{ "metadata": { "name": "", "signature": "sha256:4c5c82048733cdbf75467f019dd99a78107b5c684747f698fcf7390509004f84" }, "nbformat": 3, "nbformat_minor": 0, "worksheets": [ { "cells": [ { "cell_type": "heading", "level": 1, "metadata": {}, "source": [ "Introduction to scikit-learn" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`scikit-learn` is a Machine Learning library for Python, in this tutorial we will learn about the `scikit-learn` interface walking through a regression problem, in details:\n", "\n", "* load a dataset in `.csv` format with `pandas`\n", "* preprocess the data to make them suitable for `scikit-learn`\n", "* fit a Decision Tree to the data\n", "* compare the predicted and true values\n", "* plot the learning curve\n", "* use cross validation to tune model parameters" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "First install `scikit-learn` with `conda` (recommended) or `pip`:\n", "\n", " conda install sklearn\n", "\n", "or:\n", "\n", " pip install sklearn" ] }, { "cell_type": "code", "collapsed": false, "input": [ "%matplotlib inline\n", "import matplotlib.pyplot as plt\n", "import numpy as np\n", "import pandas as pd" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 1 }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "The dataset: predict Abalone age" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Abalone is a mollusc with a peculiar ear-shaped shell lined of mother of pearl.\n", "Its age can be estimated counting the number of rings in their shell with a microscope, but it is a time consuming process, in this tutorial we will use Machine Learning to predict the age using physical measurements." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Picture of the shell of an Abalone:\n", "![Abalone](http://upload.wikimedia.org/wikipedia/commons/0/0b/AbaloneInside.jpg)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The dataset is available from the [University of California Irvine Machine Learning data repository](http://archive.ics.uci.edu/ml/datasets/Abalone), we can download the [data in Comma-Separated Values format](http://archive.ics.uci.edu/ml/machine-learning-databases/abalone/abalone.data) in the current folder." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "There are no column labels in the data, so we copy them from the documentation and use `pandas` to read and print few lines of the dataset." ] }, { "cell_type": "code", "collapsed": false, "input": [ "column_names = [\"sex\", \"length\", \"diameter\", \"height\", \"whole weight\", \n", " \"shucked weight\", \"viscera weight\", \"shell weight\", \"rings\"]\n", "data = pd.read_csv(\"abalone.data\", names=column_names)\n", "print(\"Number of samples: %d\" % len(data))\n", "data.head()" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "Number of samples: 4177\n" ] }, { "html": [ "
\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
sexlengthdiameterheightwhole weightshucked weightviscera weightshell weightrings
0 M 0.455 0.365 0.095 0.5140 0.2245 0.1010 0.150 15
1 M 0.350 0.265 0.090 0.2255 0.0995 0.0485 0.070 7
2 F 0.530 0.420 0.135 0.6770 0.2565 0.1415 0.210 9
3 M 0.440 0.365 0.125 0.5160 0.2155 0.1140 0.155 10
4 I 0.330 0.255 0.080 0.2050 0.0895 0.0395 0.055 7
\n", "

5 rows \u00d7 9 columns

\n", "
" ], "metadata": {}, "output_type": "pyout", "prompt_number": 2, "text": [ " sex length diameter height whole weight shucked weight viscera weight \\\n", "0 M 0.455 0.365 0.095 0.5140 0.2245 0.1010 \n", "1 M 0.350 0.265 0.090 0.2255 0.0995 0.0485 \n", "2 F 0.530 0.420 0.135 0.6770 0.2565 0.1415 \n", "3 M 0.440 0.365 0.125 0.5160 0.2155 0.1140 \n", "4 I 0.330 0.255 0.080 0.2050 0.0895 0.0395 \n", "\n", " shell weight rings \n", "0 0.150 15 \n", "1 0.070 7 \n", "2 0.210 9 \n", "3 0.155 10 \n", "4 0.055 7 \n", "\n", "[5 rows x 9 columns]" ] } ], "prompt_number": 2 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Sex is either *Male* (\"M\"), *Female* (\"F\") or *Infant* (\"I\"), this is not suitable for regression algorithms, so we create a binary/boolean feature for each of the 3 options:" ] }, { "cell_type": "code", "collapsed": false, "input": [ "# for more complicated cases use sklearn.feature_extraction.DictVectorizer\n", "for label in \"MFI\":\n", " data[label] = data[\"sex\"] == label\n", "del data[\"sex\"]" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 3 }, { "cell_type": "code", "collapsed": false, "input": [ "data.head()" ], "language": "python", "metadata": {}, "outputs": [ { "html": [ "
\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
lengthdiameterheightwhole weightshucked weightviscera weightshell weightringsMFI
0 0.455 0.365 0.095 0.5140 0.2245 0.1010 0.150 15 True False False
1 0.350 0.265 0.090 0.2255 0.0995 0.0485 0.070 7 True False False
2 0.530 0.420 0.135 0.6770 0.2565 0.1415 0.210 9 False True False
3 0.440 0.365 0.125 0.5160 0.2155 0.1140 0.155 10 True False False
4 0.330 0.255 0.080 0.2050 0.0895 0.0395 0.055 7 False False True
\n", "

5 rows \u00d7 11 columns

\n", "
" ], "metadata": {}, "output_type": "pyout", "prompt_number": 4, "text": [ " length diameter height whole weight shucked weight viscera weight \\\n", "0 0.455 0.365 0.095 0.5140 0.2245 0.1010 \n", "1 0.350 0.265 0.090 0.2255 0.0995 0.0485 \n", "2 0.530 0.420 0.135 0.6770 0.2565 0.1415 \n", "3 0.440 0.365 0.125 0.5160 0.2155 0.1140 \n", "4 0.330 0.255 0.080 0.2050 0.0895 0.0395 \n", "\n", " shell weight rings M F I \n", "0 0.150 15 True False False \n", "1 0.070 7 True False False \n", "2 0.210 9 False True False \n", "3 0.155 10 True False False \n", "4 0.055 7 False False True \n", "\n", "[5 rows x 11 columns]" ] } ], "prompt_number": 4 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Data in `sklearn` are structured as:\n", "\n", "* a 2D `numpy` array (n_samples, n_features), where each column is a feature (e.g. sex and physical dimensions) for each sample, e.g. each measured Abalone.