LoLFy4cqpZRKwjgaJESEf/ied5u1balkdfFudVDjIoAbmTvyb2EwPgSEAE8vnyl9IQRUEp9GAAeIKITfN+/KhSxPwGA44wx64Um0WVTGW9xV8vOa9mNKCNjLfUYb0ZlVF+yCAEhkFACSqnrAeCzRHSZ7/unRBiUUl8DgEt4PtBa71oLHhHA1dEbTwFMRFemUql5o9Vs5cqVSwcHB9kKQJIQEAItRkAEcIt1iFRnYhOIffRspbX+eyiAnwKATq11X6Wtq0TczZw5c72Ojo4jLcvaHwB6iGg6ALwd7kJc4vv+vdH9Cz6o1qqW1jr+bkCl1OcAgHe0twSASQDgE9F17373uy9fsGBBNl5AVGcuQyn1BQD4DwBwAOBfAHDH0NDQGdFHQYX1KIqvEkZcgOM4uyDiZ4iI/7sJAEwGAD5ffbsx5vxSixS9vb1bBEFwOgB8BAA2AgD+sBkgol/7vv+jeOV6e3s3C4LgbADYJ8qLiA8i4vkDAwML4nlnz56dtm3bB4D7crncJ23bPhcRDwaAjQHg8tB8Gx3HOQwRTwpZrgsArwCAh4i/9TzvykrHluQXAkKgdgJKKTZNnklEn/J9/zcxAXwbABwEAN/TWvO7oOokArg6dOMpgAHgP7XW51ZXM7lKCAiBZhMQAdzsHpD7T2gClZpEhY0t2ylTJeLOdd19iOguAFgCABoAXkPE9xHRDnxfRPyC53k/C0X5BgDw/UigFZ6fjZ0xsxzHuRkRPw0A7NCLTfr4jNv2AMAC+/da648DgIl9+OUddyHixUT01fCaVwFgJwDYEAAeDHdESClVbj1KjpNKGIUC+HFE3IKInrUsa5CIWABvE7bnmbfffnuHZcuWrSgQ9YcCwLW8kAEAzwDAc4i4PhHN5TZprVNRfsdxtkLE+wBgfQDoB4C/I+JmRLQjAGSJ6FDf938b5Y8EMCI+QkRTQ+H7F/6diJ7wff88pRT31alhH/yV/4uIM4iIz5kv11qnJ/SDJJUXAhOQgOu63UT0Ilc9m81Of/7551+OvQf5ncfvt/201vxerjqJAK4OnQjg6rjJVUIgCQREACehl6WN40bAdV2XiHhXkMUK+746BgB4RyB/1hcR0+xkCgDuBoD8hxIRveb7/pnlVKoScee67uYsigYGBh6Ol51Op7e2LOtP7I06lUptHHPSwueTR3WC5bru14mIHXfda4w5IpPJ8Ecd8Fnn1atX30RELH6/rLW+olAAA8BLRLSf7/u8A87nmfljkB3BpBFxD8/zuE75NFY9RmNVCSMux3Gcj2Wz2YfipmnpdHoSIs5DxM8DwNla6+/F2tMDANwG7t8jPM/7daw+vECwr+/7/xv+jT1+s0DuQ8TzPM/7Nnd52P7PAMBNAPCWMSYdsYztAHO2v+Zyuf0XL17Mu+X51N3dPWXq1KlvAMDSSZMmbfvMM8/w/+fTNtts0/HWW2/toLV+sJzxJHmEgBCoHwHXdQ8holuIyPd9X0UlO47Ti4gLeWEwl8utH3+eq7l7rQK4r6+vc2hoaG/LsvYOFy95wYwX23hRMwMALNB/rLV+rbB+Sil24nTUGPUuuqjL766urq7jjTEHIiJbQE0LLWeeIqKbfN+/jp2HFSs75t0/7/RLKcXvYV4E3AsAZoTlPGyMuTCTyTxapN759+4YqaJd3IIF74qujc0n7weAAwGAj0sxk/cCwBARLbMsixc353me9/hYFeff0+n0rrZtH2WM+RDP/TxHhQvg/byoDQB3ep73Vgm+bJ3A/bpduFDDC9ts0XT7O++8c0XhInBsbJ+IiHmLo1QqNWPFihXLOzo6voyIhwDA7HDh9xta6wuia9Lp9Ed5AR4R+V7TETEgIv6WeIkto4Ig+N9MJpNf9JUkBBpFQARwo0jLfdqegOu6WxIRmz2fpbU+nxvsOM6FiPh1Y8ymmUxmaaUQKhV3pcp3Xfc8IjrbGHNAJpP5XZRvNOG52267pZYtW/ZSaPI8q/DjaPPNN9+oo6PjBd5t1lqzaXQ+xbwy847zNfE6xUzER3w81EMAl2o7EW0difDR+IcOzPiD8HGtNe9wR+3hs9zHFZ7xK1aW4zh7IeI9/FGpteYPthEfd47j3IGIB/Ciie/7vKsLcQFcrK7RLhN/0FQaRqvS8Sb5hYAQKE7AcZyjEZFD3NWUqvG4X6sALlPE/hMRD/Q876GCd3ZVAlgp9UEAYDNwPspRKj2WzWYPiO+cF84jbHlDRHdalnU9EU0pUhC/Y4/QWt9SUO+WE8AxK62xxtCo4prnqpUrV/4CEccKqzhCiPJNwwXVXwHAx0apxBLbtvfr7+9/tjCP4zjDApjnKwC4ni2qCvIN39dxHF5Y/uIYDX5Ra83HkSQJgYYREAHcMNRyo3Yn4LruV1kkGWN2jFaklVLzeXVVa80roxWnKgSwnU6n97AsaydEnB6a9/J9+QwumyCfqrX+QVSR0YRn+AHzGO8OaK33K1b5MOTTnBUrVrxr6dKlKzlPVOcgCN63aNEiNsceTo7j7I+ILMCv0lqfUE49xoI2VhikXC737cWLF/8jXs7s2bM3TaVS+bPSANBFRFb4O69ir9Jas/lyPimlFgPA5gDQq7UeGK0+juN8BxG/Fe7+fqswr+u6nyKiW/mDzvd93gWIC+AlWuv3FSmfz1Kz9QDvoH/Dtu0b+/v7eWFCkhAQAg0iMJEFsOM4NyAiH8G4DREfC4LgH7Zt54wxHGN+TyJiyxc+3vFqEARbLFq0iP0LRO+/jW3bnhYEAYt/9nL9uG3bbOk0nIhoaGBggI/d5FNPT89cYwxb+7wr9FXAO4YPEtE/iWhDy7IOAAB+//PRkfldXV0fLuVLAgCe5Hcv7xayUzEiepytrcLdbLak4iMsb8atargO7LMhm812W5b1B/43EX0zlUrdEa/30NDQK/G2jjWUat0BDuc/tgL6HwD4ExF5RPSWbdsbGWPmIiL7y4iE4GFa65sL68QL0y+99NIfQ8sy/pl3bedZlvVEEAQrEbEbEXcGAJ7LfhrfieXM0SJsyGSBZVmXIiLPaxuwlRcAHBne89VsNju3cHEiLoAB4GkA4ONEv7As69YgCF5BxJls5cQ+R6L5jstDxCeMMT/hNiPiv2zbXo+IeonoowDwfq31rLH4y+9CoJ4ERADXk6aUlWgCSqnbAWCP7u7uaffff3+ur69vai6XY3PVG7TWIz4YygVViQBOp9ObWJbFEyubWJVK52qt/zP6cTQB7Lrup4mIV4rLSZtorfMm3jEnWCwqR6zCxz4gRsSyrMcOcLk7K47jnI6I/wUAHSUaltNaD/+mlOLwJbYxZkomk1k9Ggyl1NXsLCw8bz1i95uvi4VG4V1m3iGJC+CHtNYfKlZ+uLPMH07vCX9fzOFIjDE3+r7/x3I6SPIIASFQPYFZs2atyyafUQmWZa1vjMnvlvJHPCLGF/vYnJP9HXyJhU78rl1dXYsKxd5Ytap1BzidTs/OZDK8kFd0VzQUrHx0Zuooi3f54zJlhEHiBTs+MsJWQWwRtWcx0+pwN5SPjljF3pcxE2jmuyAIgj0KTcld1z2CiG4I+X1Na31pnOV4ngEeywt0MXE9e/bsDYlodSmT+JkzZ06eNGnS3aG4ZcsqXqAd0Weu655BRJF58S1dXV1HlhhPluu60z3PWxYxcRznk4gYHeG5q7u7+wD+VokzcxznK4j4w/Bv12mtR5i/FwhgFraf9TzvlyXmrV+x/xA+IrBy5cr3R4vkhXnnzJmz/nPPPff6WM+B/C4E6klABHA9aUpZiSJQzY5AGR8PIxhWIoAdx7kbEfmMFztYujCXy3mLFy/m8z/GcZzjEZHDMZVteqyUYsdPfM7XR8QR54qLdPRp0UdO3At0Yb5mC2ClFAtMPi/LHpz/I5fL/cWyrJcjYauU4p2P9eNOraoUwMdGDsfiDGIC+G9aaz4PFRfA92mt9yz1EPEHeEdHx/5ExJ6l+Vz5ppwXEW/0PI9X7iUJASHQIALpdPrjlmXdyTucWmv2Cp9PSik2+c0fd2HBvHDhwv+rtUq1CuBy7u84Du8EnszOAX3fZ+d+I1K5cYBjVj78bnq/53m8S1g0KaXYbJl3Ktda/IsL4FHKYbHNrLt5d1tr/Yn4jcZTAJfBtKozwo7jbI+I+TPNxpgtMpnMc9G9Qp8P3F5eXBlcsWJFXylRWax+SilejOEoBqts255VypJIKfVAeEaZF383ji9gFAhgdoJZ0pQ6Vs5NWuvDy2AmWYRAwwiIAG4YarlRuxEIxRSH+eHEE9K+AMATlxf+jU3OFBH9EhGjVdaBQpOk0biUK4Bj51df1VrzB1jh2dMLEPGMCgVwJBbX+rAop84FoZTyl7SAAOa4nF8joi/6vv/f8XaEO/a8YBAUCOC6mUBHK/AlTKBHFcCFzHt6enYyxtzKH3+8AxUPc9Vuz5q0Rwi0GgHXdS8KHSCOeD9GjrEAYFG9vLPXWwDPnTt32sqVK/mYx+SOjo78d2AQBJ8EAA7rY7q6uiYX7iqWK4CVUj8FAJ4XvXAHs2TXKaV4h/zH7ARKa83ne4fnrZgAfibuY6KwMKUUWz2xCPu71nqriSSAecfXsiw2CZ/a0dGRP4aTy+XeFQlgIjrC9/0bozaF7/y81QEinul5HjuoLCvxvTo7O9m5Ipu6/1Zrzf1dNIVhDznqAadPaK35LHc+FZwBHlG/In0TLXC8nMvltis8ilRWxSWTEBgnAiKAxwmsFJssAo7jnICILKh2iTzyKqU47uv7PM/brFoa5QrgWDiOp7TW7JhiOIXOrHgVmb2UjliVVkqxw6a9giBwFi1axN5AhxN7RubdUTbB4rNipeLjFpn08mGQKhHAo9VjLHblMuJyXNf9GREdQ0T7xzw352/huu7niYjNlgsFcN4JFgBcqrX+2mj1GcsJllIqHxu0hBOsigRwWOfoI3zE2e6xmMnvQkAI1EbAdd2Hw9BmI549x3EuQ0QO/zbimEctd6uHAGYzZyI6hYh4oZZD2JVMQRBsVHg2tgIBzGd2RwjRctpujNkw8ozP+aP3OvtM8H2fd4mLJqUUe5Lmc6uLC31tjPMOcFU7vK7rdhljeHx8BhH5bLNdqm0cRtD3/cgcmZlECwa8O7xTJpPhc9ZlpZiTThbPZ3ied1GpC8PoFnl/F4j4bc/zvhvljQtg27bnFnOUFeVVSrHvkHyEBETk0IK3E9E9RPSg7/u8sCxJCDSNgAjgpqGXG7cTAdd1f0lEnzDGrMfmtOFqK5vZ3qq1jpxKVNzkcsVdKHI5hAWvHu8W8+RpK6UuBoBTwpsXCuC8h89igpDzO47zTUTkye+Ptm1/ob+/n70+D6dwUmXnUMMeOKsxgY68lJaqx2jgymUUflSx0xT20H1bd3f3IdH5p/AMHJ+l5Z38EQI4DGnCH3X8vvyM1prPekeJwx7to7X+ffiH4TBIAPAdrTXvqOTPcIUOQfhMdakwSEUFMIe3IqIPr1ix4ta4uVs4xthU7YN8zqogPFPFY00uEAJCoDwCBbtp22ut2VlgPimlOFY6O4s6XmvNu6E1p1oFsOu6xxIRL9AOxysfrVK2bc8s8q4v6wxw6LCPTZIrSoX3LAyDVKqwmIfrtUIxtZoA7u3tdYwx7MCqmLPDtZpYuMurlDon3KWHcn1eRIUWOPA6Wmsd7fCudd/wvDt/v/D568t834++H0bsALPDrsj3xyj9w9dyWEF2VhZPfGb+d7ZtXzmaiK5oEElmIVABARHAFcCSrEJglJc8exl+XmvNZzPz8fksy7q/1o+g2EcAOxUp6oAJEW/2PO8ypdRZAMDOndiMjM/6sFMJDufD59N+FjpkGSGAHcf5RBgv8E1emWXvjFx/rXVk2s2Cjp2MHMb3Z0+ORMRtZY/E7LWRvSPfobXmmIL5VI0ALqMeJQdfJQKYnZDYts1n0pgJr0BzvEV2LMX9xs5BODbj9LgJNN/Ydd3PEhF7QeUPSL5+IRFNQ0R29LJhPL/jOFsh4n1hPESOBcqOYNgKgL1wZ4noUN/3+Zx2PsXCIJUSwNsS0d94BZ09oCLiUiJi76rs6XMDRHxk6tSpu1bqWEeeZiEgBMojED+PWt4VRXNVvSNciwAO4+dybHJ+d71CRBcT0Z8mTZo0OHny5Lei90bMAqaouKpgB5jPPPP79SHbtk8sl9eUKVO8+DusHQVwbHGEF0V5cYQXRAeMMa9FfijCMEXvhNxGhDGKC+BiixSjsS4QwEdprXnnvGiaOXPmep2dnfl486MJ4HLPuPf19U3P5XLsp4J9XPA8+O7YjQ0inhvfZS53zEg+IVALARHAtdCTa4XAGnHEO3Tslfc8z/PyoW9c1/0WEXFInB7P86IzwRXzKufDi4gu933/5FB8fg4ATkZEl4g4LNFf2YTJGPOBMIblWmZbruuy4xM28eVQTZO4nELz5fDsKoti3tlYFwB4t5l3g9nxFgvw4TZWI4BDZmPWoxjASgRwKDg3tW2bz06x6OVzcIs5jIPv+z9QSnGb1hLAfF06nd7asqyvh9fxAgCH9BhARN7lnxevW29v72ZBEHwTANhhFX8M8sLCA4h4vud5LLqH01gCmFfjbds+DhF3B4C+sLy3QwF/3dDQ0DWDg4OrKh5ccoEQEAJlESjnPVxGQU0RwI7jRP4fAiKa6/t+f4n36KkAkI9NXmx3sQIBzMdt+D3FXozdMrgUzdJuAjhcGGVLorXMiuMA4k7UOOxd3GdIq5lAlyuACzrYdl13ayLiM8gnRWLYsqx9BgYG8iGrJAmBRhAQAdwIynIPISAEhIAQEAJCYEIS6OnpUYjIzoPyyRhzFZ//RcSLLcsa3kkLguB0AOBFyN/bts1OB4dTEARvjGUuWgpOjTvAHHed468/4XneNqXuoZTiMGvs+b+oAHYc5898vGasSAZhzOG8V/pKdykLhGDel8RYZ6lHM4EOFyK5HN7JPMb3fT7yU3WqJQ5wPGSTbduqv7/fL1YR13UPJKLomM0IAZxOp3e0LCsfkWGcnWDxsa1oXJd0glWlAB5udrw9RPQz3/ePrbpz5EIhUCEBEcAVApPsQkAICAEhIASEQGIJ8LEQPl6ybqEjophjrK9ordm7cV1SjQL4rtAShSMQsNOltVJoospHQtYpJYCVUlE5j2qtOcJB0RTzgs2/z9Nas+OmilM9doA333zzjTo6OvJhqIjoRN/3ORRg1akWAayU4ni6kQDfSmvNR0jAr+wAAAUySURBVGPWSjHO/NsIARyGQeIjSOzE7PkVK1bMqTYMUiqV2rxUiK5ot5+9c48WBqlWAcwNdF33HSKaEo+MUHUHyYVCoAICIoArgCVZhYAQEAJCQAgIgeQSSKfTH7AsawEArOzq6lo3Ore6ySabrDNlyhQ+6tAxVvzbSunVKIDZi/BX1mxcm50zmUw+xmyUwjOnHEqI48PmUwkT6LwH/TDuMQuwvHO/IokdL/KZYxbbhIjHeZ7H3vWLpt7e3i2y2ezmmUyGd6qHUz0EcOgcks/TdvJuved5fISl6lSLAI6HMCoVUUApxQ6jfhCr4AgBzH93HOd0RIw8ON/S1dV1ZAn/D1Zvb+9G8Vi/URi+sPzfd3d3Hxg5gozuqZRis+Qrwn9fp7Vm4T6c4l6gxxLASqlDh4aGbi91RCcMJfkgFx4/ylV1B8mFQqACAiKAK4AlWYWAEBACQkAICIHkEgh9JlwKAH/WWvO5/HxyXXd3ImLnd69rrdlHQCmBWDG8AgH8EBFdXUYhT/m+/5RS6oMAEHmpZsdGHDrtYURchYjbGGNOQUSHnVaFjvWKCmClFPuAyHu1ZsdItm3fkM1m804TOzs7s3Gv0aFXfTbVnRrm/wMi3khEHhFlbdvekIg4XN/+oVOkS7TWp8XbVA8BzOUppf4atot9NvBCwFPGmCz/Nnny5Nefe+453s0vK9UigNkiXCnFThE5HCGPDXYuebMx5v8sy2JTbzadZ2eSw/1QuAMcVpLLYSeXu4T/5jPd8yzLeiKXy620LGsGIu5IROy48ur4GWLO7zjOHYh4QHjt40R0qW3bXi6Xe49lWYdzVIjwt1ez2ezc559/nkMhDqcKBTDvvncSEYf/e9CyLD8IghWIuIFlWbsQ0ZcBYD12Dmnb9gfEG3RZw1Ay1YmACOA6gZRihIAQEAJCQAgIgfYmEMXyLhJTnUOenTMeppwFArhcwMMOD5VS3+b6jnLhJUT0bOgosagA7uvrm5rL5dhsl73/F6a1QhBxiDxjzK9DcT1Wnc/RWn8nnqleAthxnI8hIu8uF/verSiWb40CmMU4L0bcGzqSXIsJn9PO5XIH2bbNZs6c1toB5j9yX2Sz2V/GhGwpvmtdH+74s/fpj43SKUts296vmCCtQgCzE8iSKYxucKzW+uaxBon8LgTqSUAEcD1pSllCQAgIASEgBIRAuxJApdQrYfixPTzP4524fFJK8e4v7wifprW+pJ4AahXAYf32Q8SvEhGLMA6jxu14jOMD+75/r+M4R48mgLkMPlObSqW+gYgfRcTN+Oxm2M61BDD/PTRB5l3FgwGAHXC9ly2sQw/6nmVZfw2C4LZMJvNEIa96CWAu13XdjxDRVzlmeliHjvB+DRXAYV04agQz3JuIZgDAmwDgI+ItQRBcmcvlsLOzkyM4cCoqgCNWjuPsZVnW0UTEIfFYaPKZ3ReJ6DnLsm595513flfqjLBSivuEd3u3C8Mavo2IHJLptnXWWWfe008/HYViGtE1lQjgcNzuxeGPiKgPEbmO0wCAQ/pxm+/N5XJXLl68OBL89XxspCwhMCoBEcAyQISAEBACQkAICAEhIASEgBAQAkIgEQREACeim6WRQkAICAEhIASEgBAQAkJACAgBISACWMaAEBACQkAICAEhIASEgBAQAkJACCSCgAjgRHSzNFIICAEhIASEgBAQAkJACAgBISAERADLGBACQkAICAEhIASEgBAQAkJACAiBRBAQAZyIbpZGCgEhIASEgBAQAkJACAgBISAEhMD/A0spEQ+31xVnAAAAAElFTkSuQmCC\" width=\"960\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "2f18f097-5405-42ff-9a09-20b8ea1e559e",
        "id": "jB0mn3L9GHOb",
        "colab": {}
      },
      "source": [
        "for i in ind:\n",
        "    print(\"({}, {})\".format(i, np.round(expl_var[i-1], 2)))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(1, 0.23)\n",
            "(2, 0.26)\n",
            "(4, 0.3)\n",
            "(8, 0.34)\n",
            "(20, 0.38)\n",
            "(60, 0.44)\n",
            "(100, 0.47)\n",
            "(200, 0.53)\n",
            "(300, 0.57)\n",
            "(400, 0.61)\n",
            "(500, 0.64)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "LQ7ZH-6qGHOe"
      },
      "source": [
        "\n",
        "> I think 500 dimensions is good enough \n",
        "\n",
        "---------\n",
        "\n",
        "-  By just taking __(20 to 30)__ latent factors, explained variance that we could get is __20 %__. \n",
        "\n",
        "- To take it to __60%__, we have to take  __almost 400 latent factors__. It is not fare.\n",
        "\n",
        "\n",
        "\n",
        "- It basically is the __gain of variance explained__, if we ___add one additional latent factor to it.___\n",
        "\n",
        "\n",
        "- By adding one by one latent factore too it, the ___gain in expained variance__ with that addition is decreasing. (Obviously, because they are sorted that way).\n",
        "- ___LHS Graph___:\n",
        "    - __x__ --- ( No of latent factos ),\n",
        "    - __y__ --- ( The variance explained by taking x latent factors)\n",
        "\n",
        "\n",
        "\n",
        "- __More decrease in the line (RHS graph) __:\n",
        "    - We  are getting more expained variance than before.\n",
        "- __Less decrease in that line (RHS graph)__  :\n",
        "    - We are not getting benifitted from adding latent factor furthur. This is what is shown in the plots.\n",
        "\n",
        "\n",
        "- ___RHS Graph___:\n",
        "    - __x__ --- ( No of latent factors ),\n",
        "    - __y__ --- ( Gain n Expl_Var by taking one additional latent factor) "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "5f7de336-1593-41ee-8d5e-522ed4b6555d",
        "id": "NQfBtVUrGHOe",
        "colab": {}
      },
      "source": [
        "# Let's project our Original U_M matrix into into 500 Dimensional space...\n",
        "start = datetime.now()\n",
        "trunc_matrix = train_sparse_matrix.dot(netflix_svd.components_.T)\n",
        "print(datetime.now()- start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0:00:45.670265\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "a5194761-c6da-4f25-a31e-89cc55033f45",
        "id": "RFFxNjL0GHOh",
        "colab": {}
      },
      "source": [
        "type(trunc_matrix), trunc_matrix.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(numpy.ndarray, (2649430, 500))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "3R5wziskGHOj"
      },
      "source": [
        "* Let's convert this to actual sparse matrix and store it for future purposes"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "v0DQr_zLGHOj",
        "colab": {}
      },
      "source": [
        "if not os.path.isfile('/content/drive/My Drive/Netflix_recommender/data_folder/trunc_sparse_matrix.npz'):\n",
        "    # create that sparse sparse matrix\n",
        "    trunc_sparse_matrix = sparse.csr_matrix(trunc_matrix)\n",
        "    # Save this truncated sparse matrix for later usage..\n",
        "    sparse.save_npz('trunc_sparse_matrix', trunc_sparse_matrix)\n",
        "else:\n",
        "    trunc_sparse_matrix = sparse.load_npz('/content/drive/My Drive/Netflix_recommender/data_folder/trunc_sparse_matrix.npz')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "0bb095f5-eb23-417e-febf-1038260dd6a4",
        "id": "8Qq5w0tYGHOk",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "trunc_sparse_matrix.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(2649430, 500)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "b3a387d8-3fb5-4a3d-e0de-9a5f02a7926f",
        "id": "7ymbMKtKGHOm",
        "colab": {}
      },
      "source": [
        "start = datetime.now()\n",
        "trunc_u_u_sim_matrix, _ = compute_user_similarity(trunc_sparse_matrix, compute_for_few=True, top=50, verbose=True, \n",
        "                                                 verb_for_n_rows=10)\n",
        "print(\"-\"*50)\n",
        "print(\"time:\",datetime.now()-start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Computing top 50 similarities for each user..\n",
            "computing done for 10 users [  time elapsed : 0:02:09.746324  ]\n",
            "computing done for 20 users [  time elapsed : 0:04:16.017768  ]\n",
            "computing done for 30 users [  time elapsed : 0:06:20.861163  ]\n",
            "computing done for 40 users [  time elapsed : 0:08:24.933316  ]\n",
            "computing done for 50 users [  time elapsed : 0:10:28.861485  ]\n",
            "Creating Sparse matrix from the computed similarities\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "--------------------------------------------------\n",
            "time: 0:10:52.658092\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "FHEcJOfrGHOn"
      },
      "source": [
        "**: This is taking more time for each user than Original one.**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "eMLuVBD5GHOn"
      },
      "source": [
        "- from above plot, It took almost __12.18__ for computing simlilar users for __one user__\n",
        "    \n",
        "    \n",
        "- We have __405041 users__ with us in training set.\n",
        "\n",
        "\n",
        "- ${ 405041 \\times 12.18 ==== 4933399.38 \\sec } ====  82223.323 \\min ==== 1370.388716667 \\text{ hours}\n",
        "==== 57.099529861 \\text{ days}...$\n",
        "\n",
        "    - Even we run on 4 cores parallelly (a typical system now a days), It will still take almost __(14 - 15) __ days.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "_v7kZ4IFGHOo"
      },
      "source": [
        "- __Why did this happen...??__\n",
        "\n",
        "\n",
        "    - Just think about it. It's not that difficult.\n",
        "\n",
        "---------------------------------_( sparse & dense..................get it ?? )_-----------------------------------"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ASXSsaGmGHOo"
      },
      "source": [
        "__Is there any other way to compute user user similarity..??__"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "7ehSsq3gGHOq"
      },
      "source": [
        "-An alternative is to compute similar users for a particular user,  whenenver required (**ie., Run time**)\n",
        "    - We maintain a binary Vector for users, which tells us whether we already computed or not..\n",
        "    - ***If not*** : \n",
        "        - Compute top (let's just say, 1000) most similar users for this given user, and add this to our datastructure, so that we can just access it(similar users) without recomputing it again.\n",
        "        - \n",
        "    - ***If It is already Computed***:\n",
        "        - Just get it directly from our datastructure, which has that information.\n",
        "        - In production time, We might have to recompute similarities, if it is computed a long time ago. Because user preferences changes over time. If we could maintain some kind of Timer, which when expires, we have to update it ( recompute it ). \n",
        "        - \n",
        "    - ***Which datastructure to use:***\n",
        "        - It is purely implementation dependant. \n",
        "        - One simple method is to maintain a **Dictionary Of Dictionaries**.\n",
        "            - \n",
        "            - **key    :** _userid_ \n",
        "            - __value__: _Again a dictionary_\n",
        "                - __key__  : _Similar User_\n",
        "                - __value__: _Similarity Value_"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": true,
        "colab_type": "text",
        "id": "oYouSYH0GHOr"
      },
      "source": [
        "<h3> 3.4.2 Computing Movie-Movie Similarity matrix </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "7_72kq4YGHOs",
        "colab": {}
      },
      "source": [
        "start = datetime.now()\n",
        "if not os.path.isfile('/content/drive/My Drive/Netflix_recommender/data_folder/m_m_sim_sparse.npz'):\n",
        "    print(\"It seems you don't have that file. Computing movie_movie similarity...\")\n",
        "    start = datetime.now()\n",
        "    m_m_sim_sparse = cosine_similarity(X=train_sparse_matrix.T, dense_output=False)\n",
        "    print(\"Done..\")\n",
        "    # store this sparse matrix in disk before using it. For future purposes.\n",
        "    print(\"Saving it to disk without the need of re-computing it again.. \")\n",
        "    sparse.save_npz(\"m_m_sim_sparse.npz\", m_m_sim_sparse)\n",
        "    print(\"Done..\")\n",
        "else:\n",
        "    print(\"It is there, We will get it.\")\n",
        "    m_m_sim_sparse = sparse.load_npz(\"/content/drive/My Drive/Netflix_recommender/data_folder/m_m_sim_sparse.npz\")\n",
        "    print(\"Done ...\")\n",
        "\n",
        "print(\"It's a \",m_m_sim_sparse.shape,\" dimensional matrix\")\n",
        "\n",
        "print(datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "cd481df2-f6ea-4f01-d4d1-b500ab74f838",
        "id": "TXIu4Xx1GHOv",
        "colab": {}
      },
      "source": [
        "m_m_sim_sparse.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(17771, 17771)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "FwunajS_GHOy"
      },
      "source": [
        "- Even though we have similarity measure of each movie, with all other movies, We generally don't care much about least similar movies.\n",
        "\n",
        "\n",
        "- Most of the times, only top_xxx similar items matters. It may be 10 or 100.\n",
        "\n",
        "\n",
        "- We take only those top similar movie ratings and store them  in a saperate dictionary."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "am27Q4cNGHOy",
        "colab": {}
      },
      "source": [
        "movie_ids = np.unique(m_m_sim_sparse.nonzero()[1])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "7e9d1c69-1ceb-4c26-e81f-83291f44aaa5",
        "id": "9b8fpLlxGHO0",
        "colab": {}
      },
      "source": [
        "start = datetime.now()\n",
        "similar_movies = dict()\n",
        "for movie in movie_ids:\n",
        "    # get the top similar movies and store them in the dictionary\n",
        "    sim_movies = m_m_sim_sparse[movie].toarray().ravel().argsort()[::-1][1:]\n",
        "    similar_movies[movie] = sim_movies[:100]\n",
        "print(datetime.now() - start)\n",
        "\n",
        "# just testing similar movies for movie_15\n",
        "similar_movies[15]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0:00:33.411700\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 8279,  8013, 16528,  5927, 13105, 12049,  4424, 10193, 17590,\n",
              "        4549,  3755,   590, 14059, 15144, 15054,  9584,  9071,  6349,\n",
              "       16402,  3973,  1720,  5370, 16309,  9376,  6116,  4706,  2818,\n",
              "         778, 15331,  1416, 12979, 17139, 17710,  5452,  2534,   164,\n",
              "       15188,  8323,  2450, 16331,  9566, 15301, 13213, 14308, 15984,\n",
              "       10597,  6426,  5500,  7068,  7328,  5720,  9802,   376, 13013,\n",
              "        8003, 10199,  3338, 15390,  9688, 16455, 11730,  4513,   598,\n",
              "       12762,  2187,   509,  5865,  9166, 17115, 16334,  1942,  7282,\n",
              "       17584,  4376,  8988,  8873,  5921,  2716, 14679, 11947, 11981,\n",
              "        4649,   565, 12954, 10788, 10220, 10963,  9427,  1690,  5107,\n",
              "        7859,  5969,  1510,  2429,   847,  7845,  6410, 13931,  9840,\n",
              "        3706])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 62
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "X4Uf1_uZGHO2"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "XXdHrIN7GHO3"
      },
      "source": [
        "<h3> 3.4.3 Finding most similar movies using similarity matrix </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "g2S87T-0GHO4"
      },
      "source": [
        "__ Does Similarity really works as the way we expected...? __ <br>\n",
        "_Let's pick some random movie and check for its similar movies...._"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "2d737e9e-a444-4ec5-aba0-6ad6363aa24d",
        "id": "VVQVMv5ZGHO4",
        "colab": {}
      },
      "source": [
        "# First Let's load the movie details into soe dataframe..\n",
        "# movie details are in 'netflix/movie_titles.csv'\n",
        "\n",
        "movie_titles = pd.read_csv(\"data_folder/movie_titles.csv\", sep=',', header = None,\n",
        "                           names=['movie_id', 'year_of_release', 'title'], verbose=True,\n",
        "                      index_col = 'movie_id', encoding = \"ISO-8859-1\")\n",
        "\n",
        "movie_titles.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Tokenization took: 4.50 ms\n",
            "Type conversion took: 165.72 ms\n",
            "Parser memory cleanup took: 0.01 ms\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>year_of_release</th>\n",
              "      <th>title</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>movie_id</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2003.0</td>\n",
              "      <td>Dinosaur Planet</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2004.0</td>\n",
              "      <td>Isle of Man TT 2004 Review</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1997.0</td>\n",
              "      <td>Character</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1994.0</td>\n",
              "      <td>Paula Abdul's Get Up &amp; Dance</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>2004.0</td>\n",
              "      <td>The Rise and Fall of ECW</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          year_of_release                         title\n",
              "movie_id                                               \n",
              "1                  2003.0               Dinosaur Planet\n",
              "2                  2004.0    Isle of Man TT 2004 Review\n",
              "3                  1997.0                     Character\n",
              "4                  1994.0  Paula Abdul's Get Up & Dance\n",
              "5                  2004.0      The Rise and Fall of ECW"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "x_dc-9T1GHO6"
      },
      "source": [
        "<p style='font-size:15px'><b>Similar Movies for 'Vampire Journals'</b></p>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "1665280e-124d-4e73-f508-808853ebba2d",
        "id": "hJNsE740GHO7",
        "colab": {}
      },
      "source": [
        "mv_id = 67\n",
        "\n",
        "print(\"\\nMovie ----->\",movie_titles.loc[mv_id].values[1])\n",
        "\n",
        "print(\"\\nIt has {} Ratings from users.\".format(train_sparse_matrix[:,mv_id].getnnz()))\n",
        "\n",
        "print(\"\\nWe have {} movies which are similarto this  and we will get only top most..\".format(m_m_sim_sparse[:,mv_id].getnnz()))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Movie -----> Vampire Journals\n",
            "\n",
            "It has 270 Ratings from users.\n",
            "\n",
            "We have 17284 movies which are similarto this  and we will get only top most..\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "GXMvP8xyGHO-",
        "colab": {}
      },
      "source": [
        "similarities = m_m_sim_sparse[mv_id].toarray().ravel()\n",
        "\n",
        "similar_indices = similarities.argsort()[::-1][1:]\n",
        "\n",
        "similarities[similar_indices]\n",
        "\n",
        "sim_indices = similarities.argsort()[::-1][1:] # It will sort and reverse the array and ignore its similarity (ie.,1)\n",
        "                                               # and return its indices(movie_ids)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "2e929724-0612-4d54-cab4-e8a4b6081b37",
        "id": "igXbZvu4GHPA",
        "colab": {}
      },
      "source": [
        "plt.plot(similarities[sim_indices], label='All the ratings')\n",
        "plt.plot(similarities[sim_indices[:100]], label='top 100 similar movies')\n",
        "plt.title(\"Similar Movies of {}(movie_id)\".format(mv_id), fontsize=20)\n",
        "plt.xlabel(\"Movies (Not Movie_Ids)\", fontsize=15)\n",
        "plt.ylabel(\"Cosine Similarity\",fontsize=15)\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "eODAeoKbGHPE"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "YhHN7FLwGHPF"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "4FxioJYTGHPF"
      },
      "source": [
        "__Top 10 similar movies__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "eaebe922-cee7-4ef5-a41a-c7f6de5732d6",
        "id": "JxYjiCF2GHPF",
        "colab": {}
      },
      "source": [
        "movie_titles.loc[sim_indices[:10]]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>year_of_release</th>\n",
              "      <th>title</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>movie_id</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>323</th>\n",
              "      <td>1999.0</td>\n",
              "      <td>Modern Vampires</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4044</th>\n",
              "      <td>1998.0</td>\n",
              "      <td>Subspecies 4: Bloodstorm</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1688</th>\n",
              "      <td>1993.0</td>\n",
              "      <td>To Sleep With a Vampire</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13962</th>\n",
              "      <td>2001.0</td>\n",
              "      <td>Dracula: The Dark Prince</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12053</th>\n",
              "      <td>1993.0</td>\n",
              "      <td>Dracula Rising</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16279</th>\n",
              "      <td>2002.0</td>\n",
              "      <td>Vampires: Los Muertos</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4667</th>\n",
              "      <td>1996.0</td>\n",
              "      <td>Vampirella</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1900</th>\n",
              "      <td>1997.0</td>\n",
              "      <td>Club Vampire</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13873</th>\n",
              "      <td>2001.0</td>\n",
              "      <td>The Breed</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15867</th>\n",
              "      <td>2003.0</td>\n",
              "      <td>Dracula II: Ascension</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          year_of_release                     title\n",
              "movie_id                                           \n",
              "323                1999.0           Modern Vampires\n",
              "4044               1998.0  Subspecies 4: Bloodstorm\n",
              "1688               1993.0   To Sleep With a Vampire\n",
              "13962              2001.0  Dracula: The Dark Prince\n",
              "12053              1993.0            Dracula Rising\n",
              "16279              2002.0     Vampires: Los Muertos\n",
              "4667               1996.0                Vampirella\n",
              "1900               1997.0              Club Vampire\n",
              "13873              2001.0                 The Breed\n",
              "15867              2003.0     Dracula II: Ascension"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Jz4p_iYqGHPJ"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "hZXYanT2GHPJ"
      },
      "source": [
        " > Similarly, we can ___find similar users___ and compare how similar they are. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Nx6WfTBfGHPK"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ZfMOhLKQGHPN"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "4pFk8qDmGHPO"
      },
      "source": [
        " <h1> 4.  Machine Learning Models </h1>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "7oIEoF0WGHPQ"
      },
      "source": [
        "<img src='images/models.jpg' width=500px>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "UGjGNEqdGHPQ",
        "colab": {}
      },
      "source": [
        "def get_sample_sparse_matrix(sparse_matrix, no_users, no_movies, path, verbose = True):\n",
        "    \"\"\"\n",
        "        It will get it from the ''path'' if it is present  or It will create \n",
        "        and store the sampled sparse matrix in the path specified.\n",
        "    \"\"\"\n",
        "\n",
        "    # get (row, col) and (rating) tuple from sparse_matrix...\n",
        "    row_ind, col_ind, ratings = sparse.find(sparse_matrix)\n",
        "    users = np.unique(row_ind)\n",
        "    movies = np.unique(col_ind)\n",
        "\n",
        "    print(\"Original Matrix : (users, movies) -- ({} {})\".format(len(users), len(movies)))\n",
        "    print(\"Original Matrix : Ratings -- {}\\n\".format(len(ratings)))\n",
        "\n",
        "    # It just to make sure to get same sample everytime we run this program..\n",
        "    # and pick without replacement....\n",
        "    np.random.seed(15)\n",
        "    sample_users = np.random.choice(users, no_users, replace=False)\n",
        "    sample_movies = np.random.choice(movies, no_movies, replace=False)\n",
        "    # get the boolean mask or these sampled_items in originl row/col_inds..\n",
        "    mask = np.logical_and( np.isin(row_ind, sample_users),\n",
        "                      np.isin(col_ind, sample_movies) )\n",
        "    \n",
        "    sample_sparse_matrix = sparse.csr_matrix((ratings[mask], (row_ind[mask], col_ind[mask])),\n",
        "                                             shape=(max(sample_users)+1, max(sample_movies)+1))\n",
        "\n",
        "    if verbose:\n",
        "        print(\"Sampled Matrix : (users, movies) -- ({} {})\".format(len(sample_users), len(sample_movies)))\n",
        "        print(\"Sampled Matrix : Ratings --\", format(ratings[mask].shape[0]))\n",
        "\n",
        "    print('Saving it into disk for furthur usage..')\n",
        "    # save it into disk\n",
        "    sparse.save_npz(path, sample_sparse_matrix)\n",
        "    if verbose:\n",
        "            print('Done..\\n')\n",
        "    \n",
        "    return sample_sparse_matrix"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "xWynlzd4GHPT"
      },
      "source": [
        "<h2> 4.1 Sampling Data </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "vCBwI1ObGHPV"
      },
      "source": [
        "<h3>4.1.1 Build sample train data from the train data</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "8424501c-0460-4559-ccc2-bc0737cc5c75",
        "id": "bvKU0iwfGHPW",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "start = datetime.now()\n",
        "path = \"/content/drive/My Drive/Netflix_recommender/sample_train_sparse_matrix.npz\"\n",
        "if os.path.isfile(path):\n",
        "    print(\"It is present in your pwd, getting it from disk....\")\n",
        "    # just get it from the disk instead of computing it\n",
        "    sample_train_sparse_matrix = sparse.load_npz(path)\n",
        "    print(\"DONE..\")\n",
        "else: \n",
        "    # get 10k users and 1k movies from available data \n",
        "    sample_train_sparse_matrix = get_sample_sparse_matrix(train_sparse_matrix, no_users=13000, no_movies=1500,\n",
        "                                             path = path)\n",
        "\n",
        "print(datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Original Matrix : (users, movies) -- (405041 17424)\n",
            "Original Matrix : Ratings -- 80384405\n",
            "\n",
            "Sampled Matrix : (users, movies) -- (13000 1500)\n",
            "Sampled Matrix : Ratings -- 239403\n",
            "Saving it into disk for furthur usage..\n",
            "Done..\n",
            "\n",
            "0:01:04.277437\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "M9imqSJAGHPY"
      },
      "source": [
        "<h3>4.1.2 Build sample test data from the test data</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "3ba55b84-5d0a-477c-916f-b30f185d34ba",
        "id": "UlgBQjH8GHPZ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "start = datetime.now()\n",
        "\n",
        "path = \"/content/drive/My Drive/Netflix_recommender/sample_test_sparse_matrix.npz\"\n",
        "if os.path.isfile(path):\n",
        "    print(\"It is present in your pwd, getting it from disk....\")\n",
        "    # just get it from the disk instead of computing it\n",
        "    sample_test_sparse_matrix = sparse.load_npz(path)\n",
        "    print(\"DONE..\")\n",