\n", "* a 1D `numpy` array (n_samples) of the value we aim to predict, e.g. the number of rings, standard variable name `y`" ] }, { "cell_type": "code", "collapsed": false, "input": [ "y = data.rings.values" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 5 }, { "cell_type": "code", "collapsed": false, "input": [ "del data[\"rings\"] # remove rings from data, so we can convert all the dataframe to a numpy 2D array.\n", "X = data.values.astype(np.float)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 6 }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Split data in Training and Test sets" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can split the data into training and validation sets and use Machine Learning to create an estimator that can learn from the training set and then check its performance on the test set." ] }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn import cross_validation\n", "train_X, test_X, train_y, test_y = cross_validation.train_test_split(X, y) # splits 75%/25% by default" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 7 }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Fit a Decision Tree to the data" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "One of the simpler method is a Decision Tree, see this example:\n", "![decision_tree_example.png](files/decision_tree_example.png)\n", "A Decision Tree learns from available data the best conditions to set on the features, e.g. the purpose and the programming language, to predict an output value, e.g. what format should be used." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The `DecisionTreeRegressor` is a similar algorithm used to estimate a continous variable instead of a discrete one:" ] }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn.tree import DecisionTreeRegressor\n", "# create an estimator, optionally specifying parameters\n", "model = DecisionTreeRegressor()\n", "# fit the estimator to the data\n", "model.fit(train_X, train_y)\n", "# apply the model to the test and training data\n", "predicted_test_y = model.predict(test_X)\n", "predicted_train_y = model.predict(train_X)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 8 }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can visualize the results with a scatter-plot of the true number of rings against the predicted number of rings:" ] }, { "cell_type": "code", "collapsed": false, "input": [ "def scatter_y(true_y, predicted_y):\n", " \"\"\"Scatter-plot the predicted vs true number of rings\n", " \n", " Plots:\n", " * predicted vs true number of rings\n", " * perfect agreement line\n", " * +2/-2 number dotted lines\n", "\n", " Returns the root mean square of the error\n", " \"\"\"\n", " fig, ax = plt.subplots(figsize=(8, 8))\n", " ax.plot(true_y, predicted_y, '.k')\n", " \n", " ax.plot([0, 30], [0, 30], '--k')\n", " ax.plot([0, 30], [2, 32], ':k')\n", " ax.plot([2, 32], [0, 30], ':k')\n", " \n", " rms = (true_y - predicted_y).std()\n", " \n", " ax.text(25, 3,\n", " \"Root Mean Square Error = %.2g\" % rms,\n", " ha='right', va='bottom')\n", "\n", " ax.set_xlim(0, 30)\n", " ax.set_ylim(0, 30)\n", " \n", " ax.set_xlabel('True number of rings')\n", " ax.set_ylabel('Predicted number of rings')\n", " \n", " return rms" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 9 }, { "cell_type": "code", "collapsed": false, "input": [ "scatter_y(train_y, predicted_train_y)\n", "plt.title(\"Training data\")\n", "scatter_y(test_y, predicted_test_y)\n", "plt.title(\"Test data\");" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAH4CAYAAAC1/7+LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n", "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAH4CAYAAAC1/7+LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEzgREZEN+n8n2sqwm59T6AAAAABJRU5ErkJggg==\n", "text": [ "" ] } ], "prompt_number": 10 }, { "cell_type": "markdown", "metadata": {}, "source": [ "The Decision Tree overfits the training set, i.e. its parameters are fine tuned to reproduce the results of the training set but generalized badly to data not seen previously.\n", "\n", "To prevent this issue we specify a maximum depth of the decision tree of $10$, so that the estimator does not \"specialize\" too much on the training data." ] }, { "cell_type": "code", "collapsed": false, "input": [ "model = DecisionTreeRegressor(max_depth=10)\n", "# fit the estimator to the data\n", "model.fit(train_X, train_y)\n", "# apply the model to the test and train data\n", "predicted_test_y = model.predict(test_X)\n", "predicted_train_y = model.predict(train_X)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 11 }, { "cell_type": "code", "collapsed": false, "input": [ "scatter_y(train_y, predicted_train_y)\n", "plt.title(\"Training data\")\n", "rms_decision_tree = scatter_y(test_y, predicted_test_y)\n", "plt.title(\"Test data\");" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAH4CAYAAAC1/7+LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3XlcVPX+P/DXLCAgu+KCuJaAioaK4gq4UEGRdv1iZaap\nRVrZNW+Ztqh1M72pt+3mVXP9ZlaaZuZS17qgggvivqS4sJgOKgooChyGOb8/+M78BgRmzixnBng9\nHw8eD2fmfM77M8PI+3w+57MoRFEUQURERPWK0tEVICIiIumYwImIiOohJnAiIqJ6iAmciIioHmIC\nJyIiqoeYwImIiOohJnCiei4+Ph5ff/21zY+1llKpxKVLl2SJRdQYKTgPnEh+np6eUCgUAIC7d+/C\nzc0NKpUKALB8+XI888wzjqyeTSiVSly4cAGdOnWq87js7Gx06tQJWq0WSiXbFETmUju6AkSNUXFx\nseHfHTt2xMqVKzF06ND7jtNqtVCrG8d/U7YliKTh5S6RE0lJSUFQUBA+/vhjtG7dGpMmTUJhYSEe\nf/xxtGjRAv7+/khISMCVK1cMZWJiYrBy5UoAwJo1azBo0CC8+eab8Pf3R6dOnfDLL79YdGxWVhai\noqLg7e2N2NhYvPLKK3juuedqrfvChQsRGBiIoKAgrFq1qspr27dvR8+ePeHj44N27drh/fffN7wW\nFRUFAPD19YWXlxcOHjyIixcvYujQoWjevDkCAgIwduxYFBUVWfHJEjU8TOBETubatWsoKChAbm4u\nli1bBp1Oh0mTJiE3Nxe5ublwd3fHq6++ajheoVAYuuMBID09HaGhobh58yZmzJiBSZMmWXTsmDFj\n0K9fP9y6dQtz587FunXrqpQ19ssvv2Dx4sX47bffkJmZid9++63K656enli3bh2Kioqwfft2/Pvf\n/8ZPP/0EANi7dy8AoKioCHfu3EFkZCQA4J133oFGo8Eff/yBy5cvY+7cuRZ+okQNExM4kZNRKpV4\n//334eLiAjc3N/j7++PJJ5+Em5sbPD098fbbb2P37t21lm/fvj0mTZoEhUKBcePGQaPR4