        "else:\n",
        "    # get 5k users and 500 movies from available data \n",
        "    sample_test_sparse_matrix = get_sample_sparse_matrix(test_sparse_matrix, no_users=5000, no_movies=500,\n",
        "                                                 path = \"/content/drive/My Drive/Netflix_recommender/sample_test_sparse_matrix.npz\")\n",
        "print(datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Original Matrix : (users, movies) -- (349312 17757)\n",
            "Original Matrix : Ratings -- 20096102\n",
            "\n",
            "Sampled Matrix : (users, movies) -- (5000 500)\n",
            "Sampled Matrix : Ratings -- 7333\n",
            "Saving it into disk for furthur usage..\n",
            "Done..\n",
            "\n",
            "0:00:14.324827\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "hOXpmSIgGHPb"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "o4p-sL3_GHPb"
      },
      "source": [
        "<h2>4.2 Finding Global Average of all movie ratings, Average rating per User, and Average rating per Movie (from sampled train)</h2>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "RqfAqSqtGHPc",
        "colab": {}
      },
      "source": [
        "sample_train_averages = dict()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "8LN1-CcPGHPe"
      },
      "source": [
        "<h3>4.2.1 Finding Global Average of all movie ratings</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "37d5653c-0b80-48a1-ba8e-4583c818926b",
        "id": "XFMPS01AGHPl",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# get the global average of ratings in our train set.\n",
        "global_average = sample_train_sparse_matrix.sum()/sample_train_sparse_matrix.count_nonzero()\n",
        "sample_train_averages['global'] = global_average\n",
        "sample_train_averages"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'global': 3.5437651157253667}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "cpf0Qd3fGHPn"
      },
      "source": [
        "<h3>4.2.2 Finding Average rating per User</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "a9b3fcd3-0e18-4a48-fd7f-a5d1faeb07dc",
        "id": "cw6hNYaWGHPo",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "sample_train_averages['user'] = get_average_ratings(sample_train_sparse_matrix, of_users=True)\n",
        "print('\\nAverage rating of user 1515220 :',sample_train_averages['user'][1515220])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Average rating of user 1515220 : 3.953488372093023\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "inFN6jsVGHPq"
      },
      "source": [
        "<h3>4.2.3 Finding Average rating per Movie</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "b4b840fe-da31-4d84-8720-89b82e2eb394",
        "id": "ZguYkJAgGHPq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "sample_train_averages['movie'] =  get_average_ratings(sample_train_sparse_matrix, of_users=False)\n",
        "print('\\n AVerage rating of movie 15153 :',sample_train_averages['movie'][15153])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            " AVerage rating of movie 15153 : 2.6\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "6OjxppiKGHPt"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "d5ihzVhvGHPt"
      },
      "source": [
        "<h2> 4.3 Featurizing data </h2>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "08147a89-1b03-4a6f-8c64-5cfbd3ea1942",
        "id": "NouTdnsyGHPu",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "print('\\n No of ratings in Our Sampled train matrix is : {}\\n'.format(sample_train_sparse_matrix.count_nonzero()))\n",
        "print('\\n No of ratings in Our Sampled test  matrix is : {}\\n'.format(sample_test_sparse_matrix.count_nonzero()))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            " No of ratings in Our Sampled train matrix is : 239403\n",
            "\n",
            "\n",
            " No of ratings in Our Sampled test  matrix is : 7333\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "JCk3bRYkGHPw"
      },
      "source": [
        "<h3> 4.3.1 Featurizing data for regression problem </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "9hkCl_jxGHPx"
      },
      "source": [
        "<h4> 4.3.1.1 Featurizing train data </h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "6S69hvjAGHPx",
        "colab": {}
      },
      "source": [
        "# get users, movies and ratings from our samples train sparse matrix\n",
        "sample_train_users, sample_train_movies, sample_train_ratings = sparse.find(sample_train_sparse_matrix)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "51abb095-433c-4404-be97-bbe8311b02f4",
        "id": "eA3eLkgVGHP0",
        "colab": {}
      },
      "source": [
        "############################################################\n",
        "# It took me almost 10 hours to prepare this train dataset.#\n",
        "############################################################\n",
        "start = datetime.now()\n",
        "if os.path.isfile('sample/small/reg_train.csv'):\n",
        "    print(\"File already exists you don't have to prepare again...\" )\n",
        "else:\n",
        "    print('preparing {} tuples for the dataset..\\n'.format(len(sample_train_ratings)))\n",
        "    with open('sample/small/reg_train.csv', mode='w') as reg_data_file:\n",
        "        count = 0\n",
        "        for (user, movie, rating)  in zip(sample_train_users, sample_train_movies, sample_train_ratings):\n",
        "            st = datetime.now()\n",
        "        #     print(user, movie)    \n",
        "            #--------------------- Ratings of \"movie\" by similar users of \"user\" ---------------------\n",
        "            # compute the similar Users of the \"user\"        \n",
        "            user_sim = cosine_similarity(sample_train_sparse_matrix[user], sample_train_sparse_matrix).ravel()\n",
        "            top_sim_users = user_sim.argsort()[::-1][1:] # we are ignoring 'The User' from its similar users.\n",
        "            # get the ratings of most similar users for this movie\n",
        "            top_ratings = sample_train_sparse_matrix[top_sim_users, movie].toarray().ravel()\n",
        "            # we will make it's length \"5\" by adding movie averages to .\n",
        "            top_sim_users_ratings = list(top_ratings[top_ratings != 0][:5])\n",
        "            top_sim_users_ratings.extend([sample_train_averages['movie'][movie]]*(5 - len(top_sim_users_ratings)))\n",
        "        #     print(top_sim_users_ratings, end=\" \")    \n",
        "\n",
        "\n",
        "            #--------------------- Ratings by \"user\"  to similar movies of \"movie\" ---------------------\n",
        "            # compute the similar movies of the \"movie\"        \n",
        "            movie_sim = cosine_similarity(sample_train_sparse_matrix[:,movie].T, sample_train_sparse_matrix.T).ravel()\n",
        "            top_sim_movies = movie_sim.argsort()[::-1][1:] # we are ignoring 'The User' from its similar users.\n",
        "            # get the ratings of most similar movie rated by this user..\n",
        "            top_ratings = sample_train_sparse_matrix[user, top_sim_movies].toarray().ravel()\n",
        "            # we will make it's length \"5\" by adding user averages to.\n",
        "            top_sim_movies_ratings = list(top_ratings[top_ratings != 0][:5])\n",
        "            top_sim_movies_ratings.extend([sample_train_averages['user'][user]]*(5-len(top_sim_movies_ratings))) \n",
        "        #     print(top_sim_movies_ratings, end=\" : -- \")\n",
        "\n",
        "            #-----------------prepare the row to be stores in a file-----------------#\n",
        "            row = list()\n",
        "            row.append(user)\n",
        "            row.append(movie)\n",
        "            # Now add the other features to this data...\n",
        "            row.append(sample_train_averages['global']) # first feature\n",
        "            # next 5 features are similar_users \"movie\" ratings\n",
        "            row.extend(top_sim_users_ratings)\n",
        "            # next 5 features are \"user\" ratings for similar_movies\n",
        "            row.extend(top_sim_movies_ratings)\n",
        "            # Avg_user rating\n",
        "            row.append(sample_train_averages['user'][user])\n",
        "            # Avg_movie rating\n",
        "            row.append(sample_train_averages['movie'][movie])\n",
        "\n",
        "            # finalley, The actual Rating of this user-movie pair...\n",
        "            row.append(rating)\n",
        "            count = count + 1\n",
        "\n",
        "            # add rows to the file opened..\n",
        "            reg_data_file.write(','.join(map(str, row)))\n",
        "            reg_data_file.write('\\n')        \n",
        "            if (count)%10000 == 0:\n",
        "                # print(','.join(map(str, row)))\n",
        "                print(\"Done for {} rows----- {}\".format(count, datetime.now() - start))\n",
        "\n",
        "\n",
        "print(datetime.now() - start)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "preparing 129286 tuples for the dataset..\n",
            "\n",
            "Done for 10000 rows----- 0:53:13.974716\n",
            "Done for 20000 rows----- 1:47:58.228942\n",
            "Done for 30000 rows----- 2:42:46.963119\n",
            "Done for 40000 rows----- 3:36:44.807894\n",
            "Done for 50000 rows----- 4:28:55.311500\n",
            "Done for 60000 rows----- 5:24:18.493104\n",
            "Done for 70000 rows----- 6:17:39.669922\n",
            "Done for 80000 rows----- 7:11:23.970879\n",
            "Done for 90000 rows----- 8:05:33.787770\n",
            "Done for 100000 rows----- 9:00:25.463562\n",
            "Done for 110000 rows----- 9:51:28.530010\n",
            "Done for 120000 rows----- 10:42:05.382141\n",
            "11:30:13.699183\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "6d4OEPwJGHP2"
      },
      "source": [
        "__Reading from the file to make a Train_dataframe__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "32649084-d35a-4184-ba1c-2faa04c8ab6d",
        "id": "GgXpqzZ1GHP3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "reg_train = pd.read_csv('/content/drive/My Drive/Netflix_recommender/reg_train_25.csv', names = ['user', 'movie', 'GAvg', 'sur1', 'sur2', 'sur3', 'sur4', 'sur5','smr1', 'smr2', 'smr3', 'smr4', 'smr5', 'UAvg', 'MAvg', 'rating'], header=None)\n",
        "reg_train.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>174683</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.882353</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>233949</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.692308</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>555770</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.795455</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>767518</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.884615</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>894393</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg  sur1  sur2  ...  smr4  smr5      UAvg      MAvg  rating\n",
              "0  174683     10  3.587581   5.0   5.0  ...   3.0   2.0  3.882353  3.611111       5\n",
              "1  233949     10  3.587581   4.0   4.0  ...   3.0   3.0  2.692308  3.611111       3\n",
              "2  555770     10  3.587581   4.0   5.0  ...   4.0   4.0  3.795455  3.611111       4\n",
              "3  767518     10  3.587581   2.0   5.0  ...   4.0   3.0  3.884615  3.611111       5\n",
              "4  894393     10  3.587581   3.0   5.0  ...   4.0   4.0  4.000000  3.611111       4\n",
              "\n",
              "[5 rows x 16 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "-HewozfoGHP6"
      },
      "source": [
        "-----------------------\n",
        "\n",
        "- __GAvg__ : Average rating of all the ratings \n",
        "\n",
        "\n",
        "- __Similar users rating of this movie__:\n",
        "    - sur1, sur2, sur3, sur4, sur5 ( top 5 similar users who rated that movie.. )\n",
        "    \n",
        "\n",
        "\n",
        "- __Similar movies rated by this user__:\n",
        "    - smr1, smr2, smr3, smr4, smr5 ( top 5 similar movies rated by this movie.. )\n",
        "\n",
        "\n",
        "- __UAvg__ : User's Average rating\n",
        "\n",
        "\n",
        "- __MAvg__ : Average rating of this movie\n",
        "\n",
        "\n",
        "- __rating__ : Rating of this movie by this user.\n",
        "\n",
        "-----------------------"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "dijPar5jGHP6"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "71zxlWvxGHP7"
      },
      "source": [
        "<h4> 4.3.1.2 Featurizing test data </h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "kOXRQ442GHP8",
        "colab": {}
      },
      "source": [
        "# get users, movies and ratings from the Sampled Test \n",
        "sample_test_users, sample_test_movies, sample_test_ratings = sparse.find(sample_test_sparse_matrix)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "bcb3bc92-eaee-4d85-a547-56a6de5ae9fd",
        "id": "wqg0S9JcGHP-",
        "colab": {}
      },
      "source": [
        "sample_train_averages['global']"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "3.581679377504138"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "767f43c5-ad21-43db-c636-4bc220000fa2",
        "id": "jBuDLV9uGHQB",
        "colab": {}
      },
      "source": [
        "start = datetime.now()\n",
        "\n",
        "if os.path.isfile('sample/small/reg_test.csv'):\n",
        "    print(\"It is already created...\")\n",
        "else:\n",
        "\n",
        "    print('preparing {} tuples for the dataset..\\n'.format(len(sample_test_ratings)))\n",
        "    with open('sample/small/reg_test.csv', mode='w') as reg_data_file:\n",
        "        count = 0 \n",
        "        for (user, movie, rating)  in zip(sample_test_users, sample_test_movies, sample_test_ratings):\n",
        "            st = datetime.now()\n",
        "\n",
        "        #--------------------- Ratings of \"movie\" by similar users of \"user\" ---------------------\n",
        "            #print(user, movie)\n",
        "            try:\n",
        "                # compute the similar Users of the \"user\"        \n",
        "                user_sim = cosine_similarity(sample_train_sparse_matrix[user], sample_train_sparse_matrix).ravel()\n",
        "                top_sim_users = user_sim.argsort()[::-1][1:] # we are ignoring 'The User' from its similar users.\n",
        "                # get the ratings of most similar users for this movie\n",
        "                top_ratings = sample_train_sparse_matrix[top_sim_users, movie].toarray().ravel()\n",
        "                # we will make it's length \"5\" by adding movie averages to .\n",
        "                top_sim_users_ratings = list(top_ratings[top_ratings != 0][:5])\n",
        "                top_sim_users_ratings.extend([sample_train_averages['movie'][movie]]*(5 - len(top_sim_users_ratings)))\n",
        "                # print(top_sim_users_ratings, end=\"--\")\n",
        "\n",
        "            except (IndexError, KeyError):\n",
        "                # It is a new User or new Movie or there are no ratings for given user for top similar movies...\n",
        "                ########## Cold STart Problem ##########\n",
        "                top_sim_users_ratings.extend([sample_train_averages['global']]*(5 - len(top_sim_users_ratings)))\n",
        "                #print(top_sim_users_ratings)\n",
        "            except:\n",
        "                print(user, movie)\n",
        "                # we just want KeyErrors to be resolved. Not every Exception...\n",
        "                raise\n",
        "\n",
        "\n",
        "\n",
        "            #--------------------- Ratings by \"user\"  to similar movies of \"movie\" ---------------------\n",
        "            try:\n",
        "                # compute the similar movies of the \"movie\"        \n",
        "                movie_sim = cosine_similarity(sample_train_sparse_matrix[:,movie].T, sample_train_sparse_matrix.T).ravel()\n",
        "                top_sim_movies = movie_sim.argsort()[::-1][1:] # we are ignoring 'The User' from its similar users.\n",
        "                # get the ratings of most similar movie rated by this user..\n",
        "                top_ratings = sample_train_sparse_matrix[user, top_sim_movies].toarray().ravel()\n",
        "                # we will make it's length \"5\" by adding user averages to.\n",
        "                top_sim_movies_ratings = list(top_ratings[top_ratings != 0][:5])\n",
        "                top_sim_movies_ratings.extend([sample_train_averages['user'][user]]*(5-len(top_sim_movies_ratings))) \n",
        "                #print(top_sim_movies_ratings)\n",
        "            except (IndexError, KeyError):\n",
        "                #print(top_sim_movies_ratings, end=\" : -- \")\n",
        "                top_sim_movies_ratings.extend([sample_train_averages['global']]*(5-len(top_sim_movies_ratings)))\n",
        "                #print(top_sim_movies_ratings)\n",
        "            except :\n",
        "                raise\n",
        "\n",
        "            #-----------------prepare the row to be stores in a file-----------------#\n",
        "            row = list()\n",
        "            # add usser and movie name first\n",
        "            row.append(user)\n",
        "            row.append(movie)\n",
        "            row.append(sample_train_averages['global']) # first feature\n",
        "            #print(row)\n",
        "            # next 5 features are similar_users \"movie\" ratings\n",
        "            row.extend(top_sim_users_ratings)\n",
        "            #print(row)\n",
        "            # next 5 features are \"user\" ratings for similar_movies\n",
        "            row.extend(top_sim_movies_ratings)\n",
        "            #print(row)\n",
        "            # Avg_user rating\n",
        "            try:\n",
        "                row.append(sample_train_averages['user'][user])\n",
        "            except KeyError:\n",
        "                row.append(sample_train_averages['global'])\n",
        "            except:\n",
        "                raise\n",
        "            #print(row)\n",
        "            # Avg_movie rating\n",
        "            try:\n",
        "                row.append(sample_train_averages['movie'][movie])\n",
        "            except KeyError:\n",
        "                row.append(sample_train_averages['global'])\n",
        "            except:\n",
        "                raise\n",
        "            #print(row)\n",
        "            # finalley, The actual Rating of this user-movie pair...\n",
        "            row.append(rating)\n",
        "            #print(row)\n",
        "            count = count + 1\n",
        "\n",
        "            # add rows to the file opened..\n",
        "            reg_data_file.write(','.join(map(str, row)))\n",
        "            #print(','.join(map(str, row)))\n",
        "            reg_data_file.write('\\n')        \n",
        "            if (count)%1000 == 0:\n",
        "                #print(','.join(map(str, row)))\n",
        "                print(\"Done for {} rows----- {}\".format(count, datetime.now() - start))\n",
        "    print(\"\",datetime.now() - start)  "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "preparing 7333 tuples for the dataset..\n",
            "\n",
            "Done for 1000 rows----- 0:04:29.293783\n",
            "Done for 2000 rows----- 0:08:57.208002\n",
            "Done for 3000 rows----- 0:13:30.333223\n",
            "Done for 4000 rows----- 0:18:04.050813\n",
            "Done for 5000 rows----- 0:22:38.671673\n",
            "Done for 6000 rows----- 0:27:09.697009\n",
            "Done for 7000 rows----- 0:31:41.933568\n",
            " 0:33:12.529731\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "l1OSGIpEGHQF"
      },
      "source": [
        "__Reading from the file to make a test dataframe __"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "c7819189-2f17-428b-be85-e5af5b1a5ed4",
        "id": "maMuye0CGHQF",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 173
        }
      },
      "source": [
        "reg_test_df = pd.read_csv('/content/drive/My Drive/Netflix_recommender/reg_test_25.csv', names = ['user', 'movie', 'GAvg', 'sur1', 'sur2', 'sur3', 'sur4', 'sur5',\n",
        "                                                          'smr1', 'smr2', 'smr3', 'smr4', 'smr5',\n",
        "                                                          'UAvg', 'MAvg', 'rating'], header=None)\n",
        "reg_test_df.head(4)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>808635</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>941866</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1737912</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1849204</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      user  movie      GAvg      sur1  ...      smr5      UAvg      MAvg  rating\n",
              "0   808635     71  3.581679  3.581679  ...  3.581679  3.581679  3.581679       5\n",
              "1   941866     71  3.581679  3.581679  ...  3.581679  3.581679  3.581679       4\n",
              "2  1737912     71  3.581679  3.581679  ...  3.581679  3.581679  3.581679       3\n",
              "3  1849204     71  3.581679  3.581679  ...  3.581679  3.581679  3.581679       4\n",
              "\n",
              "[4 rows x 16 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "-q9xwxPUGHQH"
      },
      "source": [
        "-----------------------\n",
        "\n",
        "- __GAvg__ : Average rating of all the ratings \n",
        "\n",
        "\n",
        "- __Similar users rating of this movie__:\n",
        "    - sur1, sur2, sur3, sur4, sur5 ( top 5 simiular users who rated that movie.. )\n",
        "    \n",
        "\n",
        "\n",
        "- __Similar movies rated by this user__:\n",
        "    - smr1, smr2, smr3, smr4, smr5 ( top 5 simiular movies rated by this movie.. )\n",
        "\n",
        "\n",
        "- __UAvg__ : User AVerage rating\n",
        "\n",
        "\n",
        "- __MAvg__ : Average rating of this movie\n",
        "\n",
        "\n",
        "- __rating__ : Rating of this movie by this user.\n",
        "\n",
        "-----------------------"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "cYhXG7-DGHQI"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "77EkmNItGHQJ"
      },
      "source": [
        "<h3> 4.3.2 Transforming data for Surprise models</h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "tB-SzdjkGHQJ",
        "colab": {}
      },
      "source": [
        "from surprise import Reader, Dataset"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "4aqgl9VxGHQL"
      },
      "source": [
        "<h4> 4.3.2.1 Transforming train data </h4>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "v3vQZFeDGHQL"
      },
      "source": [
        "- We can't give raw data (movie, user, rating) to train the model in Surprise library.\n",
        "\n",
        "\n",
        "- They have a saperate format for TRAIN and TEST data, which will be useful for training the models like SVD, KNNBaseLineOnly....etc..,in Surprise.\n",
        "\n",
        "\n",
        "- We can form the trainset from a file, or from a Pandas  DataFrame. \n",
        "http://surprise.readthedocs.io/en/stable/getting_started.html#load-dom-dataframe-py "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "c1FdSfO2GHQM",
        "colab": {}
      },
      "source": [
        "# It is to specify how to read the dataframe.\n",
        "# for our dataframe, we don't have to specify anything extra..\n",
        "reader = Reader(rating_scale=(1,5))\n",
        "\n",
        "# create the traindata from the dataframe...\n",
        "train_data = Dataset.load_from_df(reg_train[['user', 'movie', 'rating']], reader)\n",
        "\n",
        "# build the trainset from traindata.., It is of dataset format from surprise library..\n",
        "trainset = train_data.build_full_trainset() "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "JxUW5EQqGHQT"
      },
      "source": [
        "<h4> 4.3.2.2 Transforming test data </h4>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "7Pf9lRjdGHQV"
      },
      "source": [
        "- Testset is just a list of (user, movie, rating) tuples. (Order in the tuple is impotant) "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "3f2972ec-a005-4895-8735-b90c4579e293",
        "id": "IacqW5k_GHQX",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "testset = list(zip(reg_test_df.user.values, reg_test_df.movie.values, reg_test_df.rating.values))\n",
        "testset[:3]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[(808635, 71, 5), (941866, 71, 4), (1737912, 71, 3)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "O8tfA9mFGHQa"
      },
      "source": [
        "<h2> 4.4 Applying Machine Learning models </h2>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "iQYKXNIcGHQb"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "haY7YB-KGHQb"
      },
      "source": [
        "-  Global dictionary that stores rmse and mape for all the models....\n",
        "\n",
        "    - It stores the metrics in a dictionary of dictionaries\n",
        "\n",
        "    > __keys__ : model names(string)\n",
        "\n",
        "    > __value__: dict(__key__ : metric, __value__ : value ) "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "7cf70b8c-8dd1-4233-e42c-3b77bc5ec2f0",
        "id": "fRaxkZt-GHQc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "models_evaluation_train = dict()\n",
        "models_evaluation_test = dict()\n",
        "\n",
        "models_evaluation_train, models_evaluation_test"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "({}, {})"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "SCooXq6hGHQf"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "73iu2FP-GHQg"
      },
      "source": [
        " > __Utility functions for running regression models__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "wjX8V_FMGHQh",
        "colab": {}
      },
      "source": [
        "# to get rmse and mape given actual and predicted ratings..\n",
        "def get_error_metrics(y_true, y_pred):\n",
        "    rmse = np.sqrt(np.mean([ (y_true[i] - y_pred[i])**2 for i in range(len(y_pred)) ]))\n",
        "    mape = np.mean(np.abs( (y_true - y_pred)/y_true )) * 100\n",
        "    return rmse, mape\n",
        "\n",
        "###################################################################\n",
        "###################################################################\n",
        "def run_xgboost(algo,  x_train, y_train, x_test, y_test, verbose=True):\n",
        "    \"\"\"\n",
        "    It will return train_results and test_results\n",
        "    \"\"\"\n",
        "    \n",
        "    # dictionaries for storing train and test results\n",
        "    train_results = dict()\n",
        "    test_results = dict()\n",
        "    \n",
        "    \n",
        "    # fit the model\n",
        "    print('Training the model..')\n",
        "    start =datetime.now()\n",
        "    algo.fit(x_train, y_train, eval_metric = 'rmse')\n",
        "    print('Done. Time taken : {}\\n'.format(datetime.now()-start))\n",
        "    print('Done \\n')\n",
        "\n",
        "    # from the trained model, get the predictions....\n",
        "    print('Evaluating the model with TRAIN data...')\n",
        "    start =datetime.now()\n",
        "    y_train_pred = algo.predict(x_train)\n",
        "    # get the rmse and mape of train data...\n",
        "    rmse_train, mape_train = get_error_metrics(y_train.values, y_train_pred)\n",
        "    \n",
        "    # store the results in train_results dictionary..\n",
        "    train_results = {'rmse': rmse_train,\n",
        "                    'mape' : mape_train,\n",
        "                    'predictions' : y_train_pred}\n",
        "    \n",
        "    #######################################\n",
        "    # get the test data predictions and compute rmse and mape\n",
        "    print('Evaluating Test data')\n",
        "    y_test_pred = algo.predict(x_test) \n",
        "    rmse_test, mape_test = get_error_metrics(y_true=y_test.values, y_pred=y_test_pred)\n",
        "    # store them in our test results dictionary.\n",
        "    test_results = {'rmse': rmse_test,\n",
        "                    'mape' : mape_test,\n",
        "                    'predictions':y_test_pred}\n",
        "    if verbose:\n",
        "        print('\\nTEST DATA')\n",
        "        print('-'*30)\n",
        "        print('RMSE : ', rmse_test)\n",
        "        print('MAPE : ', mape_test)\n",
        "        \n",
        "    # return these train and test results...\n",
        "    return train_results, test_results\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "pf0xjvJRGHQj"
      },
      "source": [
        "> __Utility functions for Surprise modes__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "yis8mCQqGHQk",
        "colab": {}
      },
      "source": [
        "# it is just to makesure that all of our algorithms should produce same results\n",
        "# everytime they run...\n",
        "\n",
        "my_seed = 15\n",
        "random.seed(my_seed)\n",
        "np.random.seed(my_seed)\n",
        "\n",
        "##########################################################\n",
        "# get  (actual_list , predicted_list) ratings given list \n",
        "# of predictions (prediction is a class in Surprise).    \n",
        "##########################################################\n",
        "def get_ratings(predictions):\n",
        "    actual = np.array([pred.r_ui for pred in predictions])\n",
        "    pred = np.array([pred.est for pred in predictions])\n",
        "    \n",
        "    return actual, pred\n",
        "\n",
        "################################################################\n",
        "# get ''rmse'' and ''mape'' , given list of prediction objecs \n",
        "################################################################\n",
        "def get_errors(predictions, print_them=False):\n",
        "\n",
        "    actual, pred = get_ratings(predictions)\n",
        "    rmse = np.sqrt(np.mean((pred - actual)**2))\n",
        "    mape = np.mean(np.abs(pred - actual)/actual)\n",
        "\n",
        "    return rmse, mape*100\n",
        "\n",
        "##################################################################################\n",
        "# It will return predicted ratings, rmse and mape of both train and test data   #\n",
        "##################################################################################\n",
        "def run_surprise(algo, trainset, testset, verbose=True): \n",
        "    '''\n",
        "        return train_dict, test_dict\n",
        "    \n",
        "        It returns two dictionaries, one for train and the other is for test\n",
        "        Each of them have 3 key-value pairs, which specify ''rmse'', ''mape'', and ''predicted ratings''.\n",
        "    '''\n",
        "    start = datetime.now()\n",
        "    # dictionaries that stores metrics for train and test..\n",
        "    train = dict()\n",
        "    test = dict()\n",
        "    \n",
        "    # train the algorithm with the trainset\n",
        "    st = datetime.now()\n",
        "    print('Training the model...')\n",
        "    algo.fit(trainset)\n",
        "    print('Done. time taken : {} \\n'.format(datetime.now()-st))\n",
        "    \n",
        "    # ---------------- Evaluating train data--------------------#\n",
        "    st = datetime.now()\n",
        "    print('Evaluating the model with train data..')\n",
        "    # get the train predictions (list of prediction class inside Surprise)\n",
        "    train_preds = algo.test(trainset.build_testset())\n",
        "    # get predicted ratings from the train predictions..\n",
        "    train_actual_ratings, train_pred_ratings = get_ratings(train_preds)\n",
        "    # get ''rmse'' and ''mape'' from the train predictions.\n",
        "    train_rmse, train_mape = get_errors(train_preds)\n",
        "    print('time taken : {}'.format(datetime.now()-st))\n",
        "    \n",
        "    if verbose:\n",
        "        print('-'*15)\n",
        "        print('Train Data')\n",
        "        print('-'*15)\n",
        "        print(\"RMSE : {}\\n\\nMAPE : {}\\n\".format(train_rmse, train_mape))\n",
        "    \n",
        "    #store them in the train dictionary\n",
        "    if verbose:\n",
        "        print('adding train results in the dictionary..')\n",
        "    train['rmse'] = train_rmse\n",
        "    train['mape'] = train_mape\n",
        "    train['predictions'] = train_pred_ratings\n",
        "    \n",
        "    #------------ Evaluating Test data---------------#\n",
        "    st = datetime.now()\n",
        "    print('\\nEvaluating for test data...')\n",
        "    # get the predictions( list of prediction classes) of test data\n",
        "    test_preds = algo.test(testset)\n",
        "    # get the predicted ratings from the list of predictions\n",
        "    test_actual_ratings, test_pred_ratings = get_ratings(test_preds)\n",
        "    # get error metrics from the predicted and actual ratings\n",
        "    test_rmse, test_mape = get_errors(test_preds)\n",
        "    print('time taken : {}'.format(datetime.now()-st))\n",
        "    \n",
        "    if verbose:\n",
        "        print('-'*15)\n",