Pr165KO\nzc3NRUZGBj744AOo1WoMHDgQTzzxRK3d3Bs2bMDEiRPRtWtXeHh4VGlhA0B0dDS6desGAOjevTue\nfvppw3uo6ZwPPPAAhg0bBhcXFzRv3hyvv/56ne+ZqDFiAidyMgEBAXB1dTU8vnfvHl566SV06NAB\nPj4+iI6ORlFRUa3JtFWrVoZ/e3h4AKh6z92cY69evQp/f3+4ubkZXm/btm2tddZoNFVeb9euXZXX\nDx48iCFDhqBFixbw9fXFsmXLcPPmzVrPd+3aNTz99NMICgqCj48PnnvuuTqPJ2qMmMCJnEz1burF\nixcjMzMT6enpKCoqwu7duyGKol0HfbVu3Rq3bt1CSUmJ4bnc3Nw6jzd+vfqxY8aMwciRI/Hnn3+i\nsLAQkydPhk6nA3D/+wWAt99+GyqVCqdOnUJRURG+/vprw/FEVIkJnMjJFRcXw93dHT4+Prh169Z9\n3dP20L59e0RERGDu3LkoLy/H/v37sW3btlrvgY8ePRpr1qzBH3/8gXv37t1Xx+LiYvj5+cHV1RXp\n6elYv3694VwBAQFQKpW4ePFileObNm0Kb29vXLlyBQsXLrTfmyWqp5jAiZxM9SQ5bdo0lJSUoHnz\n5hgwYADi4uJqTaTVB6nVdD5zj/3mm2+wf/9+NGvWDO+99x6eeuqpKl37xh599FFMmzYNQ4cORXBw\nMIYNG1blXEuWLMHs2bPh7e2Nv//973jqqacMr3l4eOCdd97BwIED4e/vj/T0dMyZMwdHjhyBj48P\nEhISMGrUqFrfB1FjZbeFXEpLSxEdHY2ysjIIgoARI0Zg/vz5uHXrFp566ink5OSgQ4cO2LBhA3x9\nfe1RBSKyoaeeegpdu3bFnDlzHF0VIoIdW+Bubm5ITk7GsWPHcOLECSQnJyM1NRULFixAbGwsMjMz\nMWzYMCxYsMBeVSAiK2RkZODixYvQ6XTYuXMntm7dipEjRzq6WkT0f+zaha4f1SoIAioqKuDn54et\nW7di/PjxAIDx48djy5Yt9qwCEVkoLy8PQ4YMgZeXF15//XUsXboUDz30kKOrRUT/x65roet0OvTq\n1QsXL17ElClT8PHHH8PPzw8FBQUAKud/+vv7Gx4TERGReey6yLJSqcSxY8dQVFSERx55BMnJyVVe\nr2kQDQA8+OCDVUakEhERNXQPPPAALly4YPbxsoxC9/HxwWOPPYbDhw+jZcuWyMvLA1C5+EOLFi3u\nO/7ixYuGea78qftnzpw5Dq9Dffjh58TPip8TPydTPzqdDuvWrUNZWZlsMa9evYr+/fsjLy9PcsPV\nbgk8Pz8fhYWFAICSkhLs2rULPXv2xBNPPIG1a9cCANauXctBMURE5BQUCgWeffbZWqdL2ppGo8GQ\nIUPw+OOPo2XLlpLL260LXaPRYPz48dDpdNDpdHjuuecwbNgw9OzZE6NHj8bKlSsN08iIiIgcJSkp\nCS+++CL69OkjW0x98h43bhzefvtti85htwTevXt3HDly5L7n/f3979upiCwXExPj6CrUC/yczMfP\nyjz8nMxTHz6nqVOnIiQkRLZ4tkjegJ1HoVtKoVDACatFREQNRHl5OdRqtUNW+Nu4cSMuXLiAWbNm\nVXleau5jAiciokbn3XffRfv27fHiiy86uioGTOBEREQm3L59G25ubrINWDMHEzgREVENBEHA7du3\n0bx5c0dXpUZScx93IyMiokZhy5Yt+PDDD2WNqdFokJGRYZdzswVORESNRkVFBVQqlSyx9KPNX3zx\nRfztb38zeTxb4ERERP9HEAQcPHjQ8Fju5D1u3DizkrclmMCJiKjBunjxIr766itZY9pqnrcp7EIn\nIiKykfLycoSHh+PZZ5+VnLw5Cp2IiBo1QRCwatUqJCUlQamUv6P5zJkz6Nq1q+RyvAdORESNWnl5\nOa5evQqtVuuQ+JYkb0uwBU5EROQE2AInIqJGR6fTYdq0abh27ZqscR3Z2GQCJyKiek+pVCIqKgp+\nfn6yxdRoNOjbty+uXLkiW0xjTOBERNQg/OUvf5FtbXP9VLEnn3wSbdq0kSVmdUzgRERUb02ePBl7\n9+6VNaZc87xN4SA2IiKqtzIzM9GhQwfZW972SN6cB05ERA2aIAhQqVSyLYtqbNu2bTh16hRmzpxp\n83MzgRMRUYP297//Hb6+vpg6daqjq2JTTOBERNSg3bt3D2q1WrZuc7lwHjgRETU4giBAo9EAADw8\nPBpc8rYEEzgRETm9X3/9FfPmzZM1pkajQWpqqqwxpWAXOhER1Qs6nU62zUn0o80nTZqEN998U5aY\n7EInIqIGQRAE7N692/BY7uQ9btw42ZK3JZjAiYjIKV2+fBnffPONrD2yjlqkxZL3yC50IiIiAFqt\nFr169cLTTz8ta/LOycnBpEmT8Pvvv3MaGRER1U+CIGDJkiV49dVXoVarZY9/4cIFPPjgg7LGFEUR\nFy9eROfOnXkPnIiI6iedTod79+5Bp9M5JL6cybu4uBhAZaPVkrhsgRMREcmstLQUERERSEtLg4+P\nDwCOQicionpGFEW88soruHz5sqxxKyoqZI1nzM3NDQcPHjQkb0swgRMRkUMpFAo89thjaNmypWwx\nNRoNevfujaysLNliAsD58+cNreymTZtadS4mcCIicgjj7uL4+HjZtwQdPXo0OnbsKEtMoPL9vvHG\nG/jjjz9scj7eAyciIoeYMmUKRo0aheHDh8sW01HzvPVEUYRCoajxNe5GRkRE9UJ2djYCAwNlb3nL\nnbwPHDiADh06oFWrVnUeJzX3yT/JjoiIGi1BEAAArq6u6NChg6yx//jjD0ycOBEzZsyQNW5aWhrK\nyspMJnCp2AInIiLZLFy4EACceo1xR2EXOhEROS3jFnhDlpqaisLCQjz++ONml+E8cCIiciqCICAn\nJwdAZeJu6MkbqHyfbm5udo3BFjgREdnVrl27sGnTJixdulS2mBqNBqdOnUJsbKxsMa3FLnQiInI6\ndU2fsrWxY8di8+bNaNu2LQ4ePAhfX19Z4h4+fBjff/89Pv74Y4vKswudiIgcThAE/Prrr4bHciVv\njUaDzZs3o6SkBJmZmUhKSpIlLgCEhITgySeflC0eEzgREdnctWvX8OOPP8ram6qf5922bVsAQERE\nBJYvX273uPqd0zw9PdG/f3+7x9NjAiciIptr27Ytli5dKlvLW6fT4bHHHsO4ceNw8OBBJCYmYteu\nXXbvPr9y5QqioqIcsjEK74ETEZFNCIKATz75BNOmTUOTJk1kj5+Tk4P27dvX27i8B05ERA6hUCig\nVqtla3VXJ2fyLiwsdEhcY2yBExERSSAIAiIiIpCcnIxmzZrZ7LxsgRMRkWxEUcSUKVNw6dIlWeNq\ntVpZ4xlzdXVFenq6TZO3JZjAiYjIYgqFAomJiQgKCpItpkajQc+ePZGZmSlbTKByMxT9YDV7r7Jm\nDiZwIiKSzLird+jQobJvCfrMM88gODhYlph6H3zwAU6dOiVrzLrwHjgREUk2depUPPzww0hISJAt\npqP289az92pyXEqViIjs7sqVKwgICJC95S138t6/fz8CAwNlGWkuNfep7VgXIiJqQARBQEVFBdzd\n3dGmTRtZY2dlZWHSpEmy7yN+7NgxlJSUOGyqWF3YAiciIrN88cUXKCwsxHvvvefoqjRI7EInIiK7\nqKioQEVFRYPfzzstLQ1XrlzB6NGjZY3LeeBERGQzgiDg/PnzAACVSiVb8k5KSkJMTAzi4+OrrHom\nBx8fH/j7+8sa0xJsgRMRUa327t2Lr7/+WpZdvfQ0Gg1iY2Nx+vRpAEBiYiI2bNggW3xHYRc6ERHZ\nlL2nTxnTjzYXRRGZmZmIiIiQZVex48ePY/ny5fjyyy/tGqcu7EInIiKrCIKALVu2GB7Lnbzl3hIU\nAEJCQjB+/Hi7x7ElJnAiIqri1q1b+P3336HT6WSLWX2et6+vLzZs2GD35G28NGrfvn3tGsvW2IVO\nREQOJYoiIiMjMXLkSFkXabl+/Toee+wx7N+/H2q145dF4T1wIiKSrLy8HB999BHeeOMNNG3aVPb4\nV69eRWBgYKOJWxPeAyciIslUKhWaN28OFxcXh8SXM4nevHnTIXFtjS1wIiJqNLRaLfr27YudO3ei\nZcuWjq5OFWyBExGR2ZKSknD27FlZYwqCIGs8Y2q1GgcOHHC65G0JJnAiokbs+eefR6dOnWSLp9Fo\nEB4eLvu+2idPnjRcODSUpWCZwImIGhn91CkAGDBggOxbgo4dOxZhYWGyxNT77LPPcPLkSVlj2pvd\nEvjly5cxZMgQdOvWDWFhYfj8888BAHPnzkVQUBB69uyJnj174pdffrFXFYiIqAZvvPEGNm7cKGtM\nR+3nrbdixQr07t1b9rj2ZLdBbHl5ecjLy0N4eDiKi4vRu3dvbNmyBRs2bICXlxemT59ee6U4iI2I\nyG6uX78OX19f2VvecifvtLQ0BAQEIDg4WLaY1pCa++w2c71Vq1Zo1aoVAMDT0xNdunTBlStXAIDJ\nmYhIZoIgoKysDF5eXmjRooWssfPy8pCUlFRnw80eLl68iNLS0nqTwKWSZRpZdnY2oqOjcfr0aSxe\nvBirV6+Gj48PIiIisHjx4vuWymMLnIjItpYvX47c3Fx8+OGHjq4K1cLpVmIrLi5GTEwM3n33XYwc\nORLXr19HQEAAAOC9996DRqPBypUrq1ZKocCcOXMMj2NiYhATE2PPahIRNWg6nQ4VFRUOW6hFLmlp\nacjMzMSECRMcXRWTUlJSkJKSYnj8/vvvO08CLy8vx+OPP464uDhMmzbtvtezs7ORkJBw38hAtsCJ\niKwnCALOnz+Pbt26Oboqsrlw4QJyc3MxdOhQR1dFMqdZyEUURUyaNAldu3atkrw1Go3h3z/++CO6\nd+9uryoQETVqx48fx6effiprTI1Gg++++07WmMYefPDBepm8LWG3FnhqaiqioqLQo0cPw16yH330\nEb799lscO3YMCoUCHTt2xLJly+5bEYctcCKi+kc/2nz8+PGYNWuWbHFPnTplGF9VnzndPXBLMIET\nEVlGEARs3rwZTz31lKHxJAdHzvMuLy/HyZMn0atXL1nj2prTdKETEZH87ty5g/T09Cqrrdmbo5J3\neXk5AMDFxaXeJ29LsAVOREQWE0URgwcPRnx8vKzJ+9atWxg6dCjS09MbzNrm7EInImpktFot5syZ\ngxkzZsDHx0f2+Ddu3DBMD24Mce2FXehERI2MSqVCp06d4O7u7pD4cibRvLw8h8R1RmyBExFRvVBR\nUYEBAwZg8+bNaNOmjaOrY3NsgRMRNRJJSUk4fvy4rDFLSkpkjWdMpVIhLS2tQSZvSzCBExHVUy+9\n9BK6dOkiWzyNRoOePXvi8OHDssUEgKNHjxouHNRqu+3BVe8wgRMR1SNardbQzdq7d2+HbAkq977a\na9aswYkTJ2SNWR/wHjgRUT0yc+ZMdO3aFePGjZMtpiMXaWlMOI2MiKgBKygoQNOmTWVteXfo0AEK\nhQJeXl7IyMhA+/bt7R43LS0Nnp6eeOihh+wey1lwEBsRUQMjCAIKCgoAAH5+frIuXFJQUAClUomy\nsjLk5+dj0KBBssTNy8vDjRs3ZIlVX7EFTkTk5L7++mucOHECCxcudEj8gIAA5Ofnw8PDA2fOnJGl\nBd4YsQudiKiBEUURFRUVDhuBnZOTg0GDBiE1NdWuyXv//v04fPgwXn31VbvFcGZScx/H4xMROSFB\nEHDq1Cn06tULCoXCodOn2rdvj8uXL9s9TlBQEEpLS+0ep6HgPXAiIid09uxZLFmyRNaYGo0