        "        print('Test Data')\n",
        "        print('-'*15)\n",
        "        print(\"RMSE : {}\\n\\nMAPE : {}\\n\".format(test_rmse, test_mape))\n",
        "    # store them in test dictionary\n",
        "    if verbose:\n",
        "        print('storing the test results in test dictionary...')\n",
        "    test['rmse'] = test_rmse\n",
        "    test['mape'] = test_mape\n",
        "    test['predictions'] = test_pred_ratings\n",
        "    \n",
        "    print('\\n'+'-'*45)\n",
        "    print('Total time taken to run this algorithm :', datetime.now() - start)\n",
        "    \n",
        "    # return two dictionaries train and test\n",
        "    return train, test"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "jmKDnEaYGHQn"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "T1new257GHQo"
      },
      "source": [
        "<h3> 4.4.1 XGBoost with initial 13 features </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "k9O7ECAiGHQo",
        "colab": {}
      },
      "source": [
        "import xgboost as xgb"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "e2dc745d-84d8-4468-cba2-6417fc5e6277",
        "id": "ul00N5aoGHQs",
        "colab": {}
      },
      "source": [
        "# prepare Train data\n",
        "x_train = reg_train.drop(['user','movie','rating'], axis=1)\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# Prepare Test data\n",
        "x_test = reg_test_df.drop(['user','movie','rating'], axis=1)\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "# initialize Our first XGBoost model...\n",
        "first_xgb = xgb.XGBRegressor(silent=False, n_jobs=13, random_state=15, n_estimators=100)\n",
        "train_results, test_results = run_xgboost(first_xgb, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "models_evaluation_train['first_algo'] = train_results\n",
        "models_evaluation_test['first_algo'] = test_results\n",
        "\n",
        "xgb.plot_importance(first_xgb)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model..\n",
            "Done. Time taken : 0:00:01.795787\n",
            "\n",
            "Done \n",
            "\n",
            "Evaluating the model with TRAIN data...\n",
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.0761851474385373\n",
            "MAPE :  34.504887593204884\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Ib3tzN3-GHQu"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "nByY611nGHQv"
      },
      "source": [
        "<h3> 4.4.2 Suprise BaselineModel </h3>\n",
        "    \n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "t6t9A1gJGHQv",
        "colab": {}
      },
      "source": [
        "from surprise import BaselineOnly "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "NvKt4O0RGHQx"
      },
      "source": [
        "__Predicted_rating : ( baseline prediction ) __\n",
        "\n",
        "    -  http://surprise.readthedocs.io/en/stable/basic_algorithms.html#surprise.prediction_algorithms.baseline_only.BaselineOnly \n",
        " >$   \\large {\\hat{r}_{ui} = b_{ui} =\\mu + b_u + b_i} $\n",
        "\n",
        "\n",
        "- $\\pmb \\mu $ : Average of all trainings in training data.\n",
        "- $\\pmb b_u$ : User bias\n",
        "- $\\pmb b_i$ : Item bias (movie biases) "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "57zPul1RGHQy"
      },
      "source": [
        "__Optimization function ( Least Squares Problem ) __\n",
        "\n",
        "    - http://surprise.readthedocs.io/en/stable/prediction_algorithms.html#baselines-estimates-configuration \n",
        "\n",
        "> $ \\large \\sum_{r_{ui} \\in R_{train}} \\left(r_{ui} - (\\mu + b_u + b_i)\\right)^2 +\n",
        "\\lambda \\left(b_u^2 + b_i^2 \\right).\\text {        [mimimize } {b_u, b_i]}$ "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "118c4d00-a202-460b-dd78-a89e664d5570",
        "id": "dW7WNz8UGHQy",
        "colab": {}
      },
      "source": [
        "\n",
        "# options are to specify.., how to compute those user and item biases\n",
        "bsl_options = {'method': 'sgd',\n",
        "               'learning_rate': .001\n",
        "               }\n",
        "bsl_algo = BaselineOnly(bsl_options=bsl_options)\n",
        "# run this algorithm.., It will return the train and test results..\n",
        "bsl_train_results, bsl_test_results = run_surprise(my_bsl_algo, trainset, testset, verbose=True)\n",
        "\n",
        "\n",
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "models_evaluation_train['bsl_algo'] = bsl_train_results \n",
        "models_evaluation_test['bsl_algo'] = bsl_test_results"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            "Estimating biases using sgd...\n",
            "Done. time taken : 0:00:00.822391 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:00:01.116752\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.9347153928678286\n",
            "\n",
            "MAPE : 29.389572652358183\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.074418\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.0730330260516174\n",
            "\n",
            "MAPE : 35.04995544572911\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:00:02.014073\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "mfGIOsfVGHQ0"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "CvPjDdN8GHQ1"
      },
      "source": [
        "<h3> 4.4.3 XGBoost with initial 13 features + Surprise Baseline predictor </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "eKPn3wK0GHQ2"
      },
      "source": [
        "__Updating Train Data__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "ad5d0309-054c-406d-dc87-c8384a5fbf12",
        "id": "KqCj5GvzGHQ3",
        "colab": {}
      },
      "source": [
        "# add our baseline_predicted value as our feature..\n",
        "reg_train['bslpr'] = models_evaluation_train['bsl_algo']['predictions']\n",
        "reg_train.head(2) "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>bslpr</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>53406</td>\n",
              "      <td>33</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.370370</td>\n",
              "      <td>4.092437</td>\n",
              "      <td>4</td>\n",
              "      <td>3.898982</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>99540</td>\n",
              "      <td>33</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.555556</td>\n",
              "      <td>4.092437</td>\n",
              "      <td>3</td>\n",
              "      <td>3.371403</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    user  movie      GAvg  sur1  sur2  sur3  sur4  sur5  smr1  smr2  smr3  \\\n",
              "0  53406     33  3.581679   4.0   5.0   5.0   4.0   1.0   5.0   2.0   5.0   \n",
              "1  99540     33  3.581679   5.0   5.0   5.0   4.0   5.0   3.0   4.0   4.0   \n",
              "\n",
              "   smr4  smr5      UAvg      MAvg  rating     bslpr  \n",
              "0   3.0   1.0  3.370370  4.092437       4  3.898982  \n",
              "1   3.0   5.0  3.555556  4.092437       3  3.371403  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Yrfrqx8oGHQ5"
      },
      "source": [
        "__Updating Test Data__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "194d8d7f-a65a-4b67-f2a8-fd80daf2c5f4",
        "id": "KPfnztxzGHQ5",
        "colab": {}
      },
      "source": [
        "# add that baseline predicted ratings with Surprise to the test data as well\n",
        "reg_test_df['bslpr']  = models_evaluation_test['bsl_algo']['predictions']\n",
        "\n",
        "reg_test_df.head(2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>bslpr</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>808635</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5</td>\n",
              "      <td>3.581679</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>941866</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "      <td>3.581679</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg      sur1      sur2      sur3      sur4      sur5  \\\n",
              "0  808635     71  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "1  941866     71  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "\n",
              "       smr1      smr2      smr3      smr4      smr5      UAvg      MAvg  \\\n",
              "0  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "1  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "\n",
              "   rating     bslpr  \n",
              "0       5  3.581679  \n",
              "1       4  3.581679  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 45
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "e51e11b3-f92e-4af9-dc65-f981eabab9c7",
        "id": "x7NO8ManGHQ7",
        "colab": {}
      },
      "source": [
        "# prepare train data\n",
        "x_train = reg_train.drop(['user', 'movie','rating'], axis=1)\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# Prepare Test data\n",
        "x_test = reg_test_df.drop(['user','movie','rating'], axis=1)\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "# initialize Our first XGBoost model...\n",
        "xgb_bsl = xgb.XGBRegressor(silent=False, n_jobs=13, random_state=15, n_estimators=100)\n",
        "train_results, test_results = run_xgboost(xgb_bsl, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "models_evaluation_train['xgb_bsl'] = train_results\n",
        "models_evaluation_test['xgb_bsl'] = test_results\n",
        "\n",
        "xgb.plot_importance(xgb_bsl)\n",
        "plt.show()\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model..\n",
            "Done. Time taken : 0:00:02.388635\n",
            "\n",
            "Done \n",
            "\n",
            "Evaluating the model with TRAIN data...\n",
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.0763419061709816\n",
            "MAPE :  34.491235560745295\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAoAAAAHgCAYAAAA10dzkAAAgAElEQVR4nOzde5gU9Zk2/lI07+7qZl83TthwmO6u+t4Pg6NyMgKCZng1bIxAEiIaAxolUaJhE9S8YDzhcc1qYpQchF9iwLy7m4RIZBNQiIeMibCCvcYDMOARBEUnIjCA0KHl+f1B9djT9AxjeoaZp77357qe63K6q6tr5na+c09NdRMERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERETUOgAXioiWG+fcdzrreaMoOkVEbkin0/+7s56jEiKyHsC8rj6Ov5Zz7rj465vu6mMhIiKibqZQAAFc6JwbVjxhGFZ34vN+U0S0uxYU59wg51zU1cfx13LOnR2X+LquPhYiIiLqZgoFMAzDkw7x83ZKAezTp8/fBkFwWEfu05gjgyA4ggWQiIiIWvUBCuBhzrnLADwDYDeArQDuF5GwZH+fBPBfADYB2APgJQBzevXqdWxhGxG5oZU/OdfF96uI3FB6AKV/li0cu3NuNICfisif41L5N0EQBGEYAsB/AmgUkZyINDjnvtaer0vpcznn6uIzpV8E8G8isllEdgL4bSaT6Xnsscf+vXPu/xORt0XkbQBzq6qqji7Zp4rIDwBMAfCCiOQArBGRL5Q+fxRFx8dfx63x1/EZEflS8TaFY3LOnS8i3xWR1wHsA/DVcl9fABe2N6PinJxztQB+DmA7gLcA/DQMw38oOeTDAfxL0f8f2wA86ZwbV/J5nQvgvwHsir9+S51zg9qTCREREXWQoj8BDw2C4IiSaRaXm784577jnPtnAOeJSAOANzOZTM+i/X3VOXeVc24sgNNE5IK4vKwN9p+dCpxzfQDMiovJ54r+7PzhIPjgBTAuMnOiKPoUgM8HQdDDOXccgG0i8pxz7nwAn3TOfQfAe865mQf7urRWAOPb58ZfgykiskNEHgPwOwB3xOVqOoA8gFkl+1QArwFYLSJfcM6NFZGH4pJ1dmG7TCbTT0SaALwUl7tPA/jP+PHTS48p/vx/FX/Nz8pkMj1F5Fvxfi8r+vpWtTej+HgLRX2tiNwoImc45y6PS+NPSz63nwHYJyI/ds6Ni6LoUyJyNYCvF21zdVxQ7wVwloh8DsByEdnpnDvuYJkQERFRB2nrRSBBXAKdc8PiMnFF8WPjIvcugH9rZfeHBUFwRBiG1fHjm88GtfUn4L/iDOB9ZT6vJQA2Fkpl0TF/H8Du6urqY9r4srR1BvA3Jc/zvfj2u0se/4CIbCn9vAC8W1yYgyDoERfpF4v2+XMAe6Io6lvy+AcB7CqcfSs6psdLj/8D/Am41YwKBRDA/y05jh8C2B0/NhCRU+PtbmntSaIo6isie0tLcVVV1dHx2dRfHuQ4iYiIqKMUlajzwzA8qXiKtrkFwL4oij4alJwljP+ct6KwbRRFHwUwG8BGAO+V/Il3RtE+O7IAtvgzYzqd/puistHieKMoOlNENIqiM9v6urRxBvCSku0uKfwZuvh259y/iogW/xk4Lkm/LfNchT+19ok/r7cALC7dzjl3Tnzsnyo+puKzbEXbtloA25tR4bgymUy/4sfHZz61UGQLn2sqlfrYAV/I94/nK/L+pQYtMhGRXwB4q7XHEhERUQdDO64BFJEft3aWMC4gL8ebHh5fA9YI4F+cc3VhGH4cwNDSUteRBTAMw4+XbNe7reMtFN62vi6tFcDiP9W29fUrlKeSax9VRH5c+lxF1+wNiD/Oi8hPSrcLw3BkfAwTS45pQum2bRTAdmdU7nMo/pwL2YnIjwHkgzZefAPgmoP8P/Rea48lIiKiDtbOAnhbfAZwROlZwnhOCIIgyGQyJ8Y/0Fu8WCGKIvdBCiCAPSJyW5nj2NFKAWxx7L169fq7+Bq8ua0c70m9e/f+SFtfl84qgB1xBtA5989tHVN8X9kC+EEyam8BbM8ZwMJZQwCfby2T1h5LREREHaw9BTCKohFxmTinrX2FYXhCXCJavKoVwB1lCuC/xIWgf+l+RGRtaQGKouj/xNvPa8+xA3gYwDNBEHyorWNuTScWwNauAXypaJ//CWB33759e5U816Jy1wC2UgDHlvtT9wfJ6AOcASxcA3hT6XEUpNPptOz/s/z01rYhIiKiQ6Q9BTDebg6AXQBuBzAmiqJR2P+WKD9yzl0ab3YkgJfi8nSec+6fReQHIrKutFwUXb822zk3PAzDk4499ti/j5/rmvjVojeFYXh6XBbXYf9bi8xrz7HH/xLGOwBWYP+bXNc558Y65y4XkccO9nXpxAJY9lXAURSdW9iu8CpgEVnnnJsYRdGZAP5dSl6Q0VYBTKVSmfhr/kAYhiOLznq2O6P2FsB425/Fmc2Jv86jnXMzAPxL0Tbfikvg7CiKPuuc+4Rz7hzn3HdE5MaDZUJEREQdpL0FMN72IgBPyv73b3sX+9+m5L4oioYUbdMfwO/iAvOOiMyPXwF6wHV98Z8OXy+8EKHoz5UfAvBvcVl6F0C9iAwoLWUHO/Z0Op3G/rcc2SQif8H+9wNcBuCag32unVUAReQHzrlL4xL2l/js3xdLnz9+H8DfYP9b2eTi6/YuLPn6tVoA42P7hoi8El9TWPw+gO3K6IMUwCAIDnfOTQPwfHy82wAsBzCm5Jg/I/vfNmd7/Kf+9QB+FYbh6eU+ByIiIiLTCgWwq4+DiIiIiA4RFkAiIiIiz7AAEhERERERERERERERERERERERERERERERERERdScrVqz4SDab/eLKlStPyWazgzkcDofD4XT/iX9uf3HFihVt/tvhRGVls9kvZrNZ5XA4HA6HY3IO+Nd6iA4q/g1CGxsbddeuXRwD09TUpNlsVpuamrr8WDjMLKnDzOyNb5m9/fbbms1mdeXKlad0dZcgg7L7TyVrU1OTkg35fF6z2azm8/muPhRqJ2ZmDzOzx7fMdu3aVTgDOLiruwQZxAJoj2+LXBIwM3uYmT2+ZcYCSBVhAbTHt0UuCZiZPczMHt8yYwGkirAA2uPbIpcEzMweZmaPb5mxAFJFWADt8W2RSwJmZg8zs8e3zFgAqSIsgPb4tsglATOzh5nZ41tmLIBUERZAe3xb5JKAmdnDzOzxLTMWQKoIC6A9vi1yScDM7GFm9viWGQsgVYQF0B7fFrkkYGb2MDN7fMuMBZAqwgJoj2+LXBIwM3uYmT2+ZcYCSBVhAbTHt0UuCZiZPczMHt8yYwGkirAA2uPbIpcEzMweZmaPb5mxAFJFWADt8W2RSwJmZg8zs8e3zFgAqSIsgPb4tsglATOzh5nZ41tmLIBUERZAe3xb5JKAmdnDzOzxLTMWQKoIC6A9vi1yScDM7GFm9viWGQsgVYQF0B7fFrkkYGb2MDN7fMuMBZAqwgJoj2+LXBIwM3uYmT2+ZcYCSBVhAbTHt0UuCZiZPczMno7MbPbs2Tp+/HgdOHCgDhs2TC+99FJ9+eWXW2yTy+X0pptu0pNPPlkHDBigU6ZM0c2bNzffv2DBAhWRsvP2229XfIwsgFQRFkB7+IPJHmZmDzOzpyMzmzx5si5YsEBfeOEFbWho0EsuuUTr6up0165dzdtcf/31euqpp+qyZct09erVev755+u4ceOan3/37t3a2NjYYiZPnqyTJk2q+PhUWQBNAFAP4K7S26Mo+qyIaPFtffr0+VsAW0XknT59+vxtZx8bC6A9/MFkDzOzh5nZ05mZbdmyRUVEV65cqaqqTU1NWltbq4sXL27e5s0339Samhr9wx/+0Oo+amtr9YEHHuiQY2IBNOCDFEDn3CQR+aOIPOGcm9jZx1YogHW3LdHUjEUcDofD4Xg5bVm/fr2KiK5bt05VVZcvX64iotu2bWux3dixY/Xuu+8uu497771XhwwZort372YB9MUHKYAAfg9gCoCvishjJff9XER+UbKbI0XkbQAXBUEQHHvssX8P4D8A7BKRzc65y1t7/iBgAeRwOBwOJzWj9QK4b98+nTJlip533nnNt/3mN7/R2traA7a96KKL9Lrrriu7n09/+tM6c+bMiotfAQugAe0tgM65CMCe6urqY/r06fOPAPaISFi0nzEA3q2qqjq65LbdzrkPB0EQiMiPRWR9GIanR1F0PIBfi0gTCyCHw+FwOK1PPp8vOzNnztS6ujrdtGlT820LFy7U2traA7b90pe+pNdee+0Bt2ezWRURffbZZ1t9ng86TU1NLIDd3QcogLeKyAOFj0VkIYBbih5ypIj82Tl3ftG+/1NE5gfB/rN/IvIX59zZhfvDMPwHALtYADkcDofDaX3iMtVipk6dqsOGDdMHH3ywxe333XefiojW19e3uP2MM87Qq6666oD9XHLJJfrJT36y7HN0wLAAdlftLIA9AGwC8PnC/c65swFsDIKgR9G+fgRgSRAEQc+ePY8CsMs595kgCAIRGSAiGoZhdcnzP80CyOFwOBxO61N8dm3v3r16ww036MiRI/Wll1464Ozb1q1b9bjjjtPf/va3zbdt3rxZa2pqtL6+vsW227dv14EDB+p9993XYWf/eAbQCAC/ATC3zO0XAtgeBEEgIp8WEQWQLx4R0SiKziw8JoqiEQDyURR91Dk3UUS2BEHwoXh/A+Pt+5Y8z59YADkcDofDaX2KzZw5U4cMGaIrVqxo8TYuxS/guP766/W0007T5cuX6+rVq/WCCy5o8TYwBfPnz9cTTjjhgBeM8BpADwC4HcDzpbeLyA9FZGW8zQIAP4+i6PjiAfDvAO4vethhIvIqgH8RkQcBzC7cUfgTcMlZxA+LyM6DFUC+DYwdfHsKe5iZPczMno7MrLU3cF6wYEHzNnv27Gl+I+gTTzxRp0yZom+88cYB+zr33HP1iiuuqPiYSrEAGpBOp9MA3o0L34AwDMU59zUAe5xzE5xzVSLylyiKPlX6WACfjK/rqyrc5py7FcBqEdkbhuHI4u3jF4G8EkXRKOdcLYD74xeBfK/csbEA2sMfTPYwM3uYmT2+ZcYCaEQURUMALAHwFoDtIvKUiHwhCIJARK4EsDUIgiPLPPQIEdninLuicINz7rj4t5H1QRAcVrxxK28Ds0JEbit3XCyA9vi2yCUBM7OHmdnjW2YsgNSm+IUi25xzXy53PwugPb4tcknAzOxhZvb4lhkLILXgnBsE4DznXOScGxy/lcy2Xr16HVtuexZAe3xb5JKAmdnDzOzxLTMWQGohLoD/IyI7ReQdAA+HYXhCa9uzANrj2yKXBMzMHmZmj2+ZsQBSRVgA7fFtkUsCZmYPM7PHt8xYAKkiLID2+LbIJQEzs4eZ2eNbZiyAVBEWQHt8W+SSgJnZw8zs8S0zFkCqCAugPb4tcknAzOxhZvb4lhkLIFWEBdAe3xa5JGBm9jAze3zLjAWQKsICaI9vi1wSMDN7mJk9vmXGAkgVYQG0x7dFLgmYmT3MzB7fMmMBpIqwANrj2yKXBMzMHmZmj2+ZsQBSRVgA7fFtkUsCZmYPM7PHt8xYAKkiLID2+LbIJQEzs4eZ2eNbZiyAVBEWQHt8W+SSgJnZw8zs8S0zFkCqCAugPb4tcknAzOxpK7OVK1fqlClTdMSIESoi+vDDD7e4f8aMGSoiLWbChAkttvnFL36hkyZN0kGDBqmI6Pbt2zv18/GBb99nLIBUERZAe3xb5JKAmdnTVmb19fV655136tKlS1stgF/+8pe1sbGxebZu3dpim7lz5+qcOXN0zpw5LIAdxLfvMxbAbgzAPBFRALPL3Pej+L55xbdHUXQKgPcALDkUx8gCaI9vi1wSMDN72ptZawXw0ksvbdfzPPnkkyyAHcS37zMWwG4MwDwArwHY1qdPn78t3J5Op/8GwFYAG0oLoIj8BMBdIrIzDMPqzj5GFkB7fFvkkoCZ2VNpARwyZIgOGzZMR48erddcc42+/fbbZR/PAthxfPs+YwHsxuIzgAtF5Dnn3MSi278oIs+JyMLiAtizZ8+jRKQpk8n0E5FfOOeuL9rd4QA2Afhq8XM45wbH15iEQRAE6XS6RkSeALAHwBoROUNENIqiz5Y7xkIBrLttiaZmLOJwOBxOybSlXAFcvHix/v73v9d169bpo48+quPGjdOzzjpLc7ncAY9nAew4LIDUbRQKoHPuchF5pHC7iDzinJtWWgCdc5NF5Kn4sWNE5NUgCA4ruv87IvLH4udwzn0HwPL4w8NFZC2A34nIgDAMRwJYwQLI4XA4f/3k8/lWR0R0yZIlbW6zefNmPe644/Shhx464L7ly5eriOg777zT5j44B59cLqfZbFZzuVyXH8uhmKamJhbA7qpQAHv16nUsgD3pdDqdyWRSAHb36tXr2NICKCLLAHwj/vAIEfmziJxRuN85NwjAvkwmk4pvOhzAJufcZUEQBFEUfUpE9qbT6X8q2ifPAHI4HE4FE/+QLTsiovfcc0+b22SzWT311FP1+uuvP+D2n/3sZyoi+vjjjx90HxxOK8MC2N0UCmD83wucczNF5AYA9wdBEBQXwPjPvnszmUzPwuNF5AcA/rNkn2ucc1cFQRBEUTRKRP7Sq1evY+P7viEirxRv75z7MAsgh8Ph/PXT1lmY9pwBfPvtt/X444/XBQsWHHAfzwB23PAMIHUbJQXwLBF5NZ5PB0HLAgjg9vhVwfmieQ/Anurq6mOK9nmNiDwXP/7HAH5buM85Nw3Ay8XHwALI4XA4lU2pnTt36po1a3TNmjUqIjp37lxds2aNvv7667pz50799re/rU8//bRu3LhRn3zyST333HP11FNP1R07djTvo7GxUdesWaPz589XEdGnnnpK16xZc8DbxVD75fO8BpC6ieICGARBDxF5XUReD4KgRxC0KIBHAHjTOXdFFEXHF4+IrBORqYV9plKpTFzohgDYKiJfKNxX+BNw8VnEMAxPb08B5KuA7fBtkUsCZmZPW5kVXrhROjNmzNDdu3fr5MmTddiwYVpbW6t1dXU6Y8YMfeONN1rsY9asWWX3sWDBgkP1KSaOb99nLIDdWEkBDJxzH3bOfbjwcaEARlH0WRHJhWH4D6X7cM7dCuBPxbfF1wo+IyI7it9eJthfMtcCWJLJZE6MomgEgCdFRJ1znyl3jCyA9vi2yCUBM7OHmdnjW2YsgN1YaQEsVSiAAH4LYHG5bQpv8+KcG1x022XxbfeVbl94GxgRyYlIQ/xqYnXO/XO5/bMA2uPbIpcEzMweZmaPb5mxAFKboigaERfAqNz9LID2+LbIJQEzs4eZ2eNbZiyA1IKIfA7AJ9PpdFpEzgCwWkSeaG17FkB7fFvkkoCZ2cPM7PEtMxZAakFELgDwYvwvgWwCMK93794faW17FkB7fFvkkoCZ2cPM7PEtMxZAqggLoD2+LXJJwMzsYWb2+JYZCyBVhAXQHt8WuSRgZvYwM3t8y4wFkCrCAmiPb4tcEjAze5iZPb5lxgJIFWEBtMe3RS4JmJk9zMwe3zJjAaSKsADa49silwTMzB5mZo9vmbEAUkVYAO3xbZFLAmZmDzOzx7fMWACpIiyA9vi2yCUBM7OHmdnjW2YsgFQRFkB7fFvkkoCZ2cPM7PEtMxZAqggLoD2+LXJJwMzsYWb2+JYZCyBVhAXQHt8WuSRgZvYwM3t8y4wFkCrCAmiPb4tcEjAze5iZPb5lxgJIFWEBtMe3RS4JmJk9zMwe3zJjAaSKsADa09oiN2vWLBWRFnPKKac037906VKdPHmynnzyySoiumbNmkN96N7y7QdTEjAze3zLjAWQWgBwDYDlAN4FsO1g27MA2tNWATzrrLO0sbGxebZs2dJ8/wMPPKDf//73df78+SyAh5hvP5iSgJnZ41tmLIBU8KEgCAIRudE5d7mIfJcFMJnaKoDjxo076OM3btzIAniI+faDKQmYmT2+ZcYCaJhz7mwAzwPYLSJbROSRnj17HgWgHsBdxduKyEIA84o+Xg/gWgDzAGx3zt1XvD2ACz9IAay7bYmmZiziGJtis2bN0gEDBuiIESN01KhROm3aNH3ttdcOWDRYAA89334wJQEzs8e3zFgAjUqlUh8Tkb3OucvT6XQ6DMMTnHOXVVVVHf0BCuB2AN+MoshFUeSKt2cB9GOK1dfX65IlS3Tt2rW6bNkynTRpkp5yyin6zjvvtNiOBfDQ8+0HUxIwM3t8y4wF0Cjn3GAR0Uwmkyq9r70FUEQeaG3/LIB+TD6fb3Wampp0+PDh+pOf/KTF7Rs2bFAR0eeff77Nx3M6bnK5nGazWc3lcl1+LBxmltTxLbOmpiYWQKN6iMgjItIE4FcicnF1dfUxQdD+AgjgmtZ2zgLox8Tf/K3O+PHj9dJLL21x20MPPaQiovfff/9BH8/hcDicbj8sgAYdFkXRCBG5UUSeA9CYSqUyIvIYgLuLNwSwuLQAOuemtbZjFkA/pq3fDt9991099dRTddasWS1u5xnAQz++nZlIwjAze+NbZjwDmBw9AGxyzl0hIr8Ukfkl923ozALIVwHbkc+Xv87l29/+tq5YsUJfe+01feaZZ3TKlCk6aNAg3bRpk6qqbt26VdesWaP19fUqIrp48WJds2aNNjY2dsWn4ZXWMqPui5nZ41tmvAbQKABDReTqMAxPCsOw2jk3QURyURSdCWAKgF0Azkqn0zUA5sQv+JhXeHxrBTAMw2oAA51z14vIDgADAQysqqo6utxxsADa09oiN23aNB0xYoTW1tbqyJEjderUqfriiy82379gwYID3ihaRHTWrFmH+lPwjm8/mJKAmdnjW2YsgEYB6A9gCYBGAHtEZJ2ITI3vPhLAj0RkC4C3nHNXlbsGsFwBBDCv3A9551xdueNgAbTHt0UuCZiZPczMHt8yYwGkirAA2uPbIpcEzMweZmaPb5mxAFJFWADt8W2RSwJmZg8zs8e3zFgAqSIsgPb4tsglATOzh5nZ41tmLIBUERZAe3xb5JKAmdnDzOzxLTMWQKoIC6A9vi1yScDM7GFm9viWGQsgVYQF0B7fFrkkYGb2MDN7fMuMBZAqwgJoj2+LXBIwM3uYmT2+ZcYCSBVhAbTHt0UuCZiZPczMHt8yYwGkirAA2uPbIpcEzMweZmaPb5mxAFJFWADt8W2RSwJmZg8zs8e3zFgAqSIsgPb4tsglATOzh5nZ41tmLIBUERZAe3xb5JKAmdnDzOzxLTMWQKoIC6A9vi1yScDM7GFm9viWGQsgVYQF0B7fFrkkYGb2MDN7fMuMBZAqwgJoj2+LXBIwM3uYmT2+ZcYCSBVhAezeZs+erSKit9xyS/NthUVu7969+uUvf1lFRB9++OEuPEo6GN9+MCUBM7PHt8xYAKkiLIDd17PPPqujRo3SsWPHli2A9957r37lK19hATTAtx9MScDM7PEtMxZAapZOp9MA7hWRVwHsBvCyiNwYBMGHWnsMC2D3tHPnTh09erQuW7ZMJ02adEABXLBggZ522mna2NjIAmiAbz+YkoCZ2eNbZiyAVPChKIo+BWCuc260iITOuXEA3nLOfae1B7EAdk/Tp0/XW2+9VVX1gAK4Y8cOHTVqlC5dulRVlQXQAN9+MCUBM7PHt8xYAA1zzp0N4HkAu0Vki4g80rNnz6MA1AO4q3hbEVkIYF7Rx+sBXAtgHoDtzrn7yj0HgP8rIq+0dgyFAlh32xJNzVjE6YIptWjRIh0zZozu2bNHVQ8sgNdcc41ecsklzYscC2D359sPpiRgZvb4lhkLoFGpVOpjIrLXOXd5Op1Oh2F4gnPusqqqqqM/QAHcDuCbURS5KIpcuecBcIuIZFs7DhbArp98Pt88Gzdu1OHDh+uqVauab5s4caLefPPNms/n9Xe/+52eccYZ+sQTT2gul9N8Pq8iokuWLGmxH073mlwup9lstjkzTvcfZmZvfMusqamJBdAi59xgEdFMJpMqva+9BVBEHjjIc0Tx2cGvtLYNC2DXT/wNrNlsVu+55x4VEa2pqWkeEdF+/fppTU2NTp06tfm/S+//zGc+02JfHA6Hw/FiWACN6SEij4hIE4BficjF1dXVxwRB+wsggGta23nfvn17AXhRRH7S1kFkWQC7fIp/o9u+fbs2NDS0mPHjx+uVV16pDQ0NunnzZl21apU+8MADumrVKm1oaFAR0blz5+r69eu7/DdSTvnx7cxEEoaZ2RvfMuMZQNsOi6JohIjcKCLPAWhMpVIZEXkMwN3FGwJYXFoAnXPTyu20b9++vURknYj8LAiCw9s6ABbArp+DKfcq4Gz2/etceA1g91eaGXV/zMwe3zLjNYDJ0QPAJufcFSLySxGZX3LfhvYUQBHpDeAFAD8PgqDHwZ40y1cBd3ssgPb59oMpCZiZPb5lxgJoFIChInJ1GIYnhWFY7ZybICK5KIrOBDAFwC4AZ6XT6RoAc+IXfMwrPL5cASz6s++jItI7nU7/U2FaOw4WQHt8W+SSgJnZw8zs8S0zFkCjAPQHsARAI4A98Z9sp8Z3HwngRyKyJX4fv6vKXQNYWgABXCgiWm5aOw4WQHt8W+SSgJnZw8zs8S0zFkCqCAugPb4tcknAzOxhZvb4lhkLIFWEBdAe3xa5JGBm9jAze3zLjAWQKsICaI9vi1wSMDN7mJk9vmXGAkgVYQG0x7dFLgmYmT3MzB7fMmMBpIqwANrj2yKXBMzMHmZmj2+ZsQBSRVgA7fFtkUsCZmYPM7PHt8xYAKkiLID2+LbIJQEzs4eZ2eNbZiyAVBEWQHt8W+SSgJnZw8zs8S0zFkCqCAugPb4tcknAzOxhZvb4lhkLIFWEBdAe3xa5JGBm9jAze3zLjAWQKsICaI9vi1wSMDN7mOmI/IoAACAASURBVJk9vmXGAkgVYQG0x7dFLgmYmT3MzB7fMmMBpIqwANrj2yKXBMzMHmZmj2+ZsQBSRVgA7fFtkUsCZmYPM7PHt8xYAKkiLID2+LbIJQEzs4eZ2eNbZiyAVBEWwO5t9uzZKiJ6yy23NN9WWOT27t2rX/7yl1VE9OGHH+7Co6SD8e0HUxIwM3t8y4wFkFoA8BsArwHYIyKbReT/9e3bt1dr27MAdl/PPvusjho1SseOHVu2AN577736la98hQXQAN9+MCUBM7PHt8xYAKngQ0EQBM65y51zwzKZTCqKolMALAewvLUHsQB2Tzt37tTRo0frsmXLdNKkSQcUwAULFuhpp52mjY2NLIAG+PaDKQmYmT2+ZcYCaJhz7mwAzwPYLSJbROSRnj17HgWgHsBdxduKyEIA84o+Xg/gWgDzAGx3zt3XynOMA7AvCIIjy91fKIB1ty3R1IxFnC6YcqZPn6633nqrquoBBXDHjh06atQoXbp0qaoqC6ABvv1gSgJmZo9vmbEAGpVKpT4mInudc5en0+l0GIYnOOcuq6qqOvoDFMDtAL4ZRZGLosiVPkefPn3+UUR+KSJPtHYcLIBdP6UWLVqkY8aM0T179qjqgQXwmmuu0UsuuaR5kWMB7P58+8GUBMzMHt8yYwE0yjk3WEQ0k8mkSu9rbwEUkQfK7RvAvwHYJSIK4L979+79kdaOgwWw6yefzzfPxo0bdfjw4bpq1arm2yZOnKg333yz5vN5/d3vfqdnnHGGPvHEE5rL5TSfz6uI6JIlS1rsh9O9JpfLaTabbc6M0/2Hmdkb3zJrampiATSqh4g8IiJNAH4lIhdXV1cfEwTtL4AArim34169eh0bhqEA+KSIPAFgcRAEh5XblgWw6yf+BtZsNqv33HOPiojW1NQ0j4hov379tKamRqdOndr836X3f+Yzn2mxLw6Hw+F4MSyABh0WRdEIEblRRJ4D0JhKpTIi8hiAu4s3BLC4tAA656Yd7Amcc31ERJ1zw8vdn2UB7PIp/o1u+/bt2tDQ0GLGjx+vV155pTY0NOjmzZt11apV+sADD+iqVau0oaFBRUTnzp2r69ev7/LfSDnlx7czE0kYZmZvfMuMZwCToweATc65K+Lr9uaX3LfhrymAURT1jQtgXbn7s3wVcLdX7lXA2ez717nwGsDurzQz6v6YmT2+ZcZrAI0CMFRErg7D8KQwDKudcxNEJBdF0ZkApgDYBeCsdDpdA2BO/IKPeYXHlyuAInKyiEwFMDB+G5hRIvJHAC855/5XueNgAez+WADt8+0HUxIwM3t8y4wF0CgA/QEsAdAYv2nzOhGZGt99JIAficgWAG85564qdw1gaQEMw/AEEXksftweEXkVwD0i0ru142ABtMe3RS4JmJk9zMwe3zJjAaSKsADa49silwTMzB5mZo9vmbEAUkVYAO3xbZFLAmZmDzOzx7fMWACpIiyA9vi2yCUBM7OHmdnjW2YsgFQRFkB7fFvkkoCZ2cPM7PEtMxZAqggLoD2+LXJJwMzsYWb2+JYZCyBVhAXQHt8WuSRgZvYwM3t8y4wFkCrCAmiPb4tcEjAze5iZPb5lxgJIFWEBtMe3RS4JmJk9zMwe3zJjAaSKsADa49silwTMzB5mZo9vmbEAUkVYAO3xbZFLAmZmDzOzx7fMWACpIiyA9vi2yCUBM7OHmdnjW2YsgFQRFkB7fFvkkoCZ2cPM7PEtMxZAqggLoD2+LXJJwMzsYWb2+JYZCyBVhAXQHt8WuSRgZvYwM3t8y4wFkCrCAmiPb4tcEjAze5iZPb5lxgJIFWEB7D5mz56tIqK33HKLqqpu3bpVb7rpJh09erSeeOKJ+olPfEJvvvlm3bp1q1eLXBL49oMpCZiZPb5lxgJIFWEB7B6effZZHTVqlI4dO7a5AK5bt06nTp2qjz76qG7YsEGXL1+uo0eP1qlTp3q1yCWBbz+YkoCZ2eNbZiyA1IKIrBcRLZlvt7Y9C2DX27lzp44ePVqXLVumkyZNai6A5Tz44INaW1urK1as8GaRSwLffjAlATOzx7fMWACp4ENBsL8AArgunU7/U2GqqqqObu1BhQJYd9sSTc1YxOnkKWf69Ol66623qqoetADOnz9fhw4d6tUilwS+/WBKAmZmj2+ZsQAa5pw7G8DzAHaLyBYReaRnz55HAagHcFfxtiKyEMC8oo/XA7gWwDwA251z9xVud85Na+8xsAB2bQFctGiRjhkzRvfs2aOqbRfAd955R+vq6vS73/2uV4tcEvj2gykJmJk9vmXGAniIpNPp/92R+0ulUh8Tkb3OucvT6XQ6DMMTnHOXVVVVHf0BCuB2AN+MoshFUeQKt4vIZhHZAuAZANcE8dnBclgAD+3k8/nm2bhxow4fPlxXrVrVfNvEiRP15ptvbrFdPp/Xbdu26dlnn62TJ0/WnTt3ajab1Vwud8B2nO45uVyOmRkbZmZvfMusqamJBbCjOedmRFF0buFjEZkP4D0ReV1EBnTQcwwWEc1kMqnS+9pbAEXkgTL7vdw594lMJnOic+4rIvJnEflJa8fBAnhoJ/5m1Ww2q/fcc4+KiNbU1DSPiGi/fv20pqZGV65cqdlsVv/4xz/qWWedpZ/73Od0+fLlLfbB4XA4HO+HBbCjiMgrURSdEgRBAOCTALY650aLyE8A/K6DnqaHiDwiIk0AfiUiF1dXVx8TP2d7zwBec7AnAfB5EdHevXt/pNz9WRbAQzrFv71t375dGxoaWsz48eP1yiuv1IaGBs3n95/5mzBhgk6cOFF37Njh5W+5SRhmZm+Ymb3xLTOeAewEAHZHUdQ3/u+7AcwJgiAIw1AAbO3ApzosiqIRInKjiDwHoDGVSmVE5DEAd5cc0+LSAtiea/1EpLeIKICh5e5nATy0czDF1wDu2LFDJ0yYoGPGjNENGzZoY2OjNjY26ubNm3XlypWaz/txnUsS5PN+XZuUBMzMHt8y4zWAnUBE3iicARSRdc65CUEQBJlMpp+INHXS0/YAsMk5d4WI/FJE5pfct+GvKYAAxoiIhmFYXe7+LN8GplspLoBPPvlk6dv5NM9DDz3kzSKXBL79YEoCZmaPb5mxAHYCEflB/CfWh0Xk7cLbqERRdC6ApzviOQAMFZGrwzA8KQzDaufcBBHJRVF0JoApAHYBOCudTtcAmBO/4GNe0TEeUACdc8Odc5cDGJhKpTLOuXNE5HUA/9XacbAA2uPbIpcEzMweZmaPb5mxAHaOIwF8E8DdzrlBhRudc9Occ1/piCcA0B/AEgCNAPaIyDoRmVr0/D+KX8n7lnPuqnLXAJYpgIMBPAlgW/zWMmtF5IZevXr9XWvHwQJoj2+LXBIwM3uYmT2+ZcYCSBVhAbTHt0UuCZiZPczMHt8yYwHsJM6580XkCRF5o/BWLfEZwM909bF1JBZAe3xb5JKAmdnDzOzxLTMWwE7gnLtURP4M4BoA74pIGARBAOBCAL/v6uPrSCyA9vi2yCUBM7OHmdnjW2YsgJ0AwJooij4bBEEgIjsKBTCKouNF5O2uPbqOxQJoj2+LXBIwM3uYmT2+ZcYC2AkA7C782be4AIZhCAC7u/boOhYLoD2+LXJJwMzsYWb2+JYZC2AnALCmcK1fcQEE8HUA/9O1R9exWADt8W2RSwJmZg8zs8e3zFgAOwGAiwBsiqLoXBHZKSJfAHBN4b+7+vg6EgugPb4tcknAzOxhZvb4lhkLYCcRkYvjf31jXzwbnXNf7urj6mgsgPb4tsglATOzh5nZ41tmLIAd77BMJpPq06fP3wZBEPTq1evYKIo+2tUH1VlYAO3xbZFLAmZmDzOzx7fMWAA73uEi8pcwDNHVB3IosADa49silwTMzB5mZo9vmbEAdgIAq51zw7r6OA4FFkB7fFvkkoCZ2cPM7PEtMxbATgDgLBH5YxRFx3f1sXQ2FkB7fFvkkoCZ2cPM7PEtMxbATgBgq4jkALwHYLeIvFM8XX18HYkF0B7fFrkkYGb2MDN7fMuMBbATiMiX2pquPr6OxAJoj2+LXBIwM3uYmT2+ZcYCSBVhAew6s2fPVhHRW265pfm2XC6nN910k5588sk6YMAAnTJlim7evLnF43xb5JKAmdnDzOzxLTMWwE4QhmF1W9PVx9eRWAC7xrPPPqujRo3SsWPHtiiA119/vZ566qm6bNkyXb16tZ5//vk6bty4Fguab4tcEjAze5iZPb5lxgLYCeI3fn6vtenq42sP59z/AvCMiCiAga1txwJ46O3cuVNHjx6ty5Yt00mTJjUXwKamJq2trdXFixc3b/vmm29qTU2N/uEPf2i+zbdFLgmYmT3MzB7fMmMB7AQiMqB4wjA8SUQuFpEGAOO7+vha8aHiDwDcLSIPsgB2P9OnT9dbb71VVbVFAVy+fLmKiG7btq3F9mPHjtW77767+WPfFrkkYGb2MDN7fMuMBfAQAnAWgPqO2p9z7mwAz8evNN4iIo/07NnzKAD1AO4q3lZEFgKYV/TxegDXApgHYLtz7r7CfVEUnSkiDc6549pbAOtuW6KpGYs4HTylFi1apGPGjNE9e/aoassC+Jvf/EZra2sPeMxFF12k1113XfPHvi1yScDM7GFm9viWGQvgIRSGIQDs6oh9pVKpj4nIXufc5el0Oh2G4QnOucuqqqqO/gAFcDuAb0ZR5KIockEQBJlMpieATWEYnpROp9MsgF07+Xy+eTZu3KjDhw/XVatWNd82ceJEvfnmmzWfz+vChQu1tra2xWPy+bx+6Utf0muvvbb541wup9lsVnO53AHbcrrnMDN7w8zsjW+ZNTU1sQB2NOfch4snDMN/SKfTNSLyCwDPdNBzDBYRzWQyqdL72lsAReSBkoceJiIPAbg2CIKABbDrJ/7m1Gw2q/fcc4+KiNbU1DSPiGi/fv20pqZG582bpyKi9fX1LR53xhln6FVXXdXiNg6Hw+Fw4mEB7CitvAhkH4ANzrnhHfQ0PUTkERFpAvArEbm4urr6mPj523sG8JqS4/66iCwLgqBHELAAdocp/m1t+/bt2tDQ0GLGjx+vV155pTY0NOjWrVv1uOOO09/+9rfNj9m8ebPW1NRofX29t7/lJmGYmb1hZvbGt8x4BrATOOc+UTwicmo6na4JguCIDn6qw6IoGiEiN4rIcwAaU6lURkQeA3B38YYAFpcWQOfctOJt4pL4HoB8YeICmC++RrAYC2DnzsEUXwOouv9tYE477TRdvny5rl69Wi+44AK+DUwCMDN7mJk9vmXGawA7AYDTgvJl74j4vs7QA8Am59wVIvJLEZlfct+GgxXAMAyroyg6vjDOudFxAfy8c65PuSfN8lXAXaq0AO7Zs6f5jaBPPPFEnTJlir7xxhstHuPbIpcEzMweZmaPb5mxAHYCAO9FUfTR0tt79+79kY56H0AAQ0Xk6jAMTwrDsNo5N0FEclEUnQlgCoBdAM5Kp9M1AObEL/iYV3h8uQJY6oP8CZgF0A7fFrkkYGb2MDN7fMuMBbATANjnnKsqvT0MQxGRpg56jv4AlgBoBLBHRNaJyNT47iMB/EhEtgB4yzl3VblrAFkA/eTbIpcEzMweZmaPb5mxAHYgAL+O5734mrtfF81/icirAJZ09XF2JBZAe3xb5JKAmdnDzOzxLTMWwA4EYG48++K3fJlbNHNE5Fu9evU6tquPsyOxANrj2yKXBMzMHmZmj2+ZsQB2AufczJ49ex7V1cdxKLAA2uPbIpcEzMweZmaPb5mxAFJFWADt8W2RSwJmZg8zs8e3zFgAO4lz7mwRmQ/gSQBPF09XH1tHYgG0x7dFLgmYmT3MzB7fMmMB7ATxv6ixQ0R+ICI5ALMBPAxgm3Pu1q4+vo7EAmiPb4tcEjAze5iZPb5lxgLYCURkLYDz4v/eISJhEAQBgJtE5Adde3QdiwXQHt8WuSRgZvYwM3t8y4wFsBMAeDeTyaTi/24UkQFBEARhGEJEtnTt0XUsFkB7fFvkkoCZ2cPM7PEtMxbATiAirzjnBsf//RSAKUEQBPE/rfZO1x5dx2IBtMe3RS4JmJk9zMwe3zJjAewEIvIT59zMIAgCAF8F8G58DeBWAPd29fF1JBZAe3xb5JKAmdnDzOzxLTMWwM5xeBAERxQ+cM6dA2AWgK8HQfChrjusjscCaI9vi1wSMDN7mJk9vmXGAkgVYQG0x7dFLgmYmT3MzB7fMmMB7CQiciqAfwfw3yLSOwiCwDl3fhiGI7v62DoSC6A9vi1yScDM7GFm9viWGQtgJwDweQDvisiPAewpvA2Mc+4yEXmwq4+vI7EA2uPbIpcEzMweZmaPb5mxAHYCAH8SkQuC4ID3ARwI4M2uPbqOxQJoj2+LXBIwM3uYmT2+ZcYC2AkAvJtOp9NB0LIAikgIYE+XHlwHYwHsfP/xH/+hY8aM0UGDBumgQYP0nHPO0fr6+ub7N2zYoJdddpkOHTpUBw0apF//+tf1z3/+c6v7822RSwJmZg8zs8e3zFgAOwGAl0XkjCA4oABeAGBN1x5d29Lp9N8AmAfgeQB5EVnY1vYsgJ3v0Ucf1fr6en3llVf0lVde0TvvvFNra2v1hRde0F27dunpp5+uX/va13Tt2rW6du1avfTSS/Xzn/+8vvfee2X359silwTMzB5mZo9vmbEAdgIA0wGsBjBURJrCMBzpnJsY/6sgU7v6+NpwZM+ePY8CcI+IXAxgCQtg9/Txj39c58+fr3/84x+1pqZGd+zY0Xzftm3bVER02bJlZR/r2yKXBMzMHmZmj2+ZsQB2EufcrfEbQO+LZzeAmyvc59nxmbndIrJFRB6JC9s8EVkoIlcDeAvAtviNqI8AcIeIvANgk3NucmFf6XQ6LSIav0dhPYA9AC4qfr7Cfts6pkIBrLttiaZmLOJ00LQmn8/rokWLtLa2Vl988UV99NFHtX///prL5Zq32b17t9bU1OisWbNa3YdPi1wSMDN7mJk9vmXGAtiB4j/1Hlb4uFevXn8XhuFJInJyVVXV0ZXsO5VKfUxE9jrnLk+n0+kwDE9wzl1WVVV1dFzUmkTkB5lMpp9zbrKIaHwG7+owDAHgWhH5SxRFfYPg/QIoIq8CGJ9KpTJ9+/btVfycLIDdpwCuXbtWBw4cqP3799chQ4Y0XwO4ZcsWHTx4sN5yyy367rvv6q5du/TGG29UEdHrrruu7De9b4tcEjAze5iZPb5lxgLYgQC8F0XRRwsfi8gvM5lMz47Yt3NusIhoJpNJlXneeSKyPtj/L5AUnnstgD8UbdZDRHaKyBeC4P0CCOAbbXw+LIBdNPl8vsW8++67+vLLL+szzzyjt99+uw4dOlTXrl2r+XxeH3/8cT399NO1X79+2r9/f73yyiv1s5/9rF5//fUH7Cefz2sul9NsNqu5XK7s/ZzuN8zM3jAze+NbZk1NTSyAHQXAvpIC2PwCkA7QQ0QeEZEmAL8SkYurq6uPiZ93HoDFJcfyuIj8sOS2DfE/R9dcAKMoGtHG58MC2EUTf1O2Op/73Of0q1/9aovbHnvsMX388cc1m83qxz/+cZ05c+ZB98PhcDgc74cFsFKdXACDIAgOi6JohIjcKCLPAWhMpVKZckUtvq7vruLbRGS9c25aELQ4Aziwjc+HBbCL5mC/uZ1//vk6ffr0svc98cQT2q9fP33xxRf5W25ChpnZG2Zmb3zLjGcAOxCA95xzVYWPRWRHKpXKdNLT9Yhf2HFFdyiAfBVw5/nud7+rTz31lG7cuFHXrl2rd955p9bU1OgTTzyhqqr333+//ulPf9INGzbowoUL9eSTT9bbbrut1f3l835d55IEzMweZmaPb5nxGsAOFL/adzGAXwP4tYjsBbC08HFh/sp9D41f0HFSGIbVzrkJIpKLoujMji6Azrnj4n+15DcAfh//d9miyALY+b71rW/pqFGjtLa2VocNG6Zf+tKXmsufquodd9yhp5xyitbW1uro0aP1pz/9qe7bt6/V/fm2yCUBM7OHmdnjW2YsgB0IwNz2zF+57/4AlgBojP994XWF9xTs6AIoIuvjVwi3mHLHxQJoj2+LXBIwM3uYmT2+ZcYCSBVhAbTHt0UuCZiZPczMHt8yYwGkirAA2uPbIpcEzMweZmaPb5mxAFJFWADt8W2RSwJmZg8zs8e3zFgAqSIsgPb4tsglATOzh5nZ41tmLIBUERZAe3xb5JKAmdnDzOzxLTMWQKoIC6A9vi1yScDM7GFm9viWGQsgVYQF0B7fFrkkYGb2MDN7fMuMBZAqwgJoj2+LXBIwM3uYmT2+ZcYCSBVhAbTHt0UuCZiZPczMHt8yYwGkirAA2uPbIpcEzMweZmaPb5mxAFJFWADt8W2RSwJmZg8zs8e3zFgAqSIsgPb4tsglATOzh5nZ41tmLIBUERZAe3xb5JKAmdnDzOzxLTMWQKoIC6A9vi1yScDM7GFm9viWGQsgVYQF0B7fFrkkYGb2MDN7fMuMBZAq4mMBnD17to4fP14HDhyow4YN00svvVRffvnlA7Z7+umn9fzzz9cBAwbokCFDdNKkSbp79+4uOOKWfFvkkoCZ2cPM7PEtMxZAqoiPBXDy5Mm6YMECfeGFF7ShoUEvueQSraur0127djVv8/TTT+vgwYN1zpw5+sILL+irr76qDz30kOZyuS488v18W+SSgJnZw8zs8S0zFkBqwTlXB+C/RGQzgF0AnnHOTWxtex8LYKktW7aoiOjKlSubb5swYYJ+73vf68Kjap1vi1wSMDN7mJk9vmXGAkjFjhSRqwHcHEXRKc65CMDXAbznnBtb7gEsgKrr169XEdF169apqurbb7+tIqI/+9nP9Nxzz9Xhw4frxIkT9amnnuriI93Pt0UuCZiZPczMHt8yYwE0xDl3NoDnAewWkS0i8kjPnj2PAjBPRBbG5e0tANucczODIDgCwB0i8g6ATc65yYV9pdPptIioc+4cAPUA9gC4qNzzAlgM4Kfl7isUwLrblmhqxqLETmv27dunU6ZM0fPOO6/5tj/96U8qInryySfr/fffr6tXr9Zbb71Va2tr9dVXXz0E39Zt822RSwJmZg8zs8e3zFgAjUilUh8Tkb3OucvT6XQ6DMMTnHOXVVVVHR0XwCYR+UEmk+nnnJssIgpgiYhcHYYhAFwrIn+JoqhvELxfAEXkVQDjU6lUpm/fvr3KPbeIPOGc+065+3wpgPl8vuzMnDlT6+rqdNOmTc23PfXUUyoiescdd7TYdsyYMXr77be3uq9DNblcTrPZrOZyuS4/Fg4zS+owM3vjW2ZNTU0sgBY45waLiGYymVTpfXEBXB8EweGF20RkLYA/FG3WQ0R2isgXguD9AgjgGwd53rNFJOecqy13vy8FMP4maTFTp07VYcOG6YMPPtji9gcffFBFRO+6664Wt19wwQV64YUXlt0Xh8PhcDhdNCyA3VwPEXlERJoA/EpELq6urj4mCPYXQACLizcG8LiI/LDktg0Avh4E7xfAKIpGtPaEzrm6uDRe0No2WU8KYPFvTXv37tUbbrhBR44cqS+99NIBv1Xt3btXR44cqXfeeWeL28eNG3fAWUH+lsthZskcZmZvfMuMZwBtOSyKohEicqOIPAegMZVKZQrXABZvGF/Xd1fxbSKy3jk3LQhanAEcWO6JnHOfEJEdInJJWwfkSwEsNnPmTB0yZIiuWLFCGxsbm6f4Pf7mzp2rgwcP1oceekjXr1+v3/ve9/SEE07QDRs2HOrLPA6Qz/t1nUsSMDN7mJk9vmXGawDt6hG/sOOKji6AhTN/zrmvHewgsh6+Cji+dvKAWbBgQYvt5syZo6eddpoOGDBAzz33XL4KmP5qzMweZmaPb5mxABoBYGj8go6TwjCsds5NEJFcFEVndmQBjN8HcJdz7l/T6fQ/FaZPnz7/WO64fCyA1vm2yCUBM7OHmdnjW2YsgEYA6A9gCYBGAHtEZJ2ITI3v67ACGO/rgLNbAOrLHRcLoD2+LXJJwMzsYWb2+JYZCyBVhAXQHt8WuSRgZvYwM3t8y4wFkCrCAmiPb4tcEjAze5iZPb5lxgJIFWEBtMe3RS4JmJk9zMwe3zJjAaSKsADa49silwTMzB5mZo9vmbEAUkVYAO3xbZFLAmZmDzOzx7fMWACpIiyA9vi2yCUBM7OHmdnjW2YsgFQRFkB7fFvkkoCZ2cPM7PEtMxZAqggLoD2+LXJJwMzsYWb2+JYZCyBVhAXQHt8WuSRgZvYwM3t8y4wFkCrCAmiPb4tcEjAze5iZPb5lxgJIFWEBtMe3RS4JmJk9zMwe3zJjAaSKsADa49silwTMzB5mZo9vmbEAUkVYAO3xbZFLAmZmDzOzx7fMWACpIiyA9vi2yCUBM7OHmdnjW2YsgFQRFkB7fFvkkoCZ2cPM7PEtMxZAqohPBXD27Nk6fvx4HThwoA4bNkwvvfRSffnll1tsM2nSJBWRFjNt2rQuOuLyfFvkkoCZ2cPM7PEtMxZAaiGTyfQD8HsAbwHYIyKvALglCIIjy23vUwGcPHmyLliwQF944QVtaGjQSy65ROvq6nTXrl3N20yaNEmvvfZabWxsbJ7u9rXxbZFLAmZmDzOzx7fMWACp2JEiEgK4SEQGZDKZlHNuHIC3nHP/Wu4BPhXAUlu2bFER0ZUrVzbfNmnSJL3lllu68KgOzrdFLgmYmT3MzB7fMmMBNMQ5dzaA5wHsFpEtIvJIz549jwIwT0QWisjV8Zm7bc65mUEQHAHgDhF5B8Am59zkwr7S6XRaRNQ5dw6AegB7AFxU7nlF5E4R+WO5+woFsO62JZqasShRczDr169XEdF169Y13zZp0iQdOnSonnzyyfrpT39av/3tb+uOHTs683v4A/NtkUsCZmYPM7PHt8xYAI1IpVIfE5G9zrnL0+l0OgzDE5xzl1VVVR0dF8AmEflBJpPp55ybLCIKYImIXB2GIQBcKyJ/iaKobxC8y/sOawAAHAtJREFUXwBF5FUA41OpVKZv3769Sp83iiIHYE38Z+AD+FoA9+3bp1OmTNHzzjuvxe2//OUvddmyZbpu3TpdtGiRjho1Si+88MLO/B7+wHxb5JKAmdnDzOzxLTMWQCOcc4NFRDOZTKr0vrgArg+C4PDCbSKyFsAfijbrISI7ReQLQfB+AQTwjXLPB2B5fA2gAphTvO9iSS6A+Xy+1Zk5c6bW1dXppk2b2tzu2WefVRHR5557rs3tDuXkcjnNZrOay+W6/Fg4zCypw8zsjW+ZNTU1sQAa0UNEHhGRJgC/EpGLq6urjwmC/QUQwOLijQE8LiI/LLltA4CvB8H7BTCKohHlniyKor7OueMAnAdgE4Dp5bZLcgGMvzEOmKlTp+qwYcP0wQcfbHWbwjz11FPav39//f73v3/QbTkcDofD6YJhATTgsCiKRojIjSLyHIDGVCqVKVwDWLxhfF3fXcW3ich659y0IGhxBnDgwZ7UOTcJwLtBEPQovS+b4AJY+tvS3r179YYbbtCRI0fqSy+91K7fsBoaGlRE9Mknn+zy3/Z8/S03CcPM7A0zsze+ZcYzgHb1iF/YccUhKIDni8jeIAiOKL0v69GrgGfOnKlDhgzRFStWtHibl927d6uq6oYNG/T73/++Pvfcc7px40atr6/XT33qU/rZz35W8/nuc01JPu/XdS5JwMzsYWb2+JYZrwE0AsDQ+AUdJ4VhWO2cmyAiuSiKzuzIAuicmxi/Mri/iITOuQnxn4D/vdxx+VQAS9/guTALFixQVdU33nhDJ06cqCeffLLW1tbqGWecoTfffLNu3bq1i4+8Jd8WuSRgZvYwM3t8y4wF0AgA/QEsAdAYvzhjnYhMje/rsAIYRdG5AP5HRHaIyE4Aq0XkW+l0+m/KHZdPBTApfFvkkoCZ2cPM7PEtMxZAqggLoD2+LXJJwMzsYWb2+JYZCyBVhAXQHt8WuSRgZvYwM3t8y4wFkCrCAmiPb4tcEjAze5iZPb5lxgJIFWEBtMe3RS4JmJk9zMwe3zJjAaSKsADa49silwTMzB5mZo9vmbEAUkVYAO3xbZFLAmZmDzOzx7fMWACpIiyA9vi2yCUBM7OHmdnjW2YsgFQRFkB7fFvkkoCZ2cPM7PEtMxZAqggLoD2+LXJJwMzsYWb2+JYZCyBVhAXQHt8WuSRgZvYwM3t8y4wFkCrCAmiPb4tcEjAze5iZPb5lxgJIFWEBtMe3RS4JmJk9zMwe3zJjAaSKsADa49silwTMzB5mZo9vmbEAUkVYAO3xbZFLAmZmDzOzx7fMWACpIiyA9vi2yCUBM7OHmdnjW2YsgFSR7lQAV65cqVOmTNERI0aoiOjDDz/c6rbXXXediojOnTv30B1gN+HbIpcEzMweZmaPb5mxAFJFulMBrK+v1zvvvFOXLl3aZgF8+OGHddy4cTpy5EgWQDKBmdnDzOzxLTMWQGpVFEVORHYA2NbaNt2pABZrrQC++eabeuqpp+oLL7ygo0aNYgEkE5iZPczMHt8yYwGkYkcW/7eIPCUiD7anANbdtkRTMxYd0mlLuQL43nvv6fnnn6/z5s1TVWUB9GSRSwJmZg8zs8e3zFgADXHOnQ3geQC7RWSLiDzSs2fPowDME5GFInI1gLcAbHPOzQyC4AgAd4jIOwA2OecmF/aVTqfTIqLOuXMA1APYA+Ciwv0A/k1E/h+AC5NSAGfPnq0XXXSR7tu3T1VZAH1Z5JKAmdnDzOzxLTMWQCNSqdTHRGSvc+7ydDqdDsPwBOfcZVVVVUfHBbBJRH6QyWT6Oecmi4gCWCIiV4dhCADXishfoijqGwTvF0AReRXA+FQqlenbt2+vIAiCKIr+j4i84pz7cHcugPl8vtUREV2yZEnzx88++6wOHz5c33jjjebb6urq9Kc//Wmb+0ni5HI5zWazmsvluvxYOMwsqcPM7I1vmTU1NbEAWuCcGywimslkUqX3xQVwfRAEhxduE5G1AP5QtFkPEdkpIl8IgvcLIIBvFO+rd+/eHwHwGoDT4n132wIY/49bdkRE77nnnuaPb775Zu3Xr5/W1NQ0j4hov3799JRTTmlzXxwOh8PhJHhYALu5HiLyiIg0AfiViFxcXV19TBDsL4AAFhdvDOBxEflhyW0bAHw9CN4vgFEUjSjZ5tci8u2ij7ttAWzrN5vSM4Bvv/22NjQ0tJiRI0fq7bffri+++GKX/ybG33I5zCxZw8zsjW+Z8QygLYdFUTRCRG4UkecANKZSqUzhGsDiDePr+u4qvk1E1jvnpgVBizOAA0setw1Avmjei7fLF19DWNCVBbDUzp07dc2aNbpmzZrm9/hbs2aNvv7662Wvf+A1gH5c55IEzMweZmaPb5nxGkC7esQv7Liigwtg/yiKji8MgGtEpCmKouMLZxyLZbvR28A8+eSTGl/X2GJmzJhRdnsWQD8WuSRgZvYwM3t8y4wF0AgAQ+MXdJwUhmG1c26CiOSiKDqzIwtgmedt15+Au0MBpPbxbZFLAmZmDzOzx7fMWACNANAfwBIAjQD2iMg6EZka38cCSO3m2yKXBMzMHmZmj2+ZsQBSRVgA7fFtkUsCZmYPM7PHt8xYAKkiLID2+LbIJQEzs4eZ2eNbZiyAVBEWQHt8W+SSgJnZw8zs8S0zFkCqCAugPb4tcknAzOxhZvb4lhkLIFWEBdAe3xa5JGBm9jAze3zLjAWQKsICaI9vi1wSMDN7mJk9vmXGAkgVYQG0x7dFLgmYmT3MzB7fMmMBpIqwANrj2yKXBMzMHmZmj2+ZsQBSRVgA7fFtkUsCZmYPM7PHt8xYAKkiLID2+LbIJQEzs4eZ2eNbZiyAVBEWQHt8W+SSgJnZw8zs8S0zFkCqCAugPb4tcknAzOxhZvb4lhkLIFWEBdAe3xa5JGBm9jAze3zLjAWQKsICaI9vi1wSMDN7mJk9vmXGAkgVOZQFcOXKlTplyhQdMWKEiog+/PDDLe7ft2+fzpo1S0eMGKEnnHCCTpo0SV944YVOPy5rfFvkkoCZ2cPM7PEtMxZAqsihLID19fV655136tKlS8sWwDlz5uigQYN06dKlum7dOp02bZqOGDFCd+zY0enHZolvi1wSMDN7mJk9vmXGAkgtpNPptIho6URR9Kly23fVn4BLC+C+fft0xIgROmfOnObbcrmcDhkyRH/+858f0mPr7nxb5JKAmdnDzOzxLTMWQCp2ZKEAhmF4ejqd/qfCBEHwoXIP6C4F8LXXXlMR0dWrV7fY7qtf/apOnz79kB5bd+fbIpcEzMweZmaPb5mxABrinDsbwPMAdovIFhF5pGfPnkcBmCciC0XkagBvAdjmnJsZBMERAO4QkXcAbHLOTS7sq1D0nHPnAKj//9u7+yCryvsO4A+h0aSaFwPbjbvs3nvO+X2/O1RMUaxGTBUbOjZt2ia0VYG+BExMauvrdEJCrdEME9KaqVQ0SDNTTYxpMtGOiSJIEseXsQN47SgmmxBb1CpCtpLKyiILi6d/cC57uO4b3r179znP9zPz/ME9z7377H7Zh+99OQcA+wEsqd4OYPZY1lQtgPNWbkhLy+4f1zGS2gL45JNPpiTTXbt2HTXv2muvTZcuXdqQXx5fhbbJFYEy848y809omakAeqJUKp1M8qCZXV0ul8txHJ9qZpe1tLScmBXAXpK3RFHUZWZLsxK3geTyOI4B4FqSB5Ik6XDuqLd6nwOwoFQqRR0dHW25Avg/AHpIPm5mfzLcuhpZAAcGBoYdJNMNGzYc+fMTTzyRkkx37tx51Lzly5enS5cuHfGxQhv9/f1ppVJJ+/v7m74WDWVW1KHM/BuhZdbb26sC6AMzO51kGkVRqfZYVgCfd869rXobyZ8BeDQ3bSrJvSQvdm6wAAK4Mv9YbW1t083sapJnxnF8BoAvAjhkZn821LoaWQCzv5hDDpLpmjVrjvx53bp1Kcn07rvvPmrewoUL00suuWTEx9LQ0NDQ0Ah4qABOclNJ/pBkL4DvkvxUZ2fnSc4dLoAA1uUnA3iE5K01t70A4ArnBgtgkiTnjPaFzWw1ya1DHatMklcADx48mM6dOzddu3btkdv27duXzpkzJ73rrrua/kxrMo3QnuUWYSgz/4Yy82+ElpleAfTLlCRJziF5A8mtAHpKpVJU/QxgfmL2ub5V+dtIPm9mVzl31CuAo37Wz8wWA3h9qGONLIC19u7dm3Z3d6fd3d0pyfT2229Pu7u70x07dqRpevgyMHPmzEk3btyYbtu2Lb3mmmt0GZghDAyE9TmXIlBm/lFm/gktM30G0F9TsxM7rpmAAvgVktuHOlaZwLOAN23a9KbL05BMly1blqbp0ReCnjVrVrp48eJ027ZtDV+Xb0Lb5IpAmflHmfkntMxUAD0B4KzshI4z4jjuNLM/JdmfJMlHxrMAkvxLAIsAzIyiqAvA35I8YGZXD7WuiSyAMj5C2+SKQJn5R5n5J7TMVAA9AWAmgA0AegDsJ7mN5N9kx8a7AHYD6MvOLK4MdwKIcyqAPgptkysCZeYfZeaf0DJTAZS6qAD6J7RNrgiUmX+UmX9Cy0wFUOqiAuif0Da5IlBm/lFm/gktMxVAqYsKoH9C2+SKQJn5R5n5J7TMVAClLiqA/gltkysCZeYfZeaf0DJTAZS6qAD6J7RNrgiUmX+UmX9Cy0wFUOqiAuif0Da5IlBm/lFm/gktMxVAqYsKoH9C2+SKQJn5R5n5J7TMVAClLiqA/gltkysCZeYfZeaf0DJTAZS6qAD6J7RNrgiUmX+UmX9Cy0wFUOqiAuif0Da5IlBm/lFm/gktMxVAqYsKoH9C2+SKQJn5R5n5J7TMVAClLiqA/gltkysCZeYfZeaf0DJTAZS6qAD6J7RNrgiUmX+UmX9Cy0wFUOqiAuif0Da5IlBm/lFm/gktMxVAqYsKoH9C2+SKQJn5R5n5J7TMVAA9AuBhAKsm+r4jGe8CeP7556ck3zSuv/76cXl8CW+TKwJl5h9l5p/QMlMB9EgIBXD37t1pT0/PkfH444+nJNNNmzaNy+NLeJtcESgz/ygz/4SWmQqgRyZBAXx77Q2Nfgt4xYoV6fz589M33nijIY8fotA2uSJQZv5RZv4JLTMVQI8AeJjkLSRvAfAqyd0AVjjnpjjnnJldBuBZAPsB/ALA3fn75gsgyecB/D2Ab5HcS/JlAJfnvx7JFMBnAHwPQB/JG2rXVC2A81ZuSEvL7j/mMZL+/v70zDPPTNesWTNBvw5hCG2TKwJl5h9l5p/QMlMB9EhWAF8DsCqKoi4zW5wVs0/FcXwGgAEAC6MoKpnZaQCuyN+3tgCS7DWzz8VxTACXZ/f/ndycFMAvzGwpyTiKolLtmhpZANetW5fOnDkz3bVr1wT9OoQhtE2uCJSZf5SZf0LLTAXQI1mJ63bZK37OOUfyywC6ASwAsGf69OnvGuG+tQVwfX4OyW+TfCD35xTATSOtqd4CODAwMOxYsmRJeumll444R+PYR39/f1qpVNL+/v6mr0VDmRV1KDP/RmiZ9fb2qgD6Iitx/5q/zcz+iOTBOI7fQ3Iryf8leaeZLW5ra/vVmvseVQDN7Lqax7+S5HO5OamZLR5pTfUWwOwv35vG+vXr066urvS2224bdo6GhoaGhoZG3UMFcLIbqQA656Y6536F5HwA/wjgvwE8Wy6X35u771gK4PbcnDRJko+NtKZKg14BXLVqVTp37tx0//79TX+WVLQR2rPcIgxl5t9QZv6N0DLTK4Aeyb0FfATJlbW3Oedca2vrCSQPAliQu2/tW8AP5O8D4N9q3wIeawEcz7OADx06lM6bNy+98cYbx+0xZdDAQFifcykCZeYfZeaf0DLTZwA9Uj0JhOQ/RVHUBWAhyb0APg3gowCuADA7OwnkrwAcMrNTqvcd4izgPQA+G8cxzeyvAQyY2QW5OU0pgI899lhKMt2+ffu4PaYMCm2TKwJl5h9l5p/QMlMB9EhWAG8FsAbAHpK/JLnSOTcljuMPZcd/CWAfyafN7ML8fYd6C5jkdwD0AdgF4Mr812tWAZTGCm2TKwJl5h9l5p/QMlMBDFRWAK+q93FUAP0T2iZXBMrMP8rMP6FlpgIYKBXAcIW2yRWBMvOPMvNPaJmpAAZKBTBcoW1yRaDM/KPM/BNaZiqAUhcVQP+EtskVgTLzjzLzT2iZqQBKXVQA/RPaJlcEysw/ysw/oWWmAih1UQH0T2ibXBEoM/8oM/+ElpkKoNRFBdA/oW1yRaDM/KPM/BNaZiqAUhcVQP+EtskVgTLzjzLzT2iZqQBKXVQA/RPaJlcEysw/ysw/oWWmAih1UQH0T2ibXBEoM/8oM/+ElpkKoNRFBdA/oW1yRaDM/KPM/BNaZiqAUhcVQP+EtskVgTLzjzLzT2iZqQBKXVQA/RPaJlcEysw/ysw/oWWmAih1UQH0T2ibXBEoM/8oM/+ElpkKoNRFBdA/oW1yRaDM/KPM/BNaZiqAUhcVQP+EtskVgTLzjzLzT2iZqQBKXVQA/RPaJlcEysw/ysw/oWWmAih1UQH0T2ibXBEoM/8oM/+ElpkKoNRFBdA/oW1yRaDM/KPM/BNaZiqAUhcVQP+EtskVgTLzjzLzT2iZqQBKXbZs2TK3UqmkPT09aV9fn4YHo7e3N62W9mavRUOZFXUoM/9GaJm98soraaVSSbds2TK32V1CPFSpVBZlzyA0NDQ0NDQ0/BuLmt0lxEObN2+eVqlUFmWvBJ6uoaGhoaGhMflH9u/2os2bN09rdpcQEREREREREREREREREREREREREZEJYGaXkXwOwH4AT5L8rWavSQ4jeT3JND8A7MpNmZLNeRnA6wAeNrNTmrbgAAE4F8B9JF8mmSZJ8rGaKaNm1NnZeRLJOwHsAbCH5J3lcvm9E/htBGW0zADcMcTv3ab8HDM73sxWk3wFQB+A75vZjIn9TsJB8vMknyD5GoAekvdGUdSVnzOWTOI47gRwH4C+bN7NzrnjJvSbEZkMkiS5iOQBM/skgJkAVpHcG8dxZ7PXJocLIIAfl8vl91eHmbVUj5vZMpK9ABYkSTKL5LdJvjx9+vR3NXPdIUmS5CMAVgBYMFSZGEtGJNcDeMbMzjazswE8A+C+if9uwjBaZlkBXJ//vZsxY8b7auasAfASyflmdhrJhwA85ZybOqHfTCAAbADwCTM7heRvALgfwAutra0n5OaMlslUAM+QfCg7Pp/kDjNb3ZzvSqSJAGwGsCZ/G8mfklzZrDXJoKwAPjXM4Skkd5rZsuoNZnY8gFcBfHqClig5Q5SJUTMCMDN7hems3JwPkkxrX+GQ8TdCAbx3uPvEcfwekgeSJLmoeltHR0cbgENmdkEj1yuHmVlL9ntzrnNjyyQr/oc6OjraqnNIXgxgv5m9e+K/C5HmOQ7AAMmP528E8M8AHmnWomRQVgD7sreqnstePYqzYzHJ1MxOy98HwPfM7OvNWXHYasvEWDIys6UAXq19rKwkLmn8qsM2XAHMfv49AH5O8mtJkvxa9XiSJL9NMu3s7Dyp5rGeJnnDRK09ZEmSWJbdrOzPo2YC4Iskn84fzz5+kSZJcv7ErV6kyTo6Otqyv/hH/fcxJJeT3Nasdcmg7BnrH8dxfCrJ+QAeBrCrvb19WpIkc0mm+WezzjlnZv8C4MFmrTlktWViLBmRXA7g57WPlRWPzzd+1WEbqgAmSXIRgN9PkmSWmf0BgKcA/NjMjnfOOQCLSPbXPhaAjQDWTtTaAzYFwPdJPla9YSyZZL93G2vnkOwHsLCxSxaZRKoF0MzOzt8O4O9I/qxZ65Lhtba2ngBgl5ldUy0XpVLp5Pwckl8DsKFZawzZcAVwpIyGe8IF4Fkz+1zjVx22YU7cOUqpVDqZ5AEAC5wbsWz8AMBtjVqrHEbyVpLP50/wGEsmwz05JnmA5MWNXbXI5KK3gD2UbWhr9Bbw5KO3gP0zlgLo3JFCvsw5vQXcTGa2GsCLpVIpyt+ut4BFjlF2EshXa27r1kkgk1N2AsFLZnady04wAPDZ3JTjdBJI8wx3EshIGVVPAiF5ZnUCgLN0EsjEGEsBbG9vnwZgP8m/cG7whAMzu7A6p1QqnayTQBpqCslbSO6I4xi1B8eSSfUkkPwr8tnb/ToJRMKTuwzM0uwyMDeR3BtFUanZaxPnzOwrZnZeqVSKAJyFw9cu663mY2bLALxK8uNJkswC8C1dBmZitbS0nAhgNoDZ2at9VwOYXb2U0lgyIrme5NNm9sHsDOCtugxM44yUWUtLy4nZ793Z5XK5bGbzAPwHgJfymWWXHHkxjuMPZ5cU+ZEuA9M4AL4K4FUzO6/m8jzvzM0ZLZPqZWB+aGanxXH8YQAv6jIwEqzsQtDPZx+EfbJ6Wr00X/WacdlnVHYAuMfMfj03pXqR4Z04fCHvR6pnxcnEMLN5rLlocHZ5ijuyKaNmNGPGjPcB+CbJXh6+ZuA3dSHoxhkpsxkzZrwTwIM4fLHhAwBeAHBHkiQd+ccol8vvyC46vBvAPgD31c6R8TNUXllmn6jOGUsm2YWg7wewj+RuM1tdPblHRERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERERLwE4I6h/qurJEms2WsTERERkQbICuD6/H90Xy6X3+8G/yP7Znt7sxcgIiIiUihZAbx3rPOjKCoBuA/A/wHoA/ATkr9XPW5mpwBYR7KX5GskHzOzJDv8NjO7DsBLJPsBPJUkye9W71sul8skUzO7EMDDAPYDWOKcc0mSzAXwKIDXAbwI4ObW1tYTxu0HISIiIhKKYy2AAO4HsDGO41NJxgA+CuBc55wj2U5yN4B74jg+I45jAlgSRVGXc86Z2dUA9pC8OIqiLgD/QPJAHMdwbrAAknwOwIJSqRR1dHS0ZV/rNTO7Ko5jZGXwPwHc3pifioiIiEiBAbgDwADJvdUB4LvDzSe51cy+MNQxM/sSye1umLdtSe4gubzmti0kb3VusAACuLJmzjcArM3fFsfxhwAcKpfL7xjjtyoiIiIizh0pgD9IksSqo1QqnTzcfDP7JMmDJB8neUMURR+oHiP5gJl9fZj7vTt7e/e8mq9/E8mHnBssgEmSnFMz5yck+2tKal9WFmfW9xMQERERCcyxvgXsnHNJknQA+AyAfyd5AMDl2WPdM1oBrL5dnPv6q0j+yLmjXgGcnZ9D8qcAbs6X1Opwzh13TN+wiIiISOjeSgHMI7mS5FbnnDOzL7zFt4BvcW74AgjgrmpJFBEREZE6vYWTQFaZ2QWlUikys9MBbCb5Heeca29vn0byldxJIDCzP8+dBHIVgD1JklwURVEXyS8PdRJIbQGMougDAPaRvBXA7Oxx/9DMVo/nz0JEREQkCMdaAM1sNYD/yi7R0kPyG+3t7dOqx7Oy9mD2Gb1eAI+SjLPD+cvAHBjuMjC1BdA55+I4/k0AG7NLy+wl+XTtq4kiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiIiInIM/h/l2wCevj2RiwAAAABJRU5ErkJggg==\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "FI9B0KtGGHQ-"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "HH8TUxKnGHQ_"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "6bz3TFwUGHRA"
      },
      "source": [
        "<h3> 4.4.4 Surprise KNNBaseline predictor </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "BTf7-oX_GHRA",
        "colab": {}
      },
      "source": [
        "from surprise import KNNBaseline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "pWpJHruUGHRD"
      },
      "source": [
        "- KNN BASELINE\n",
        "    - http://surprise.readthedocs.io/en/stable/knn_inspired.html#surprise.prediction_algorithms.knns.KNNBaseline "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "evfRoMO4GHRJ"
      },
      "source": [
        "- PEARSON_BASELINE SIMILARITY\n",
        "    - http://surprise.readthedocs.io/en/stable/similarities.html#surprise.similarities.pearson_baseline "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "AsWYa_1vGHRJ"
      },
      "source": [
        "- SHRINKAGE\n",
        "    - _2.2 Neighborhood Models_ in http://courses.ischool.berkeley.edu/i290-dm/s11/SECURE/a1-koren.pdf "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "1_aY9mirGHRK"
      },
      "source": [
        "- __predicted Rating__ : ( ___ based on User-User similarity ___ )\n",
        "\n",
        "\\begin{align} \\hat{r}_{ui} = b_{ui} + \\frac{ \\sum\\limits_{v \\in N^k_i(u)}\n",
        "\\text{sim}(u, v) \\cdot (r_{vi} - b_{vi})} {\\sum\\limits_{v \\in\n",
        "N^k_i(u)} \\text{sim}(u, v)} \\end{align}\n",
        "\n",
        "- $\\pmb{b_{ui}}$ -  _Baseline prediction_ of (user,movie) rating\n",
        "\n",
        "- $ \\pmb {N_i^k (u)}$ - Set of __K similar__ users (neighbours) of __user (u)__ who rated __movie(i)__  \n",
        "\n",
        "- _sim (u, v)_ - __Similarity__ between users __u and v__  \n",
        "    - Generally, it will be cosine similarity or Pearson correlation coefficient. \n",
        "    - But we use __shrunk Pearson-baseline correlation coefficient__, which is based on the pearsonBaseline similarity ( we take base line predictions instead of mean rating of user/item)\n",
        "       "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "-6IOidY5GHRL"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "756ewezkGHRL"
      },
      "source": [
        "- __ Predicted rating __ ( based on Item Item similarity ):\n",
        " \\begin{align} \\hat{r}_{ui} = b_{ui} + \\frac{ \\sum\\limits_{j \\in N^k_u(i)}\\text{sim}(i, j) \\cdot (r_{uj} - b_{uj})} {\\sum\\limits_{j \\in N^k_u(j)} \\text{sim}(i, j)} \\end{align}\n",
        "\n",
        "    -  ___Notations follows same as above (user user based predicted rating ) ___"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "E7TBmajbGHRM"
      },
      "source": [
        "  <h4> 4.4.4.1 Surprise KNNBaseline with user user similarities</h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "3b216a14-66a1-41be-ccdd-586d890a0efd",
        "id": "5YxpLxKfGHRM",
        "colab": {}
      },
      "source": [
        "# we specify , how to compute similarities and what to consider with sim_options to our algorithm\n",
        "sim_options = {'user_based' : True,\n",
        "               'name': 'pearson_baseline',\n",
        "               'shrinkage': 100,\n",
        "               'min_support': 2\n",
        "              } \n",
        "# we keep other parameters like regularization parameter and learning_rate as default values.\n",
        "bsl_options = {'method': 'sgd'} \n",
        "\n",
        "knn_bsl_u = KNNBaseline(k=40, sim_options = sim_options, bsl_options = bsl_options)\n",
        "knn_bsl_u_train_results, knn_bsl_u_test_results = run_surprise(knn_bsl_u, trainset, testset, verbose=True)\n",
        "\n",
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "models_evaluation_train['knn_bsl_u'] = knn_bsl_u_train_results \n",
        "models_evaluation_test['knn_bsl_u'] = knn_bsl_u_test_results\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            "Estimating biases using sgd...\n",
            "Computing the pearson_baseline similarity matrix...\n",
            "Done computing similarity matrix.\n",
            "Done. time taken : 0:00:30.173847 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:01:35.970614\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.33642097416508826\n",
            "\n",
            "MAPE : 9.145093375416348\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.075213\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.0726493739667242\n",
            "\n",
            "MAPE : 35.02094499698424\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:02:06.220108\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ivjFheRnGHRO"
      },
      "source": [
        "<h4> 4.4.4.2 Surprise KNNBaseline with movie movie similarities</h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "6e1164ed-6151-49d1-d55a-b6c8e109cf23",
        "id": "iN8JIbgvGHRP",
        "colab": {}
      },
      "source": [
        "# we specify , how to compute similarities and what to consider with sim_options to our algorithm\n",
        "\n",
        "# 'user_based' : Fals => this considers the similarities of movies instead of users\n",
        "\n",
        "sim_options = {'user_based' : False,\n",
        "               'name': 'pearson_baseline',\n",
        "               'shrinkage': 100,\n",
        "               'min_support': 2\n",
        "              } \n",
        "# we keep other parameters like regularization parameter and learning_rate as default values.\n",
        "bsl_options = {'method': 'sgd'}\n",
        "\n",
        "\n",
        "knn_bsl_m = KNNBaseline(k=40, sim_options = sim_options, bsl_options = bsl_options)\n",
        "\n",
        "knn_bsl_m_train_results, knn_bsl_m_test_results = run_surprise(knn_bsl_m, trainset, testset, verbose=True)\n",
        "\n",
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "models_evaluation_train['knn_bsl_m'] = knn_bsl_m_train_results \n",
        "models_evaluation_test['knn_bsl_m'] = knn_bsl_m_test_results\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            "Estimating biases using sgd...\n",
            "Computing the pearson_baseline similarity matrix...\n",
            "Done computing similarity matrix.\n",
            "Done. time taken : 0:00:01.093096 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:00:07.964272\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.32584796251610554\n",
            "\n",
            "MAPE : 8.447062581998374\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.075229\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.072758832653683\n",
            "\n",
            "MAPE : 35.02269653015042\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:00:09.133017\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "BXnyPGztGHRR"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "4MyrRggKGHRR"
      },
      "source": [
        "<h3> 4.4.5 XGBoost with initial 13 features + Surprise Baseline predictor + KNNBaseline predictor </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "giJoi5O5GHRS"
      },
      "source": [
        "- - - First we will run XGBoost with predictions from both KNN's ( that uses User\\_User and Item\\_Item similarities along with our previous features.\n",
        "\n",
        " \n",
        "- - - Then we will run XGBoost with just predictions form both knn models and preditions from our baseline model. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "QKnprtpsGHRS"
      },
      "source": [
        "__Preparing Train data __"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "e3025e82-1136-4f0a-8b79-524498c5b46f",
        "id": "kKOAq9JcGHRT",
        "colab": {}
      },
      "source": [
        "# add the predicted values from both knns to this dataframe\n",
        "reg_train['knn_bsl_u'] = models_evaluation_train['knn_bsl_u']['predictions']\n",
        "reg_train['knn_bsl_m'] = models_evaluation_train['knn_bsl_m']['predictions']\n",
        "\n",
        "reg_train.head(2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>bslpr</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>53406</td>\n",
              "      <td>33</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.370370</td>\n",
              "      <td>4.092437</td>\n",
              "      <td>4</td>\n",
              "      <td>3.898982</td>\n",
              "      <td>3.93002</td>\n",
              "      <td>3.867958</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>99540</td>\n",
              "      <td>33</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.555556</td>\n",
              "      <td>4.092437</td>\n",
              "      <td>3</td>\n",
              "      <td>3.371403</td>\n",
              "      <td>3.17733</td>\n",
              "      <td>3.076302</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "    user  movie      GAvg  sur1  sur2  sur3  sur4  sur5  smr1  smr2  smr3  \\\n",
              "0  53406     33  3.581679   4.0   5.0   5.0   4.0   1.0   5.0   2.0   5.0   \n",
              "1  99540     33  3.581679   5.0   5.0   5.0   4.0   5.0   3.0   4.0   4.0   \n",
              "\n",
              "   smr4  smr5      UAvg      MAvg  rating     bslpr  knn_bsl_u  knn_bsl_m  \n",
              "0   3.0   1.0  3.370370  4.092437       4  3.898982    3.93002   3.867958  \n",
              "1   3.0   5.0  3.555556  4.092437       3  3.371403    3.17733   3.076302  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Pf5oI7XyGHRV"
      },
      "source": [
        "__Preparing Test data  __"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "7137cabd-a1cb-46a6-b64f-937a1d7e407f",
        "id": "suF6egPpGHRV",
        "colab": {}
      },
      "source": [
        "reg_test_df['knn_bsl_u'] = models_evaluation_test['knn_bsl_u']['predictions']\n",
        "reg_test_df['knn_bsl_m'] = models_evaluation_test['knn_bsl_m']['predictions']\n",
        "\n",
        "reg_test_df.head(2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>bslpr</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>808635</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>941866</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg      sur1      sur2      sur3      sur4      sur5  \\\n",
              "0  808635     71  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "1  941866     71  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "\n",
              "       smr1      smr2      smr3      smr4      smr5      UAvg      MAvg  \\\n",
              "0  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "1  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "\n",
              "   rating     bslpr  knn_bsl_u  knn_bsl_m  \n",
              "0       5  3.581679   3.581679   3.581679  \n",
              "1       4  3.581679   3.581679   3.581679  "
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "fc95b5f1-6eaf-4398-8015-729096e3629c",
        "id": "HrbMGwSHGHRY",
        "colab": {}
      },
      "source": [
        "# prepare the train data....\n",
        "x_train = reg_train.drop(['user', 'movie', 'rating'], axis=1)\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# prepare the train data....\n",
        "x_test = reg_test_df.drop(['user','movie','rating'], axis=1)\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "# declare the model\n",
        "xgb_knn_bsl = xgb.XGBRegressor(n_jobs=10, random_state=15)\n",
        "train_results, test_results = run_xgboost(xgb_knn_bsl, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "models_evaluation_train['xgb_knn_bsl'] = train_results\n",
        "models_evaluation_test['xgb_knn_bsl'] = test_results\n",
        "\n",
        "\n",
        "xgb.plot_importance(xgb_knn_bsl)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model..\n",
            "Done. Time taken : 0:00:02.092387\n",
            "\n",
            "Done \n",
            "\n",
            "Evaluating the model with TRAIN data...\n",
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.0763602465199797\n",
            "MAPE :  34.48862808016984\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "dkCFYIfxGHRa"
      },
      "source": [
        "<h3> 4.4.6 Matrix Factorization Techniques </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "_jSmfkoTGHRb"
      },
      "source": [
        "<h4> 4.4.6.1 SVD Matrix Factorization User Movie intractions </h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "JNLdtZ30GHRb",
        "colab": {}
      },
      "source": [
        "from surprise import SVD"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "MSyKGXPkGHRc"
      },
      "source": [
        "http://surprise.readthedocs.io/en/stable/matrix_factorization.html#surprise.prediction_algorithms.matrix_factorization.SVD "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "kFy5-aZAGHRd"
      },
      "source": [
        "- __ Predicted Rating : __\n",
        "    - \n",
        "    - $ \\large  \\hat r_{ui} = \\mu + b_u + b_i + q_i^Tp_u $\n",
        "    \n",
        "        - $\\pmb q_i$ - Representation of item(movie) in latent factor space\n",
        "        \n",
        "        - $\\pmb p_u$ - Representation of user in new latent factor space\n",
        "        \n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "zHTRaqUHGHRd"
      },
      "source": [
        "- A BASIC MATRIX FACTORIZATION MODEL in  https://datajobs.com/data-science-repo/Recommender-Systems-[Netflix].pdf"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "_uQA4NX7GHRe"
      },
      "source": [
        "- __Optimization problem with user item interactions and regularization (to avoid overfitting)__\n",
        "    - \n",
        "    - $\\large \\sum_{r_{ui} \\in R_{train}} \\left(r_{ui} - \\hat{r}_{ui} \\right)^2 +\n",
        "\\lambda\\left(b_i^2 + b_u^2 + ||q_i||^2 + ||p_u||^2\\right) $"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "26accf59-7e7b-4973-9e3f-6a47d2a7c32d",
        "id": "E5SvFo1BGHRe",
        "colab": {}
      },
      "source": [
        "# initiallize the model\n",
        "svd = SVD(n_factors=100, biased=True, random_state=15, verbose=True)\n",
        "svd_train_results, svd_test_results = run_surprise(svd, trainset, testset, verbose=True)\n",
        "\n",
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "models_evaluation_train['svd'] = svd_train_results \n",
        "models_evaluation_test['svd'] = svd_test_results"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            "Processing epoch 0\n",
            "Processing epoch 1\n",
            "Processing epoch 2\n",
            "Processing epoch 3\n",
            "Processing epoch 4\n",
            "Processing epoch 5\n",
            "Processing epoch 6\n",
            "Processing epoch 7\n",
            "Processing epoch 8\n",
            "Processing epoch 9\n",
            "Processing epoch 10\n",
            "Processing epoch 11\n",
            "Processing epoch 12\n",
            "Processing epoch 13\n",
            "Processing epoch 14\n",
            "Processing epoch 15\n",
            "Processing epoch 16\n",
            "Processing epoch 17\n",
            "Processing epoch 18\n",
            "Processing epoch 19\n",
            "Done. time taken : 0:00:07.297438 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:00:01.305539\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.6574721240954099\n",
            "\n",
            "MAPE : 19.704901088660474\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.067811\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.0726046873826458\n",
            "\n",
            "MAPE : 35.01953535988152\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:00:08.671347\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "LYtjizREGHRg"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "igvo-ChaGHRh"
      },
      "source": [
        "  <h4> 4.4.6.2 SVD Matrix Factorization with implicit feedback from user ( user rated movies ) </h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "8dbSFL97GHRi",
        "colab": {}
      },
      "source": [
        "from surprise import SVDpp"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Lt2dMwUBGHRj"
      },
      "source": [
        "- ----->  2.5 Implicit Feedback in http://courses.ischool.berkeley.edu/i290-dm/s11/SECURE/a1-koren.pdf"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "MtJ96Iu8GHRk"
      },
      "source": [
        "- __ Predicted Rating : __\n",
        "    - \n",
        "    - $ \\large \\hat{r}_{ui} = \\mu + b_u + b_i + q_i^T\\left(p_u +\n",
        "    |I_u|^{-\\frac{1}{2}} \\sum_{j \\in I_u}y_j\\right) $ "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ENt1oW5rGHRk"
      },
      "source": [
        " - $ \\pmb{I_u}$ --- the set of all items rated by user u\n",
        "\n",
        "- $\\pmb{y_j}$ --- Our new set of item factors that capture implicit ratings.  "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "HAq-DaWoGHRl"
      },
      "source": [
        "- __Optimization problem with user item interactions and regularization (to avoid overfitting)__\n",
        "    - \n",
        "    - $ \\large \\sum_{r_{ui} \\in R_{train}} \\left(r_{ui} - \\hat{r}_{ui} \\right)^2 +\n",
        "\\lambda\\left(b_i^2 + b_u^2 + ||q_i||^2 + ||p_u||^2 + ||y_j||^2\\right) $ "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "d4a9169f-e5da-4c71-9394-2de56687bad3",
        "id": "VXSVAgBnGHRl",
        "colab": {}
      },
      "source": [
        "# initiallize the model\n",
        "svdpp = SVDpp(n_factors=50, random_state=15, verbose=True)\n",
        "svdpp_train_results, svdpp_test_results = run_surprise(svdpp, trainset, testset, verbose=True)\n",
        "\n",
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "models_evaluation_train['svdpp'] = svdpp_train_results \n",
        "models_evaluation_test['svdpp'] = svdpp_test_results\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            " processing epoch 0\n",
            " processing epoch 1\n",
            " processing epoch 2\n",
            " processing epoch 3\n",
            " processing epoch 4\n",
            " processing epoch 5\n",
            " processing epoch 6\n",
            " processing epoch 7\n",
            " processing epoch 8\n",
            " processing epoch 9\n",
            " processing epoch 10\n",
            " processing epoch 11\n",
            " processing epoch 12\n",
            " processing epoch 13\n",
            " processing epoch 14\n",
            " processing epoch 15\n",
            " processing epoch 16\n",
            " processing epoch 17\n",
            " processing epoch 18\n",
            " processing epoch 19\n",
            "Done. time taken : 0:01:56.765007 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:00:06.387920\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.6032438403305899\n",
            "\n",
            "MAPE : 17.49285063490268\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.071642\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.0728491944183447\n",
            "\n",
            "MAPE : 35.03817913919887\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:02:03.225068\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "JY6alxu4GHRn"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "PL2V49faGHRn"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "OChrCmcGGHRo"
      },
      "source": [
        "<h3> 4.4.7 XgBoost with 13 features + Surprise Baseline + Surprise KNNbaseline + MF Techniques </h3>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "jhzF7uV0GHRp"
      },
      "source": [
        "__Preparing Train data__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "c33595e6-d05e-4807-e813-1ecd42da5e38",
        "id": "egHxbWBNGHRp",
        "colab": {}
      },
      "source": [
        "# add the predicted values from both knns to this dataframe\n",
        "reg_train['svd'] = models_evaluation_train['svd']['predictions']\n",
        "reg_train['svdpp'] = models_evaluation_train['svdpp']['predictions']\n",
        "\n",
        "reg_train.head(2) "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>...</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>bslpr</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>svd</th>\n",
              "      <th>svdpp</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>53406</td>\n",
              "      <td>33</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.370370</td>\n",
              "      <td>4.092437</td>\n",
              "      <td>4</td>\n",
              "      <td>3.898982</td>\n",
              "      <td>3.93002</td>\n",
              "      <td>3.867958</td>\n",
              "      <td>4.303764</td>\n",
              "      <td>3.621664</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>99540</td>\n",
              "      <td>33</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>...</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.555556</td>\n",
              "      <td>4.092437</td>\n",
              "      <td>3</td>\n",
              "      <td>3.371403</td>\n",
              "      <td>3.17733</td>\n",
              "      <td>3.076302</td>\n",
              "      <td>3.278967</td>\n",
              "      <td>3.538247</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2 rows × 21 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "    user  movie      GAvg  sur1  sur2  sur3  sur4  sur5  smr1  smr2    ...     \\\n",
              "0  53406     33  3.581679   4.0   5.0   5.0   4.0   1.0   5.0   2.0    ...      \n",
              "1  99540     33  3.581679   5.0   5.0   5.0   4.0   5.0   3.0   4.0    ...      \n",
              "\n",
              "   smr4  smr5      UAvg      MAvg  rating     bslpr  knn_bsl_u  knn_bsl_m  \\\n",
              "0   3.0   1.0  3.370370  4.092437       4  3.898982    3.93002   3.867958   \n",
              "1   3.0   5.0  3.555556  4.092437       3  3.371403    3.17733   3.076302   \n",
              "\n",
              "        svd     svdpp  \n",
              "0  4.303764  3.621664  \n",
              "1  3.278967  3.538247  \n",
              "\n",
              "[2 rows x 21 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 59
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "AYR2mci0GHRr"
      },
      "source": [
        "__Preparing Test data  __"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "39997a27-2a18-46b8-8b1f-f43f5b502151",
        "id": "7kLmXTQjGHRr",
        "colab": {}
      },
      "source": [
        "reg_test_df['svd'] = models_evaluation_test['svd']['predictions']\n",
        "reg_test_df['svdpp'] = models_evaluation_test['svdpp']['predictions']\n",
        "\n",
        "reg_test_df.head(2) "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>...</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>bslpr</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>svd</th>\n",
              "      <th>svdpp</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>808635</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>...</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>941866</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>...</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2 rows × 21 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg      sur1      sur2      sur3      sur4      sur5  \\\n",
              "0  808635     71  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "1  941866     71  3.581679  3.581679  3.581679  3.581679  3.581679  3.581679   \n",
              "\n",
              "       smr1      smr2    ...         smr4      smr5      UAvg      MAvg  \\\n",
              "0  3.581679  3.581679    ...     3.581679  3.581679  3.581679  3.581679   \n",
              "1  3.581679  3.581679    ...     3.581679  3.581679  3.581679  3.581679   \n",
              "\n",
              "   rating     bslpr  knn_bsl_u  knn_bsl_m       svd     svdpp  \n",
              "0       5  3.581679   3.581679   3.581679  3.581679  3.581679  \n",
              "1       4  3.581679   3.581679   3.581679  3.581679  3.581679  \n",
              "\n",
              "[2 rows x 21 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 60
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "uobp0yUdGHRs"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "83ccbda0-527a-4e09-c9cf-ae86ac1cd7c1",
        "id": "Etf4AipqGHRt",
        "colab": {}
      },
      "source": [
        "# prepare x_train and y_train\n",
        "x_train = reg_train.drop(['user', 'movie', 'rating',], axis=1)\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# prepare test data\n",
        "x_test = reg_test_df.drop(['user', 'movie', 'rating'], axis=1)\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "\n",
        "\n",
        "xgb_final = xgb.XGBRegressor(n_jobs=10, random_state=15)\n",
        "train_results, test_results = run_xgboost(xgb_final, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "models_evaluation_train['xgb_final'] = train_results\n",
        "models_evaluation_test['xgb_final'] = test_results\n",
        "\n",
        "\n",
        "xgb.plot_importance(xgb_final)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model..\n",
            "Done. Time taken : 0:00:04.203252\n",
            "\n",
            "Done \n",
            "\n",
            "Evaluating the model with TRAIN data...\n",
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.0763580984894978\n",
            "MAPE :  34.487391651053336\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "n82AAbjKGHRu"
      },
      "source": [
        "<h3> 4.4.8 XgBoost with Surprise Baseline + Surprise KNNbaseline + MF Techniques </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "99e75da3-58bf-4b1f-bd65-d8162ee50f20",
        "id": "3Q1gteaAGHRv",
        "colab": {}
      },
      "source": [
        "# prepare train data\n",
        "x_train = reg_train[['knn_bsl_u', 'knn_bsl_m', 'svd', 'svdpp']]\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# test data\n",
        "x_test = reg_test_df[['knn_bsl_u', 'knn_bsl_m', 'svd', 'svdpp']]\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "\n",
        "xgb_all_models = xgb.XGBRegressor(n_jobs=10, random_state=15)\n",
        "train_results, test_results = run_xgboost(xgb_all_models, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "models_evaluation_train['xgb_all_models'] = train_results\n",
        "models_evaluation_test['xgb_all_models'] = test_results\n",
        "\n",
        "xgb.plot_importance(xgb_all_models)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model..\n",
            "Done. Time taken : 0:00:01.292225\n",
            "\n",
            "Done \n",
            "\n",
            "Evaluating the model with TRAIN data...\n",
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.075480663561971\n",
            "MAPE :  35.01826709436013\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "/* Put everything inside the global mpl namespace */\n",
              "window.mpl = {};\n",
              "\n",
              "\n",
              "mpl.get_websocket_type = function() {\n",
              "    if (typeof(WebSocket) !== 'undefined') {\n",
              "        return WebSocket;\n",
              "    } else if (typeof(MozWebSocket) !== 'undefined') {\n",
              "        return MozWebSocket;\n",
              "    } else {\n",
              "        alert('Your browser does not have WebSocket support.' +\n",
              "              'Please try Chrome, Safari or Firefox ≥ 6. ' +\n",
              "              'Firefox 4 and 5 are also supported but you ' +\n",
              "              'have to enable WebSockets in about:config.');\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure = function(figure_id, websocket, ondownload, parent_element) {\n",
              "    this.id = figure_id;\n",
              "\n",
              "    this.ws = websocket;\n",
              "\n",
              "    this.supports_binary = (this.ws.binaryType != undefined);\n",
              "\n",
              "    if (!this.supports_binary) {\n",
              "        var warnings = document.getElementById(\"mpl-warnings\");\n",
              "        if (warnings) {\n",
              "            warnings.style.display = 'block';\n",
              "            warnings.textContent = (\n",
              "                \"This browser does not support binary websocket messages. \" +\n",
              "                    \"Performance may be slow.\");\n",
              "        }\n",
              "    }\n",
              "\n",
              "    this.imageObj = new Image();\n",
              "\n",
              "    this.context = undefined;\n",
              "    this.message = undefined;\n",
              "    this.canvas = undefined;\n",
              "    this.rubberband_canvas = undefined;\n",
              "    this.rubberband_context = undefined;\n",
              "    this.format_dropdown = undefined;\n",
              "\n",
              "    this.image_mode = 'full';\n",
              "\n",
              "    this.root = $('<div/>');\n",
              "    this._root_extra_style(this.root)\n",
              "    this.root.attr('style', 'display: inline-block');\n",
              "\n",
              "    $(parent_element).append(this.root);\n",
              "\n",
              "    this._init_header(this);\n",
              "    this._init_canvas(this);\n",
              "    this._init_toolbar(this);\n",
              "\n",
              "    var fig = this;\n",
              "\n",
              "    this.waiting = false;\n",
              "\n",
              "    this.ws.onopen =  function () {\n",
              "            fig.send_message(\"supports_binary\", {value: fig.supports_binary});\n",
              "            fig.send_message(\"send_image_mode\", {});\n",
              "            if (mpl.ratio != 1) {\n",
              "                fig.send_message(\"set_dpi_ratio\", {'dpi_ratio': mpl.ratio});\n",
              "            }\n",
              "            fig.send_message(\"refresh\", {});\n",
              "        }\n",
              "\n",
              "    this.imageObj.onload = function() {\n",
              "            if (fig.image_mode == 'full') {\n",
              "                // Full images could contain transparency (where diff images\n",
              "                // almost always do), so we need to clear the canvas so that\n",
              "                // there is no ghosting.\n",
              "                fig.context.clearRect(0, 0, fig.canvas.width, fig.canvas.height);\n",
              "            }\n",
              "            fig.context.drawImage(fig.imageObj, 0, 0);\n",
              "        };\n",
              "\n",
              "    this.imageObj.onunload = function() {\n",
              "        fig.ws.close();\n",
              "    }\n",
              "\n",
              "    this.ws.onmessage = this._make_on_message_function(this);\n",
              "\n",
              "    this.ondownload = ondownload;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_header = function() {\n",
              "    var titlebar = $(\n",
              "        '<div class=\"ui-dialog-titlebar ui-widget-header ui-corner-all ' +\n",
              "        'ui-helper-clearfix\"/>');\n",
              "    var titletext = $(\n",
              "        '<div class=\"ui-dialog-title\" style=\"width: 100%; ' +\n",
              "        'text-align: center; padding: 3px;\"/>');\n",
              "    titlebar.append(titletext)\n",
              "    this.root.append(titlebar);\n",
              "    this.header = titletext[0];\n",
              "}\n",
              "\n",
              "\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(canvas_div) {\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_canvas = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var canvas_div = $('<div/>');\n",
              "\n",
              "    canvas_div.attr('style', 'position: relative; clear: both; outline: 0');\n",
              "\n",
              "    function canvas_keyboard_event(event) {\n",
              "        return fig.key_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    canvas_div.keydown('key_press', canvas_keyboard_event);\n",
              "    canvas_div.keyup('key_release', canvas_keyboard_event);\n",
              "    this.canvas_div = canvas_div\n",
              "    this._canvas_extra_style(canvas_div)\n",
              "    this.root.append(canvas_div);\n",
              "\n",
              "    var canvas = $('<canvas/>');\n",
              "    canvas.addClass('mpl-canvas');\n",
              "    canvas.attr('style', \"left: 0; top: 0; z-index: 0; outline: 0\")\n",
              "\n",
              "    this.canvas = canvas[0];\n",
              "    this.context = canvas[0].getContext(\"2d\");\n",
              "\n",
              "    var backingStore = this.context.backingStorePixelRatio ||\n",
              "\tthis.context.webkitBackingStorePixelRatio ||\n",
              "\tthis.context.mozBackingStorePixelRatio ||\n",
              "\tthis.context.msBackingStorePixelRatio ||\n",
              "\tthis.context.oBackingStorePixelRatio ||\n",
              "\tthis.context.backingStorePixelRatio || 1;\n",
              "\n",
              "    mpl.ratio = (window.devicePixelRatio || 1) / backingStore;\n",
              "\n",
              "    var rubberband = $('<canvas/>');\n",
              "    rubberband.attr('style', \"position: absolute; left: 0; top: 0; z-index: 1;\")\n",
              "\n",
              "    var pass_mouse_events = true;\n",
              "\n",
              "    canvas_div.resizable({\n",
              "        start: function(event, ui) {\n",
              "            pass_mouse_events = false;\n",
              "        },\n",
              "        resize: function(event, ui) {\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "        stop: function(event, ui) {\n",
              "            pass_mouse_events = true;\n",
              "            fig.request_resize(ui.size.width, ui.size.height);\n",
              "        },\n",
              "    });\n",
              "\n",
              "    function mouse_event_fn(event) {\n",
              "        if (pass_mouse_events)\n",
              "            return fig.mouse_event(event, event['data']);\n",
              "    }\n",
              "\n",
              "    rubberband.mousedown('button_press', mouse_event_fn);\n",
              "    rubberband.mouseup('button_release', mouse_event_fn);\n",
              "    // Throttle sequential mouse events to 1 every 20ms.\n",
              "    rubberband.mousemove('motion_notify', mouse_event_fn);\n",
              "\n",
              "    rubberband.mouseenter('figure_enter', mouse_event_fn);\n",
              "    rubberband.mouseleave('figure_leave', mouse_event_fn);\n",
              "\n",
              "    canvas_div.on(\"wheel\", function (event) {\n",
              "        event = event.originalEvent;\n",
              "        event['data'] = 'scroll'\n",
              "        if (event.deltaY < 0) {\n",
              "            event.step = 1;\n",
              "        } else {\n",
              "            event.step = -1;\n",
              "        }\n",
              "        mouse_event_fn(event);\n",
              "    });\n",
              "\n",
              "    canvas_div.append(canvas);\n",
              "    canvas_div.append(rubberband);\n",
              "\n",
              "    this.rubberband = rubberband;\n",
              "    this.rubberband_canvas = rubberband[0];\n",
              "    this.rubberband_context = rubberband[0].getContext(\"2d\");\n",
              "    this.rubberband_context.strokeStyle = \"#000000\";\n",
              "\n",
              "    this._resize_canvas = function(width, height) {\n",
              "        // Keep the size of the canvas, canvas container, and rubber band\n",
              "        // canvas in synch.\n",
              "        canvas_div.css('width', width)\n",
              "        canvas_div.css('height', height)\n",
              "\n",
              "        canvas.attr('width', width * mpl.ratio);\n",
              "        canvas.attr('height', height * mpl.ratio);\n",
              "        canvas.attr('style', 'width: ' + width + 'px; height: ' + height + 'px;');\n",
              "\n",
              "        rubberband.attr('width', width);\n",
              "        rubberband.attr('height', height);\n",
              "    }\n",
              "\n",
              "    // Set the figure to an initial 600x600px, this will subsequently be updated\n",
              "    // upon first draw.\n",
              "    this._resize_canvas(600, 600);\n",
              "\n",
              "    // Disable right mouse context menu.\n",
              "    $(this.rubberband_canvas).bind(\"contextmenu\",function(e){\n",
              "        return false;\n",
              "    });\n",
              "\n",
              "    function set_focus () {\n",
              "        canvas.focus();\n",
              "        canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    window.setTimeout(set_focus, 100);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items) {\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) {\n",
              "            // put a spacer in here.\n",
              "            continue;\n",
              "        }\n",
              "        var button = $('<button/>');\n",
              "        button.addClass('ui-button ui-widget ui-state-default ui-corner-all ' +\n",
              "                        'ui-button-icon-only');\n",
              "        button.attr('role', 'button');\n",
              "        button.attr('aria-disabled', 'false');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "\n",
              "        var icon_img = $('<span/>');\n",
              "        icon_img.addClass('ui-button-icon-primary ui-icon');\n",
              "        icon_img.addClass(image);\n",
              "        icon_img.addClass('ui-corner-all');\n",
              "\n",
              "        var tooltip_span = $('<span/>');\n",
              "        tooltip_span.addClass('ui-button-text');\n",
              "        tooltip_span.html(tooltip);\n",
              "\n",
              "        button.append(icon_img);\n",
              "        button.append(tooltip_span);\n",
              "\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    var fmt_picker_span = $('<span/>');\n",
              "\n",
              "    var fmt_picker = $('<select/>');\n",
              "    fmt_picker.addClass('mpl-toolbar-option ui-widget ui-widget-content');\n",
              "    fmt_picker_span.append(fmt_picker);\n",
              "    nav_element.append(fmt_picker_span);\n",
              "    this.format_dropdown = fmt_picker[0];\n",
              "\n",
              "    for (var ind in mpl.extensions) {\n",
              "        var fmt = mpl.extensions[ind];\n",
              "        var option = $(\n",
              "            '<option/>', {selected: fmt === mpl.default_extension}).html(fmt);\n",
              "        fmt_picker.append(option)\n",
              "    }\n",
              "\n",
              "    // Add hover states to the ui-buttons\n",
              "    $( \".ui-button\" ).hover(\n",
              "        function() { $(this).addClass(\"ui-state-hover\");},\n",
              "        function() { $(this).removeClass(\"ui-state-hover\");}\n",
              "    );\n",
              "\n",
              "    var status_bar = $('<span class=\"mpl-message\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.request_resize = function(x_pixels, y_pixels) {\n",
              "    // Request matplotlib to resize the figure. Matplotlib will then trigger a resize in the client,\n",
              "    // which will in turn request a refresh of the image.\n",
              "    this.send_message('resize', {'width': x_pixels, 'height': y_pixels});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_message = function(type, properties) {\n",
              "    properties['type'] = type;\n",
              "    properties['figure_id'] = this.id;\n",
              "    this.ws.send(JSON.stringify(properties));\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.send_draw_message = function() {\n",
              "    if (!this.waiting) {\n",
              "        this.waiting = true;\n",
              "        this.ws.send(JSON.stringify({type: \"draw\", figure_id: this.id}));\n",
              "    }\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    var format_dropdown = fig.format_dropdown;\n",
              "    var format = format_dropdown.options[format_dropdown.selectedIndex].value;\n",
              "    fig.ondownload(fig, format);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.figure.prototype.handle_resize = function(fig, msg) {\n",
              "    var size = msg['size'];\n",
              "    if (size[0] != fig.canvas.width || size[1] != fig.canvas.height) {\n",
              "        fig._resize_canvas(size[0], size[1]);\n",
              "        fig.send_message(\"refresh\", {});\n",
              "    };\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_rubberband = function(fig, msg) {\n",
              "    var x0 = msg['x0'] / mpl.ratio;\n",
              "    var y0 = (fig.canvas.height - msg['y0']) / mpl.ratio;\n",
              "    var x1 = msg['x1'] / mpl.ratio;\n",
              "    var y1 = (fig.canvas.height - msg['y1']) / mpl.ratio;\n",
              "    x0 = Math.floor(x0) + 0.5;\n",
              "    y0 = Math.floor(y0) + 0.5;\n",
              "    x1 = Math.floor(x1) + 0.5;\n",
              "    y1 = Math.floor(y1) + 0.5;\n",
              "    var min_x = Math.min(x0, x1);\n",
              "    var min_y = Math.min(y0, y1);\n",
              "    var width = Math.abs(x1 - x0);\n",
              "    var height = Math.abs(y1 - y0);\n",
              "\n",
              "    fig.rubberband_context.clearRect(\n",
              "        0, 0, fig.canvas.width, fig.canvas.height);\n",
              "\n",
              "    fig.rubberband_context.strokeRect(min_x, min_y, width, height);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_figure_label = function(fig, msg) {\n",
              "    // Updates the figure title.\n",
              "    fig.header.textContent = msg['label'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_cursor = function(fig, msg) {\n",
              "    var cursor = msg['cursor'];\n",
              "    switch(cursor)\n",
              "    {\n",
              "    case 0:\n",
              "        cursor = 'pointer';\n",
              "        break;\n",
              "    case 1:\n",
              "        cursor = 'default';\n",
              "        break;\n",
              "    case 2:\n",
              "        cursor = 'crosshair';\n",
              "        break;\n",
              "    case 3:\n",
              "        cursor = 'move';\n",
              "        break;\n",
              "    }\n",
              "    fig.rubberband_canvas.style.cursor = cursor;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_message = function(fig, msg) {\n",
              "    fig.message.textContent = msg['message'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_draw = function(fig, msg) {\n",
              "    // Request the server to send over a new figure.\n",
              "    fig.send_draw_message();\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_image_mode = function(fig, msg) {\n",
              "    fig.image_mode = msg['mode'];\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Called whenever the canvas gets updated.\n",
              "    this.send_message(\"ack\", {});\n",
              "}\n",
              "\n",
              "// A function to construct a web socket function for onmessage handling.\n",
              "// Called in the figure constructor.\n",
              "mpl.figure.prototype._make_on_message_function = function(fig) {\n",
              "    return function socket_on_message(evt) {\n",
              "        if (evt.data instanceof Blob) {\n",
              "            /* FIXME: We get \"Resource interpreted as Image but\n",
              "             * transferred with MIME type text/plain:\" errors on\n",
              "             * Chrome.  But how to set the MIME type?  It doesn't seem\n",
              "             * to be part of the websocket stream */\n",
              "            evt.data.type = \"image/png\";\n",
              "\n",
              "            /* Free the memory for the previous frames */\n",
              "            if (fig.imageObj.src) {\n",
              "                (window.URL || window.webkitURL).revokeObjectURL(\n",
              "                    fig.imageObj.src);\n",
              "            }\n",
              "\n",
              "            fig.imageObj.src = (window.URL || window.webkitURL).createObjectURL(\n",
              "                evt.data);\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "        else if (typeof evt.data === 'string' && evt.data.slice(0, 21) == \"data:image/png;base64\") {\n",
              "            fig.imageObj.src = evt.data;\n",
              "            fig.updated_canvas_event();\n",
              "            fig.waiting = false;\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        var msg = JSON.parse(evt.data);\n",
              "        var msg_type = msg['type'];\n",
              "\n",
              "        // Call the  \"handle_{type}\" callback, which takes\n",
              "        // the figure and JSON message as its only arguments.\n",
              "        try {\n",
              "            var callback = fig[\"handle_\" + msg_type];\n",
              "        } catch (e) {\n",
              "            console.log(\"No handler for the '\" + msg_type + \"' message type: \", msg);\n",
              "            return;\n",
              "        }\n",
              "\n",
              "        if (callback) {\n",
              "            try {\n",
              "                // console.log(\"Handling '\" + msg_type + \"' message: \", msg);\n",
              "                callback(fig, msg);\n",
              "            } catch (e) {\n",
              "                console.log(\"Exception inside the 'handler_\" + msg_type + \"' callback:\", e, e.stack, msg);\n",
              "            }\n",
              "        }\n",
              "    };\n",
              "}\n",
              "\n",
              "// from http://stackoverflow.com/questions/1114465/getting-mouse-location-in-canvas\n",
              "mpl.findpos = function(e) {\n",
              "    //this section is from http://www.quirksmode.org/js/events_properties.html\n",
              "    var targ;\n",
              "    if (!e)\n",
              "        e = window.event;\n",
              "    if (e.target)\n",
              "        targ = e.target;\n",
              "    else if (e.srcElement)\n",
              "        targ = e.srcElement;\n",
              "    if (targ.nodeType == 3) // defeat Safari bug\n",
              "        targ = targ.parentNode;\n",
              "\n",
              "    // jQuery normalizes the pageX and pageY\n",
              "    // pageX,Y are the mouse positions relative to the document\n",
              "    // offset() returns the position of the element relative to the document\n",
              "    var x = e.pageX - $(targ).offset().left;\n",
              "    var y = e.pageY - $(targ).offset().top;\n",
              "\n",
              "    return {\"x\": x, \"y\": y};\n",
              "};\n",
              "\n",
              "/*\n",
              " * return a copy of an object with only non-object keys\n",
              " * we need this to avoid circular references\n",
              " * http://stackoverflow.com/a/24161582/3208463\n",
              " */\n",
              "function simpleKeys (original) {\n",
              "  return Object.keys(original).reduce(function (obj, key) {\n",
              "    if (typeof original[key] !== 'object')\n",
              "        obj[key] = original[key]\n",
              "    return obj;\n",
              "  }, {});\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.mouse_event = function(event, name) {\n",
              "    var canvas_pos = mpl.findpos(event)\n",
              "\n",
              "    if (name === 'button_press')\n",
              "    {\n",
              "        this.canvas.focus();\n",
              "        this.canvas_div.focus();\n",
              "    }\n",
              "\n",
              "    var x = canvas_pos.x * mpl.ratio;\n",
              "    var y = canvas_pos.y * mpl.ratio;\n",
              "\n",
              "    this.send_message(name, {x: x, y: y, button: event.button,\n",
              "                             step: event.step,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "\n",
              "    /* This prevents the web browser from automatically changing to\n",
              "     * the text insertion cursor when the button is pressed.  We want\n",
              "     * to control all of the cursor setting manually through the\n",
              "     * 'cursor' event from matplotlib */\n",
              "    event.preventDefault();\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    // Handle any extra behaviour associated with a key event\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.key_event = function(event, name) {\n",
              "\n",
              "    // Prevent repeat events\n",
              "    if (name == 'key_press')\n",
              "    {\n",
              "        if (event.which === this._key)\n",
              "            return;\n",
              "        else\n",
              "            this._key = event.which;\n",
              "    }\n",
              "    if (name == 'key_release')\n",
              "        this._key = null;\n",
              "\n",
              "    var value = '';\n",
              "    if (event.ctrlKey && event.which != 17)\n",
              "        value += \"ctrl+\";\n",
              "    if (event.altKey && event.which != 18)\n",
              "        value += \"alt+\";\n",
              "    if (event.shiftKey && event.which != 16)\n",
              "        value += \"shift+\";\n",
              "\n",
              "    value += 'k';\n",
              "    value += event.which.toString();\n",
              "\n",
              "    this._key_event_extra(event, name);\n",
              "\n",
              "    this.send_message(name, {key: value,\n",
              "                             guiEvent: simpleKeys(event)});\n",
              "    return false;\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onclick = function(name) {\n",
              "    if (name == 'download') {\n",
              "        this.handle_save(this, null);\n",
              "    } else {\n",
              "        this.send_message(\"toolbar_button\", {name: name});\n",
              "    }\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.toolbar_button_onmouseover = function(tooltip) {\n",
              "    this.message.textContent = tooltip;\n",
              "};\n",
              "mpl.toolbar_items = [[\"Home\", \"Reset original view\", \"fa fa-home icon-home\", \"home\"], [\"Back\", \"Back to  previous view\", \"fa fa-arrow-left icon-arrow-left\", \"back\"], [\"Forward\", \"Forward to next view\", \"fa fa-arrow-right icon-arrow-right\", \"forward\"], [\"\", \"\", \"\", \"\"], [\"Pan\", \"Pan axes with left mouse, zoom with right\", \"fa fa-arrows icon-move\", \"pan\"], [\"Zoom\", \"Zoom to rectangle\", \"fa fa-square-o icon-check-empty\", \"zoom\"], [\"\", \"\", \"\", \"\"], [\"Download\", \"Download plot\", \"fa fa-floppy-o icon-save\", \"download\"]];\n",
              "\n",
              "mpl.extensions = [\"eps\", \"jpeg\", \"pdf\", \"png\", \"ps\", \"raw\", \"svg\", \"tif\"];\n",
              "\n",
              "mpl.default_extension = \"png\";var comm_websocket_adapter = function(comm) {\n",
              "    // Create a \"websocket\"-like object which calls the given IPython comm\n",
              "    // object with the appropriate methods. Currently this is a non binary\n",
              "    // socket, so there is still some room for performance tuning.\n",
              "    var ws = {};\n",
              "\n",
              "    ws.close = function() {\n",
              "        comm.close()\n",
              "    };\n",
              "    ws.send = function(m) {\n",
              "        //console.log('sending', m);\n",
              "        comm.send(m);\n",
              "    };\n",
              "    // Register the callback with on_msg.\n",
              "    comm.on_msg(function(msg) {\n",
              "        //console.log('receiving', msg['content']['data'], msg);\n",
              "        // Pass the mpl event to the overriden (by mpl) onmessage function.\n",
              "        ws.onmessage(msg['content']['data'])\n",
              "    });\n",
              "    return ws;\n",
              "}\n",
              "\n",
              "mpl.mpl_figure_comm = function(comm, msg) {\n",
              "    // This is the function which gets called when the mpl process\n",
              "    // starts-up an IPython Comm through the \"matplotlib\" channel.\n",
              "\n",
              "    var id = msg.content.data.id;\n",
              "    // Get hold of the div created by the display call when the Comm\n",
              "    // socket was opened in Python.\n",
              "    var element = $(\"#\" + id);\n",
              "    var ws_proxy = comm_websocket_adapter(comm)\n",
              "\n",
              "    function ondownload(figure, format) {\n",
              "        window.open(figure.imageObj.src);\n",
              "    }\n",
              "\n",
              "    var fig = new mpl.figure(id, ws_proxy,\n",
              "                           ondownload,\n",
              "                           element.get(0));\n",
              "\n",
              "    // Call onopen now - mpl needs it, as it is assuming we've passed it a real\n",
              "    // web socket which is closed, not our websocket->open comm proxy.\n",
              "    ws_proxy.onopen();\n",
              "\n",
              "    fig.parent_element = element.get(0);\n",
              "    fig.cell_info = mpl.find_output_cell(\"<div id='\" + id + \"'></div>\");\n",
              "    if (!fig.cell_info) {\n",
              "        console.error(\"Failed to find cell for figure\", id, fig);\n",
              "        return;\n",
              "    }\n",
              "\n",
              "    var output_index = fig.cell_info[2]\n",
              "    var cell = fig.cell_info[0];\n",
              "\n",
              "};\n",
              "\n",
              "mpl.figure.prototype.handle_close = function(fig, msg) {\n",
              "    var width = fig.canvas.width/mpl.ratio\n",
              "    fig.root.unbind('remove')\n",
              "\n",
              "    // Update the output cell to use the data from the current canvas.\n",
              "    fig.push_to_output();\n",
              "    var dataURL = fig.canvas.toDataURL();\n",
              "    // Re-enable the keyboard manager in IPython - without this line, in FF,\n",
              "    // the notebook keyboard shortcuts fail.\n",
              "    IPython.keyboard_manager.enable()\n",
              "    $(fig.parent_element).html('<img src=\"' + dataURL + '\" width=\"' + width + '\">');\n",
              "    fig.close_ws(fig, msg);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.close_ws = function(fig, msg){\n",
              "    fig.send_message('closing', msg);\n",
              "    // fig.ws.close()\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.push_to_output = function(remove_interactive) {\n",
              "    // Turn the data on the canvas into data in the output cell.\n",
              "    var width = this.canvas.width/mpl.ratio\n",
              "    var dataURL = this.canvas.toDataURL();\n",
              "    this.cell_info[1]['text/html'] = '<img src=\"' + dataURL + '\" width=\"' + width + '\">';\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.updated_canvas_event = function() {\n",
              "    // Tell IPython that the notebook contents must change.\n",
              "    IPython.notebook.set_dirty(true);\n",
              "    this.send_message(\"ack\", {});\n",
              "    var fig = this;\n",
              "    // Wait a second, then push the new image to the DOM so\n",
              "    // that it is saved nicely (might be nice to debounce this).\n",
              "    setTimeout(function () { fig.push_to_output() }, 1000);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._init_toolbar = function() {\n",
              "    var fig = this;\n",
              "\n",
              "    var nav_element = $('<div/>')\n",
              "    nav_element.attr('style', 'width: 100%');\n",
              "    this.root.append(nav_element);\n",
              "\n",
              "    // Define a callback function for later on.\n",
              "    function toolbar_event(event) {\n",
              "        return fig.toolbar_button_onclick(event['data']);\n",
              "    }\n",
              "    function toolbar_mouse_event(event) {\n",
              "        return fig.toolbar_button_onmouseover(event['data']);\n",
              "    }\n",
              "\n",
              "    for(var toolbar_ind in mpl.toolbar_items){\n",
              "        var name = mpl.toolbar_items[toolbar_ind][0];\n",
              "        var tooltip = mpl.toolbar_items[toolbar_ind][1];\n",
              "        var image = mpl.toolbar_items[toolbar_ind][2];\n",
              "        var method_name = mpl.toolbar_items[toolbar_ind][3];\n",
              "\n",
              "        if (!name) { continue; };\n",
              "\n",
              "        var button = $('<button class=\"btn btn-default\" href=\"#\" title=\"' + name + '\"><i class=\"fa ' + image + ' fa-lg\"></i></button>');\n",
              "        button.click(method_name, toolbar_event);\n",
              "        button.mouseover(tooltip, toolbar_mouse_event);\n",
              "        nav_element.append(button);\n",
              "    }\n",
              "\n",
              "    // Add the status bar.\n",
              "    var status_bar = $('<span class=\"mpl-message\" style=\"text-align:right; float: right;\"/>');\n",
              "    nav_element.append(status_bar);\n",
              "    this.message = status_bar[0];\n",
              "\n",
              "    // Add the close button to the window.\n",
              "    var buttongrp = $('<div class=\"btn-group inline pull-right\"></div>');\n",
              "    var button = $('<button class=\"btn btn-mini btn-primary\" href=\"#\" title=\"Stop Interaction\"><i class=\"fa fa-power-off icon-remove icon-large\"></i></button>');\n",
              "    button.click(function (evt) { fig.handle_close(fig, {}); } );\n",
              "    button.mouseover('Stop Interaction', toolbar_mouse_event);\n",
              "    buttongrp.append(button);\n",
              "    var titlebar = this.root.find($('.ui-dialog-titlebar'));\n",
              "    titlebar.prepend(buttongrp);\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._root_extra_style = function(el){\n",
              "    var fig = this\n",
              "    el.on(\"remove\", function(){\n",
              "\tfig.close_ws(fig, {});\n",
              "    });\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._canvas_extra_style = function(el){\n",
              "    // this is important to make the div 'focusable\n",
              "    el.attr('tabindex', 0)\n",
              "    // reach out to IPython and tell the keyboard manager to turn it's self\n",
              "    // off when our div gets focus\n",
              "\n",
              "    // location in version 3\n",
              "    if (IPython.notebook.keyboard_manager) {\n",
              "        IPython.notebook.keyboard_manager.register_events(el);\n",
              "    }\n",
              "    else {\n",
              "        // location in version 2\n",
              "        IPython.keyboard_manager.register_events(el);\n",
              "    }\n",
              "\n",
              "}\n",
              "\n",
              "mpl.figure.prototype._key_event_extra = function(event, name) {\n",
              "    var manager = IPython.notebook.keyboard_manager;\n",
              "    if (!manager)\n",
              "        manager = IPython.keyboard_manager;\n",
              "\n",
              "    // Check for shift+enter\n",
              "    if (event.shiftKey && event.which == 13) {\n",
              "        this.canvas_div.blur();\n",
              "        event.shiftKey = false;\n",