Gq1at\nkjWmsbZt22LIkCEOi1/fMIETETmhHj16YMWKFbLF008VM15rXA5nz57F6NGjedvUAkzgREROQhAE\nrF69WvZk5sh53sHBwXj33XehUChkjdsQMIETETmJkpISZGZmQqvVyhbTUcm7rKwMAKBUKtGjRw/Z\n4jYkHIVORNSIDR06FMOHD5c1ed++fRsDBw5Eenq6w7ZAdUacRkZEVI9UVFTgrbfewqxZs9CsWTPZ\n49+6dQv+/v6NJq4z40psRET1iFKpRK9eveDl5eWQ+HIm0T///NMhcRsqtsCJiMjuRFFEdHQ01q5d\ni44dOzq6Ok6JLXAionogKSkJ6enpssa8e/eurPGMKRQKpKSkMHnbEBM4EZEDvPbaawgPD5ctnkaj\nQe/evbFv3z7ZYgLA4cOHcfv2bQCVtwvIdvhpEhHJpLy8HDqdDgAQFhYm65aggYGBOHfuHAYOHIid\nO3fKEhcANm3ahJMnT8oWrzHhPXAiIpm89957CAoKwksvvSRbTP0873PnzlV5nn9jnQ/vgRMROakZ\nM2ZgwoQJssUzXqTF2I4dO+waNzU1Ffv377drDGICJyKyK0EQcOPGDQCAl5eXbN3mQOXKbq+++ire\nfvttQ9LesWMH4uLi7Bq3uLgYxcXFdo1B7EInIrKrH374Aampqfj0008dXRVyclyJjYjIyVRUVECl\nUjm6GnZ14MAB7NmzBzNmzHB0Veot3gMnInIwQRCq3ANu6MkbADp16oR+/fo5uhqNChM4EZGNXbp0\nCatWrZI1pkajwRdffCFrTOD/j2Zv0aIFoqKiZI/fmDGBExHZWGhoKL766ivZ4ulHm9+5c0e2mABw\n4cIFJCQk8JangzCBExHZgCAIWLJkiWGhFrk4aj9vAHjggQewcOFCKBQKWeNSJSZwIiIbKC8vx/Xr\n16HVamWL6ajkXVJSAqBy0FWXLl1ki0tVcRQ6EVE9FR8fj0GDBsmavO/evYvIyEgcOHAAnp6essVt\nDDiNjIhIJjqdDq+//jpmzZqFVq1ayR7/9u3b8Pb2bjRxGzpOIyMikolSqURMTAz8/f0dEl/OJJqV\nlWVILkzezoEJnIhIIuNW0pNPPinb8qiurq5QKBRQKpVITU2VJSZQ+X4nT56M8+fPyxaTTGMXOhGR\nRJMnT8aYMWNknfdcVFQEX19fw2OlUomKigrZ4ouiyNHmdsYudCIiO5s+fbqsq45pNBpERkZWeW73\n7t12j5ueno6bN28CAJO3E2ICJyIyQ3l5uaHFGxwcLFu3ufFUsb1790KpVGLv3r0YNGiQ3WP/5z//\nwZkzZ+wehyzDLnQiIjPMmzcPnp6e+Otf/ypbTEcu0kLy4zQyIiI7KCkpgUqlckjLW87knZqairKy\nMgwbNky2mFRJau5T27EuRET1miAIyM/PR2BgINzd3WWNLYoipk2bhsmTJ8saV6fTyb4cLFmGLXAi\nolps374d27dvx5IlSxxdFWoE2IVORGRDOp0OSmXDHu976NAhbNu2De+//76jq9KocRoZEZEVBEFA\ncnKy4XFDT94A0LlzZ8TGxjq6GiRRw/9mEhFJ8Oeff+Lbb7+VtRdQo9HgH//4h+w9j/p4vr6+skxL\nI9tiAiciMtKpUycsX75ctoVL9KPNKyoqZF0sJScnB8OHD+eAtXqMCZyIGj1BEPDJJ5+gvLxc1riO\nnOfdrl07LFu2rFHcImio+JsjokZPFEWUlpbK3m3uiORdXFwMoHLA1IMPPihbXLI9jkInInKAJ598\nEn369JE1eZeWliIiIgKpqalVNkYh58BR6EREZhBFEa+88gouX74sa1ylUgmFQoEtW7agZ8+essZ2\nc3PDwYMHmbwbCCZwImqUFAoFEhIS0LJlS1njGrew4uPjZYmZmZlpGKzWtGlTWWKS/TGBE1GjYpxA\nH330UdnWNq/Jjh07ZInz1ltv4Y8//pAlFsmHCZyIGpWXX34Zu3btkjXmrVu3DC1gfdLesWMH4uLi\nZIm/efNmdOvWTZZYJB8OYiOiRiU7OxuBgYGy7yr26aef4tFHH5UlJgAcOHAA7du3R+vWrWWLSdbh\nbmRERNUIggAAcHV1RYcOHWSLazxVTM7kDQD79+9HaWkpE3gDxhY4ETV4ixYtgk6nw4wZM2SL6chF\nWqh+4m5kRETVGLfA5eCo5J2WloabN2/iiSeekC0m2Q670ImIUJm0r169ig4dOsg+0tzFxQVvvPEG\nXnjhBVnjNmnSBO7u7rLGJMdhC5yIGqTff/8dGzduxNKlSx1dFSKzsAudiOj/iKIo6w5fjnDkyBGs\nX78eixYtcnRVyEpcSpWIGi1BELBz507D44aevAEgJCQEo0aNcnQ1yAHslsAvX76MIUOGoFu3bggL\nC8Pnn38OoHJBg9jYWAQHB+Phhx9GYWGhvapARI3M9evX8fPPP8u+q9js2bNl7zU0Xhq1f//+ssYm\n52C3BO7i4oJPPvkEp0+fxoEDB/Dll1/ijz/+wIIFCxAbG4vMzEwMGzYMCxYssFcViKiRCQoKwpIl\nS2RreetHm7u5ucna2r969SoGDx6MiooK2WKS87FbAm/VqhXCw8MBAJ6enujSpQuuXLmCrVu3Yvz4\n8QCA8ePHY8uWLfaqAhE1AoIgYP78+SgtLZU1riPneQcGBuLbb7+FSqWSNS45F1mmkWVnZ+Po0aOI\njIzEtWvXDLv/tGzZEteuXZOjCkTUQCmVSri5uUGplG9Ij4uLC7RaLQAgKipKtriFhYWGrUDbtWsn\nW1xyTnZP4MXFxRg1ahQ+++wzeHl5VXlNoVDU2u00d+5cw79jYmIQExNjx1oSUX2lVqvx+uuvyxpT\nn7wBIDo6WpaubEEQEBUVheTkZDRr1szu8cj+UlJSkJKSYnF5u04jKy8vx+OPP464uDhMmzYNABAa\nGoqUlBS0atXK0AV19uzZqpXiNDIiqoMoipgyZQrefPNNPPDAA7LHN2547N27F4MGDZIlbmlpKdzc\n3GSJRfJzmmlkoihi0qRJ6Nq1qyF5A8ATTzyBtWvXAgDWrl2LkSNH2qsKRNRAKRQKPP3002jbtq1D\n4u/duxdKpVKW5H3mzBlDC5/Jm4zZrQWempqKqKgo9OjRw3C1On/+fPTt2xejR49Gbm4uOnTogA0b\nNhju6RgqxRY4UaORlJSEzMxMeHh4YP369ff9PTDmLAuzSKlzdaGhocjLy4OLiwsyMjLQvn37Oo9/\n9tlnMWPGDDz00EPWVpucHFdiI6J6JSYmBrt37wYAJCYmYsOGDbUe+9prr2H48OGybtaRn58PPz+/\nKiO+pdS5Ol9fXxQVFQGonPZ2+fJl21aY6i2n6UInIjKHh4cHACAiIgLLly+v89iZM2fKuq+2RqPB\noEGDsG3btirPS6lzdS4uLoZzpKam1njMvn37kJ2dLb3C1KgwgRORQ61fvx6JiYnYtWtXjV3RgiCg\npKQEQOX8Z0dsCTpixAhJda5LRkYGgoKCcObMmVq7z0+ePImsrCyL606NA7vQicip/etf/8KtW7cw\ne/Zs2WI6cpEWarx4D5yIGpSKigpUVFQ4pOUtZ/JOTU3Fn3/+iaefflq2mORcpOY+WVZiIyKSQhAE\nZGdnIzg4GCqVStYlQz08PDBr1izDks9y8fPzQ1lZmawxqX5jC5yInE5aWhrWrFmDr776ytFVIZKN\nzbvQi4uL4e7uDpVKhXPnzuHcuXOIi4szjKS0ByZwInKWOd/2dPz4cSxduhT//ve/HV0VcgI2n0YW\nFRWFsrIyXLlyBY888gi+/vprPP/889bUkYjoPoIg4McffzQ8bujJG6hc1GXixImOrgbVUyYTuCiK\n8PDwwObNm/Hyyy9j48aNOHXqlBx1I6JGpKCgAP/973+h0+lki6nRaPDmm2/KGhOAYWnUJk2aoE+f\nPrLGpobDrEFs+/fvxzfffIOVK1cCgOxfdiJquKovSyplW1BrljQ1buF37NgRL7/8sixxn3vuOWzZ\nsgUDBw7Ed999J3keOZGBaEJKSoqYkJAgLliwQBRFUbxw4YI4depUU8WsYka1iKgBEARBbN++vQhA\nBCAmJiZKKh8dHW1R2atXrxrK6X/kiGttWWrYpH4PTbbAo6OjER0dbXj8wAMP4PPPP7fxZQQRNUYq\nlQpNmzYFYNmypJYsaaqf523syy+/tHvc/Px8NG/e3KplWImMmRyFnpCQUGVknEKhgI+PDyIiIvDS\nSy/ZZXspWaZxAAAgAElEQVQ7jkInajwKCwuRlJSE5cuXS+5OtqTshAkT0LlzZ/j6+uKVV17Bl19+\nKan73JK4Wq0Wffv2xY4dO+Dm5mbx+6WGzebTyF577TXk5+fjmWeegSiK+P777+Ht7Q2lUonbt2/j\n66+/trrS91WKCZyoQUtKSsK0adPQtWtX2WMLgiDbqm7OEJfqD5sn8IiICGRkZNT4XLdu3XD69GnL\nalpXpZjAiRq0/fv3o3fv3lYnNDn35bbEiRMnEBoaysRNZrH5PPC7d+8iJyfH8DgnJwd3794FAH4p\nichs+qlTANC/f3+b/P3IzMzE7t27sXPnTiQlJUkqm5eXh6KiIuTn52PQoEFW16Um//rXv3DixAm7\nnJvI5CC2xYsXY/DgwejUqRMA4NKlS1iyZAnu3r0r+1rBRFR/vfnmm+jXrx9Gjx5ts3OaGhB2/fp1\n+Pr61nixYM6+3NbiIDWyJ7PWQi8tLcXZs2ehUCgQEhJil4FrVSrFLnQih7CmS9qUGzduwMfH575k\nak1Xdl2DyfSjzT/88EP8z//8z31ln3nmGWzevBn9+/fHli1bJL1XV1dXlJeXQ6FQYM+ePVVa8Glp\naWjWrBlCQ0PNPh8RYKftRPft24esrCxotVrD4gfjxo2zvJamKsUETuQQMTEx2L17NwAgMTERGzZs\nsOp8giCgrKwMXl5etR7j6+uLoqIiAEBQUBAuX75sVUzAvC1BrXmvxovAKJXKKrcH1q1bh9atW2PY\nsGEW1p4aK5tvJzp27FhcunQJ4eHhVbb0s2cCJyLHsPUc5bVr1yI7Oxvz5s2r9ZjS0lIAlYlw+/bt\nks5fU4+Buft52+q9btu2rcrjsWPHWnwuIklMrfQSGhoq6nQ6SavDWMuMahGRHRQUFIiJiYliQUGB\nTc6n0+lEQRDqPCYyMtJmq5ppNBoxJCREnDdvnsmy1rzXhx56qErctLQ0ccWKFZLPQ2RMau4zOQo9\nLCwMGo3GntcQROQkfH19sWHDBqvufQuCYNjwSKFQmNx62N/fH4BtVmLz9vbGe++9V2fLW8+a9xoY\nGFglbsuWLdGxY0fJ5yGyhsl74DExMTh27Bj69u2LJk2aVBZSKLB161b7VYr3wInqrYyMDCxduhQr\nVqww63i5V2KzBUfFpYbN5oPYUlJSanw+JiZGSr0kYQInImd3+vRpfPzxx1i7dq2jq0INhF1GocuN\nCZyofhEEAT/88AOeeeaZKiO07c041sSJEw1bHstBq9Xi5MmT6Nmzp2wxqWGz2UpsAwcOBAB4enrC\ny8uryo+3t7f1NSWiBqO4uBhHjhypMp3K3qqPzVm1apUsccvLywEAarWayZscii1wIqp39FPFzp07\nZ3hOjhZ4QUEBYmJikJ6ebhgTRGQrNl0LXavVcjUhIqqRVqvFrFmzUFhYKGtc43neEydOBCBf97mf\nnx9+++03Jm9yCnUu5KJWqxESEoKcnBy77NRDRPWXSqVCcHCwYSqXHGpapEWOxJ2Xl4dWrVoBAAIC\nAuwej8gcJrvQBw8ejKNHj6Jv375o2rRpZSFOIyMiG5GyFvrLL7+MoKAgs+Z5m2Luuu86nQ4DBgzA\nDz/8gKCgIKvjEtVGlmlkCoUC0dHRkitnLiZwIueVlJSEl19+GeHh4TY5n5S10LVaLdRqkytAm0XK\nWui2jEtUG5uvhW7P+d5EVP9MmTIFXbt2tdn5pGzracskamot9KNHjyIkJAQeHh5M3uSUTC6lSkSk\n1WoNLYOePXvWuL+2pTIyMhAUFIQzZ87IOtZm/fr1SExMxK5du2rsPl+7di1OnjwpW32IpOI0MiIy\naebMmQgNDcXzzz8vW8y8vDx4e3vLOkiOyJFsNo1Mv5ftjBkzrK8VEdVrM2fOxJgxY2SLp9FoEBMT\ngy1btsgWEwBSU1Nx9OhRWWMSWarWGzsajQb79u3D1q1b8fTTT0MUxSrLFvbq1UuWChKRYwiCgLt3\n78LPz0/WDTuMp4rJedEAADdu3DDsT07k7GrtQt+4cSNWrlyJtLQ0RERE3Pd6cnKy/SrFLnQih1u3\nbh2OHTuGRYsWyRazpnneRI2FzaeRffDBB5g9e7bVFZOCCZzIMXx9fVFcXAylUolDhw6hW7duZo/A\nNndedW3lVCoVLl++jOeff17W5L1//34cOnQIr732mtllLH2v1nJUXJKHXXYj++mnn7Bnzx7D/O+E\nhASrKmmyUkzgRA6hUqmg0+kAAG5ubigpKTG7rJR51bWV69evH/bv3y+x1ta5cuUKzp8/L2nKrKXv\n1VqOikvysPk88JkzZ+LQoUN49tlnIYoiPv/8c+zbtw/z58+3qqJE5Hz0yRsAdu3aJamsqXnVtdEn\nJABo2bKlpJjWqN6alcLS92otR8UlJyWaEBYWJmq1WsNjrVYrhoWFmSpmFTOqRUR20Lt3bxGACEBM\nTEyUVLagoEBMTEwUCwoKJJXTx9P/yOHs2bNi8+bNZX+v1nJUXJKH1O+/yYVcFApFld2GCgsLq4xG\nJ6L6TRAErFq1CqIookWLFgAsa+H5+vpiw4YNVt2XTUxMtLisFJ07dzbstOio92oJR8Ul52Qygc+a\nNQu9evXC888/j/Hjx6N3794cHUrUgJSWluLChQvQarUmVyezFY1Gg4kTJ6K8vNyQtOW4p1tWVgYA\nUCqV+Pnnn2V5r0T2YtYgtqtXr+LQoUNQKBTo06cPWrdubd9KcRAbUYPlqKlit2/fxoABA3Do0CG4\nu7vLFpfIXHYZhS43JvCGi9NgnENFRQXeeustzJw5E82bN5ctbk3J25rvhJStSAGgoKAAfn5+Vsfl\n95jsQXLus+0teNtw0mqRDURHR1s8cKi+efHFF8Xo6GgxLi5O8qAja8qa69tvvxXLyspsFtfHx0dU\nqVSii4uLePz48ftev3r1qhgSEiLOmzevyvMwGsA2ZMgQSTFdXFwMZQMDA2s8Jjc3t8bn3d3dDWXj\n4+MlxW3VqpWh7MiRIyWVJaqN1NznlJmSCbzhiouLEwGIERERDX4krTUXK4660LEmrkqlMpR1c3O7\n7/Xp06ffl7xFsWoCVygUkmKq1eo6k7BOpxOjoqLEixcv1lnf2pJ/bfz8/AxlR4wYIaksUW1smsDL\ny8vF4OBgqypkCSbwhqsxTYOx5mLFXhc6SUlJ4v79++0SV98aVigUNbbAKyoqaixnnMD37t0rKab+\ngqN79+611re2uM2aNRMBiO7u7mJ2drakuMOHDxcBiOHh4Y3iu0zysHkL/IknnpD85bYWEzg1BNZc\nrNjrQufUqVP3dZvbKu7x48dFNze3GpN3Xfbu3SsqlUrJyVsUa6/voUOHxKKiojrLZmdni0FBQRb9\nfWtMF6IkH6m5z+QgtsGDB+Po0aPo27cvmjZtarjRvnXrVvNvtEvEQWxEtlNeXg6VSgWl0uSsUVkH\nlOkZr7+ekZGBHj16mB2zNu+88w7i4+MxcOBAq89FJBebj0JPSUmpMUh0dLTkypmLCZzIdmbPno3W\nrVtjypQpJo+1Zq1tX19fFBUVAQBat26Ns2fPwtvb22Q5tVqNiooKANLXXydqSGy+FnpMTAyys7Nx\n4cIFDB8+HPfu3YNWq7WqkkQknxkzZsDV1dWsY61Za9vFxQUA4O7uDnd3d/z4448YP368yXJKpRIV\nFRVQKBQ4ePCgpJjGUlNToVQqMWDAAIvPQVSfmOxTW758ORITE/HSSy8BAP788088+eSTdq8YEVlO\nEARcv34dAODp6Wl2ArdmJbaMjAy0bt0arVu3xqRJk8xK3vpybm5uOHbsmFXd53fv3sXdu3ctLk9U\n35jsQn/ooYeQnp6Ofv364ejRowCA7t274+TJk/arFLvQiayyadMm7NmzB5999plsMR21whpRQyE1\n95lsgTdp0gRNmjQxPNZqtdzMhMjJjRo1Cv/85z8ll0tKSkJMTAzi4+OrbGJkSn5+PgIDA3Hu3Dm8\n8847khK4pTEB4ODBg1iwYIGkMkQNhckEHh0djXnz5uHevXvYtWsXEhMTkZCQIEfdiEgCQRCwb98+\nw2OVSiX5HJmZmdi9ezd27tyJpKQks8tV726fP3++3WMCQKdOnXjPmxotkwl8wYIFCAgIQPfu3bFs\n2TLEx8fjww8/lKNuRCRBVlYWVq9ebdU5LB3EplZXHQ87a9Ysu8bUdzMGBAQgKirK7FhEDYlZm5mU\nlZXh7NmzUCgUCA0NNXtAjMWV4j1wIocoLCxEUlISli9fLnkQ29tvv4358+dj1qxZ+Oijj+wW8+LF\ni5g6dSq2b9/O23nUoNh8Hvj27dsxefJkdOrUCQBw6dIlQ0vcXpjAicwjCAK++uorTJ482aIu8/pI\nFEWcO3cOoaGhjq4KkU3ZfBDb9OnTkZycjN27d2P37t1ISUnB66+/blUlicg2tFot8vPzDQuhyEGj\n0WDMmDEoKyuTLSYA3Lt3DwAMPYFEjZ3JBO7t7Y0HH3zQ8LhTp05mra5ERPbn4eGBOXPm2P22lp5+\nqlhYWFiV2Sn2du/ePfTt2xd37tyRLSaRs6s1gW/atAmbNm1CREQE4uPjsWbNGqxZswaPP/44IiIi\n5KwjkVWsmabkKHXVWafT4bXXXoNGo5E1rql53kqlEgqFAgqFAjt37rRJTD0PDw/s378fXl5e0t6Q\nndTH7xQ1PLUupfrzzz8bBoi0aNHCsD5yQEAASktL5akdkQ3opykBlX94pazv7Sh11VmpVGLYsGFo\n1qyZbHHNWaTF+N5dfHy82ffy6nqvly5dQseOHaFQKJwmeQP18ztFDU+tCXzNmjUyVoPIfqxZ39tR\naqqzKIqGi+oRI0bIFhcAlixZImmFtR07dlgdUxRFvPLKK/j0008REhJi9vnkUB+/U9TwmByFfunS\nJXzxxRfIzs42bGLC7USpPrFmapQ122taU1apVBr+D+zYsQNxcXF46aWXMGbMGJM7AVq6rScAdO7c\nGdnZ2fD29saRI0cMZY0vHmpjvKvY3r17MWjQILNi+vj44Pbt21Cr1Th8+HCV9dBNxXV1dUV5eTkU\nCgX27NljdkzAut+PNd8potpIzn2mNgzv3r27+Nlnn4m///67mJycLCYnJ4spKSlmbTY+YcIEsUWL\nFmJYWJjhuTlz5oht2rQRw8PDxfDwcHHnzp33lTOjWkSyiI6OFgGIAMTExETZyurL6X9EURTPnz8v\nlpWVmSzr4+NjKBcUFCQprjVljeurVCrNLqdSqQzl3NzcxAMHDog3btywa0xRtO73Q2QPUnOfye1E\n3dzc8Nprr5l/RWBkwoQJmDp1KsaNG2d4TqFQYPr06Zg+fbpF5ySSkzVdpbbqZv35558BoMpskLro\nt/X08PBAamqqpFjWlDVuPejvD5uj+nai27ZtgyAIGDx4sN1iAuwGp/rP5DSyqVOnYu7cudi/fz+O\nHDli+DHH4MGD4efnd9/zIrvHSUbWjBgOCAhAQECARd2k1mzNGRQUBKByM6H09HRJZTMyMhAUFIQz\nZ85I6j4HgOHDh8PFxQXh4eHw8fGRVFZfZ29vb7Rt21ZSfY23E3377bfNSt4A8NhjjwEAIiMjERYW\nJqm+1vxuOQqdnIKpJvpbb70lBgYGilFRUWJMTIzhx1xZWVlVutDnzp0rtm/fXuzRo4c4ceJEsaCg\n4L4yZlSLyGyO6ga3hnFXdps2bWSL269fP4vfrzXd73v37hX/85//SCojio773bL7nexBau4z2YW+\nceNGZGVl2WyhiClTpmD27NkAgPfeew9/+9vfsHLlSpucm6gmztANLoUgCLh9+7bh8XfffSdLXI1G\ngwMHDhge3717V1J540VWvv32W7PLJSUl4dChQ3BxcUGfPn0ktYgd9btl9zs5BVMZfsSIEWJeXp7F\nVxTVW+DmvAZAnDNnjuEnOTnZ4vhEBQUFYmJiYo29PfYsa6nt27eLLi4uFrdmLXH16lUxJCSkxsFz\n5vLy8rKozta0Zh31u3XE94IanuTk5Cq5Tur/OZPTyKKjo3HixAn06dPHsHSilGlk2dnZSEhIwMmT\nJwFUXuW3bt0aAPDJJ5/g0KFDWL9+fZUynEZGjV3z5s1x8+ZNeHh4WHQvW4pbt25hwIABGDduHN55\n5x3D8yNGjMCWLVvMPk9AQADy8/PNrvOhQ4ewdetWHD58GDt37kRERIRF4wWIGgqb70aWkpJS4/Mx\nMTEmT/7MM89g9+7dyM/PR8uWLfH+++8jJSUFx44dg0KhQMeOHbFs2TK0bNnSqjdBVN8JgoDU1FQM\nHToUAJCTk4NBgwYhNTXVrskbqFya9ddff0VcXBxGjhyJn376SXLyBqTXuaioCKdOnUK3bt04p5oI\ndkjgjsAETo1NVlYW5s+fj2XLljX4Pa5FMxaFIWqMbJ7APT09Df/ZBEFAeXk5PD09qwyysTUmcKKG\nKTc3F+PHj8fvv/8OpdLkLFaiRsXm+4EXFxfjzp07uHPnDkpKSrB582a8/PLLVlWSiCoviP/5z3+i\nvLzc0VWRTbt27bBy5UombyIbkPS/SKlUYuTIkfjll1/sVR8ip2LPBTtEUYQgCDVecVsT19fXF2q1\nGq6urjhx4kSV165evYonn3wS9+7dk1zWlLrqbDzFrFOnTpLOa2lMoobO5DzwTZs2Gf6t0+lw+PBh\nuLu727VSRM7CnttGNmnSBDNnzrR53OLiYlRUVKCiogKRkZEoKSkBUJm8hwwZgvHjxxvmMZtb1hy1\n1bm0tBT9+/dHamqqzQepcVtPasxMJnDjfcHVajU6dOiAn376ye4VI3IGtl6wQxRFvPrqq5gxY0ad\nI7WtiVt9bXGgavKua0vQmsqaq7Y6u7m5IT09vdaLBmtwQRVq1GwwF93mnLRa1AjZY8GOX375xeSu\nYtbEPX78uOjm5iYeP35cFEVRvHLlihgcHCzOmzdPclkpqtf57NmzYkVFheTzWBOTqD6TmvtMjkK/\nfv06vvrqq/v2A1+1apXdLio4Cp0aGlHi1Clr9vSuvpf4sWPHIIpinS3v2srGxcWZHbe6v/zlL/jg\ngw8kbzJC1FjZfBpZ//79ERUVhd69extGjioUCowaNcq6mtZVKSZwagCMk3BcXByeffZZPPLII2aV\ndXV1NYxODwwMxJUrV8yOW/1CQafTmX3xUP04Kf8Pk5KSkJmZCQ8PD6xfvx4+Pj52n+9dPSYXgqH6\nTGruM3kPvKSkBP/4xz+sqhRRY5SXl4eioiIAwK5du7BixQqzyxpPLevYsaPFdVi9erXFSXThwoWS\njnfEgDIOYqPGzOQ0sscffxzbt2+Xoy5EDYp+BLdSqcSvv/5q8Y5+GRkZFtfhxRdftLjsm2++Kel4\n/YCy4OBg2QaUcRAbNWZmrcR27949uLq6wsXFpbKQQsGV2IhMaN++PXJzcwEAiYmJklqHxq3mFi1a