              "        // Send a \"J\" for go to next cell\n",
              "        event.which = 74;\n",
              "        event.keyCode = 74;\n",
              "        manager.command_mode();\n",
              "        manager.handle_keydown(event);\n",
              "    }\n",
              "}\n",
              "\n",
              "mpl.figure.prototype.handle_save = function(fig, msg) {\n",
              "    fig.ondownload(fig, null);\n",
              "}\n",
              "\n",
              "\n",
              "mpl.find_output_cell = function(html_output) {\n",
              "    // Return the cell and output element which can be found *uniquely* in the notebook.\n",
              "    // Note - this is a bit hacky, but it is done because the \"notebook_saving.Notebook\"\n",
              "    // IPython event is triggered only after the cells have been serialised, which for\n",
              "    // our purposes (turning an active figure into a static one), is too late.\n",
              "    var cells = IPython.notebook.get_cells();\n",
              "    var ncells = cells.length;\n",
              "    for (var i=0; i<ncells; i++) {\n",
              "        var cell = cells[i];\n",
              "        if (cell.cell_type === 'code'){\n",
              "            for (var j=0; j<cell.output_area.outputs.length; j++) {\n",
              "                var data = cell.output_area.outputs[j];\n",
              "                if (data.data) {\n",
              "                    // IPython >= 3 moved mimebundle to data attribute of output\n",
              "                    data = data.data;\n",
              "                }\n",
              "                if (data['text/html'] == html_output) {\n",
              "                    return [cell, data, j];\n",
              "                }\n",
              "            }\n",
              "        }\n",
              "    }\n",
              "}\n",
              "\n",
              "// Register the function which deals with the matplotlib target/channel.\n",
              "// The kernel may be null if the page has been refreshed.\n",
              "if (IPython.notebook.kernel != null) {\n",
              "    IPython.notebook.kernel.comm_manager.register_target('matplotlib', mpl.mpl_figure_comm);\n",
              "}\n"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<img src=\"data:image/png;base64,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\" width=\"640\">"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "OTH9n3xoGHRw"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "BVDDPizEGHRx"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "kw26rJ3SGHRy"
      },
      "source": [
        "<h2> 4.5 Comparision between all models </h2>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "colab_type": "code",
        "outputId": "9124baa4-ff9f-4ab4-c9f4-4d4af70f46f7",
        "id": "vSx1wSCDGHRy",
        "colab": {}
      },
      "source": [
        "# Saving our TEST_RESULTS into a dataframe so that you don't have to run it again\n",
        "pd.DataFrame(models_evaluation_test).to_csv('sample/small/small_sample_results.csv')\n",
        "models = pd.read_csv('sample/small/small_sample_results.csv', index_col=0)\n",
        "models.loc['rmse'].sort_values()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "svd               1.0726046873826458\n",
              "knn_bsl_u         1.0726493739667242\n",
              "knn_bsl_m          1.072758832653683\n",
              "svdpp             1.0728491944183447\n",
              "bsl_algo          1.0730330260516174\n",
              "xgb_knn_bsl_mu    1.0753229281412784\n",
              "xgb_all_models     1.075480663561971\n",
              "first_algo        1.0761851474385373\n",
              "xgb_bsl           1.0763419061709816\n",
              "xgb_final         1.0763580984894978\n",
              "xgb_knn_bsl       1.0763602465199797\n",
              "Name: rmse, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "WGZU19iSGHR0"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "530dd853-d632-4187-b1f8-be58f60b5e28",
        "id": "g0OB4lW0GHR1",
        "colab": {}
      },
      "source": [
        "print(\"-\"*100)\n",
        "print(\"Total time taken to run this entire notebook ( with saved files) is :\",datetime.now()-globalstart)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "----------------------------------------------------------------------------------------------------\n",
            "Total time taken to run this entire notebook ( with saved files) is : 0:42:08.302761\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Av5WS6n2xrsZ",
        "colab_type": "text"
      },
      "source": [
        "<h1> 5. Assignment </h1>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": true,
        "id": "JkK0HIzgxrsd",
        "colab_type": "text"
      },
      "source": [
        "1.Instead of using 10K users and 1K movies to train the above models, use 25K users and 3K movies (or more) to train all of the above models. Report the RMSE and MAPE on the test data using larger amount of data and provide a comparison between various models as shown above.\n",
        "\n",
        "NOTE: Please be patient as some of the code snippets make take many hours to compelte execution.\n",
        "\n",
        "2.Tune hyperparamters of all the Xgboost models above to improve the RMSE."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C-s6Laykxrsd",
        "colab_type": "code",
        "outputId": "83116885-7860-48e2-befd-8eca1bd66286",
        "colab": {}
      },
      "source": [
        "%%javascript\n",
        "// Converts integer to roman numeral\n",
        "// https://github.com/kmahelona/ipython_notebook_goodies\n",
        "// https://kmahelona.github.io/ipython_notebook_goodies/ipython_notebook_toc.js\n",
        "function romanize(num) {\n",
        "    var lookup = {M:1000,CM:900,D:500,CD:400,C:100,XC:90,L:50,XL:40,X:10,IX:9,V:5,IV:4,I:1},\n",
        "\troman = '',\n",
        "\t    i;\n",
        "\tfor ( i in lookup ) {\n",
        "\t    while ( num >= lookup[i] ) {\n",
        "\t\troman += i;\n",
        "\t\tnum -= lookup[i];\n",
        "\t    }\n",
        "\t}\n",
        "\treturn roman;\n",
        " }\n",
        "\n",
        "// Builds a <ul> Table of Contents from all <headers> in DOM\n",
        "function createTOC(){\n",
        "    var toc = \"\";\n",
        "    var level = 0;\n",
        "    var levels = {}\n",
        "    $('#toc').html('');\n",
        "\n",
        "    $(\":header\").each(function(i){\n",
        "\t    if (this.id=='tocheading'){return;}\n",
        "        \n",
        "\t    var titleText = this.innerHTML;\n",
        "\t    var openLevel = this.tagName[1];\n",
        "\n",
        "\t    if (levels[openLevel]){\n",
        "\t\tlevels[openLevel] += 1;\n",
        "\t    } else{\n",
        "\t\tlevels[openLevel] = 1;\n",
        "\t    }\n",
        "\n",
        "\t    if (openLevel > level) {\n",
        "\t\ttoc += (new Array(openLevel - level + 1)).join('<ul class=\"toc\">');\n",
        "\t    } else if (openLevel < level) {\n",
        "\t\ttoc += (new Array(level - openLevel + 1)).join(\"</ul>\");\n",
        "\t\tfor (i=level;i>openLevel;i--){levels[i]=0;}\n",
        "\t    }\n",
        "\n",
        "\t    level = parseInt(openLevel);\n",
        "\n",
        "\n",
        "\t    if (this.id==''){this.id = this.innerHTML.replace(/ /g,\"-\")}\n",
        "\t    var anchor = this.id;\n",
        "        \n",
        "\t    toc += '<li><a style=\"text-decoration:none\", href=\"#' + encodeURIComponent(anchor) + '\">' + titleText + '</a></li>';\n",
        "        \n",
        "\t});\n",
        "\n",
        "    \n",
        "    if (level) {\n",
        "\ttoc += (new Array(level + 1)).join(\"</ul>\");\n",
        "    }\n",
        "\n",
        " \n",
        "    $('#toc').append(toc);\n",
        "\n",
        "};\n",
        "\n",
        "// Executes the createToc function\n",
        "setTimeout(function(){createTOC();},100);\n",
        "\n",
        "// Rebuild to TOC every minute\n",
        "setInterval(function(){createTOC();},60000);"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "application/javascript": [
              "// Converts integer to roman numeral\n",
              "// https://github.com/kmahelona/ipython_notebook_goodies\n",
              "// https://kmahelona.github.io/ipython_notebook_goodies/ipython_notebook_toc.js\n",
              "function romanize(num) {\n",
              "    var lookup = {M:1000,CM:900,D:500,CD:400,C:100,XC:90,L:50,XL:40,X:10,IX:9,V:5,IV:4,I:1},\n",
              "\troman = '',\n",
              "\t    i;\n",
              "\tfor ( i in lookup ) {\n",
              "\t    while ( num >= lookup[i] ) {\n",
              "\t\troman += i;\n",
              "\t\tnum -= lookup[i];\n",
              "\t    }\n",
              "\t}\n",
              "\treturn roman;\n",
              " }\n",
              "\n",
              "// Builds a <ul> Table of Contents from all <headers> in DOM\n",
              "function createTOC(){\n",
              "    var toc = \"\";\n",
              "    var level = 0;\n",
              "    var levels = {}\n",
              "    $('#toc').html('');\n",
              "\n",
              "    $(\":header\").each(function(i){\n",
              "\t    if (this.id=='tocheading'){return;}\n",
              "        \n",
              "\t    var titleText = this.innerHTML;\n",
              "\t    var openLevel = this.tagName[1];\n",
              "\n",
              "\t    if (levels[openLevel]){\n",
              "\t\tlevels[openLevel] += 1;\n",
              "\t    } else{\n",
              "\t\tlevels[openLevel] = 1;\n",
              "\t    }\n",
              "\n",
              "\t    if (openLevel > level) {\n",
              "\t\ttoc += (new Array(openLevel - level + 1)).join('<ul class=\"toc\">');\n",
              "\t    } else if (openLevel < level) {\n",
              "\t\ttoc += (new Array(level - openLevel + 1)).join(\"</ul>\");\n",
              "\t\tfor (i=level;i>openLevel;i--){levels[i]=0;}\n",
              "\t    }\n",
              "\n",
              "\t    level = parseInt(openLevel);\n",
              "\n",
              "\n",
              "\t    if (this.id==''){this.id = this.innerHTML.replace(/ /g,\"-\")}\n",
              "\t    var anchor = this.id;\n",
              "        \n",
              "\t    toc += '<li><a style=\"text-decoration:none\", href=\"#' + encodeURIComponent(anchor) + '\">' + titleText + '</a></li>';\n",
              "        \n",
              "\t});\n",
              "\n",
              "    \n",
              "    if (level) {\n",
              "\ttoc += (new Array(level + 1)).join(\"</ul>\");\n",
              "    }\n",
              "\n",
              " \n",
              "    $('#toc').append(toc);\n",
              "\n",
              "};\n",
              "\n",
              "// Executes the createToc function\n",
              "setTimeout(function(){createTOC();},100);\n",
              "\n",
              "// Rebuild to TOC every minute\n",
              "setInterval(function(){createTOC();},60000);"
            ],
            "text/plain": [
              "<IPython.core.display.Javascript object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RqU78-8S6ER_",
        "colab_type": "code",
        "outputId": "47a85ab7-ae6c-4e11-b747-426f9a110d1e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "tuned_models_evaluation_train = dict()\n",
        "tuned_models_evaluation_test = dict()\n",
        "\n",
        "tuned_models_evaluation_train, tuned_models_evaluation_test"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "({}, {})"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZFCtXdAh6Dcy",
        "colab_type": "text"
      },
      "source": [
        ""
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "16u9RhlD6LZW"
      },
      "source": [
        "<h3> 5.1 XGBoost with hyperparameter tuning with initial 13 features </h3>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iqJ7mlWKoHVV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "param_grid = dict(\n",
        "    learning_rate = [0.0001,0.001,0.01,0.1],\n",
        "    max_depth = [3,4,5,6],\n",
        "    n_estimators = [100,150,200,250]\n",
        "\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HpFO-ggkoHVX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.model_selection import GridSearchCV\n",
        "from xgboost.sklearn import XGBRegressor"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IeRYzpW8oHVY",
        "colab_type": "text"
      },
      "source": [
        "### 5.1 Tuning XGBoost with initial 13 features"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NlFNg8JRoHVZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# prepare Train data\n",
        "x_train = reg_train.drop(['user','movie','rating'], axis=1)\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# Prepare Test data\n",
        "x_test = reg_test_df.drop(['user','movie','rating'], axis=1)\n",
        "y_test = reg_test_df['rating']\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NAcgFLWboHVd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# making model ready for tuning\n",
        "xgb_model = XGBRegressor(silent=False, n_jobs=-1,random_state=15,verbosity=1,nthread=-1)\n",
        "xgb_knn_bsl = GridSearchCV(xgb_model,param_grid,n_jobs=-1,cv = 3,verbose=10,return_train_score=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "id": "WJxy3amqoHVf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_results, test_results = run_xgboost(xgb_knn_bsl, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "tuned_models_evaluation_train['first_algo'] = train_results\n",
        "tuned_models_evaluation_test['first_algo'] = test_results"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1PCfMLyeJVzS",
        "colab_type": "text"
      },
      "source": [
        "Done. Time taken : 0:34:43.292529\n",
        "\n",
        "Done \n",
        "\n",
        "Evaluating the model with TRAIN data...\n",
        "Evaluating Test data\n",
        "\n",
        "> TEST DATA Scores\n",
        "------------------------------\n",
        "*   RMSE :  1.074079123678349\n",
        "*   MAPE :  34.74543952747899\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iQA0HwZoLZG9",
        "colab_type": "code",
        "outputId": "7fac3368-8ac8-4ba2-ad17-a3ec4459f4d0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "print(x_train.columns)\n",
        "print(xgb_knn_bsl.best_estimator_.feature_importances_)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Index(['GAvg', 'sur1', 'sur2', 'sur3', 'sur4', 'sur5', 'smr1', 'smr2', 'smr3',\n",
            "       'smr4', 'smr5', 'UAvg', 'MAvg'],\n",
            "      dtype='object')\n",
            "[0.         0.09263241 0.07177851 0.07284065 0.05912598 0.06930886\n",
            " 0.18233553 0.07926813 0.03691885 0.02542087 0.01437838 0.1986779\n",
            " 0.0973139 ]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EgWg9WRPMJkD",
        "colab_type": "code",
        "outputId": "0f0c0651-670f-4761-e035-0f32380bdfb8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "import xgboost as xgb\n",
        "xgb.plot_importance(xgb_knn_bsl.best_estimator_)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEWCAYAAACnlKo3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nERFJmZqIiIikTE1ERERS9v8BNqWBvRR6J2wAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "yx-NFQWyb84m",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ELnx7ypkoHVm",
        "colab_type": "text"
      },
      "source": [
        "### 5.2 Tuning XGBoost with initial 13 features + Surprise Baseline predictor"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "TCybeBBdOLOY",
        "colab": {}
      },
      "source": [
        "from surprise import BaselineOnly "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ramGwIIxOLOm"
      },
      "source": [
        "__Predicted_rating : ( baseline prediction ) __\n",
        "\n",
        "    -  http://surprise.readthedocs.io/en/stable/basic_algorithms.html#surprise.prediction_algorithms.baseline_only.BaselineOnly \n",
        " >$   \\large {\\hat{r}_{ui} = b_{ui} =\\mu + b_u + b_i} $\n",
        "\n",
        "\n",
        "- $\\pmb \\mu $ : Average of all trainings in training data.\n",
        "- $\\pmb b_u$ : User bias\n",
        "- $\\pmb b_i$ : Item bias (movie biases) "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "C6vhbbSAOLOo"
      },
      "source": [
        "__Optimization function ( Least Squares Problem ) __\n",
        "\n",
        "    - http://surprise.readthedocs.io/en/stable/prediction_algorithms.html#baselines-estimates-configuration \n",
        "\n",
        "> $ \\large \\sum_{r_{ui} \\in R_{train}} \\left(r_{ui} - (\\mu + b_u + b_i)\\right)^2 +\n",
        "\\lambda \\left(b_u^2 + b_i^2 \\right).\\text {        [mimimize } {b_u, b_i]}$ "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_3fCU9RAoHVo",
        "colab_type": "code",
        "outputId": "552c14c6-e0fc-4aef-e518-0e6d0ededcad",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# create the traindata from the dataframe...\n",
        "train_data = Dataset.load_from_df(reg_train[['user', 'movie', 'rating']], reader)\n",
        "\n",
        "# build the trainset from traindata.., It is of dataset format from surprise library..\n",
        "trainset = train_data.build_full_trainset() \n",
        "\n",
        "testset = list(zip(reg_test_df.user.values, reg_test_df.movie.values, reg_test_df.rating.values))\n",
        "testset[:3]"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[(808635, 71, 5), (941866, 71, 4), (1737912, 71, 3)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a5-T-79BoHVq",
        "colab_type": "code",
        "outputId": "522da646-1d78-451b-8623-88c64806c0b1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        }
      },
      "source": [
        "# options are to specify.., how to compute those user and item biases\n",
        "bsl_options = {'method': 'sgd',\n",
        "               'learning_rate': .001\n",
        "               }\n",
        "bsl_algo = BaselineOnly(bsl_options=bsl_options)\n",
        "# run this algorithm.., It will return the train and test results..\n",
        "bsl_train_results, bsl_test_results = run_surprise(bsl_algo, trainset, testset, verbose=True)\n",
        "\n",
        "\n",
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "tuned_models_evaluation_train['bsl_algo'] = bsl_train_results \n",
        "tuned_models_evaluation_test['bsl_algo'] = bsl_test_results"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            "Estimating biases using sgd...\n",
            "Done. time taken : 0:00:01.676659 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:00:02.528766\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.9552562205339807\n",
            "\n",
            "MAPE : 30.353233415224267\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.070667\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.0731636807255809\n",
            "\n",
            "MAPE : 34.989136338308896\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:00:04.278548\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "D4Yj8qwaPNTG"
      },
      "source": [
        "__Updating Train Data__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "owzIgPVEoHVz",
        "colab_type": "code",
        "outputId": "a114673e-e359-480c-ccff-40e6ab91d278",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 111
        }
      },
      "source": [
        "# add our baseline_predicted value as our feature..\n",
        "reg_train['bslpr'] = tuned_models_evaluation_train['bsl_algo']['predictions']\n",
        "reg_train.head(2) \n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>bslpr</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>174683</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.882353</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>5</td>\n",
              "      <td>4.996793</td>\n",
              "      <td>4.977554</td>\n",
              "      <td>3.672848</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>233949</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.692308</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>3</td>\n",
              "      <td>3.062453</td>\n",
              "      <td>3.051725</td>\n",
              "      <td>3.688917</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg  sur1  ...  rating  knn_bsl_u  knn_bsl_m     bslpr\n",
              "0  174683     10  3.587581   5.0  ...       5   4.996793   4.977554  3.672848\n",
              "1  233949     10  3.587581   4.0  ...       3   3.062453   3.051725  3.688917\n",
              "\n",
              "[2 rows x 19 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 54
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "b2LKcf0FPSi5"
      },
      "source": [
        "__Updating Train Data__"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E8m67TjMOr8J",
        "colab_type": "code",
        "outputId": "0544e615-59b2-4d4f-a008-cb3884f5808c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 131
        }
      },
      "source": [
        "# add that baseline predicted ratings with Surprise to the test data as well\n",
        "reg_test_df['bslpr']  = tuned_models_evaluation_test['bsl_algo']['predictions']\n",
        "reg_test_df.head(2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>bslpr</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>808635</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>941866</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg      sur1  ...  rating  knn_bsl_u  knn_bsl_m     bslpr\n",
              "0  808635     71  3.581679  3.581679  ...       5   3.579949   3.579949  3.579949\n",
              "1  941866     71  3.581679  3.581679  ...       4   3.579949   3.579949  3.579949\n",
              "\n",
              "[2 rows x 19 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C5FcFpDIO7FE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#here"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pJ3IvsMFoHV1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# prepare train data\n",
        "x_train = reg_train.drop(['user', 'movie','rating'], axis=1)\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# Prepare Test data\n",
        "x_test = reg_test_df.drop(['user','movie','rating'], axis=1)\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "# making model ready for tuning\n",
        "xgb_model = XGBRegressor(silent=False, n_jobs=-1,random_state=15,verbosity=1,nthread=-1)\n",
        "xgb_bsl = GridSearchCV(xgb_model,param_grid,n_jobs=-1,cv = 3,verbose=10,return_train_score=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "id": "BIuMvqu1oHV4",
        "colab_type": "code",
        "outputId": "404de5e1-117e-47b9-c424-96cd8c8ab705",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 666
        }
      },
      "source": [
        "train_results, test_results = run_xgboost(xgb_bsl, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "tuned_models_evaluation_train['xgb_bsl'] = train_results\n",
        "tuned_models_evaluation_test['xgb_bsl'] = test_results"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model..\n",
            "Fitting 3 folds for each of 64 candidates, totalling 192 fits\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
            "[Parallel(n_jobs=-1)]: Done   5 tasks      | elapsed:   49.6s\n",
            "[Parallel(n_jobs=-1)]: Done  10 tasks      | elapsed:  1.6min\n",
            "[Parallel(n_jobs=-1)]: Done  17 tasks      | elapsed:  2.5min\n",
            "/usr/local/lib/python3.6/dist-packages/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
            "  \"timeout or by a memory leak.\", UserWarning\n",
            "[Parallel(n_jobs=-1)]: Done  24 tasks      | elapsed:  4.0min\n",
            "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  5.9min\n",
            "[Parallel(n_jobs=-1)]: Done  42 tasks      | elapsed:  8.0min\n",
            "[Parallel(n_jobs=-1)]: Done  53 tasks      | elapsed: 10.5min\n",
            "[Parallel(n_jobs=-1)]: Done  64 tasks      | elapsed: 12.2min\n",
            "[Parallel(n_jobs=-1)]: Done  77 tasks      | elapsed: 14.6min\n",
            "[Parallel(n_jobs=-1)]: Done  90 tasks      | elapsed: 17.8min\n",
            "[Parallel(n_jobs=-1)]: Done 105 tasks      | elapsed: 21.1min\n",
            "[Parallel(n_jobs=-1)]: Done 120 tasks      | elapsed: 23.9min\n",
            "[Parallel(n_jobs=-1)]: Done 137 tasks      | elapsed: 28.3min\n",
            "[Parallel(n_jobs=-1)]: Done 154 tasks      | elapsed: 32.0min\n",
            "[Parallel(n_jobs=-1)]: Done 173 tasks      | elapsed: 35.5min\n",
            "[Parallel(n_jobs=-1)]: Done 192 out of 192 | elapsed: 41.1min finished\n",
            "/usr/local/lib/python3.6/dist-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
            "  if getattr(data, 'base', None) is not None and \\\n",
            "/usr/local/lib/python3.6/dist-packages/xgboost/core.py:588: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
            "  data.base is not None and isinstance(data, np.ndarray) \\\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "[13:10:57] WARNING: /workspace/src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
            "Done. Time taken : 0:41:17.495814\n",
            "\n",
            "Done \n",
            "\n",
            "Evaluating the model with TRAIN data...\n",
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.0743350336895106\n",
            "MAPE :  34.68804607716922\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0j1wguEOQLA5",
        "colab_type": "code",
        "outputId": "54af9f69-cf5d-4f99-d5ee-3072176e6269",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "xgb.plot_importance(xgb_bsl.best_estimator_)\n",
        "plt.show()\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEWCAYAAACnlKo3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Y_Uvkqq9dTuD",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "5bE97Eqpb_vA"
      },
      "source": [
        "### 5.3 Tuning XGBoost with initial 13 features + Surprise Baseline predictor + KNNBaseline predictor"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "yOaogYbRcHtW",
        "colab": {}
      },
      "source": [
        "from surprise import KNNBaseline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "un30QyKncHtb"
      },
      "source": [
        "- KNN BASELINE\n",
        "    - http://surprise.readthedocs.io/en/stable/knn_inspired.html#surprise.prediction_algorithms.knns.KNNBaseline "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "VXUki-DZcHtd"
      },
      "source": [
        "- PEARSON_BASELINE SIMILARITY\n",
        "    - http://surprise.readthedocs.io/en/stable/similarities.html#surprise.similarities.pearson_baseline "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Cb2tlnR_cHte"
      },
      "source": [
        "- SHRINKAGE\n",
        "    - _2.2 Neighborhood Models_ in http://courses.ischool.berkeley.edu/i290-dm/s11/SECURE/a1-koren.pdf "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "BuIMn6H4cHtf"
      },
      "source": [
        "- __predicted Rating__ : ( ___ based on User-User similarity ___ )\n",
        "\n",
        "\\begin{align} \\hat{r}_{ui} = b_{ui} + \\frac{ \\sum\\limits_{v \\in N^k_i(u)}\n",
        "\\text{sim}(u, v) \\cdot (r_{vi} - b_{vi})} {\\sum\\limits_{v \\in\n",
        "N^k_i(u)} \\text{sim}(u, v)} \\end{align}\n",
        "\n",
        "- $\\pmb{b_{ui}}$ -  _Baseline prediction_ of (user,movie) rating\n",
        "\n",
        "- $ \\pmb {N_i^k (u)}$ - Set of __K similar__ users (neighbours) of __user (u)__ who rated __movie(i)__  \n",
        "\n",
        "- _sim (u, v)_ - __Similarity__ between users __u and v__  \n",
        "    - Generally, it will be cosine similarity or Pearson correlation coefficient. \n",
        "    - But we use __shrunk Pearson-baseline correlation coefficient__, which is based on the pearsonBaseline similarity ( we take base line predictions instead of mean rating of user/item)\n",
        "       "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "84Z2TlR5cHtg"
      },
      "source": [
        " "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "U8EjcEr_cHti"
      },
      "source": [
        "- __ Predicted rating __ ( based on Item Item similarity ):\n",
        " \\begin{align} \\hat{r}_{ui} = b_{ui} + \\frac{ \\sum\\limits_{j \\in N^k_u(i)}\\text{sim}(i, j) \\cdot (r_{uj} - b_{uj})} {\\sum\\limits_{j \\in N^k_u(j)} \\text{sim}(i, j)} \\end{align}\n",
        "\n",
        "    -  ___Notations follows same as above (user user based predicted rating ) ___"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "csAgHpeKb_vI"
      },
      "source": [
        "- Surprise KNNBaseline with user user similarities"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "luqHC5rHd8l_"
      },
      "source": [
        "  <h4> 4.4.4.1 Surprise KNNBaseline with user user similarities</h4>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "65d7d76b-6f70-428f-bfc1-0ef485221242",
        "id": "0iK9J9k9b_vJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 527
        }
      },
      "source": [
        "# we specify , how to compute similarities and what to consider with sim_options to our algorithm\n",
        "sim_options = {'user_based' : True,\n",
        "               'name': 'pearson_baseline',\n",
        "               'shrinkage': 100,\n",
        "               'min_support': 2\n",
        "              } \n",
        "# we keep other parameters like regularization parameter and learning_rate as default values.\n",
        "bsl_options = {'method': 'sgd'} \n",
        "\n",
        "knn_bsl_u = KNNBaseline(k=40, sim_options = sim_options, bsl_options = bsl_options)\n",
        "knn_bsl_u_train_results, knn_bsl_u_test_results = run_surprise(knn_bsl_u, trainset, testset, verbose=True)\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            "Estimating biases using sgd...\n",
            "Computing the pearson_baseline similarity matrix...\n",
            "Done computing similarity matrix.\n",
            "Done. time taken : 0:03:13.078602 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:09:40.598949\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.3278050914466148\n",
            "\n",
            "MAPE : 9.027978982373568\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.088058\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.072409004806261\n",
            "\n",
            "MAPE : 34.936622813459635\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:12:53.767892\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "bVd-S6qMb_vR",
        "colab": {}
      },
      "source": [
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "tuned_models_evaluation_train['knn_bsl_u'] = knn_bsl_u_train_results \n",
        "tuned_models_evaluation_test['knn_bsl_u'] = knn_bsl_u_test_results"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NThs5mlBkxBP",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pickle\n",
        "with open(\"/content/drive/My Drive/Netflix_recommender/tuned_result.pkl\",\"wb\") as f:\n",
        "    pickle.dump(tuned_models_evaluation_train,f)\n",
        "    pickle.dump(tuned_models_evaluation_test,f)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dFzjHYVKnuTx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pickle\n",
        "with open(\"/content/drive/My Drive/Netflix_recommender/tuned_result.pkl\",\"rb\") as f:\n",
        "    tuned_models_evaluation_train = pickle.load(f)\n",
        "    tuned_models_evaluation_test  = pickle.load(f)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "DxGX2xkwb_vU"
      },
      "source": [
        "- Surprise KNNBaseline with movie movie similarities"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "colab_type": "code",
        "outputId": "1c73debb-b00b-4c4c-80ca-0ef277ee9c0f",
        "id": "HV3JUUfFb_vW",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 527
        }
      },
      "source": [
        "# we specify , how to compute similarities and what to consider with sim_options to our algorithm\n",
        "\n",
        "# 'user_based' : Fals => this considers the similarities of movies instead of users\n",
        "\n",
        "sim_options = {'user_based' : False,\n",
        "               'name': 'pearson_baseline',\n",
        "               'shrinkage': 100,\n",
        "               'min_support': 2\n",
        "              } \n",
        "# we keep other parameters like regularization parameter and learning_rate as default values.\n",
        "bsl_options = {'method': 'sgd'}\n",
        "\n",
        "\n",
        "knn_bsl_m = KNNBaseline(k=40, sim_options = sim_options, bsl_options = bsl_options)\n",
        "\n",
        "knn_bsl_m_train_results, knn_bsl_m_test_results = run_surprise(knn_bsl_m, trainset, testset, verbose=True)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            "Estimating biases using sgd...\n",
            "Computing the pearson_baseline similarity matrix...\n",
            "Done computing similarity matrix.\n",
            "Done. time taken : 0:00:02.171294 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:00:19.100806\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.32316657950987937\n",
            "\n",
            "MAPE : 8.404159172327828\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.076235\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.0723215145319687\n",
            "\n",
            "MAPE : 34.92807707883072\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:00:21.350482\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "M6a-Lh2Nb_vb",
        "colab": {}
      },
      "source": [
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "tuned_models_evaluation_train['knn_bsl_m'] = knn_bsl_m_train_results \n",
        "tuned_models_evaluation_test['knn_bsl_m'] = knn_bsl_m_test_results"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "rJcv7SEZpQWE",
        "colab": {}
      },
      "source": [
        "import pickle\n",
        "with open(\"/content/drive/My Drive/Netflix_recommender/tuned_result.pkl\",\"wb\") as f:\n",
        "    pickle.dump(tuned_models_evaluation_train,f)\n",
        "    pickle.dump(tuned_models_evaluation_test,f)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "a3l5j5__pQWc",
        "colab": {}
      },
      "source": [
        "import pickle\n",
        "with open(\"/content/drive/My Drive/Netflix_recommender/tuned_result.pkl\",\"rb\") as f:\n",
        "    tuned_models_evaluation_train = pickle.load(f)\n",
        "    tuned_models_evaluation_test  = pickle.load(f)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xg0Iob7e9ZhR",
        "colab_type": "code",
        "outputId": "99208339-ab33-4294-98fa-1814fbeffbb9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "# add the predicted values from both knns to this dataframe\n",
        "reg_train['knn_bsl_u'] = tuned_models_evaluation_train['knn_bsl_u']['predictions']\n",
        "reg_train['knn_bsl_m'] = tuned_models_evaluation_train['knn_bsl_m']['predictions']\n",
        "\n",
        "reg_train.head(2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>174683</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.882353</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>5</td>\n",
              "      <td>4.996793</td>\n",
              "      <td>4.977554</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>233949</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.692308</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>3</td>\n",
              "      <td>3.062453</td>\n",
              "      <td>3.051725</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>555770</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.795455</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>4</td>\n",
              "      <td>5.000000</td>\n",
              "      <td>5.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>767518</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.884615</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>5</td>\n",
              "      <td>4.834040</td>\n",
              "      <td>4.875339</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>894393</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>4</td>\n",
              "      <td>4.038446</td>\n",
              "      <td>3.904309</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg  sur1  ...      MAvg  rating  knn_bsl_u  knn_bsl_m\n",
              "0  174683     10  3.587581   5.0  ...  3.611111       5   4.996793   4.977554\n",
              "1  233949     10  3.587581   4.0  ...  3.611111       3   3.062453   3.051725\n",
              "2  555770     10  3.587581   4.0  ...  3.611111       4   5.000000   5.000000\n",
              "3  767518     10  3.587581   2.0  ...  3.611111       5   4.834040   4.875339\n",
              "4  894393     10  3.587581   3.0  ...  3.611111       4   4.038446   3.904309\n",
              "\n",
              "[5 rows x 18 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 53
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rb70N_yD_Y8a",
        "colab_type": "code",
        "outputId": "11e049f0-d05c-4621-db9e-ee4433870f68",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 131
        }
      },
      "source": [
        "reg_test_df.head(2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>bslpr</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>808635</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>941866</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg      sur1  ...  rating  knn_bsl_u  knn_bsl_m     bslpr\n",
              "0  808635     71  3.581679  3.581679  ...       5   3.579949   3.579949  3.579949\n",
              "1  941866     71  3.581679  3.581679  ...       4   3.579949   3.579949  3.579949\n",
              "\n",
              "[2 rows x 19 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "c3f81b89-d4eb-41c0-f661-caab80fee53b",
        "id": "YKNqsLCyb_vm",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 131
        }
      },
      "source": [
        "reg_test_df['knn_bsl_u'] = tuned_models_evaluation_test['knn_bsl_u']['predictions']\n",
        "reg_test_df['knn_bsl_m'] = tuned_models_evaluation_test['knn_bsl_m']['predictions']\n",
        "\n",
        "reg_test_df.head(2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>808635</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>941866</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg      sur1  ...      MAvg  rating  knn_bsl_u  knn_bsl_m\n",
              "0  808635     71  3.581679  3.581679  ...  3.581679       5   3.579949   3.579949\n",
              "1  941866     71  3.581679  3.581679  ...  3.581679       4   3.579949   3.579949\n",
              "\n",
              "[2 rows x 18 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 50
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "zYeQcgbYb_vr",
        "colab": {}
      },
      "source": [
        "# prepare the train data....\n",
        "x_train = reg_train.drop(['user', 'movie', 'rating'], axis=1)\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# prepare the train data....\n",
        "x_test = reg_test_df.drop(['user','movie','rating'], axis=1)\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "# making model ready for tuning\n",
        "xgb_model = XGBRegressor(silent=False, n_jobs=-1,random_state=15,verbosity=1,nthread=-1)\n",
        "xgb_knn_bsl_um = GridSearchCV(xgb_model,param_grid,n_jobs=-1,cv = 3,verbose=10,return_train_score=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "1CQhUcxDb_vy",
        "colab": {}
      },
      "source": [
        "train_results, test_results = run_xgboost(xgb_knn_bsl_um, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "tuned_models_evaluation_train['xgb_knn_bsl_um'] = train_results\n",
        "tuned_models_evaluation_test['xgb_knn_bsl_um'] = test_results\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "__RFAG1xcsVW",
        "colab_type": "text"
      },
      "source": [
        "Done. Time taken : 0:53:27.148315\n",
        "\n",
        "Done "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QZFRbyvAAimg",
        "colab_type": "code",
        "outputId": "3e472181-a0a1-491c-e7a9-fb969c365c3a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        }
      },
      "source": [
        "y_train_pred = xgb_knn_bsl_um.predict(x_train)\n",
        "# get the rmse and mape of train data...\n",
        "rmse_train, mape_train = get_error_metrics(y_train.values, y_train_pred)\n",
        "\n",
        "# store the results in train_results dictionary..\n",
        "train_results = {'rmse': rmse_train,\n",
        "                'mape' : mape_train,\n",
        "                'predictions' : y_train_pred}\n",
        "\n",
        "print('\\nTrain DATA')\n",
        "print('-'*30)\n",
        "print('RMSE : ', rmse_train)\n",
        "print('MAPE : ', mape_train)\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Train DATA\n",
            "------------------------------\n",
            "RMSE :  0.8643350724897788\n",
            "MAPE :  25.960364353971094\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lVRJJ1qP-yfR",
        "colab_type": "code",
        "outputId": "7f6fac9a-dbcc-4f13-efef-74c883be56b1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "source": [
        "print('Evaluating Test data')\n",
        "y_test_pred = xgb_knn_bsl_um.predict(x_test) \n",
        "rmse_test, mape_test = get_error_metrics(y_true=y_test.values, y_pred=y_test_pred)\n",
        "# store them in our test results dictionary.\n",
        "test_results = {'rmse': rmse_test,\n",
        "                'mape' : mape_test,\n",
        "                'predictions':y_test_pred}\n",
        "if verbose:\n",
        "    print('\\nTEST DATA')\n",
        "    print('-'*30)\n",
        "    print('RMSE : ', rmse_test)\n",
        "    print('MAPE : ', mape_test)\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.0743360765716838\n",
            "MAPE :  34.68765743120491\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A3pSxpN5A-mE",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tuned_models_evaluation_train['xgb_knn_bsl_um'] = train_results\n",
        "tuned_models_evaluation_test['xgb_knn_bsl_um'] = test_results"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8CSGo8rBBBrp",
        "colab_type": "code",
        "outputId": "27f63d75-347a-4d09-9e56-07ce29561c8b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        }
      },
      "source": [
        "xgb.plot_importance(xgb_knn_bsl_um.best_estimator_)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f14e9f03cc0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaoAAAEWCAYAAAA3h9P4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3XucjfX6//HXzIgwUibsHSLFZdSW\nY2WLbFE0I746aX918CO7ws62O1B7S5SIouy+cohCtCMJHRSStkLsiqSr00yOSc6NM/P7477XtIw1\nM2vGWrPudc/1fDw8rHWf1vUpXHPf92e974Ts7GyMMcYYr0qMdQHGGGNMfqxRGWOM8TRrVMYYYzzN\nGpUxxhhPs0ZljDHG06xRGWOM8TRrVMbEMRF5QUT+Ges6jImmBPselSmJRCQTqAocD1pcV1W3nsYx\nWwPTVbX6aRUXp0TkJWCzqv4j1rUYfykV6wKMiaGOqroo1kUEiEgpVT0W6zqKQkSSYl2D8S87ozIl\nkntG1TNUoxKRK4BngPrAj8B9qrrUXdcdeBCoDuwARqjqeBEpD/wClAEOuIeqCwwj6Cwj91mXW8c4\n4H8BAcoDVYCxQCvgV2C0qj6XxzheChw/cGzgOeB+nLPFe4AjwBjgXGCUqg5z9x0MXOJudx3wLdBd\nVb9w16e6tTUEtgADVXVe0OceBGoCVwF/A54Hst3P+0BVO4rIAOAud0ybgEdU9Q33GHcCPYEVQA9g\nD3Cvqr7jrq8EPA1cC5QFPlTVzu66dOBxoBbwFXC3qq4N9d/IxD+7R2VMEBGpBryF849gJZx/8F8X\nkcruJj8D6cBZQHdgtIg0VtUsoAOwVVWT3V/hXka8FUgDzgZOAPOBL4BqwNVAPxG5Nsxj/Q440913\nEDAR6AY0AVoC/xSRC4K27wTMcsc6A5grImeIyBluHe/hNJm+wCsiIkH7/hl4AqgATAVeAZ5yx97R\n3eZ793MrAo8B00Xk90HHuBxQnCb6FPCiiCS466YB5YCL3RpGA4hII2Ay8BcgBRgPzBORMmH+NzJx\nxi79mZJsrogELrUtdX9a7wa8rapvu8vfF5HVOGccL6vqW0H7fygi7+H8Q/zf06jjOVXdBCAilwOV\nVXWIu+4HEZkIdAUWhnGso8ATqnpcRF4FJgDPqup+YL2IfAVcCmS4269R1dnuZz8D/B24wl2XDAxX\n1RPAEhFZgNNUB7vr31TV5e7rQyf3MIeqzgp6+28RGQhcBrzpLvtRVSe6n/8y8H9AVbdZdQBSVHW3\nu+2H7u+9gPGqutJ9/7KIPOzWHdjG+Ig1KlOSdQ5x6a8mcJOIdAxadgbwAYCIdAAexbmsl4jzE/+6\n06xjU67PP09E9gQtSwI+CvNYO1U1MEHkoPv79qD1B3Ea0CmfraonRGQzcF5gndukAn7EOVMLVXdI\nInI70B/nEh3uZ58btMlPQZ9/wG12yThneLuCmlSwmsAdItI3aFnpoLqNz1ijMuZkm4BpqnpX7hXu\npaXXgdtxziaOishcIHCpKtQN3yycZhbwuxDbBO+3CchQ1TpFKb4IagReiEgizr23wCXLGiKSGNSs\nzge+Cdo393hPei8iNXEuPV4NfOKe5X3Ob/+98rMJqCQiZ6vqnhDrnlDVJ8I4jvEBa1TGnGw68Kl7\nT2gRztnUFcB3wF6cyRI7gGPu2dU1wJfuvtuBFBGpqKp73WWfA38XkcdxfurvV8DnrwL2i8hDOJMi\njgCpQFlV/TRCYwzWRES6APOAvwKHcSY3JOBMCnlQRJ4GWgAdgWb5HGs7UDvofXmc5rUDciaiXBJO\nUaq6TUTeAf5PRHrjTCpprqrLcJrfGyKyCOe/VzmgNbDMvcRpfMYmUxgTxL1X1Al4GOcf2E3AA0Ci\n+4/gX4HXgN04kwnmBe37NTAT577SHhE5D2dCwBdAJs7EhH8X8PnHcSZrNMS5j/QLMAlnMkI0vAnc\ngjOe24AuqnpUVY/gNKYObg3/B9zujjEvLwL13bHPVdWvcGbtfYLTxP4ALM9n/9xuw7nn9jXOJJZ+\nAKq6Gmcm4b/cur8D7izEcU2csenpxpRQ7vT0i1S1W6xrMSY/dkZljDHG06xRGWOM8TS79GeMMcbT\n7IzKGGOMp9n09Aj473//m122bNlYlxExhw8fpkwZ/6TR2Hi8y09jARtPYR04cOCXJk2aVC5oO2tU\nEZCQkEBqamqsy4iYDRs22Hg8zE/j8dNYwMZTWGvWrPkxnO3s0p8xxhhPs0ZljDHG06xRGWOM8TRr\nVMYYYzzNGpUxxhhPs0ZljDHG06xRGWOM8TRrVMYYYzzNvvBrjDGGNm3aUL58eRITE0lKSmLOnDn8\n8MMPPProoxw+fJikpCQGDx5MgwYNAFi5ciXDhg3j2LFjnHPOOUyfPj1qtcV1oxKRbOCVwPN0RKQU\nsA1YqarpQdvNBX6nqlfEplJjjPG+l19+mUqVKp30vnfv3lx11VV8+OGHjBw5kmnTprFv3z4ee+wx\nJk2axHnnncfOnTujWle8X/rLAi4RkUDQXjtgS/AGInI20ASoKCK1McYYE5aEhASysrIA2L9/P1Wq\nVAFg/vz5tGvXjvPOOw+AlJSUqNYR12dUrreBNGA2cCvOo8BbBq3vAszHeRR2V2CYiFQE1gIXqOoJ\nESmP87jr2jiPAH8ROAG8D3RQ1UvyK6B06dIRHVCs+SmrDGw8XuansUB8jufQ0eOceUYSAD169CAh\nIYFbbrmFW265hR49ejBs2DBGjBjBiRMnePXVVwHIzMzk2LFj3HbbbWRlZXH77bfTuXPnqNXoh0b1\nKjBIRBYADYDJnNyobgWG4DSq14FhqrpXRD4HrgI+ANKBhap6VESmAHep6iciMjycAhITE6k14K3I\njcgYY4pJ5vA0NmzYwJAhQ0hJSWHPnj0MHjyYpKQkPvroI2677Tb++Mc/8p///Id+/foxZMgQfvnl\nF7777juGDBnCkSNHeOihh0hOTqZatWpRqTHuG5WqrhWRWjgN6e3gdSJSFagD/EdVs0XkqIhcoqpf\nAv8GbsFpVF2B/3MvE1ZQ1U/cQ8zAaWL5OnHiBJnD0yI2JmOMKS6Hjh4/5UwwPT2dvXv3smzZMsaM\nGUNCQgL16tVj3LhxpKamkpqaygUXXECjRo0A+OMf/8jRo0cLfUa5Zs2asLaL93tUAfOAUTiX/YLd\nDJwDZIhIJlALp6EF9mkvIpVw7mEtKeqHHzlypKi7etKGDRtiXUJE2Xi8y09jgfgcz5lnJHHgwAF+\n/fVXAA4cOMDy5cupU6cOlSpVYtWqVQCsWLGCWrVqAXD11VezZs0ajh07xsGDB1m7di0XXnhh1GqM\n+zMq12Rgj6quE5HWQctvBdoHzpBE5AJgEfCIqv4qIp8CzwILVPU4sEdE9ovI5aq6EudMyxhjfG3n\nzp307t0bgOPHj5Oenk6rVq249957GTFiBMeOHaNMmTIMGTIEgAsvvJCWLVty/fXXk5iYyI033kjd\nunWjVp8vGpWqbgaeC17mXg6sCawI2i5DRPYGNaJ/A7OA1kG79gAmisgJ4ENgb3SrN8aY2KpRowbz\n5s07ZXn9+vWZM2dOyH169uxJz549o10aEOeNSlWTQyxbCix1355yZ09VGwe9ng0k5Npkvao2ABCR\nAcDqCJVrjDGmCOK6UUVJmogMxPlv8yNwZ2zLMcaYks0aVS6q+m+cS4LGGGM8wC+z/owxxviUnVEZ\nY4xHhQqKHTNmDIsXLyYxMZGUlBSefPJJqlatmrPP2rVr6dq1K8888wzt27ePYfWRE/eNyp3dtyA4\n5khEBgO/quqooKDaF1V1QGyqNMaYoskdFNuzZ0/69esHwNSpU3n++edzpo0fP36cUaNG0aJFi5jU\nGi0l4dJfO+Ab4CYRyT3Dzxhj4kpy8m+TnQ8ePEhCwm//rE2bNo1rr7026iGxxS3uz6jCcCvOl3rv\nAZoDH4tIe6CHqt4E4H5J+H5VTReRHsBDwB7gC+CwqvbJ7wMslNbbbDze5aexQOTGk19QLMDo0aOZ\nO3cuFSpUYOrUqQBs376dRYsWMXXqVNatWxeROrzC141KRM4E2gJ/Ac7GaVof46RTTBCR8qqahZP5\n96qInAf8E2gM7MeJVfqioM+xUFpjTCQFskNnzpxJ1apV2blzJ927d6d27do0a9aMv/3tb/ztb39j\n/PjxTJ8+nb/+9a888cQT3H///SQm+u9CmR8aVXY+y9OBD1T1oIi8DvxTRPqp6jEReRfoKCKzcR4T\n8iBwNfChqu4CEJFZQPRyQYwxJg+B3MBdu3YB0LBhQxYtWnTSpb/U1FSGDh1Ku3bt+Oyzz+jTx7n4\ns3//fpYsWcK2bdu44oqiPy/20KFDnsgv9EOj2okTPBusEpCBcwZ1pRtIC5ACtMF5ztSrQB9gF7Ba\nVfeLSJEKsPR0Y0wkHTp6nJo1a3LixAmSk5M5cOAA33zzDffeey9ly5bNCYddvXp1Tpr5Rx99lLP/\ngAEDaN269WnP+tuwYUNUL8+WmPR0Vf0V2CYibQDcNPT2wOc4z6U6X1VrqWotoDe/pad/iHOJ7y6c\npgXwKXCViJzjzha8IZwaLD3d22w83uWnsUDkxnPmGUns3LmTP//5z1x//fXcdNNNXHXVVbRq1Yqn\nn36a9PR0OnbsyPLly3nkkUci8ple5oczKoDbgedF5Bn3/WM4T+pdoqqHg7Z7E3hKRMqo6mH3YYt3\nAncAqOoWERkGrMI50/oaC6U1xsRAXkGxY8eOLXDf4cPDeuZr3PBFo1LVr4A/hVj1cq7tdgGVg973\nwbn8F2yGqk5wz6jeAOZGuFxjjDGFEPeX/qJgsPuY+i9x7nNZozLGmBjyxRlVJKnq/bGuwRhjzG/s\njMoYY4ynWaMyxhjjaXbpz/jWwIEDWbp0KcnJybz//vsAfP311zz66KMcOHCAatWqMWrUKJKTk1m7\ndi3//Oc/AcjOzqZv3760a9culuUbY1x2RpWLiEwWkZ9F5MtY12JOT5cuXZg0adJJyx555BH+/ve/\nM3/+fNq2bZuzvk6dOrz++uu8+eabTJo0iUGDBnHs2LFYlG2MycUalcudjg7wEs4Xhk2ca9asGRUr\nVjxpWWZmJs2aNQOgRYsWvPfeewCULVuWUqWcPwKHDx8+KZHaGBNbvrj0JyLlgdeA6kASMBQYAcwE\nOgDHgF7Ak8BFwEhVfcFNTR8K7AbqAXVVdZn7jKuwWXq6twQnT+dWp04dFi9eTNu2bXn33XfZtm1b\nzrovvviChx9+mK1bt/LUU0/lNC5jTGz55W9ie2CrqqYBiEhFnEa1UVUbishonDOlFsCZON+ResHd\ntzFwiapmFPXDLT3dWzKHp+VE2Wzfvp0TJ07kvO/RowcTJkzg6aef5rLLLiMpKSlnXenSpRk1ahSb\nNm3i2WefpUqVKp78IcQrQaGR4KexgI0nWvzSqNYBT4vICJyn/X7kBszOC1qfrKr7gf0iclhEznbX\nrTqdJgUWSus1h44ezzkrrFChAomJiTnvU1NTcyZJZGRksH79+lPOIFNTU5k+ffpJ+3lJtINCi5Of\nxgI2nsIqMaG0AKr6Dc6Z0TrgcREZ5K4K5PydCHodeB9o0lmn+/kWSusteV32A9i5cyfg/HAxbtw4\nunbtCsCmTZtyJk9s2bKFH374gWrVqkW/WGNMgXxxRuU+8HCXqk4XkT1Az1jXZGKvf//+rFq1il27\ndtGqVSv69u3LgQMHmDFjBgDt2rXjhhucgPw1a9YwceJESpUqRWJiIoMHD6ZSpUqxLN8Y4/JFowL+\nAIwUkRPAUZzHzs8uyoFEZCbQGjhXRDYDj6rqi5Eq1BSfZ55xwvRzX7644447Ttm2c+fOdO7cudhq\nM8aEzxeNSlUXAgtzLa4VtP4lnMkUgfeBdUvdX8HHuhVjjDGe4Yt7VMYYY/zLGpUxxhhPs0ZljDHG\n06xRGWOM8TRrVMYXBg4cSPPmzUlPT89ZtmHDBm6++Wb69etHly5dWLt27Un7rF27lvr16/Puu+8W\nd7nGmELwTaMSkVrhJp6LyGARsSf5+kiopPSRI0fSu3dvxowZw3333cfIkSNz1h0/fpxRo0bRokWL\n4i7VGFNIvmlU0RSUrB6SF/PgTkc8RcAcOnocCJ2UnpCQQFaWEzyyf/9+qlSpkrNu2rRpXHvttaSk\npBRfscaYIvHF96iClBKRV3DilNYDtwODgOtxEtTfU9WTzqREZCnwBXAVzn+P/6eqq0RkMHAhUBvY\nCOT5/SoLpY2d/DIWH374YXr06MGRI0dISkri1VdfBZyg2kWLFjF16lTWrVtXXKUaY4rIb41KgB6q\nulxEJgN9gf8B6qlqdlAQbW7l3JT1VsBk4BJ3eX3gSlU9GPXKTZEFJ6UfPnw45/3EiRO57bbbaNy4\nMatXr6Zfv34MGTKEp556ihtvvBFVZc+ePWzZsiWu8g29kmgdCX4aC9h4osVvjWqTqi53X08H+gOH\ngBdFZAGwII/9ZgK4z6I6K6ihzQunSVl6euzkTkovU6ZMzvsPP/yQp59+mq+//pqePXsybtw4UlNT\n+fHHH3nuuecA2L17N1988QU1a9akbdu2MRtHYfgpodtPYwEbT2GFm57ut0aVnev9UeAy4GrgRqAP\n0CaM/QLvw0pW92N6erz8ZcsvKb1KlSqsWrWKs846ixUrVlCrVi0AlixZkrPNgAEDaN26ddw0KWNK\nIr81qvNFpLmqfgL8GfgcqKiqb4vIcuCHPPa7BfhARK4E9qrqXvd5ViZOBJLSd+/enZOUPnToUIYN\nG0ZWVhZnnXUWQ4YMiXWZxpgi8FujUqC3e3/qK+BRYIGInAkk4FwKDOWQiHwGnAH8v2Kp1ERUICk9\ntzlz5uR7hjh8+PBolmWMiQDfNCpVzQTqhVh1WYhtB+daNF1V+xWwjTHGmBiw71EZY4zxNN+cURWV\nqraOdQ3GGGPyZmdUxhhjPK3En1GZ+DNw4ECWLl1KSkoKCxY4X43r168fGRkZgBOXVKFCBd58882c\nfbZu3UpaWhp9+vShR48eManbGFM01qiCiEg7YDhQGjgCPKCqS/LfyxS3Ll260K1bNx566KGcZWPG\njMl5PXz4cJKTk0/aZ/jw4bRs2bLYajTGRI5d+nO5wbO/AB1V9Q/AHcC02FZlQgkVQBuQnZ3NO++8\nc9LjPlasWEG1atWoU6dOcZVojIkg351RiUh54DWgOpAEDAVGAE1V9RcRaQqMUtXWuYNnVTU4eHY9\nUFZEyqjq4fw+09LTi8eho8fzTaIAWL16NSkpKTkpFFlZWbzxxhvMnDmTyZMnF0OVxphI812jAtoD\nW1U1DUBEKuI0qrzkFTx7A/DfgpoUWHp6cckcnpZnAG3A9OnTadasWc7yKVOm0L59ezZu3MiOHTso\nW7asJ0I2T4dXgkIjwU9jARtPtPixUa0DnhaREcACVf2ogDikU4JnReRinOZ2TfTKNEWRVwAtwLFj\nx/j000+ZM2cOv/vd7wDYvHkzH3/8MbNnz2bfvn0kJiZSo0YNunXrFpP6IyGeshgL4qexgI2nsEpq\nKC2q+o2INAauAx4XkcU4z6IK3I87M9cuJwXPikh14A3gdlX9PpzPtPT04lHQpb+PP/6Y2rVr5zQp\ngBkzZuT8ZRs7dizlypWL6yZlTEnku8kUInIecEBVpwMjcR6imAk0cTe5IZ99zwbeAgYEPS6kQH5M\nT/eiQJPq378/Xbt2JSMjg1atWjFr1iwA3n77bdLS7AcGY/zGd2dUwB+AkSJyAucxH/cAZXGeSTUU\nWJrPvn2Ai4BBIjLIXXaNqv4cxXpNIeUVQFtQwGzfvn2jUY4xJsp816hUdSGwMMSquiG2HZzr/ePA\n49GpzBhjTFH47tKfMcYYf7FGZYwxxtOsURljjPE0a1TGGGM8zXeTKYx/FSY1fffu3fz1r3/lyy+/\npHXr1owePTqWpRtjToM1qiAikgLMBpoBL6lqnxiXZIIUJjW9TJky3HfffXz77besWrWq2Gs1xkSO\nXfpzuenph4B/AvfHuBwTQmFS08uVK0fTpk0pU6ZMcZZojIkC351RRSA9/T8iclFhPtPS06OvoPik\n3Knpxhj/8F2jInLp6WGz9PToCySnh5uaHrB161aOHTvm2VioovBKonUk+GksYOOJFj82qtNOTy8s\nC6WNvkNHj5Oamhp2anrAhg0b+P777z15llhUfkro9tNYwMZTWOGmp/vuHpWqfoMTRLsOJz19EIVI\nTy8KC6WNvsKmphtj/MN3jep00tONtxU2Nb1NmzYMHz6cJUuW0KpVK7777rviLtkYEwF+vPR3Ounp\niEgmcBZQWkQ646SnfxXNgk14CpuavmTJEsB/l2OMKWl816hOJz3dXVYr8lUZY4wpKt9d+jPGGOMv\n1qiMMcZ4WqEblYicIyINolGMMcYYk1tYjUpElorIWSJSCfgvMFFEQt/ZNibCBg4cSPPmzXPikQKm\nTZtG+/btSUtL46mnnspZ/vXXX3PLLbeQlpZGx44dfff1AWNKmnAnU1RU1X0i0hOYqqqPisjaaBYW\nCyJyGTDBfZsADFbVN2JYkiF0GO2KFStYvHgx8+bNo3Tp0uzcuRNwvvz7wAMPMHLkSOrVq8fu3bvZ\nsmVLrEo3xkRAuJf+SonI74GbgQVRrCdm3FDaL3EyARviRDGNd5ebGAoVRjtz5kx69eqVk7OYkpIC\nwPLlyxER6tWrB8A555xDUlLeXxY2xnhfuP8ID8GZ8r1cVT8VkdrAt9Erq+giEEobcCaQHc5nWiht\ndOQXRJuZmcnq1asZPXo0ZcqU4cEHH6RBgwZkZGSQkJBAjx492LVrF9dddx1XXnllMVdujImksBqV\nqs4CZgW9/wHvJjycViitiFwOTAZqArep6rGCPtBCaaMjEEQLnBJGe+DAATIyMnjsscf49ttv6d27\nN+PHj2fbtm2sWLGCUaNGUaZMGQYNGuS7R314JSg0Evw0FrDxREtYjUpE6gLjgKqqeok76+96VX08\nqtUVzWmF0qrqSuBiEUkFXhaRd1T1UHRLNnkJnN3lDqM9//zzueWWW6hfvz7169dn7Nix/O53v6NB\ngwbs2bOHyy+/HIBrr72WzZs3c/vtt8dsDJHmp6QNP40FbDyFFW4obbiX/iYCDwDjAVR1rYjMADzX\nqFT1GxFpDFyHE0q7mCKE0qrqBhH5FbgEWJ3fZ1p6enTkd+mvbdu2rFy5kiuuuIKMjAyOHj3KOeec\nw5VXXsmkSZM4ePAgZ5xxBp9++ilXX311MVdujImkcCdTlFPV3M/zLvCSWCycTiitiFwQmDwhIjWB\neu6++fLb9GcvnOrDb4npocJob7jhBjZt2kR6ejr9+/dn+PDhJCQkULFiRe68805uvPFGOnfuTP36\n9WnatGmMR2KMOR3hnlH9IiIX4k4uEJEbgW1Rq+r0nE4o7ZXAABE5CpwA7lXVX6JcrylAXmG0o0aN\nCrm8U6dOdOrUKee9VxqvMaZowm1UvXG+X1RPRLYAGcD/Rq2q03A6obSqOg2YFp3KjDHGFEWBjUpE\nEnGmdrd1p34nqur+6JdmjDHGhHGPSlVPAA+6r7OsSRljjClO4V76WyQi9wP/JmiWnKruikpVxhhj\njCvcRnWL+3vvoGXZOIkOxhhjTNSEm0xxQbQLMSXbwIEDWbp0KSkpKSxY4MRJjh07ltdee41KlSoB\nzjT1q666KmefrVu3kpaWRp8+fejRo0dM6jbGRF+4yRQhv9avqlMjW44pqUIlpAPceeedeTah4cOH\n07Jly+IozxgTQ+F+4bdZ0K+WwGDg+nA/RERqiciXha4uvGO/5H6vK9ztf41GHeb0hEpIz8+iRYuo\nVq0aderUiWJVxhgvCPfSX9/g9yJyNvBqVCqKQ5aefnryi0p65ZVXmDt3LpdccgkDBgygYsWKZGVl\nMXHiRCZPnszkyZOLtVZjTPEr6rOWsoAi3bdyHxHyOjADaA6Uw3nUxhuq+qC7za/As0A6cBDopKrb\n8zlsWxEZAJwF9FfVBSJyMTAFKI1z5niDqub7aBIRaQ3cr6rp7vt/AatV9aX89rP09NMTSEnPnZDe\ntGlT/vSnP5GQkMCMGTMYOHAgffv2ZcqUKbRt25aNGzeyY8cOypYtm2/6hFcSoCPFT+Px01jAxhMt\n4d6jms9vz2ZKxHk0xqy898jzOIJzJnYn0Aho6P5+GFARGauqm4DywApVfUREngLuIv8A3FrAZTgN\n7wMRuQi4G3hWVV8RkdI4z6aKCgulPT2Hjh4nNTX1lIT0YPfccw933303qampbN68mdWrVzNz5kz2\n7dtHYmIiNWrUoFu3biGPb4nW3uWnsYCNp7AinZ4eHKp2DPhRVTcXsqbKwJtAF1X9SkQaAYtVdS+A\niHyF8wyoTcARfnuS8BqgXQHHfs39YvK3IvIDTpjsJ8AjIlIdmFPQ2dTp8GMobXH+Zcvrst/PP/9M\nlSpVAOeeVOB+1IwZM3K2GTt2LOXKlcuzSRlj4l+4jeo6VT1pOpaIjMi9rAB7gY04wa9fucsOB60/\nHlTPUVXNDrE8L7mfxJutqjNEZCWQBrwtIn9R1SUFHCf4cSBw6iNBTJT079+fVatWsXv3blq1akXf\nvn1ZtWoVX3/9NQDVqlVjyJAhMa7SGBML4TaqdkDuptQhxLL8HAH+B1gYhZl3N4nIyzj3zWrjXEas\nDfygqs+JyPlAA6CgRvUjUF9EyuAkrl8N/CfCtZoQQiWk33TTTQXu17dv3wK3McbEt3wblYjcA9wL\n1BaRtUGrKgDLC/thqpolIunA+0Q2pXwjsApnMsXdqnpIRG4GbnMf2fETMCyM+jaJyGvAlzgJ8Z9F\nsEZjjDFFUNAZ1QzgHeBJYEDQ8v2FyflT1UycJ+Wiqntwvo+Ve5v0oNfJQa9nA7PzOfadeSwfDgwP\nsTw5xObB6x/EDeE1xhgTe/k2Kneiw17gVgARqYJz3yZZRJJVdWP0SzTGGFOShTs9vSPwDHAe8DPO\n7LwNwMXRK+2UGh4Bct+0mKWqTxThWH/g1EuPh1X18qLWZ4wxJjrCnUzxOHAFsEhVG4nIn4BinQ/s\nNqRCN6U8jrUO5ztcxhhjPC7crL+jqroTSBSRRFX9AGgaxbpMCTFw4ECaN29OenrOLUrGjh1Ly5Yt\n6dSpE506deLDDz/MWTd+/HhzCmNAAAAZf0lEQVTatWvHtddey0cffRSLko0xxSzcM6o9IpIMfAS8\nIiI/E/QARb9xp7N/BQxW1VEFbW+KrjCp6d999x1vvfUWb731Ftu3b6d79+4sXLiQpKSohY4YYzwg\n3DOqTsABoB/wLvA90DFaRcWCiAQ37WdwZjuGxUJpi+bQ0eOFSk1fvHgxaWlplC5dmho1alCzZk3W\nrl1b8I7GmLgWbnp6lojUBOqo6ssiUo4oZuedDhEpD7wGVMepcSgwAmiqqr+ISFNglKq2FpHBOPmA\ntXG+i3WriHTG+Q5V2GeMFkpbNPnlI4ZKTd++fTuXXnppzjZVq1Zl+/b8soqNMX4Q7qy/u4BeQCWc\nf9irAS/gJDd4TXtgq6qmAYhIRZxGlZf6wJWqetC9vPkQThLH/VGv1BQqNX337t1s3bo1Z7s9e/aw\nZcuWAtOdvZIAHSl+Go+fxgI2nmgJ9x5Vb5x08pUAqvqt+50qL1oHPC0iI4AFqvqRE9qep3mqetB9\nPRgYraq/FrDPSSw9vWgKm5per1494LdLk4cPH6Zx48YFXqq0RGvv8tNYwMZTWOGmp4d7j+qwquZE\nhLv3c3IHwXqCqn4DNMZpWI+LyCBODpvNHTQbfInvcuApEcnEuR/3sIj0Kegz/ZieXhzyS00PCE5N\nb9OmDW+99RZHjhxh06ZNZGZm0qBBg2Kp1RgTO+GeUX0oIg8DZUWkHU7+3/zolVV0InIesEtVp4vI\nHqAnkAk0wZkgcUNe+6pqy6DjDAZ+VdV/RbXgEq4wqel16tShQ4cOXHfddSQlJTFo0CCb8WdMCRBu\noxoA9MA5S/kL8DYwKVpFnaY/ACNF5ARwFLgHJwn9RREZCiyNYW0ml8Kmpt9zzz3cc8890SzJGOMx\nBaWnn6+qG92HEk50f3maqi4EFoZYVTfEtoPzOU6e64wxxhSfgu5RzQ28EJHXo1yLMcYYc4qCGlVC\n0Ova0SzEGGOMCaWgRpWdx2tjjDGmWBTUqC4VkX0ish9o4L7eJyL7RWRfcRRo/ClUGG3A5MmTERF2\n7XKezbly5UqaNGmSE1L7r3/ZRExjSpKCHpxYoub+ikgNYCpQFecMcoKqPhvbqvwprzDabdu2sXz5\ncs4777yTljdt2pTx48cXZ4nGGI8I9wu/vud+ifkY8HdVrY/z/K3eIlI/tpX5U15htE8++SQPPPAA\nCQkJIfYyxpRE4X6PytPyCaKdCXTAaUC9gCeBi4CRqvqCiLR2t90N1FPVusA2AFXdLyIbcHINv8rv\n8y09PXyHjh7PM5Fi0aJFVKlSJScqKdjnn3/O9ddfT5UqVXjooYdy0iqMMf7ni0ZF3kG0G1W1oYiM\nBl4CWuBEKH2JE6oLTtzSJaqaEXxAEakFNMLNN8yPpaeHL3N4Wk5EU3AY7eHDhxkzZgyDBw9mw4YN\nHDlyhG+//ZazzjqLUqVKMW7cOMqWLcvq1au56667GDduXNif6ZVgzUjx03j8NBaw8USLXxpVXkG0\n84LWJ6vqfmC/iBwWkbPddatCNKlk4HWgn6rapJEIC5yxBYfRqio7d+7kwQcfBGDXrl089NBDzJo1\ni8qVK5+075QpU6hatSqVKlUK6/MsKNS7/DQWsPEUVrihtL5oVKr6jYg0Bq7DCaJd7K467P5+Iuh1\n4H1g7Cc9d0pEzsBpUq+o6pxwPt/S08OX16U/EeGTTz7Jed+mTRtmz55NpUqV2LFjB+eeey4JCQms\nXbuWEydOcM455xRn2caYGPJFo8ojiLYox0kAXgQ2qOqpIXR58GN6erR+igo0qVBhtHll/C1cuJCZ\nM2eSlJTEmWeeyTPPPGOTLYwpQXzRqAgdRDu7CMdpAdwGrBORz91lD6vq25Ep0wSECqMNtmTJkpzX\n3bp1o1u3btEuyRjjUb5oVHkE0dYKWv8SzmSKwPvAuqUEpamr6n84OTbKGGNMjNn3qIwxxniaNSpj\njDGeZo3KGGOMp1mjMsYY42m+mExhvGvgwIEsXbqUlJQUFixYAMCYMWNYvHgxiYmJpKSk8OSTT1K1\nalVWrlzJvffeS/Xq1QFo164dffr0iWX5xhgPsEYVxI1N2gCou2iFqt4du4riX6iU9J49e9KvXz8A\npk6dyvPPP8+QIUMAS0k3xpzKGpXLTU8H+F5VGxZmXwulDe3Q0eM0a9aMzZs3n7Q8OTk55/XBgwft\ny7vGmHz5rlHlk6TeVFV/EZGmwChVbS0ig4ELgdrARmBgUT7TQmlDyy9WavTo0cydO5cKFSowderU\nnOWWkm6Myc13jYq8k9TzUh+4UlUPupf+LhCRz4B9wD9U9aNoF+xnGzZsOCklPaB9+/a0b9+e2bNn\nM2bMGG699dbTTknPi1cSoCPFT+Px01jAxhMtfmxUeSWp52Weqh50X28DzlfVnSLSBJgrIhcXlKBu\nobShHTp6nNTU1JNS0nPr0aMHvXr1yrlHFVCUlPS8WKK1d/lpLGDjKaxw09N9Nz1dVb/BecbUOpwk\n9UE4D04MjPXMXLtkBe17WFV3uq/XAN8DdQv6TD+G0kZCXg9IzMzMzHm9ePFiateuDcCOHTvIzs4G\nsJR0Y0wO351R5ZGkngk0Ad4Bbshn38ruvsdFpDZQB/gh+lX7V6iU9GXLlpGRkUFCQgLVqlXjscce\nAywl3RgTmu8aFaGT1MsCL4rIUIJCaENoBQwRkaM4z6y6W1V3RbleXwuVkp7X4zwsJd0YE4rvGlUe\nSeoQ4hKeqg7O9f51nIcmGmOM8Qjf3aMyxhjjL9aojDHGeJo1KmOMMZ5mjcpEzMCBA2nevDnp6ek5\ny0aMGEH79u3p2LEjvXv3Zt++k7+StnXrVho1asSLL75Y3OUaY+KENapcRCRTRNaJyOcisjrW9cST\nLl26MGnSpJOWtWjRggULFjB//nxq1ap1SuDs8OHDadmyZXGWaYyJM9aoXEGhtAB/UtWGqto0ZgXF\noWbNmlGxYsWTll155ZWUKuX8p23YsCE//fRTzrpFixZRrVo1y/MzxuTLF9PT8wminQl0wEmm6AU8\nCVwEjFTVF0SktbvtbqAeYaRQhFLS09MPHT2eZwpFsNdff50OHToAkJWVxcSJE5k8eTKTJ08uUp3G\nmJLBF42KvINoN6pqQxEZDbwEtMCJUPoSeMHdtzFwiapmuO+zgfdEJBsYr6oTCvrwkp6enjk8LSd2\nKVQALcCsWbM4cOAAderUYcOGDUyZMoW2bduyceNGduzYQdmyZaMWfumVYM1I8dN4/DQWsPFEi18a\nVV5BtPOC1ier6n5gv4gcFpGz3XWrgpoUOEnqW0SkCvC+iHytqsuKayDxKnAWFiqAds6cOaxfv56X\nXnqJsmXLArB582ZWr17NzJkz2bdvH4mJidSoUSMqyRQWFOpdfhoL2HgKK9xQWl80KlX9RkQaA9fh\nBNEudlcddn8/EfQ68D4w9qyg5ajqFvf3n0XkDeAyIN9GVdLT0/O79Lds2TImTZrE9OnTc5oUwIwZ\nM3Jejx07lnLlyll8kjEmJF9MpnCDaA+o6nRgJM7lvKIcp7yIVAi8Bq7BuUyYr5Kenh5oUv3796dr\n165kZGTQqlUrZs2axdChQ8nKyqJ79+506tSJQYMGRaNkY4yP+eKMitBBtLOLcJyqwBvuZcNSwAxV\nfTdiVfpcYQJog/Xt2zca5RhjfMIXjSqPINpaQetfwplMEXgfWLeUoDR1Vf0BuDQaNRpjjCkaX1z6\nM8YY41/WqIwxxniaNSpjjDGeZo3KGGOMp1mjMqclVGL6O++8Q1paGvXq1WPdunU5y48cOcLAgQPp\n2LEj119/PStXroxFycaYOGONKgQRSRKRz0RkQaxr8bpQiel169Zl7NixNGvW7KTls2bNAmD+/PlM\nmTKFESNGcOLEiWKr1RgTn6xRuXKlp98HxD7gKg6ESky/8MILqV279inbfvfdd1x++eUApKSkUKFC\nBb78ssDvUxtjSjhffI8qkunpIlIdSAOeAPqH8/klNT093NT0gHr16rFkyRLS09PZtm0b69evZ9u2\nbTRo0KCopRpjSgBfNCoim54+BngQqBDuh5fU9PRAanpeielZWVlkZmbmPI/q4osv5tNPPyUtLY0q\nVapQp04dtm7dGvV0Zq8kQEeKn8bjp7GAjSda/NKoIpKeLiLpwM+qusY92wpLSQ2lPXT0OKmpqSET\n0wHKly9PrVq1Tlo+atSonNddu3alZcuWXHTRRVGt0xKtvctPYwEbT2GFm57ui3tUqvoNzpnROpz0\n9EDyaWHT01sA14tIJvAq0EZEphf0+SU1lLYwl/0ADh48yIEDBwBYvnw5SUlJUW9Sxpj454szKjc9\nfZeqTheRPUDPohxHVQcCA91jtgbuV1V79kQ++vfvz6pVq9i9ezetWrWib9++nH322QwdOpRdu3bx\nl7/8hdTUVF588UV27txJjx49SExMpGrVqjz11FOxLt8YEwd80aiIXHq6KaRQiekA7dq1O2VZ9erV\nWbgwd3awMcbkzxeNKlLp6bmOmec6Y4wxxccX96iMMcb4lzUqY4wxnmaNyhhjjKdZozJhCxVAu2fP\nHrp3784111xD9+7d2bt3LwB79+6ld+/edOzYkRtvvJFvvvkmVmUbY+KcNaogInKmiKwSkS9EZL2I\nPBbrmrwkVADthAkTaN68Oe+99x7NmzdnwoQJALzwwgukpqYyf/58RowYwRNPPBGLko0xPmCNyuWG\n0h4G2qjqpUBDoL2IXBHbyrwjVADt4sWL6dy5MwCdO3dm0aJFAHz//fdccYXzn+7CCy9ky5Yt/PLL\nL8VbsDHGF3wxPT1SobSqWhf41T3sGe6v7II+3++htPmFz+7cuZMqVaoAULlyZXbu3Ak4AbTvvfce\nTZs2Ze3atWzdupWffvqJc889N7rFG2N8xxeNigiG0opIErAGp6E9r6oFPt3P76G04eYYJiQkkJCQ\nAECvXr144okn6NSpE3Xr1iU1NZWkpMJFLhljDPinUUUklBZAVY8DDd31b4jIJapa4h+aFMj/y52U\nXqFCBZYvX06lSpXYtWsXycnJOetuv/12ALKzs+nVqxcHDhyISRKzVxKgI8VP4/HTWMDGEy2+aFSq\n+o2INAauwwmlXeyuKmwobfAx94jIBzhna/k2Kr+npwdS0oFTktLbt2/P+vXr6dWrFxMmTKBDhw6k\npqayb98+zjzzTEqXLs1rr73GH//4R5o0aRKT+i3R2rv8NBaw8RRWiUpPd0NpD6jqdGAkzuW8ohyn\ncuBMS0TKAu2Arwvaz+/p6YH7U/3796dr165kZGTQqlUrZs2aRa9evVi+fDnXXHMNH3/8Mb169QKc\nyRQdO3bk2muvZdmyZTzyyCPFPg5jjD/44oyKyIXS/h542b1PlQi8pqoLIldmfMsrgPbll18+ZVmj\nRo0sgNYYExG+aFSRCqVV1bVAo2jUaIwxpmh8cenPGGOMf1mjMsYY42nWqIwxxniaNSpjjDGeZo3K\n5KswiemLFi2iY8eOdOrUiS5durB69epYlW2M8RFrVCZfhUlMb968OfPmzePNN99k2LBh/OMf/4hF\nycYYn4laoxKRWiISleghEXlJRG4sxPa/FryVCaUwienly5fPyfo7ePBgzmtjjDkdvvgeVaz5MT29\nKInpAO+//z5PP/00u3btYvz48cVSrzHG34qlUYlIbeB1YAbQHCgHXAi8oaoPutv8CjwLpAMHgU6q\nuj2fw7YVkQHAWUB/VV0gIhcDU4DSOGeLN6jqtwXU1hp4DNiDk3DxGk6I7X1AWaCzqn6f3zH8mJ6e\nOTwtzyDa48ePnxSzdOLEiZz31atXZ/To0axfv55hw4YxZMiQ4i8+F68Ea0aKn8bjp7GAjSdaot6o\nxIkxfxW4Eyf1oaH7+2FARWSsqm4CygMrVPUREXkKuAt4PJ9D1wIuw2l4H4jIRcDdwLOq+oqIlMZ5\nNlU4LgVSgV3AD8AkVb1MRO4D+gL9CjFk38griLZKlSqkpKRQpUoVfv75ZypXrnxKcGVqairjxo2j\natWqVKpUqdhrD2ZBod7lp7GAjaewwg2ljXajqgy8CXRR1a9EpBGwWFX3AojIV0BNYBNwBAjk6q3B\nCYTNz2uqegL4VkR+AOoBnwCPiEh1YE5BZ1NBPlXVbW5N3wPvucvXAX8qaGc/pqfnd+mvTZs2zJ07\nl169ejF37lyuvvpqAH788UfOP/98EhISWL9+PUeOHOGcc84pzrKNMT4U7Vl/e4GNwJVBy4Ift3Gc\n35rlUVXNDrE8L7mfvJutqjOA63EuHb4tIm3CrDP3I0CCHw9SYDP3Y3p6URLTFy5cSHp6Op06dWLI\nkCGMHj3aJlQYY05btM+ojgD/AyyMwsy7m0TkZeACoDbOZcTawA+q+pyInA80AJZE+HNLlMIkpvfq\n1SunaRljTKRE/XtUqpqFM0HibzgTHyJlI7AKeAe4W1UPATcDX4rI58AlwNQIfp4xxpgYSMjOzn0F\nzRTWZ599lt2okX+eDmI3hL3NT+Px01jAxlNYa9asWdOkSZOmBW1nyRTGGGM8zdNf+BWRR4Cbci2e\npapPFOFYfwCm5Vp8WFUvL2p9xhhjos/TjcptSIVuSnkcax3Od7iMMcbEEbv0Z4wxxtOsURljjPE0\na1TGGGM8zRqVMcYYT7PvUUXAmjVrdgA/xroOY4yJMzWbNGlSuaCNrFEZY4zxNLv0Z4wxxtOsURlj\njPE0a1TGGGM8zRqVMcYYT7NGZYwxxtOsURljjPE0T4fSep2ItAeeBZKASao6PMYlFUhEJuM8yPJn\nVb3EXVYJ+DdQC8gEblbV3SKSgDO+64ADwJ2q+t9Y1J0XEamB84DMqkA2MEFVn43XMYnImcAyoAzO\n38/ZqvqoiFwAvAqkAGuA21T1iIiUwRl/E2AncIuqZsak+HyISBKwGtiiqunxPB4RyQT2A8eBY6ra\nNF7/vAGIyNnAJJyHzWYD/w9QPDQeO6MqIvcv3vNAB6A+cKuI1I9tVWF5CWifa9kAYLGq1gEWu+/B\nGVsd91cvYFwx1VgYx4C/q2p94Aqgt/v/IV7HdBhoo6qX4qT9txeRK4ARwGhVvQjYDfRwt+8B7HaX\nj3a386L7gA1B7+N9PH9S1YaqGnjoX7z+eQOn8byrqvWAS3H+P3lqPNaoiu4y4DtV/UFVj+D8dNgp\nxjUVSFWXAbtyLe4EvOy+fhnoHLR8qqpmq+oK4GwR+X3xVBoeVd0W+IlOVffj/CWrRpyOya3rV/ft\nGe6vbKANMNtdnns8gXHOBq52f+r1DBGpDqTh/NSOW1/cjicPcfnnTUQqAq2AFwFU9Yiq7sFj47FG\nVXTVgE1B7ze7y+JRVVXd5r7+CecyGsTZGEWkFtAIWEkcj0lEkkTkc+Bn4H3ge2CPqh5zNwmuOWc8\n7vq9OJfTvGQM8CBwwn2fQnyPJxt4T0TWiEgvd1m8/nm7ANgBTBGRz0RkkoiUx2PjsUZlTqKq2Th/\nEeOKiCQDrwP9VHVf8Lp4G5OqHlfVhkB1nDP3ejEuqchEJHA/dE2sa4mgK1W1Mc5lsN4i0ip4ZZz9\neSsFNAbGqWojIIvfLvMB3hiPNaqi2wLUCHpf3V0Wj7YHTt/d3392l8fFGEXkDJwm9YqqznEXx/WY\nANxLMB8AzXEusQQmPwXXnDMed31FnEkIXtECuN6dgPAqziW/Z4nf8aCqW9zffwbewPlhIl7/vG0G\nNqvqSvf9bJzG5anxWKMquk+BOiJygYiUBroC82JcU1HNA+5wX98BvBm0/HYRSXBv6O8NuhzgCe79\nixeBDar6TNCquByTiFR2Z2EhImWBdjj33T4AbnQ3yz2ewDhvBJa4PwF7gqoOVNXqqloL5+/IElX9\nX+J0PCJSXkQqBF4D1wBfEqd/3lT1J2CTiIi76GrgKzw2HpueXkSqekxE+gALcaanT1bV9TEuq0Ai\nMhNoDZwrIpuBR4HhwGsi0gPncSU3u5u/jTMN9Tucqajdi73ggrUAbgPWufd1AB4mfsf0e+Bld1Zp\nIvCaqi4Qka+AV0XkceAz3Jvf7u/TROQ7nEkyXWNRdBE8RHyOpyrwhvvveilghqq+KyKfEp9/3gD6\nAq+4P3D/gFNjIh4ajz3mwxhjjKfZpT9jjDGeZo3KGGOMp1mjMsYY42nWqIwxxniaNSpjjDGeZtPT\njfEoETkOrAta1NlrSeLGFAdrVMZ410E3SqlYiEipoPw9YzzDGpUxccqNtvk3cBbO3+V7VPUj9zlp\nw3C+iP6Lql7tPi9pMlAb54uavVR1rYgMBi50l28UkW44X5ZujfNMrOdVdXzxjsyYk9k9KmO8q6yI\nfO7+eiPE+j8DC92zrkuBz0WkMjARuMF9ptVN7raPAZ+pagOc5I6pQcepD7RV1Vtxnge1V1WbAc2A\nu9yHHBoTM3ZGZYx3FXTp71NgshvKO1dVPxeR1sAyVc0AUNXAs8euBG5wly0RkRQROctdN09VD7qv\nrwEaiEggh68izkPyMiI2KmMKyRqVMXFKVZe5j5hIA14SkWdwnpZbWFlBrxOAvqq6MBI1GhMJdunP\nmDglIjWB7ao6EefpuY2BFUCrwOU6994UwEfA/7rLWuPcu9p3ykGdkOV73LM0RKSumxJuTMzYGZUx\n8as18ICIHAV+BW5X1R3uU2fniEgiznOE2gGDcS4TrsWZTHFH6EMyCagF/Nd9hMoOfnsMuTExYenp\nxhhjPM0u/RljjPE0a1TGGGM8zRqVMcYYT7NGZYwxxtOsURljjPE0a1TGGGM8zRqVMcYYT/v/EFGh\nLiRAN5EAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6LbHreqRBY9Y",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from surprise import SVD"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YMhWPtuGoHW9",
        "colab_type": "text"
      },
      "source": [
        "### 5.4 XgBoost with Surprise Baseline + Surprise KNNbaseline + MF Techniques"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VV8FEgQYoHW-",
        "colab_type": "text"
      },
      "source": [
        "- svd"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "miEei8sooHW_",
        "colab_type": "code",
        "outputId": "23b0c05c-7105-4755-a872-03be002bd67d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 816
        }
      },
      "source": [
        "# initiallize the model\n",
        "svd = SVD(n_factors=100, biased=True, random_state=15, verbose=True)\n",
        "svd_train_results, svd_test_results = run_surprise(svd, trainset, testset, verbose=True)\n",
        "\n",
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "tuned_models_evaluation_train['svd'] = svd_train_results \n",
        "tuned_models_evaluation_test['svd'] = svd_test_results"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            "Processing epoch 0\n",
            "Processing epoch 1\n",
            "Processing epoch 2\n",
            "Processing epoch 3\n",
            "Processing epoch 4\n",
            "Processing epoch 5\n",
            "Processing epoch 6\n",
            "Processing epoch 7\n",
            "Processing epoch 8\n",
            "Processing epoch 9\n",
            "Processing epoch 10\n",
            "Processing epoch 11\n",
            "Processing epoch 12\n",
            "Processing epoch 13\n",
            "Processing epoch 14\n",
            "Processing epoch 15\n",
            "Processing epoch 16\n",
            "Processing epoch 17\n",
            "Processing epoch 18\n",
            "Processing epoch 19\n",
            "Done. time taken : 0:00:17.447190 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:00:03.140872\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.6616921280231969\n",
            "\n",
            "MAPE : 20.092740147412886\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.068572\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.0724398179398351\n",
            "\n",
            "MAPE : 34.91362328393122\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:00:20.659189\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YCkvVS3KoHXC",
        "colab_type": "text"
      },
      "source": [
        "- svd++"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rYTfxOeFCs94",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from surprise import SVDpp"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0R-1RuPDoHXD",
        "colab_type": "code",
        "outputId": "23b9cb4c-b1ac-41a5-a0ea-c902a96e472d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 816
        }
      },
      "source": [
        "# initiallize the model\n",
        "svdpp = SVDpp(n_factors=50, random_state=15, verbose=True)\n",
        "svdpp_train_results, svdpp_test_results = run_surprise(svdpp, trainset, testset, verbose=True)\n",
        "\n",
        "# Just store these error metrics in our models_evaluation datastructure\n",
        "tuned_models_evaluation_train['svdpp'] = svdpp_train_results \n",
        "tuned_models_evaluation_test['svdpp'] = svdpp_test_results\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model...\n",
            " processing epoch 0\n",
            " processing epoch 1\n",
            " processing epoch 2\n",
            " processing epoch 3\n",
            " processing epoch 4\n",
            " processing epoch 5\n",
            " processing epoch 6\n",
            " processing epoch 7\n",
            " processing epoch 8\n",
            " processing epoch 9\n",
            " processing epoch 10\n",
            " processing epoch 11\n",
            " processing epoch 12\n",
            " processing epoch 13\n",
            " processing epoch 14\n",
            " processing epoch 15\n",
            " processing epoch 16\n",
            " processing epoch 17\n",
            " processing epoch 18\n",
            " processing epoch 19\n",
            "Done. time taken : 0:04:21.207708 \n",
            "\n",
            "Evaluating the model with train data..\n",
            "time taken : 0:00:15.423787\n",
            "---------------\n",
            "Train Data\n",
            "---------------\n",
            "RMSE : 0.6111215770571189\n",
            "\n",
            "MAPE : 17.916043542679418\n",
            "\n",
            "adding train results in the dictionary..\n",
            "\n",
            "Evaluating for test data...\n",
            "time taken : 0:00:00.322898\n",
            "---------------\n",
            "Test Data\n",
            "---------------\n",
            "RMSE : 1.073107287201817\n",
            "\n",
            "MAPE : 34.90611294334286\n",
            "\n",
            "storing the test results in test dictionary...\n",
            "\n",
            "---------------------------------------------\n",
            "Total time taken to run this algorithm : 0:04:36.957996\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kRiyaXaWoHXF",
        "colab_type": "code",
        "outputId": "e7a3db36-038d-4310-dec5-a14aa478151c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 131
        }
      },
      "source": [
        "# add the predicted values from both knns to this dataframe\n",
        "reg_train['svd'] =tuned_models_evaluation_train['svd']['predictions']\n",
        "reg_train['svdpp'] = tuned_models_evaluation_train['svdpp']['predictions']\n",
        "\n",
        "reg_train.head(2) "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>bslpr</th>\n",
              "      <th>svd</th>\n",
              "      <th>svdpp</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>174683</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.882353</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>5</td>\n",
              "      <td>4.996793</td>\n",
              "      <td>4.977554</td>\n",
              "      <td>3.672848</td>\n",
              "      <td>4.220762</td>\n",
              "      <td>3.997242</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>233949</td>\n",
              "      <td>10</td>\n",
              "      <td>3.587581</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>2.692308</td>\n",
              "      <td>3.611111</td>\n",
              "      <td>3</td>\n",
              "      <td>3.062453</td>\n",
              "      <td>3.051725</td>\n",
              "      <td>3.688917</td>\n",
              "      <td>3.744945</td>\n",
              "      <td>3.671130</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg  sur1  ...  knn_bsl_m     bslpr       svd     svdpp\n",
              "0  174683     10  3.587581   5.0  ...   4.977554  3.672848  4.220762  3.997242\n",
              "1  233949     10  3.587581   4.0  ...   3.051725  3.688917  3.744945  3.671130\n",
              "\n",
              "[2 rows x 21 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I32shRohD7gu",
        "colab_type": "code",
        "outputId": "7fd54f94-eab6-4ea6-de49-1f88299bede8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "reg_train.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(281949, 21)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IhucL0gUoHX6",
        "colab_type": "code",
        "outputId": "80dd73af-5640-4538-a746-d189970fcbe3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 131
        }
      },
      "source": [
        "reg_test_df['svd'] = tuned_models_evaluation_test['svd']['predictions']\n",
        "reg_test_df['svdpp'] = tuned_models_evaluation_test['svdpp']['predictions']\n",
        "\n",
        "reg_test_df.head(2) "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>user</th>\n",
              "      <th>movie</th>\n",
              "      <th>GAvg</th>\n",
              "      <th>sur1</th>\n",
              "      <th>sur2</th>\n",
              "      <th>sur3</th>\n",
              "      <th>sur4</th>\n",
              "      <th>sur5</th>\n",
              "      <th>smr1</th>\n",
              "      <th>smr2</th>\n",
              "      <th>smr3</th>\n",
              "      <th>smr4</th>\n",
              "      <th>smr5</th>\n",
              "      <th>UAvg</th>\n",
              "      <th>MAvg</th>\n",
              "      <th>rating</th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>bslpr</th>\n",
              "      <th>svd</th>\n",
              "      <th>svdpp</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>808635</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>5</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>941866</td>\n",
              "      <td>71</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>3.581679</td>\n",
              "      <td>4</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "      <td>3.579949</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     user  movie      GAvg      sur1  ...  knn_bsl_m     bslpr       svd     svdpp\n",
              "0  808635     71  3.581679  3.581679  ...   3.579949  3.579949  3.579949  3.579949\n",
              "1  941866     71  3.581679  3.581679  ...   3.579949  3.579949  3.579949  3.579949\n",
              "\n",
              "[2 rows x 21 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0cNHOz-ioHX_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# prepare x_train and y_train\n",
        "x_train = reg_train.drop(['user', 'movie', 'rating',], axis=1)\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# prepare test data\n",
        "x_test = reg_test_df.drop(['user', 'movie', 'rating'], axis=1)\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "# making model ready for tuning\n",
        "xgb_model = XGBRegressor(silent=False, n_jobs=-1,random_state=15,verbosity=1,nthread=-1)\n",
        "xgb_final = GridSearchCV(xgb_model,param_grid,n_jobs=-1,cv = 3,verbose=10,return_train_score=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XbIxKBNroHYB",
        "colab_type": "code",
        "outputId": "5efd228a-962f-4e90-c804-a8c3a66d5cc1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 666
        }
      },
      "source": [
        "train_results, test_results = run_xgboost(xgb_final, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "tuned_models_evaluation_train['xgb_final'] = train_results\n",
        "tuned_models_evaluation_test['xgb_final'] = test_results"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model..\n",
            "Fitting 3 folds for each of 64 candidates, totalling 192 fits\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
            "[Parallel(n_jobs=-1)]: Done   5 tasks      | elapsed:  1.2min\n",
            "[Parallel(n_jobs=-1)]: Done  10 tasks      | elapsed:  2.4min\n",
            "/usr/local/lib/python3.6/dist-packages/joblib/externals/loky/process_executor.py:706: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
            "  \"timeout or by a memory leak.\", UserWarning\n",
            "[Parallel(n_jobs=-1)]: Done  17 tasks      | elapsed:  3.9min\n",
            "[Parallel(n_jobs=-1)]: Done  24 tasks      | elapsed:  6.2min\n",
            "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  9.1min\n",
            "[Parallel(n_jobs=-1)]: Done  42 tasks      | elapsed: 12.3min\n",
            "[Parallel(n_jobs=-1)]: Done  53 tasks      | elapsed: 16.3min\n",
            "[Parallel(n_jobs=-1)]: Done  64 tasks      | elapsed: 19.0min\n",
            "[Parallel(n_jobs=-1)]: Done  77 tasks      | elapsed: 22.7min\n",
            "[Parallel(n_jobs=-1)]: Done  90 tasks      | elapsed: 27.5min\n",
            "[Parallel(n_jobs=-1)]: Done 105 tasks      | elapsed: 32.4min\n",
            "[Parallel(n_jobs=-1)]: Done 120 tasks      | elapsed: 36.7min\n",
            "[Parallel(n_jobs=-1)]: Done 137 tasks      | elapsed: 43.3min\n",
            "[Parallel(n_jobs=-1)]: Done 154 tasks      | elapsed: 48.8min\n",
            "[Parallel(n_jobs=-1)]: Done 173 tasks      | elapsed: 54.2min\n",
            "[Parallel(n_jobs=-1)]: Done 192 out of 192 | elapsed: 63.0min finished\n",
            "/usr/local/lib/python3.6/dist-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
            "  if getattr(data, 'base', None) is not None and \\\n",
            "/usr/local/lib/python3.6/dist-packages/xgboost/core.py:588: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
            "  data.base is not None and isinstance(data, np.ndarray) \\\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "[17:21:21] WARNING: /workspace/src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
            "Done. Time taken : 1:03:19.336545\n",
            "\n",
            "Done \n",
            "\n",
            "Evaluating the model with TRAIN data...\n",
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.0761202825525833\n",
            "MAPE :  34.48325107967784\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OPC3szFJTmhn",
        "colab_type": "code",
        "outputId": "b9ff3642-0ed6-44b8-b642-e95059f36462",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 318
        }
      },
      "source": [
        "xgb.plot_importance(xgb_final.best_estimator_)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f14d1bc1320>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 82
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbcAAAEcCAYAAABETPrGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFFwAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KdGfYz1oEq3q",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pickle\n",
        "with open(\"/content/drive/My Drive/Netflix_recommender/tuned_result.pkl\",\"wb\") as f:\n",
        "    pickle.dump(tuned_models_evaluation_train,f)\n",
        "    pickle.dump(tuned_models_evaluation_test,f)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QcWi12vOoHYR",
        "colab_type": "text"
      },
      "source": [
        "### 5.5 XgBoost with Surprise Baseline + Surprise KNNbaseline + MF Techniques"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qG75DLAVoHYS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# prepare train data\n",
        "x_train = reg_train[['knn_bsl_u', 'knn_bsl_m', 'svd', 'svdpp']]\n",
        "y_train = reg_train['rating']\n",
        "\n",
        "# test data\n",
        "x_test = reg_test_df[['knn_bsl_u', 'knn_bsl_m', 'svd', 'svdpp']]\n",
        "y_test = reg_test_df['rating']\n",
        "\n",
        "\n",
        "# making model ready for tuning\n",
        "xgb_model = XGBRegressor(silent=False, n_jobs=-1,random_state=15,verbosity=1,nthread=-1)\n",
        "xgb_all_models = GridSearchCV(xgb_model,param_grid,n_jobs=-1,cv = 3,verbose=10,return_train_score=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fw2SbfQcoHYU",
        "colab_type": "code",
        "outputId": "d1a2aced-eb8c-4f00-c992-afc8925fc2ba",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 612
        }
      },
      "source": [
        "train_results, test_results = run_xgboost(xgb_all_models, x_train, y_train, x_test, y_test)\n",
        "\n",
        "# store the results in models_evaluations dictionaries\n",
        "tuned_models_evaluation_train['xgb_all_models'] = train_results\n",
        "tuned_models_evaluation_test['xgb_all_models'] = test_results"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training the model..\n",
            "Fitting 3 folds for each of 64 candidates, totalling 192 fits\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
            "[Parallel(n_jobs=-1)]: Done   5 tasks      | elapsed:   34.3s\n",
            "[Parallel(n_jobs=-1)]: Done  10 tasks      | elapsed:  1.1min\n",
            "[Parallel(n_jobs=-1)]: Done  17 tasks      | elapsed:  1.8min\n",
            "[Parallel(n_jobs=-1)]: Done  24 tasks      | elapsed:  2.6min\n",
            "[Parallel(n_jobs=-1)]: Done  33 tasks      | elapsed:  3.8min\n",
            "[Parallel(n_jobs=-1)]: Done  42 tasks      | elapsed:  4.9min\n",
            "[Parallel(n_jobs=-1)]: Done  53 tasks      | elapsed:  6.3min\n",
            "[Parallel(n_jobs=-1)]: Done  64 tasks      | elapsed:  7.6min\n",
            "[Parallel(n_jobs=-1)]: Done  77 tasks      | elapsed:  9.4min\n",
            "[Parallel(n_jobs=-1)]: Done  90 tasks      | elapsed: 11.7min\n",
            "[Parallel(n_jobs=-1)]: Done 105 tasks      | elapsed: 13.9min\n",
            "[Parallel(n_jobs=-1)]: Done 120 tasks      | elapsed: 16.2min\n",
            "[Parallel(n_jobs=-1)]: Done 137 tasks      | elapsed: 19.6min\n",
            "[Parallel(n_jobs=-1)]: Done 154 tasks      | elapsed: 22.3min\n",
            "[Parallel(n_jobs=-1)]: Done 173 tasks      | elapsed: 25.2min\n",
            "[Parallel(n_jobs=-1)]: Done 192 out of 192 | elapsed: 29.6min finished\n",
            "/usr/local/lib/python3.6/dist-packages/xgboost/core.py:587: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
            "  if getattr(data, 'base', None) is not None and \\\n",
            "/usr/local/lib/python3.6/dist-packages/xgboost/core.py:588: FutureWarning: Series.base is deprecated and will be removed in a future version\n",
            "  data.base is not None and isinstance(data, np.ndarray) \\\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "[17:59:13] WARNING: /workspace/src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
            "Done. Time taken : 0:29:43.030055\n",
            "\n",
            "Done \n",
            "\n",
            "Evaluating the model with TRAIN data...\n",
            "Evaluating Test data\n",
            "\n",
            "TEST DATA\n",
            "------------------------------\n",
            "RMSE :  1.0750524068075662\n",
            "MAPE :  35.18174702798547\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pR39eiA5ZB2H",
        "colab_type": "code",
        "outputId": "c898fb6f-1f00-4fc5-923c-e5a4f55be8f6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 318
        }
      },
      "source": [
        "xgb.plot_importance(xgb_all_models.best_estimator_)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f14d58749b0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 85
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbcAAAEcCAYAAABETPrGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzt3XtYFOe9B/DvsrCgYiAaQSLmWE1R\nIyKry02MyKXgFYJpq6bBGDVorKjhJArGo8VaFa2mPYoSNQkqafRJRASiNokioAYUchSxjShpvICg\nXCO3XXZ3zh/WaRDEBYHF4ft5njzPzrwzs795M+6Xd2Z2RyYIggAiIiIJMTF2AURERO2N4UZERJLD\ncCMiIslhuBERkeQw3IiISHIYbkREJDkMN6JuYvXq1YiJiTF2GUSdQsbvuRG1zMfHB6WlpZDL5eK8\n48ePw9bWts3bzMrKwnvvvYf09PT2KPGpExERAVtbW7zzzjvGLoUkytTYBRA9DWJjYzF27FhjlyHS\narUwNX06//nqdDpjl0DdAE9LEj2BCxcuYObMmVCpVAgMDERWVpbYdujQIUyaNAlKpRK+vr44cOAA\nAKC2thZvvfUW7ty5A6VSCaVSiZKSEkREROCDDz4Q18/KysL48ePFaR8fH+zatQvTpk2Ds7MztFot\nSkpKEBYWBnd3d/j4+GDfvn2PrPXn23+w7d27d8PDwwPjxo3DN998g7S0NAQEBMDV1RWxsbHiutu2\nbcOSJUuwbNkyKJVKBAcH4/vvvxfbCwoKEBISApVKhSlTpuDEiRON3nfNmjV466234OzsjC+++ALJ\nycn46KOPoFQqsXDhQgDArl274OfnB6VSicmTJ+Prr78Wt5GQkIBZs2YhOjoaLi4u8PHxQVpamthe\nWVmJyMhIjBs3Di4uLli0aJHYlpqaiqCgIKhUKsycObNR3SRhAhG1yNvbWzhz5kyT+cXFxYKrq6tw\n6tQpQafTCadPnxZcXV2FsrIyQRAEITU1Vbh+/bqg1+uFrKwswcnJScjLyxMEQRAyMzOFl19+udH2\nVqxYIWzdulWcfngZb29vITAwUCgqKhLq6uoEnU4nBAcHC9u2bRPUarVw48YNwcfHR0hPT292P36+\n/czMTGH48OHCtm3bBI1GIxw8eFBwc3MTwsPDhXv37gn5+fnCyJEjhRs3bgiCIAj/+7//K7z00kvC\nsWPHBI1GI+zZs0fw9vYWNBqNoNFoBD8/P2Hnzp2CWq0Wzp49Kzg7OwsFBQXi+44ePVrIzs4WdDqd\nUF9f32RfBUEQjh49KhQXFws6nU748ssvhVGjRgklJSWCIAjCoUOHhJdeekk4ePCgoNVqhU8//VTw\n9PQU9Hq9IAiC8NZbbwlLly4VKisrBY1GI2RlZQmCIAiXL18W3N3dhQsXLgharVZISEgQvL29BbVa\nbcj/enqKceRGZIDf//73UKlUUKlU4qjgyJEjGD9+PLy8vGBiYgJPT084OjqKI4oJEybghRdegEwm\ng6urKzw9PZGdnf1EdYSEhMDOzg4WFha4dOkSysvLsXjxYigUCgwcOBC//e1vcfToUYO2ZWpqirff\nfhtmZmaYPHkyKioqMHv2bFhaWuKXv/wlXnzxRVy5ckVcfsSIEZg4cSLMzMzw5ptvQqPR4OLFi7h4\n8SJqa2sRGhoKhUIBDw8PeHt748svvxTX9fX1xZgxY2BiYgJzc/Nm65k0aRJsbW1hYmKCyZMn47/+\n67+Qm5srtj///PP47W9/C7lcjuDgYNy9exelpaW4c+cO0tPTERUVBSsrK5iZmcHV1RUAcPDgQcyY\nMQOjRo0S1zMzM8OFCxfa0v30FHk6T9oTdbKYmJgm19yKiopw/PhxpKamivO0Wi3c3NwAAGlpaYiJ\nicGPP/4IvV6P+vp6ODg4PFEddnZ24uvCwkLcuXMHKpVKnKfT6RpNt8Ta2lq8ScbCwgIA0LdvX7Hd\n3NwcNTU14nT//v3F1yYmJrC1tcWdO3fENhOT//yt/Pzzz6OkpKTZuh8lMTERn3zyCQoLCwHcP31b\nUVEhtj/33HPi6x49eojLVFVVwcrKClZWVk22WVRUhMTERMTHx4vzGhoaxLpJuhhuRG1kZ2eHoKAg\nrFu3rkmbRqPBkiVLEB0dDV9fX5iZmWHRokUQ/n1zskwma7JOjx49UF9fL06XlpY2Webn69nZ2cHe\n3h5fffVVe+zOYxUXF4uv9Xo9SkpKYGNjI7bp9Xox4G7fvo1BgwY9clsP739hYSFWrVqFuLg4KJVK\nyOVyBAUFGVRX//79UVVVhZ9++gnPPPNMozY7OzssXLgQb7/9tkHbIungaUmiNgoMDERqaioyMjKg\n0+mgVquRlZWF4uJiaDQaaDQa9OnTB6ampkhLS8OZM2fEdfv27YvKykrcu3dPnDd8+HCkpaWhsrIS\nd+/exd69e1t8fycnJ/Tq1Qu7du1CfX09dDod8vPzG53Ka0+XL1/GV199Ba1Wi71790KhUGDUqFFw\ncnKChYUF9uzZg4aGBmRlZeHkyZOYPHnyI7fVt29f3Lp1S5yuq6uDTCZDnz59ANy/Gefq1asG1WVj\nY4Px48cjKioKVVVVaGhowPnz5wEAv/nNb3DgwAFcvHgRgiCgtrYWp06dQnV19RP0BD0NGG5EbWRn\nZ4cdO3bgww8/hIeHB7y8vPDRRx9Br9fD0tISq1atwrJly+Di4oKUlBT4+PiI6w4ZMgRTpkyBn58f\nVCoVSkpKEBQUhGHDhsHHxwdz585tMRwAQC6XIzY2Ft9//z18fX3h7u6OVatWddgHt6+vL44ePQoX\nFxccOXIE27Ztg5mZGRQKBWJjY5Geng53d3dERUVh06ZNGDJkyCO39etf/xrXrl0Tr2G++OKLmDt3\nLmbOnImxY8ciPz8fo0ePNri2TZs2wdTUFJMmTcLYsWPFPwxGjhyJP/7xj1i7di1cXFzg7++PhISE\nJ+4L6vr4JW4ieqxt27bh+vXr+POf/2zsUogMwpEbERFJDsONiIgkh6cliYhIcjhyIyIiyeH33DqR\nXq9HTU0NzMzMmv2eExERNfXgBKO5ubnBn50Mt05UU1OD/Px8Y5dBRPRUcnR0fOTPtz2M4daJzMzM\nAAAODg5QKBRGrqbryMvLg6Ojo7HL6HLYL02xT5on9X7RaDStHhgw3DrRg+G0QqEw+K+P7oL90Tz2\nS1Psk+axXxrjDSVERCQ5DDciIpIchhsREUkOw42IiCSH4UZERJLDcCMiIslhuBERkeQw3IiISHIY\nbkREJDkMNyIikhyGGxERSQ7DjYiIJIfhRkREksNwIyIiyWG4ERGR5DDciIhIchhuREQkOQw3IiKS\nHIYbERFJDsONiIgkh+FGRESSw3AjIiLJYbgREZHkMNyIiEhyGG5ERCQ5DDciIpIchhsREUkOw42I\niCSH4UZERJLDcCMiIslhuBERkeQw3IiISHIYbkREJDkMNyIikhyGGxERSQ7DjYiIWkWj0WDlypXw\n9vaGUqlEUFAQ0tLSxLYlS5bAx8cHQ4cORVZWVqN1f/rpJ6xYsQIeHh7w8PDAtm3bOqRGhhsZ3Zgx\nY4xdQpfEfmmKfdK8zuoXTYMOAKDVamFnZ4f9+/cjJycHy5Ytw7Jly3Dr1i0AwOjRo7Fp0yb069ev\nyTY2bNiAuro6nDx5Ep9//jmOHDmCQ4cOtXutpu2+xW4mJCQEc+fOhbe3t8HrzP/T16is0XVgVURE\n7S95SxAAoGfPnggLCxPne3t7w97eHpcvX4a9vT3mzJkDADAxaTp+OnnyJHbv3o0ePXrA3t4ev/71\nr3Ho0CG8+uqr7VorR25ERPRESktL8eOPP+LFF19s9bqCIODq1avtXlO3H7nV1dVhxYoVuHbtGkxN\nTfGLX/wCVVVVeP311+Hn5wcASE1Nxccff4z9+/fj2rVriIyMRG1tLRwcHKBWq428B0RExtPQ0IB3\n330XwcHBGDJkyGOXf/nll7Fr1y5s3LgRZWVlOHToEOrq6tq9rm4fbqdPn0ZNTQ2OHj0KAKiqqsKp\nU6eQmJgohltCQoI4ZF6+fDlCQkIQHByMCxcuYNasWUarnYios+Xk5Iiv9Xo9tm/fjrq6OkyZMqVR\nG3A/+PLz82Fq+p+omTZtGuLi4uDj4wNLS0u4uLjg7NmzTdZ9Ut0+3IYNG4aCggJERUXB1dUVEyZM\ngL+/PzZs2ICKigoAwLlz5xAdHY3q6mrk5+cjKOj+eWdnZ2c4ODgYs3wiok714OYVQRCwcuVK6PV6\n7Nu3DxYWFk2WNTMzg4ODQ5MbXry8vMTXW7duhYuLS4s3xajVauTl5bWqzm5/zW3gwIFISUmBp6cn\nvv32WwQFBcHExAS+vr5ISUlBSkoKfH190bNnT2OXSkTUZaxZswYFBQWIjY1tEmwajUa8ZNPQ0AC1\nWg1BEAAAN27cQEVFBXQ6HdLS0nDw4EG8/fbb7V5ftx+5FRcXw8rKCn5+fvD09MTLL7+MyspKBAcH\nY/369QCAlStXAgAsLS3h4OCA5ORkBAUFITc3F/n5+cYsn4io0xUWFuLgwYNQKBQYN26cOD8qKgqB\ngYGYOHEiCgsLAQDz5s0DAJw4cQL29vbIy8vD+vXrce/ePQwaNAh//vOf8ctf/rLda+z24XblyhVs\n2bIFwP3zx6GhobC1tYWtrS2qq6sBACqVSlx+06ZNiIyMxO7du+Hg4ICRI0capW4iImMZMGAArly5\n8sj2kydPPrJt8uTJmDx5ckeU1YhMeDBWpA734Lyxo6MjzM3NjV0OEVGraBp0UJjJO/192/LZ2e2v\nuZHxtfddUlLBfmmKfdK8zuoXYwRbWzHciIhIchhuREQkOQw3IiKSHIYbERFJDsONiIgkh+FGRESS\nw3AjIiLJYbgREZHkMNyIiEhyGG5ERCQ5DDciIpIchhsREUkOw42IiCSH4UZERJLDcCMiIslhuBER\nkeQw3IiISHIYbkREJDkMNyIikhyGGxERSQ7DjYiIJIfhRkREksNwIyIiyWG4ERGR5DDciIhIchhu\nREQkOQw3IiKSHIYbERFJDsONiIgkh+FGRESSw3AjIiLJYbgREZHkMNzI6MaMGWPsErok9ktTXa1P\nNA068XV8fDymT58OR0dHREREiPOTkpKgVCrF/0aNGoWhQ4ciLy9PXOby5cv43e9+B6VSibFjx2Lv\n3r2duh9SZGrsArqj+X/6GpU1uscvSERdWvKWIPG1jY0NFi1ahIyMDKjVanF+YGAgAgMDxemEhATs\n2LEDI0aMAACUl5dj/vz5iIyMxMSJE6HRaFBSUtJ5OyFRRhm5DR06FDU1NR36Hrdu3YKbm1ur1klI\nSMCSJUs6qCIikjJ/f3/4+fnB2tq6xeUOHz6MV155BTKZDAAQFxeHcePGITAwEAqFApaWlhgyZEhn\nlCxpbQ63+vp6aDSa9qyFiEjSCgsLkZ2djaCg/4z4Lly4ACsrK8ycORMeHh5YuHAhioqKjFilNBgc\nbtHR0cjNzQUAnDp1Cq6urnBxccHJkyfb/OZ6vR7r169HeHg4NBoNIiIisHr1asyePRv+/v5Yvnw5\nBEEAgBbbWrJx40ZMmzYN06ZNQ3Z2NgCgrKwMc+bMEeevX7/e4Jp9fHyQn5//yGkiokdJTEyESqXC\nwIEDxXklJSVITEzEypUrcerUKdjb2yM8PNyIVUqDwdfckpOTxVN2MTEx2Lx5M3r37o0NGzbAx8en\n1W+sVqsRGRmJAQMGYMuWLeIQ/erVq4iLi4NMJkNwcDDOnj0LT0/Px7Y1p7KyEsOGDUNERASysrIQ\nHh6Ob775BsnJyXjhhRcQFxcHAKiqqmp1/UREAJCTk9Nouri4GOXl5U3mA8DBgwcRFBTUqE2v10Op\nVKKhoQF5eXkYN24c9u/fj4yMDPTs2bPNdXR3BodbXV0devTogYqKCty8eRMBAQEA7g+z22L+/PmY\nMmUK5s2b12i+n58fzM3NAQAvvfQSbty4IQZYS23NMTMzEy/kurm5wcLCAj/88ANGjRqFuLg4REdH\nw9XVFePGjWvTPhARPXwHZ3p6OmQyWZP5OTk5+Omnn7BgwQJYWlqK852dnWFqaiouX1lZKc7v3bu3\nQTXk5OR0uTtJ25NarW50d6khDD4tOWjQICQlJeHTTz8VA6W8vBwWFhatq/Lf3NzckJGRgbq6ukbz\nH4QXAMjlcuh0OoPaWkOpVOLw4cNwdHTEkSNHMHv2bIPXlcvl0Ov14vTP74oiou5Lq9VCrVZDr9dD\np9NBrVZDq9WK7YmJifD3928UbAAwffp0fPPNN/jnP/+JhoYG7NixA2PGjDE42Kh5BofbmjVr8Le/\n/Q2ZmZlYunQpAOD06dMtjpxasnjxYowdOxbz5s1DdXV1m7bxOA0NDUhOTgYAZGdno76+HoMHD8bN\nmzdhaWmJKVOmIDIyEpcvX24UWC154YUXcOnSJQDAt99+i9LS0g6pnYieLjt37oSTkxN27dqFpKQk\nODk5YefOnQDu/xF87NgxBAcHN1nPw8MD77zzDkJDQzF27FjcuHEDW7Zs6ezyJcfg05JOTk44cOBA\no3kPf3+jtUJDQ2FhYYE5c+Zgz549bd7Oo1hbW+P7778Xt71161YoFAqcO3cOcXFxMDExgV6vR1RU\nFExMDMv5pUuXIiIiAvHx8XB3d8fzzz/f6rr2vP+rRqNQIno6aRp0UJjJAQBhYWEICwtrdjlzc3Px\nhrbmvPbaa3jttdc6pMbuSiYYcsvhv505cwZffvklysvLERsbi0uXLqG6uhoeHh4dWaNkPDhv7Ojo\nyHD7GalfL2gr9ktT7JPmSb1f2vLZafBpyf379+MPf/gDBg0ahPPnzwMALCws8Ne//rVt1RIREXUQ\ng09L7t27F3FxcbC3t8fu3bsBAIMHD8a//vWvDivOEKtXr8bFixcbzZPL5UhISGjzNtPS0rB169Ym\n88PDw+Hl5dXm7RIRUecwONxqampgZ2cHAOJ30rRaLczMzDqmMgOtXbu23bfp5eXFECMieooZfFrS\nxcUFu3btajRv3759rf79RiIioo5m8Mht1apVWLhwIT7//HPU1NQgICAAvXr1wocfftiR9REREbWa\nweH23HPP4dChQ7h06RIKCwthZ2cHJycng2+hJyIi6iwGhZtOp4NSqUR2djacnJzg5OTU0XURERG1\nmUHDLrlcjkGDBqGioqKj6yEiInpiBp+WnDZtGhYuXIjZs2ejf//+jdr4JW4iIupKDA63zz77DACw\nbdu2RvNlMhlOnDjRvlURERE9AYPD7UkeSkpERNSZeKsjERFJjsEjNy8vL/GXSR526tSp9qqHiIjo\niRkcbps3b240fffuXezbtw+TJ09u96KIiIiehMHh5urq2uy8+fPn44033mjXooiIiJ7EE11zUygU\nuHXrVnvVQkRE1C4MHrk9/Ny2+vp6pKWlYfz48e1eFBER0ZMwONyKi4sbTffo0QNvvvkmgoKC2r0o\nIiKiJ2FwuIWHh6Nfv35N5t+9e7fZ+URERMZi8DW3gICAZudPmTKl3YohIiJqDwaHmyAITeZVV1c/\n8rtvRERExvLY05IPvrytVqsxYcKERm2VlZUcuRERUZfz2HDbvHkzBEFAaGgoNm3aJM6XyWTo27cv\nBg8e3KEFEhERtdZjw+3Bl7czMzPRo0ePDi+IiIjoSRl8t2SPHj3wz3/+E9nZ2aioqGh0DW7p0qUd\nUhwREVFbGHxDycGDBzFr1ixkZmZi9+7dyM/PxyeffIIbN250ZH1EREStZnC47dmzB3v27EFMTAws\nLCwQExODv/71rzA1NXjwR0RE1CkMDreysjKoVKr7K5mYQK/Xw8vLC6mpqR1WHBERUVsYPOzq378/\nbt26BXt7ewwaNAgnTpzAs88+CzMzs46sj4iIqNUMDrf58+ejoKAA9vb2WLRoEZYuXYqGhga8//77\nHVkfERFRqxkcbtOnTxdfe3l54dy5c2hoaECvXr06pDAiIqK2atXz3CoqKpCYmIjdu3dDoVCgurq6\nydMCiIiIjM3gcDt37hwmTpyI5ORk7NixAwBw/fp1/OEPf+io2oiIHik+Ph7Tp0+Ho6MjIiIixPm3\nbt3C0KFDoVQqxf9iYmLEdo1Gg8jISIwePRqenp745JNPjFE+dTCDT0uuX78ef/nLX+Dh4QEXFxcA\nwKhRo5Cbm9thxVH3MGbMGGOX0CWxX5oaNnyE+NrGxgaLFi1CRkYG1Gp1k2XPnz/f7FeVtm3bhuvX\nryM1NRWlpaWYPXs2hgwZwgcvS4zB4VZYWAgPDw8AEJ8EYGZmBp1O1zGVSdj8P32Nyhr2G1FrJW/5\nz8OR/f39AQCXLl1CSUmJwds4fPgwNm7cCCsrK1hZWeE3v/kNDh8+zHCTGINPSw4ZMgQZGRmN5p09\nexYODg6PXXfo0KGoqalpfXWtcOvWLbi5ubVqnYSEBCxZsqSDKiIiY/L29sb48eMRGRmJ8vJyAEBV\nVRXu3r2LYcOGicsNGzYM165dM1aZ1EEMHrlFRERgwYIFmDBhAurr67F69WqcPHlSvP5GRNQVPPvs\ns/jiiy8wfPhwVFZWYu3atXjvvffw0Ucfoba2FgDQu3dvcfnevXt3+B/f1PkeG253795Fv3794Ozs\njKSkJCQlJeHVV1+FnZ0dvvjiC/Tv39/gN9Pr9di4cSNKS0uxceNGrF69GgqFAj/++COKi4vh7OyM\n6OhoyGQyREREPLKtJRs3bsSZM2cAAGvWrIFKpUJZWRn++7//G2VlZQAADw8PrFy50qCaQ0JCMGLE\nCOTm5qKwsBCzZ8+Gra0t4uPjcefOHbz33nuYNGmSwX1ARE8mJyen0XRxcTHKy8ubzL948SIAICgo\nCIsWLcLp06fFyyhnzpyBlZUVACA3NxcmJiZN1n/aPO31t7fHhltAQAC+++47AICtrS0uXryI7du3\nt/qN1Go1IiMjMWDAAGzZskUMqatXryIuLg4ymQzBwcE4e/YsPD09H9vWnMrKSgwbNgwRERHIyspC\neHg4vvnmGyQnJ+OFF15AXFwcgPunJlqjuLgY8fHxuHv3Lvz9/TFnzhwcOHAAubm5WLx4McONqBM9\nfKNNeno6ZDLZI2/AKS0tBXD/BrjevXujX79+MDc3F5c/ffo0Ro4c+VTfwJOTk/NU1/84arUaeXl5\nrVrnsdfcfv5oG+D+VwLaYv78+Rg1ahRWrFjRaPTl5+cHc3NzKBQKvPTSS42eMtBSW3PMzMwQGBgI\nAHBzc4OFhQV++OEHjBo1Cunp6YiOjkZqaip69uzZqtonTpwIExMT2NrawtraGn5+fgCAESNGoKSk\npNk7tYioY2m1WqjVauj1euh0OqjVami1Wly8eBE//PAD9Ho9KioqsG7dOri6uoqnIl955RXs3LkT\nVVVVKCgowOeff47g4GAj7w21t8eG2+NOAxrKzc0NGRkZqKurazTf3NxcfC2XyxvdfdlSW2solUoc\nPnwYjo6OOHLkCGbPnt2q9R+u48G0XC4HcP8fGRF1rp07d8LJyQm7du1CUlISnJycsHPnTty8eRPz\n58/H6NGjMW3aNCgUCmzdulVcb8mSJRg4cCC8vb0REhKCefPm8U5JCXrsaUmdTofMzExxBKfVahtN\nAxC/ItCSxYsX49NPP8W8efOwa9cuWFpaPkHZzWtoaEBycjKCgoKQnZ2N+vp6DB48GDdv3kT//v0x\nZcoUqFQq/OpXv4Jer2/39yeizhMWFoawsLBm26ZOnfrI9RQKBTZs2IANGzZ0VGnUBTw23Pr27dvo\n5gtra+tG0zKZDCdOnDDozUJDQ2FhYYE5c+Zgz549bSi3ZdbW1vj+++/FbW/duhUKhQLnzp1DXFyc\n+KieqKgomJi06pfH2tWe93/VaDRIRIapqa1Hr54Wxi6DngIy4eGLatRhHlwUdXR0ZLj9jNQvhrcV\n+6Up9knzpN4vbfnsNN7whYiIqIMY/CXurmL16tXi91cekMvlSEhIaPM209LSGl1wfiA8PBxeXl5t\n3i4RERnHUxdua9eubfdtenl5McSIiCSEpyWJiEhyGG5ERCQ5DDciIpIchhsREUkOw42IiCSH4UZE\nRJLDcCMiIslhuBERkeQw3IiISHIYbkREJDkMNyIikhyGGxERSQ7DjYiIJIfhRkREksNwIyIiyWG4\nERGR5DDciIhIchhuREQkOQw3IiKSHIYbERFJDsONiIgkh+FGRESSw3AjIiLJYbgREZHkMNyIiEhy\nGG5ERCQ5DDciIpIchhsREUkOw42IiCSH4UZGN2bMGGOX0CV1137RNOjE1/Hx8Zg+fTocHR0RERHR\n7PLbt2/H0KFDcfbsWXFeREQEHB0doVQqxf90Ol2z65M0mRq7gK4iJCQEc+fOhbe3d4e/1/w/fY3K\nGv5DI2pO8pYg8bWNjQ0WLVqEjIwMqNXqJsveuHEDf//739GvX78mbfPmzcM777zTobVS18WRGxF1\nWf7+/vDz84O1tXWz7VFRUXj33XehUCg6uTLq6iQzcqurq8OKFStw7do1mJqa4he/+AWqqqrw+uuv\nw8/PDwCQmpqKjz/+GPv378e1a9cQGRmJ2tpaODg4NPqrMCQkBMOGDcP//d//oaqqCpMmTUJ4ePhj\n24io8xw7dgwKhQJeXl7Ntn/22Wf47LPPYG9vjwULFiAgIKCTKyRjkky4nT59GjU1NTh69CgAoKqq\nCqdOnUJiYqIYbgkJCXj11VcBAMuXL0dISAiCg4Nx4cIFzJo1q9H2CgoKcODAAajVasycORNKpVI8\nZdlSGxF1vOrqanzwwQf4+OOPm20PCQnBihUr0Lt3b5w5cwbLli3Dc889122vY3ZHkgm3YcOGoaCg\nAFFRUXB1dcWECRPg7++PDRs2oKKiAgBw7tw5REdHo7q6Gvn5+QgKun9u39nZGQ4ODo2298orr8DU\n1BSmpqaYPHkyMjMzxQBrqY2InlxOTk6j6eLiYpSXl4tt8fHxcHFxQUlJCUpKSqBWq5Gfnw9zc3Nx\nnR9++AEAYGlpCQ8PD8THx3feDhjBw33W3Ukm3AYOHIiUlBRkZmYiPT0dH3zwAZKTk+Hr64uUlBQA\ngK+vL3r27Inq6mojV0tELXl4hJWeng6ZTCa2rV27FsXFxTh16hQAoLy8HDt27MD8+fMRGhraZHtJ\nSUmwsLCQ7MgtJydHsvsGAGr1/0ggAAAJfklEQVS1Gnl5ea1aRzI3lBQXF0Mul8PPzw+RkZEoLy9H\nZWUlgoODcfjwYRw+fBjTp08HcP8vOQcHByQnJwMAcnNzkZ+f32h7SUlJ0Gq1qK2txbFjx+Du7m5Q\nGxG1H61WC7VaDb1eD51OB41GA61Wi7i4OKSkpCAxMRGJiYmwsbFBVFQUfve73wEAjh8/jpqaGuj1\nepw+fRpJSUnw8fEx8t5QZ5LMyO3KlSvYsmULAECv1yM0NBS2trawtbUVR2oqlUpcftOmTYiMjMTu\n3bvh4OCAkSNHNtre4MGDMXPmTPGmkZ+fdmypjYjaz86dO7F9+3ZxOikpCYsXL0ZYWFij5eRyOays\nrNCrVy8AwL59+/D+++9DEATY29tj3bp1cHNz69TaybhkgiAIxi6iq2npO29P8n24B0NrR0fHRtcG\niOg/NA06KMzkzbZJ/fRbW0m9X9ry2SmZ05L09OKF8OZ11355VLARtYZkTku2p/3797epjYiIugaO\n3IiISHIYbkREJDkMNyIikhyGGxERSQ7DjYiIJIfhRkREksNwIyIiyWG4ERGR5DDciIhIchhuREQk\nOQw3IiKSHIYbERFJDsONiIgkh+FGRESSw3AjIiLJYbgREZHkMNyIiEhyGG5ERCQ5DDciIpIchhsR\nEUkOw42IiCSH4UZERJLDcCMiIslhuBERkeQw3IiISHIYbkREJDkMNyIikhyGGxERSQ7DjYiIJIfh\nRkREksNwIyIiyWG4ERGR5DDciIhIchhuREQkOQw3IiKSHFNjF9CdCIIAANBoNEaupOtRq9XGLqFL\nYr80xT5pnpT7pS2fmTLhwScudbh79+4hPz/f2GUQET2VHB0dYW5ubtCyDLdOpNfrUVNTAzMzM8hk\nMmOXQ0T0VHgQU+bm5gZ/djLciIhIcnhDCRERSQ7DjYiIJIfhRkREksNwIyIiyWG4ERGR5DDciIhI\nchhuREQkOQy3TvKvf/0LM2bMQEBAAGbMmIEff/zR2CUZhY+PDyZOnIigoCAEBQUhIyMDAHDhwgUE\nBgYiICAAc+fORVlZmZEr7TjR0dHw8fHB0KFDG/1iTUvHSHc4fh7VL486ZgDpHzcVFRV46623EBAQ\ngGnTpmHx4sUoLy8H0PK+S71fDCJQpwgJCRESExMFQRCExMREISQkxMgVGYe3t7dw5cqVRvN0Op3g\n5+cnnD9/XhAEQYiJiREiIiKMUV6nOH/+vFBUVNSkL1o6RrrD8fOofmnumBGE7nHcVFRUCJmZmeL0\nxo0bhcjIyBb3vTv0iyE4cusEZWVl+Mc//oGpU6cCAKZOnYp//OMf4l9g3V1eXh7Mzc2hUqkAADNn\nzsTx48eNXFXHUalUsLOzazSvpWOkuxw/zfVLS7rDcWNtbQ03Nzdx2tnZGUVFRS3ue3foF0PwqQCd\n4Pbt27C1tYVcLgcAyOVy2NjY4Pbt2+jTp4+Rq+t87777LgRBwJgxYxAeHo7bt2/j+eefF9v79OkD\nvV6PyspKWFtbG7HSztPSMSIIQrc/fh4+Zp555plud9zo9Xp89tln8PHxaXHfu1u/PApHbtSpPv30\nUyQlJeHQoUMQBAFr1641dknUxfGYue+Pf/wjevbsiddff93YpTwVGG6dwM7ODiUlJdDpdAAAnU6H\nO3futOoUjFQ82GeFQoHXXnsN3333Hezs7FBUVCQuU15eDhMTk271V2ZLx0h3P36aO2YezO8ux010\ndDSuX7+Ov/zlLzAxMWlx37tTv7SE4dYJ+vbti+HDhyMlJQUAkJKSguHDh3ebU0oP1NbW4t69ewDu\nP8Li6NGjGD58OBwdHVFfX4/s7GwAwIEDBzBx4kRjltrpWjpGuvPx86hjBkC3OW62bt2KvLw8xMTE\nQKFQAGh537tLvzwOH3nTSQoKChAREYGffvoJzzzzDKKjozF48GBjl9Wpbt68ibCwMOh0Ouj1egwZ\nMgSrVq2CjY0NvvvuO6xZswZqtRoDBgzA5s2b8dxzzxm75A6xbt06fPXVVygtLcWzzz4La2trfPnl\nly0eI93h+GmuX2JjYx95zACQ/HFz9epVTJ06FYMGDYKFhQUAwN7eHjExMS3uu9T7xRAMNyIikhye\nliQiIslhuBERkeQw3IiISHIYbkREJDkMNyIikhyGGxERSQ5/W5LoKeXj44PS0lLxNycB4Pjx47C1\ntTViVURdA8ON6CkWGxuLsWPHGrUGrVYLU1N+lFDXwtOSRBJXXl6OBQsWQKVSwdXVFa+99hr0ej2A\n+08jWLx4Mdzd3eHm5ib+KLFer8eOHTvg7e0NDw8PLF++XPwZrFu3bmHo0KH4/PPPMWHCBLzxxhsA\n7j8gc+bMmVCpVAgMDERWVpZxdpgIHLkRSd4nn3wCW1tbfPvttwCAixcvQiaTQafTYcGCBXB3d8fJ\nkychl8tx6dIlAEBCQgIOHz6Mffv2oU+fPlixYgXWrl2LzZs3i9s9f/48jh49ChMTE5SUlGDBggXY\ntGkTXn75ZXz77bdYsmQJjh071i1+A5O6Ho7ciJ5iv//976FSqaBSqbBo0aJmlzE1NcXdu3dRVFQE\nMzMzqFQqyGQy5Obm4s6dO1i+fDl69uzZ6AGXycnJmDNnDgYOHIhevXohPDwcR48ehVarFbcbFhaG\nnj17wsLCAkeOHMH48ePh5eUFExMTeHp6wtHREWlpaZ3SD0QP48iN6CkWExPz2Gtu8+bNw/bt2zF3\n7lwAwIwZMxAaGio+1LK562V37tzBgAEDxOkBAwZAq9WirKxMnNe/f3/xdVFREY4fP47U1FRxnlar\nbfQUaaLOxHAjkjhLS0tEREQgIiIC+fn5eOONNzBy5EjY2dnh9u3bzd4QYmNjg8LCQnG6qKgIpqam\n6Nu3L4qLiwEAMplMbLezs0NQUBDWrVvXOTtF9Bg8LUkkcampqbh+/ToEQUDv3r0hl8shk8ng5OSE\nfv36YcuWLaitrYVarUZOTg4AYOrUqdi7dy9u3ryJmpoafPDBB5g0adIj74oMDAxEamoqMjIyoNPp\noFarkZWVJQYhUWdjuBFJ3PXr1/Hmm29CqVRixowZmDVrFtzd3SGXyxEbG4vr16/D29sb48ePx7Fj\nxwAAr776KgIDA/H666/D19cXCoUC//M///PI97Czs8OOHTvw4YcfwsPDA15eXvjoo4/EuzKJOhuf\n50ZERJLDkRsREUkOw42IiCSH4UZERJLDcCMiIslhuBERkeQw3IiISHIYbkREJDkMNyIikhyGGxER\nSc7/A4BhG5+WDBYkAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e2XratD1aD90",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pickle\n",
        "with open(\"/content/drive/My Drive/Netflix_recommender/tuned_result.pkl\",\"wb\") as f:\n",
        "    pickle.dump(tuned_models_evaluation_train,f)\n",
        "    pickle.dump(tuned_models_evaluation_test,f)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XRtfh-xaadIv",
        "colab_type": "code",
        "outputId": "33841ee6-b88d-42cd-9120-f7ebd5136f75",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "# Saving our TEST_RESULTS into a dataframe so that you don't have to run it again\n",
        "pd.DataFrame(tuned_models_evaluation_test).to_csv('drive/My Drive/Netflix_recommender/tuned_small_sample_results.csv')\n",
        "models = pd.read_csv('drive/My Drive/Netflix_recommender/tuned_small_sample_results.csv', index_col=0)\n",
        "models.loc['rmse'].sort_values()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "knn_bsl_m         1.0723215145319687\n",
              "knn_bsl_u          1.072409004806261\n",
              "svd               1.0724398179398351\n",
              "svdpp              1.073107287201817\n",
              "bsl_algo          1.0731636807255809\n",
              "xgb_knn_bsl_um    1.0743360765716838\n",
              "xgb_all_models    1.0750524068075662\n",
              "xgb_final         1.0761202825525833\n",
              "Name: rmse, dtype: object"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 87
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "llx_FBU4ec6N",
        "colab_type": "code",
        "outputId": "5837ac3e-cb3f-4285-e838-658fbd75047d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 111
        }
      },
      "source": [
        "df_25k_train = pd.DataFrame(data=tuned_models_evaluation_train)\n",
        "df_25k_train.drop(\"predictions\",inplace = True)\n",
        "df_25k_train"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>bsl_algo</th>\n",
              "      <th>xgb_knn_bsl_um</th>\n",
              "      <th>svd</th>\n",
              "      <th>svdpp</th>\n",
              "      <th>xgb_final</th>\n",
              "      <th>xgb_all_models</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>mape</th>\n",
              "      <td>9.02798</td>\n",
              "      <td>8.40416</td>\n",
              "      <td>30.3532</td>\n",
              "      <td>25.9604</td>\n",
              "      <td>20.0927</td>\n",
              "      <td>17.916</td>\n",
              "      <td>25.9542</td>\n",
              "      <td>35.9301</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>rmse</th>\n",
              "      <td>0.327805</td>\n",
              "      <td>0.323167</td>\n",
              "      <td>0.955256</td>\n",
              "      <td>0.864335</td>\n",
              "      <td>0.661692</td>\n",
              "      <td>0.611122</td>\n",
              "      <td>0.864208</td>\n",
              "      <td>1.08742</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     knn_bsl_u knn_bsl_m  bsl_algo  ...     svdpp xgb_final xgb_all_models\n",
              "mape   9.02798   8.40416   30.3532  ...    17.916   25.9542        35.9301\n",
              "rmse  0.327805  0.323167  0.955256  ...  0.611122  0.864208        1.08742\n",
              "\n",
              "[2 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 89
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P_E_v-42eqFr",
        "colab_type": "code",
        "outputId": "9ff87c72-3a59-48ff-926a-a04a1e279c01",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 111
        }
      },
      "source": [
        "df_25k_test = pd.DataFrame(data=tuned_models_evaluation_test)\n",
        "df_25k_test.drop(\"predictions\",inplace = True)\n",
        "df_25k_test"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>knn_bsl_u</th>\n",
              "      <th>knn_bsl_m</th>\n",
              "      <th>bsl_algo</th>\n",
              "      <th>xgb_knn_bsl_um</th>\n",
              "      <th>svd</th>\n",
              "      <th>svdpp</th>\n",
              "      <th>xgb_final</th>\n",
              "      <th>xgb_all_models</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>mape</th>\n",
              "      <td>34.9366</td>\n",
              "      <td>34.9281</td>\n",
              "      <td>34.9891</td>\n",
              "      <td>34.6877</td>\n",
              "      <td>34.9136</td>\n",
              "      <td>34.9061</td>\n",
              "      <td>34.4833</td>\n",
              "      <td>35.1817</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>rmse</th>\n",
              "      <td>1.07241</td>\n",
              "      <td>1.07232</td>\n",
              "      <td>1.07316</td>\n",
              "      <td>1.07434</td>\n",
              "      <td>1.07244</td>\n",
              "      <td>1.07311</td>\n",
              "      <td>1.07612</td>\n",
              "      <td>1.07505</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     knn_bsl_u knn_bsl_m bsl_algo  ...    svdpp xgb_final xgb_all_models\n",
              "mape   34.9366   34.9281  34.9891  ...  34.9061   34.4833        35.1817\n",
              "rmse   1.07241   1.07232  1.07316  ...  1.07311   1.07612        1.07505\n",
              "\n",
              "[2 rows x 8 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 90
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vheIp8ZURGmi",
        "colab_type": "text"
      },
      "source": [
        " ## **Note:**\n",
        " ### All the results are computed on 25k data points"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tRMeajT3QkzY",
        "colab_type": "text"
      },
      "source": [
        "## Observations\n",
        "\n",
        "1. We have got slight better scores as compared to scores on 10k data.\n",
        "2. Compute time has increased drasticaly as we have increased points.\n",
        "3. We have plot feature_importance for each models that helps in understanding features.\n",
        "4. **knn_bsl_u and knn_bsl_m** have got slightly better scores as compared to other models.\n",
        "5. Surprise library features yeild similar as compared with other Self computed features."
      ]
    }
  ]
}