\n4Nq1a2aV02g0CAwMrPKcpHtqRnEVCgV0Op3ZZQsLC2Vf09wRMYnsxeZd6MXFxVZViKix6tKlC3Jz\ncy1qHfr4+KCoqAhNmjRBenq6WWU0Gg2GDBlS5TlzNh2qzZ49e8w6Li0tDfn5+RgxYoTsXdi+vr7s\nNqdGi+sZEtmQIAjIysoCAHz33XdITEy0aIvM48ePIygoCOfOnTNrBLo+eY8bN86QtGNiYpCcnCwp\n7t69e6FUKrF3714MGjTIrDLu7u52meNNRHXjbmRENvTf//4X3333naz3Y42TtzlTxYjIOXE7USIH\n08/5tmaKk7lli4qKEBkZKXvyPnr0KNatW4fFixcbnuOULiLr2CyB37p1q86C/v7+0momARM41SeC\nIOC3335DfHx8ledjYmIMU5ykDmIzt6woikhOTsbQoUMtrL1l7t69i5MnT6Jfv36G56x5v0Rkw0Fs\nvXr1MpwsNzcXfn5+AICCggK0b9/ecJ+PqLG7fv06tm3bhri4uCqjuK2Z4mRuWYVCIWvy1ul0UCqV\naNq0aZXkDXBKF5Hcah3Elp2djaysLMTGxmLbtm24efMmbt68ie3btyM2NlbOOhI5taCgICxZsuS+\nBVPWr19v8SC2CxcuQK1WIysry7AYjLlCQ0Ph6+uLgIAA5OTkSCpbF41Gg4EDB6KioqLG1615v0Qk\nncl74GFhYTh16pTJ52xaKXahkw3Z4160IAhYtGgRpk+fDjc3N5vX2ZqlVH19fQ1JPygoCJcvX7ZZ\nvS5fvoy2bdvW+JqlnzPvnRNVkpr7TE4jCwwMxIcffmhokc+bNw9t2rSxqpJEUiUlJSEmJgbx8fEo\nLCyUVFa/3ObOnTuRlJRkk7JKpRIeHh6G/QFq4uvrC7VaDVdXV5w4cUJSXP3GQQDQrVs3AJUt4Mce\newx37typs2xpaamhjlJXUazpczb+vGtL3oDln7M1vx9Hseb7SGQzprYry8/PF6dOnSqGh4eL4eHh\n4muvvSbevHlT0pZnUplRLWpkoqOjRQAiADExMVFS2bi4OBGAGBERIXnbSWvKqlQqQ53d3NwklfX2\n9jaUHTlypHj16lUxJCTErC1BIyMjLf6sqn/OgiCIPXr0EG/cuGGyrKWflTWfsaNY830kqo3U3Gf2\n0cXFxZIrYykmcKrOmj/yarVaBCAqFApx7969FpUFIO7Zs0dMSkoSz58/b1ZZfTkA4urVqyXFNS77\n0UcfmZ28rY1r/H71n1VpaalZZTt37iyq1WqxWbNmYnZ2ttkxLS1nrRdffFGMjo4W4+LiZL2wI6qN\nzRN4Wlqa2KVLFzEoKEgURVE8duyYOGXKFMtqZ26lmMCpmoKCAjExMdGiP5bGCU2pVFpVNiUlRSwr\nK5M1LgCzk3dNZaVQKpWGck2aNJFU1sfHx1BW//fCnuWsZU0r2prvI1FtbJ7A+/TpI+bk5Ijh4eGG\n57p27Sq9ZlIqxQRONqRQKO5rVTp7WeMEPGrUKIvLbtiwQVJZFxcXQ9nvvvtOUtnmzZuLAEQPDw9J\nLWlLy1mLrWhyNpIvuM25T96uXbsqj9Vqk3ugEDmNPXv2SF7fW++xxx4DAHTu3BlhYWGyxTWWl5cn\n6Xj9AiobNmxAYmKipLIJCQlQKBQYOHAgHnnkEUllMzIyEBQUhDNnzpi1fru15azFaW9U75nK8KNG\njRJTU1PF8PBwsaysTFy4cKH41FNPWXyFYQ4zqkWNjL5laMl9bBi1SL/88ktJZY0HosXGxkoqGxIS\nIvr4+IjNmzeX3LI0rrPU/w+Wvt+0tLQqZWNiYiTFtZQ1n5OjyhLZg9T/6ybngd+4cQN//etf8dtv\nv0EURTz88MP4/PPP0axZM/tcUYDzwBsyS+f8WrNPdfUFVsz5bgmCAK1Wi6ZNmxqec3V1RVlZmdlx\nlUqlIVarVq2g0WjMLmtJna0tu2LFCrz44ouGx0qlstZFW2zJUXPerSkrxzr31PhIzn2mMnxqaqpZ\nz9mSGdWiesrSgUMwahkOGTJEUkzjsmq12qwy//rXv8S5c+dWKduiRQuL4/r5+Zk8/urVq+KlS5fu\nK+vt7W1x3C5dulhcNjIyUlJZSxmPfI+Pj5dU1pr759aUtWYAHKegUW2k5j6TRxsPXqvrOVtiAm+4\nLB04pJ/bHBoaKnnAkXFSMrcbXKvVimVlZaKvr68IVM7jtmc3uH6e9xdffHFf2VatWlkcNzAwsM5j\nU1NTxW+++cbw2HgUuqmytqJPaN27d5f8u83OzhaDgoIs6gK3pqyj1haghs1mCXzfvn3iokWLxDZt\n2oiLFy8WFy1aJC5atEicM2eO2KNHD6srWmelmMAbLEun31gzbWfgwIGG1mhd5cvKysSzZ89Wec6a\nP/LGiXThwoW1HlfTIi1+fn4WXzgYJ2FT4wVOnz4t/vbbb4bHzZo1k31EeH2ckmVNnevj+yV5SM19\ntd4D3717N5KTk7Fs2TJMnjzZ8LyXlxcSEhLQuXNnMzvppeM9cLKlwsJCJCUlYfny5XXeb9y3bx9W\nr16Nr776yiZxje9F13Y/WaPRYMiQIfft552Tk4NBgwYhNTVV8shsb29vw3KrUu/tWhOXiKxjs/3A\n9XJycmT/j8wETg2B8SC2mqaSFRcXIyIi4r7kba2AgADk5+fDw8OjxqlZx48fx7///W8sXbrUZjGJ\nyHo238zkhRdeqLJY/61btyTPDyWylr02jxAEAZs2bbLZ+YwZ/0esaQMST09PrFy5ssbkrVAoDD8b\nN26UFNfb2xsA0KRJkxpfDw0NxQsvvCDpnETkfEwm8Bs3blTpdvT398e1a9fsWimi6n7++WfDjlUT\nJkyQVLau5F9YWIg9e/bUOl3KmkRqLD4+vsbnBw4caLLs6NGjJcW6efMmAKCgoKBKq1+/w1mTJk0Q\nERFRY1mlUml4vzt37pQUl4jkZTKBq1Qq5OTkGB5nZ2fXuYUikT0Yz7+Wenulru0qW7Rogc8++wwq\nlcrkeaQmUmMuLi4Wl7U0loeHB1JTUwFUXohHRkYa5lvXxvizre2ig4icg8lMPG/ePAwePBhjx47F\n2LFjERUVhY8++kiOuhEZ9O7dGwAQHh6ONWvWSCrr4eEBAIiIiMDy5ctRXl6OOXPmoLi4WNJ59EuU\nmsv4QjcjI0NSWWM7duyQdHxNS5MGBARg+/btki4kpMYlInmZTOCPPvooDh8+jKeeegpPP/00jhw5\ngkcffVSOuhEZbNy4EYmJiUhOTpa8clX1Na9VKhVatWoFV1dXk2WHDh0KAHjooYcQGxsrKa5xa/bk\nyZMYNmwYCgoKzCqrv+hQKpVo06aNpLjt27fH5cuX0b59e9y4ccPwfKtWrUyW1SftHTt2IC4uTlJc\nIpJXraPQ//jjD3Tp0gWHDx+uMjJOPzWmV69e9qsUR6GTk4iJicHu3bsBAImJiZJa4dWXNJ03b57Z\no83VarXhvrybmxtKSkrMjqun1WoRGRmJ7du3m5W8icixpOa+WrcV++c//4mvvvoKf/vb3+77QwQA\nycnJltWQSGb6tacvXbqE77//Hv379ze7rD55A0Dfvn0trsPAgQMlTRUzHlT397//XVIs47W2f/nl\nF2V3MBoAACAASURBVAQEBEgqbymu8U0kL5PzwB2BLXCypdatWxu25ExISMDWrVvNLuuITUWsLdu8\neXPDSPQRI0Zgy5YtZpe1hjW9FURkwxb4pk2bamx56/3lL3+RVjMi2GY3si+//BIvv/yyWeUqKiqq\n7Kd9/PhxaRU2EhkZaXFZa0yZMkXS8frkDcCw25a59DuDKRQK7NmzR9I+5vqBeiqVCu+++67Z5UJD\nQ5GXlwcXFxdkZGTItnCUNXHZ20DOoNYW+PPPPw+FQoHr169j3759hsE8ycnJGDBgALZt22a/SrEF\n3mBZ2kqztEX6xhtvYPHixRaVtSaus5SVev/cnOVfa+Pj44Pbt28DkLaEqzXbelrDmrjsbSB7sNlK\nbGvWrMHq1ashCALOnDmDTZs2YdOmTTh9+jQEQbBJZanxuXjxIoDK1cIWLlxo0Tk++OADs4+dOXNm\nlcfmLJxSmxEjRlhcdtasWRaXNacFnpaWhjNnztz3/P/+7/9aHFfqRbrxtLxvv/3W7HI1zVuXgzVx\nq09NJHIIU7udhISEiDqdzvC4oqJCDAkJkbRjilRmVIvqKf3OYJC4F3KTJk0M5YKCguo8tqysTCwq\nKjI8Bszf1rM643Jubm4Wl5W6NafUOn/zzTeGXcWsqXN4eLjFe1V7eXmZ/TsyZs2Ob9awJi53FCN7\nkPr3qdZ74HrDhw/HI488gjFjxkAURXz//feS58MS6enX6ba05aJUKrF9+/Y6j1m3bh0uXLhQ44JD\n5qy4VpvNmzfX+ppGo0FRURFCQ0MNzxl3h3Xr1s3iuPrPrC5jxoyp8fl+/fpJitW6dWscO3bMot9P\nkyZNcOfOHcktWv28dblZE9fX15fd5uRwJhdy+eKLLzB58mQcP34cJ06cwEsvvYQvvvhCjrpRA1R9\nURVz6ZOYTqfD7Nmz6zx2woQJeP/992t8bc6cOeZXtprakqR+S9Bff/21yvOiic1MzFXbmgv79u3D\nihUrLD5vTc6fPw+1Wo1Lly5JHgBX0wpwRGQ/Zk0jy87Oxvnz5xEbG4t79+6hoqICXl5e9qsUB7FR\nNf7+/oZVzEaOHIkff/yxyuuCIODcuXPo3r37fWXtOZistv28bRm3RYsWNW4gdOnSJWRnZxsGmNZU\nVuogNkcNKCMiO2wnunz5ciQmJmLy5MkAgD///BMjR460vIbUqIWGhsLX1xcBAQFVNskxRb8Wes+e\nPbF69er7Xj958iQ+++wzk+cJCwszv7LVVO9+ryt5V/fAAw9YHLc2nTp1ui95A1UT+K5duySd05qB\nXfba8pWIamYygX/55ZdITU01dGEGBwfj+vXrdq8YNUx5eXkoKipCfn6+pDnG2dnZUKvVyMnJqbFr\nt3fv3mZ1J1+5ckVSfY0ZT6kqKSkxO3kD/3/0vSWM75+fPn0azz33XJ3HG29YMn/+fEmxhg8fDldX\nV/Tp0wc+Pj6Syta16xsR2YGpUW59+vQRRbFydKooimJ5ebnYvXt3SSPlpDKjWlRP6UeTK5VK8fjx\n42aXg9HI6qZNm4qiWDnafN26dVVmSZgqK/W7VVfZQ4cOyRLX19fX8Hx5ebl45MgRs8v6+/tLiqtS\nqQxlY2NjJZX18PAQAYhqtVrS7zYkJET08fERmzdvLvtIdCJnIvXvhMkWeHR0NObNm4d79+5h165d\nSExMREJCgs0uIKhxCQ8PB1A5GO3DDz+06Bx3794FUDnv+NixY5IWG7GliIgIWeIUFhYa9vFWq9Xo\n2bOnpLJSGH+We/bskVRWvz6EVquVtJOZpb0yRI2dyQT+j3/8AwEBAejevTuWLVuG+Ph4i//wEvn7\n+wOwbgEM/b1of39/LFy4EGq1ydmQ9Zq3tzciIiJQVlYmuazxZizmMP4s09PTJcfT01+omcNRC7kQ\n1Xd1/uXTarUICwvD2bNneU+LbCItLQ1A5Zrkubm5Fq0hHRUVhcLCQovKxsfHSy6j16VLF4vLWhN3\nxYoVGDJkCJo0aWLW8W3btsXly5fh7e2Ntm3bSorl5uaG4uJii+bLDxw4ELt370b37t3xzTffmF0u\nIyMDgwYNQmpqquTpZ1yTnBozk9PIRowYgc8//1zWeZ2cRtZwWbrPdfUpWWVlZXB1dbWorJTvljOU\nVSgU0Ol0ZpfVb0gCAIGBgZIG7lmzD3lhYSGSkpKwfPly2RIp1ySnhsRmu5Hp3bp1C926dUPfvn3R\ntGlTQxApWzIS6ek3yFAoFDh48KBF51AoFGYn7+octcCINa1341Hl5jD+AyClKxuw7vfjiNXJuCY5\nNWYmE7j+frfxH4W6thklqktCQgJ+/PFHDBgwAO3atbPoHB9//LHF8efOnWv2sRqNxuI41VXfEc0U\n4yvx33//XVLZ5s2bIy8vD15eXliyZImksoGBgcjJyYG3t7fkaWSOsH79etlb/UTOotYu9JKSEixd\nuhQXLlxAjx49MHHiRMktAYsrxS50p2bNfUe5txO1tKx+kZZz587JGvfo0aMIDg6Gp6en4bkhQ4bg\nv//9r9lx3d3dUVpaCqDy3rupteONWdP9zvvRRNax2Ups48ePx+HDh9GjRw/s2LEDb7zxhk0qSPWf\nNQt26Bc08fHxMbmdqFarddiFXHJyMsaPHy973HXr1uHkyZP31UUKfQIGgGPHjkkqa033OxdyIZJZ\nbRPEw8LCDP8uLy83LOQihzqqRU4gLi5OBCBGRERI3k5RynaiM2fOFFetWiWKYtXFSVavXi0ppnHZ\n+Ph4i8tK/V4al9uxY4fFZYcMGSKpbLNmzUQAoru7u+SFUaKjo0UAYvfu3SX/bq35XhCR9NxX69HV\nEzYTOOlZsxeylD/yBQUFYllZmSiKorh69WqLkrcoimJ8fLxFyVsUqyZSKauLiaIo7tixw6zknZqa\nKh4+fLjKc5GRkSIAMTQ0VPLn7Kh9rrlHNpF1pOa+Wu+Bq1QqwwhPoPKeuLu7O4DKfvrbt2/bogOg\nRrwH7tysudcZHByMrKws+Pj44PDhw/eNChcEAcXFxYYFX/SM7ycvXLhQ0i0d47KxsbH4z3/+Y1FZ\nqfeTlUql4Xu8Y8eOWlcn++mnn+Dh4YHY2FjDcz4+Prh9+zbUarXhVpa5fH19UVxcDKVSiYyMDEll\neR+byHGk5j6zthOVGxO4c7Nm7q2p7SrXr1+PI0eOYNGiRVWet/eWoNeuXavxnq8zzAOXOh/bmrnc\nnFdN5Dg2306UqDpr5t7qR0crlcoaW7PPPPMMFixYYH0lzaQfbS5llLcl9Nuh6h04cACffvqpWWUH\nDx4sKZZSWfnf2pK53JxXTVR/MIGTZOvXr0diYiJ27doluYtVvy2tTqfD7NmzAVR2m2dkZACoTDqm\n1jafNWuWBbW+n/F+3tOnTzd5vLlLmdYkNze3yuN27drVOcrbOFb13ghTMjIy4ObmhmPHjknqPges\n+90SkbzYhU6yMu4a7tmzJ44cOYJTp07h008/rXM/b1t3ZRsn77r283aGLnQvLy9JY06suY/Ne+BE\njsN74OTULE1qtkykZWVlCA8Px9ixY+tM3raOe/bsWbz99tv44YcfTK5maE1ca+5j8x44keM41T3w\niRMnomXLlujevbvhuVu3biE2NhbBwcF4+OGHJe9XTA2LIy7UXF1d8f3335tM3rbWuXNnvP/++5KX\nIu7WrZuk4/WL5Xh7e5tcLKc63gMnqj/smsAnTJiAX375pcpzCxYsQGxsLDIzMzFs2DBZByyRc3F3\nd4dWq3VIbOOLSjkcP34cSqUSYWFhksvevHlT0vGCIAAAbt++jWnTpkkqy3vgRPWHXRP44MGD4efn\nV+W5rVu3GpaoHD9+PLZs2WLPKpAdJCUlISYmBvHx8Vb1oLi7u8u2vr4jGK9nLmUqV3VSLzaMl1KV\n2sMxY8YMXL9+HWPGjGHvGJGTk30U+rVr19CyZUsAQMuWLXHt2jW5q0BWsnTNa/3cZL27d+/aump2\nJfViwzhpx8TESCprvBOY8YJK5tBPWQsPD8eaNWskleV65kT1h8ntRO1JoVDUej/QeNvHmJgYyX8A\nyX4svceqUqmqPH7nnXcsij9x4kSzj7XllqAvvPCC2cfm5uZWuWDZvHmzpFhNmzZFUVERvLy88Nln\nn0kqu3HjRou32LTm/jkRSZOSkoKUlBTLT2Dpmq3mysrKqrIxSkhIiKjRaERRFMWrV6+KISEh95WR\noVpkhRYtWhjWBx85cqSksoBlm4NYUk7//bI0pqVxdTqdGPP/2rvz+JrO/A/gn3tvIhHZUEsqiH0J\n2RclCBpUW0tMpp1qB6UpbadUMfXThf7QDjXzs07RoZZSoUPbIV7U2IIKQWIPmtiaGGKNLDfJfX5/\nZO7pvclNck+Se+6Sz/v18pKce57zfXJc+d7nOc8SHS3UarVU7umnn5YVtyb3uCbkbDZDRLVL7u8n\nxbvQhw4dirVr1wIA1q5di+HDhytdBULNnmMbnq8fMFWZN998E0ePHi13XL9imCUYzvNWmkqlwt69\ne6HT6aRjHTt2lHWN//znP9LXcv99avJve+3aNQDmbfdKRFZmoQ8SQgghXn75ZeHj4yOcnZ2Fr6+v\nWL16tcjJyREDBgwQHTp0EDExMSZ3LrJwtUj8tm0kqtHS8vb2ltU6PHfunLSrGBRogetb3nPnzq1R\nTLllk5OTxYMHD6Tva9ICN4zp4uIiq2xN/m3ZAieyHrm/nyz6DHzTpk0mj//000+WDEtmqMl8Xzc3\nNzx48AAeHh4VruddVFQEjUYDtVqNrl27mjynukuiVrWkaXJyMsaMGYMPPvig3GuGg8PkGjJkSKWv\n//DDDxg8eDB69eoFAEYt8MpWmTPFyclJmmKXnJwsq2xN/m31S91yHjiRHbDQB4kasdFqOZSa7N1s\nTivtk08+EcuWLSt3HAYtS5VKZXZMKNSKtlRZua3o1NRU4erqKnsPciG4pzeRvZL7O4ZLqZJsDRo0\nQF5eHjQaDU6ePGlyw4zc3FzUq1cP9erVMzpuOOugQ4cOSE9PNytmbS1pqlary01nM7esqb3Ek5KS\nAABRUVGVlu3Ro4fJcQBERHpcC50szsvLS9pcw3BPb61Wi/v370vz/E2xhbXQa7Psnj17AJQm98rK\nuri4SFupEhGZYlNroZNjKiwsBFB+T+8dO3Zgzpw5Zl9n1KhR1Ypv2KrPysqS1bL18PCoVkwAJkdl\nx8TEmEzeZa1fv77acYmITGECJ9n0+1jrdDqjhD1ixAhZi45888031Yqvn7qmnyp2+PBhs8s+fvy4\nWjEBSN3nx44dw7x586o833CanDWmsxGRY2MCJ9lu3boFoLQrfe7cuUYJ1JJzuw0ZzvOeOnWqIjH3\n7dsHAGjXrp3JZ95lGY5Cl7sSW22tN09EjovPwEm2qKgoKWkPHjwYvr6+WLlypVnbZNbWM/BOnTrh\nj3/8o1lbglrr+bmHhwdyc3MBGI8VMAf35SaqeziIjSzOyclJGsl96NAhs1qjeoYJ0c3NzewNTcom\n0rlz55q9n3dtJfCWLVvi2rVrZu/nrdFopFa43PvUsmVL3Lx5E15eXkhNTUXr1q3NLktE9okJnMwS\nHx+P9PR0uLm5YePGjbI2vbBGi9ZWR6GbW7ZRo0ay9vWuX7++NGp9yJAhRoMFicgxcRQ6mYXbRirr\n3r17ss433NP79OnTtV0dInIATOB1VE2W2zSkXzbUEeXl5Rl9HxcXV+1rrVmzRtb5+h4RNzc3HDly\nRFZZDoAjqhvYhV5HPXjwQNae0TqdDpMmTcKMGTPQokULo9cs2SWt0+mgVqttohtcbln9inWA/EFs\n165dQ1RUFJKSkmQ//+YAOCL7xC50Mou3tzcSEhLMfvatVqsRExODp556yui4fk64JWRlZSE8PBzX\nr1+3WAxTrl69avI/kbmD1/QMV16raGOfirRu3Ro3btyo1uC12updISLbxgROlTJMZEOHDi23tvmF\nCxcsElc/z3vkyJFo1apVrV23qg8sQgj86U9/wqVLl8q91rdvX1mxGjRoIH39hz/8QVbZmti4cSPi\n4uKwZ88eWYMTici+MIFTpSZOnIj9+/dX+HpNtoZ1dnY2edxwkRZzp4qZa+PGjZW+rlKpsGPHDnTu\n3BkAEBkZCQDo2LEjtm3bJiuWfttTNzc3adMTJcjtXSEi+8QETpWaNm0aevbsaXQsNDRU+nrx4sXV\nvrbhSGs9SyRvw67v2NhYk+ccO3YMd+7cKXf+rl27EBcXh2PHjslOiCdOnICvry/Onz/PedxEVOuc\nrF0Bsj1arRZqtRpOTk5o165dudcNpzW9++671Y7TuHHjcsfS0tIwduxY/PnPf5ZVrjKGjwEWLlxo\n8px9+/ahsLAQTZo0MTqub81Wh/45NhGRJXAUOpXz2WefwdXVFe+9957J1w1XGHN1dUV+fr7Z17a3\nhVyIiJQiN/exBU7lvPfee5VuSqJP4CqVCseOHat2nG7dulWrXE3mnu/cuVP6OikpCU+ePMGgQYOq\nfT0iImvhM3ACUNptfvPmTQClreqyo80NnThxAq6urjh9+jQCAgJkxdEn7W7duuHMmTNml9Mn7V69\neskeEKZP2jt37sRzzz0nHVer1YrtnkZEVNvYhU4AgMTERPz4449Yvny5YjGzsrJw8eJF9OvXT7GY\nRES2igu5kFnKLrf53HPPYdmyZWaV9fb2hpOTE+rVq4e0tDRZcVUqlfTH398fx48fN6ucfjU2lUqF\nxMREWTENnThxAjNnzqx2eTm4pCkRWRJb4HVUTZbbNNxO1B4GsRl69OgRzpw5o8ga7lzSlIjkYAuc\nzKJfbrNJkyZYsWKFrLL658Y1HcRW0Sj3qhgORDOX/j+Fp6enYhuwcElTIrIkJvA6Sr/cZnp6Oho2\nbCirrH4zEy8vL3h5eZldTt9q1/vqq69kx/T09ETXrl3NLgcAN27cQHR0tDT1TSlc0pSILIld6HWM\nVqvF4sWLMWPGDBQXF0OlUuHgwYOIiooy+xrWmJNtWK5+/frltvqsTHx8PFJTU9G4cWNs3LhRVjKt\nV68eioqKqnWfiIjk4DxwqpJOp0NxcTGA0gTap08fxVunNWHuM/dHjx7B09MTP/74I7KzswEAY8eO\nlbWmuX65VyEE+vbtW64XgYjIWtiFXsfUq1cP06dPNzpW0aYi9qywsBC9evXCgwcPUFhYKB2X27Nj\n2PLXD0gjIrIFTOB1gBACEydORGZmpnRMo9EAKE1Q5k7lMqWyPbIt1VoNDg6u8hwXFxckJyfD29tb\n2nwlKCgIX3/9taxYBw8ehFqtxqFDh9h9TkQ2hQm8DlCpVIiNjcXTTz8tHRs2bBhUKhV69uxZo/22\nd+zYYfJ4VlYWgoODceXKlQrLLliwwOw4+ufWrq6ulXaBX7p0SXocUL9+fQDAli1bEBcXh3379ske\nTBYVFYWSkhImbyKyOUzgDsywuzgmJsZoedScnBwIIXD48GHEx8dXO8aYMWPKHdNvCfryyy+jffv2\nFZadNm2a2XH8/f0BAAUFBZWWmzlzJs6fP290jPtjE5EjYgK3Y507d4a3tzeaNGmCa9eulXv97bff\nxq5du0yWNXye6+HhUe06/Oc//zH63hL7eQPA4cOHpa8rq++WLVvKbZJSk5XjuJoaEdkqTiOzY97e\n3nj48CEAwNfXt9ze09evX0fz5s1Nbkxiialg5iRvS6zEdvToUbRq1UqaK15WTVaO42pqRKQUTiOr\nQ/Sjx93c3KQdurRaLYQQcHFxMfvZ9sSJE2ulPleuXMG4cePM7hr39PSsVpy2bdsafX/8+HEUFhZW\nmMDVajVKSkqqtXIcV1MjIpslbJCNVsvmZGZmCl9fX5GZmSkd++tf/yrmzZtXZdmJEycKAGLixImy\n4wKQ/uzcuVNWWXd3dwFAuLq6ivv375tdrmXLlgKAaNKkiaxyQgiRmpoqXF1dRWpqqqxyQghx//59\nERcXJzsmEZFccnMfu9AdTFFREYQQle7nXVOJiYkYMmRIuf21zfHgwQPEx8dj5cqVsgaVlS2XlJSE\nO3fuYMSIEXKrT0Rkk+TmPiZwB6DVanHz5s1yXcuO7NSpU8jJycGzzz5r7aoQEdUK7kZWBx0+fBif\nf/65rDJVjWCvjLe3NzQaDZycnBQb1V22XHBwMJM3EdVpbIE7CCFEpauilWV4br169YyWG5VTVh+7\nOmWbNm2K27dvm1XOx8dHWs982LBh2L59u9kx4+PjkZ6eDjc3N9mbmRARKYUt8DpCq9UarYImJ3mb\nupa5srKyqh2nrLJzyCtj+AFD7s+anp6OAwcOIDExsUaL1hAR2RImcDt19+5dTJs2DX379q3xIiPm\nTufSz/OuLRVN+ypLp9NJ65kHBwdjzZo1suJwKhgROSJ2odsxwwVKYmJisHv3brPLyu0G1+l0iIiI\nQGxsLGbOnCmrbE3iZmdnY/jw4bh79y6uXbsGT09PnDx5Eq1btzY7ZnVHvhMRKYmj0B1YUVER5s+f\njylTpqB+/fpGyVDuCmPVeY5948YNtGzZ0iKruFUVt3v37pWuOkdEZO/4DNyBqdVquLu7S1uB6ldi\nq84KY4bMXYmtZcuW5Y6NGjWq2nEr243s/v37RnFNrTpHRFSXMYHbEY1Gg0mTJkmLtJw4cQKurq44\nffo0AgICZF2rc+fOAEqfQ8+bN09WWXd3dwClHyDmzp0rq6w+aS9YsABTp041eU5RURGio6Nx584d\n6diJEyfg6+uL8+fPy+o+JyJyVOxCt3FCCEycOBFTpkxBx44da+26VW3SUVRUJLV6y6pqE5XaUFhY\nCBcXl1q/LhGRrWIXuoNRqVR45ZVX4OfnV6vXNdxO9MmTJ0avZWVlISgoCBcuXDBZVp+8AUijw81V\n2UIu586dQ3FxMQAweRMRVYEtcBul0+mgVlvu85U1tgQFKm/5jx49GpMnT0ZwcLDZ1yMichRsgduZ\nilqkkydPxrZt2yotq1aroVKpoFKpkJiYWOO6mJO8a8qw5R8XF2f02tq1a5m8iYjMxBa4lVXUIs3O\nzkajRo0q3VWsNqdz/frrr2Yn79qMe/jwYTRr1gzt2rUz+xpERI6ILXA7Y7hK2NKlS6Xn0c2bN5e1\nJejOnTurXYcuXbrg5s2beOONN2S3vIcMGVLtuDt37sSFCxdkb6ZCRERM4Fa3ceNGxMXFYc+ePdi6\ndWulc6Mrc+7cuWrXYfjw4QgPD8f7778vu6zcpVX1H1g0Gg1atGiBcePGoX///rLjEhHVdexCtyE6\nnQ7FxcVmt7yVXhGtNsoaLv8qd/U4IiJHxi50O6PVanHx4kUApYPS5HSbG6puyx0AZsyYUe2ycuMa\njqyvyepxRER1HRO4laWkpGDhwoXVKmvOqmamlN0S9LvvvlMkLvDb6nGpqamyV48jIqLfsAu9jtFP\nFbt06ZLRcUvf77S0NCxdupTbeRIRVYBd6HZAq9Vi69atNb6O3HnghvO8DTVs2FBW3M6dO8Pb2xtN\nmjQxewR5586dER8fLysOERFVjAncCh4+fIikpCRpMFd1GX5Sq2o6V2WLtBju/GWO7OxsPHz4EHfv\n3kVUVFSl5+qXRq1Xrx7CwsJkxSEioooxgVtBkyZN8H//93/StqC1obJ54EIIxMbGVrhIi9yBaOZu\n7Xn37l1ERERAq9XKuj4REVWNCVwhRUVF+Oijj/D48WOj45Vt7lEV/ZxqtVqNFi1aVHieSqXCtm3b\nKlykRe72nM8++yzq1auH8PBweHl5VXjeU089hZ07d1Z7ZD0REVWMCVwhGo0Gvr6+5XbZSk9Px4ED\nB5CYmCj7GXFhYSGA0vnjkZGRlZ7bvHnzCl/7/e9/LytuVlYWtFotDhw4YLLOhvt4VxaXiIiqjwlc\nIWq1Gm+++Wa51qjhUqpyR2jrdDrp6z179lS7bmvWrJF1fmV1LikpweDBg8tNVSMiotrFBG5h8fHx\nOHv2bIWvGy6l6u3tLevaISEh0teLFy+Wvta3zCvTpUsX6Wu566hXVmeNRoOff/4ZPj4+sq5JRETy\nMIFb2Pjx49GxY8cKX/f29kZCQoLs5A0ATZs2BWDcEs7KykJQUBBOnz5daVk/P79yZc1lqs6pqanS\nBwf9IDciIrIcJnALKCkpkaZ4RUREWGwQl37u94kTJzB//nxpqthrr72GoKCgSsvu2rVLKnv06NEa\n1+XLL79EWlpaja9DRETmsdpKbH5+fvD09IRGo4GzszOSk5N/q5Sdr8Q2depUhISE4JVXXrFonLKb\ninTq1Mms/bxNlbXn+01E5Ajk5j4nC9alUiqVCvv370ejRo2sVQWLmTFjBjw8PBSN2ahRI7OTd1nV\n3Us8KSkJDRs2hL+/f7XKExFR9Vm1C92RWn1arRYPHz4EADRu3FiRuc/6XcTGjRuHDz/8UFby1ift\nnTt34rnnnqtW/Js3byI7O7taZYmIqGas1oXetm1beHl5QaPR4M0338Qbb7zxW6XssAt99erVuHz5\nMj777DPFYhp2g7do0QI3b95ULDYREdUuu+lCP3z4MHx8fHDnzh3ExMSgc+fO6N27t/T6rFmzpK+j\no6MRHR2tfCVlGDt2rLTutzXcunVLkThHjhzBmTNn8OabbyoSj4jIUe3fvx/79++vdnmb2E509uzZ\ncHd3x/vvvw/AflrgWq0WFy9erNG+1vHx8UhPT4ebmxs2btwoazqZYQvcw8MDjx49qnY9zJWZmYmM\njAz069fP4rGIiOoSu9hONC8vT1oT/MmTJ9i9eze6d+9ujarUyNmzZ40WUKkOuUupZmVlYd26deWO\nl11j3VL8/PyYvImIbIBVEvjt27fRu3dvBAUFITIyEi+88AIGDhxojarUSEhICL766qsaXUPOUqr6\ned6musuXLVtWo3pU5vz58xg1apTFrk9ERPLZRBd6Wbbcha7VarF582a8+uqr5eZSV8eDBw8QHx+P\nlStXVtp9bmo/7+XLl+Ptt9/GsmXL8NZbb9W4LhUpKSnB2bNnERgYaLEYRER1ndzcxwQu0/37jz+h\nkgAAGoVJREFU9zFv3jx89tlncHJSZgygqeStBK1Wy61AiYgUwgTuYIQQ6NevHwYOHKho8n7w4AH6\n9OmD5ORkuLq6KhaXiKiuYgK3gOLiYnz44Yf485//jIYNGyoePycnB40bN64zcYmI6iK7GIVubzQa\nDbp27YoGDRrU+rVVKpX053//939NnlNREjUsu3z58lqpz6+//lplXCIisj62wK2sJpuK1PaGJDqd\nDlFRUfj222/RqlWrGl2LiIjkYQu8FsXHxyMlJUWxeJ9++iny8vKqVbY2ppGp1WocOnSIyZuIyA4w\ngVfi7bfftvgCM59++qn09/jx4xESEmK0tWpl9Em7ptPITp48iSdPngAofVxARES2j13oZRQXF0Oj\n0dTKHG85rDVVDCjdv3zkyJF45plnFI1LRES/4Sj0Gpo5cybatm2LcePGKRbTmsmbiIhsAxN4DT18\n+BD169dXbAETw5a+n58fMjIyFIl7+PBhuLi4ICwsTJF4RERUObvZTtSWaLVaPH78GI0bN4aXl5fV\n6pGZmalYrJycHNSvX1+xeEREVLvYAgewadMmHD9+HH/9618Vi6lnrRY4ERHZFnahV4MQAiUlJYqt\nbW6oTZs2yMzMVCR5//zzzzh69Cjee+89i8YhIiL52IVuJq1Wi9TUVISHh0OlUlkleQNQtMXdunVr\nFBQUKBaPiIgsp87OA798+XKV+2/XtqysLHz55ZeKxjTk4+OD6Ohoq8UnIqLaU2cTuL+/P1atWqVY\nPP1UsXv37ikWEwDS09MxYsSIOrM0LRFRXVGnErhWq8XKlSuh0+kUjWvNed7t27fHnDlzFF+Yhuyb\nRqNBcHAwAgICEBsbi9zc3GpdZ+3atcjKyjL52pgxY9CgQQOja0+ePBlqtdqiH3Tz8vIwatQoBAQE\noHv37ujdu7e0EqG16e+7/s/8+fOtXSXJ4MGD0bBhQ7z44otVnvvdd99BrVbj5MmTCtSs7qpTCbyw\nsBDXr19HSUmJYjGtlbz1z7rVajX8/f0Vi0uOwc3NDadOnUJaWho8PT2xYsWKal3n66+/NtrhzpBK\npUKHDh3w/fffAyjdTOff//43fH19q11vcyxatAg+Pj5IS0vDmTNnsHr1ajg7O1ssXnFxsdnn6u+7\n/s/06dPLnVO2AWJug6SmDZfp06dj/fr1VZ73+PFjLFq0CD169KhRPKpanUrgHh4emDNnjkX/s5Y1\nevRoxZN3bm4uwsLCbKZVQfbtmWeewdWrVwEAp0+fRo8ePRAYGIjY2Fg8ePCgwuNbt27FiRMnMGrU\nKISEhJgcQPnSSy9h8+bNAID9+/cjKirKaD3+DRs2IDIyEsHBwZgwYYKUhN566y2Eh4ejW7dumDVr\nlnS+n58fZs2ahdDQUAQEBODSpUvlYmZnZ+Ppp5+Wvu/QoYO0cNPcuXPRqVMn9O7dG6+88goWLlwI\nAIiOjpY2Nrp79y7atGkDoHTthj59+iA0NBShoaE4evSo9LP07t0bw4YNQ7du3aDT6TBt2jREREQg\nMDBQ9vgbPz8/fPDBBwgNDcWWLVvKfb9p0yapR+GDDz6Qyrm7u2Pq1KkICgrCzz//LCtmWf3794e7\nu3uV53300Uf44IMP4OLiwkd3liZsUG1Wq7i4WEyePFncvn271q4px/379+tUXHIM7u7uQojS/z+x\nsbFi2bJlQgghunfvLg4ePCiEEOLjjz8WkydPrvR4dHS0SElJMRljzJgxYuvWraJHjx7i/v374o03\n3hAHDhwQfn5+IicnR5w/f168+OKLori4WAghxMSJE8W6deuEEELcu3dPql90dLQ4c+aMEEIIPz8/\nsXTpUiGEEMuXLxfjx48vF/f06dOiadOm4plnnhEffvihuHz5shBCiBMnToju3buL/Px88ejRI9G+\nfXuxcOHCcj/HnTt3hJ+fnxBCiLy8PFFQUCCEECI9PV2EhYUJIYTYt2+faNCggcjMzBRCCLFixQox\nZ84cIYQQBQUFIiwsTGRkZJSrm0ajEUFBQdKfhIQE6edasGCBdJ7h97du3RKtWrUSd+/eFcXFxaJ/\n//5i+/btQgghVCqV2LJli8n7v2DBAqNY+j+TJk0yeb7+53rhhRcqfD0lJUX87ne/K3fPyDxyc5/D\nTyPTaDTo2bMnvL29rRJfybjXrl1Dq1atoFKprPbzkmPIz89HcHAwbt26BT8/P0yYMAEPHz7Ew4cP\n0bt3bwClvUtxcXF49OiRyeN6oopWWGxsLDZt2oRjx45JXfVCCOzduxcpKSnScr/5+flo3rw5AGDz\n5s1YtWoViouLkZWVhfPnz6Nbt27S9QAgJCQE//znP8vFCwwMxC+//ILdu3fjp59+Qnh4OI4ePYpD\nhw4hNjYWrq6ucHV1xdChQ6u8T1qtFu+88w5SU1Oh0Whw+fJl6bWIiAi0bt0aALB7926cOXMGW7du\nBQA8evQIV65cgZ+fn9H16tevj1OnTpmM9dJLL5n8/vjx4+jXrx8aN24MABg1ahQOHjyIYcOGQaPR\nYOTIkSavN3XqVEydOrXKn9FcOp0OU6ZMwdq1a6VjVf3bU804fAIHYPTLxFEJIfD6669jxYoVaN++\nvbWrQ3ZOn0jy8/MxaNAgfP/99xgwYIDRORX9ci57vLIBlCqVCi+99BJCQ0MxZsyYcueOHj0a8+bN\nMzqWkZGBhQsX4sSJE/Dy8sLYsWONuuddXFwAlH54r+j5c4MGDTBixAiMGDECarUaO3fuhEajMaq7\n4ddOTk5S971hrL/97W/w8fHB+vXrUVJSAldXV6MYhpYuXYqYmJgK70VVyl5P/33ZxT+EENJ9dHV1\nrfD+L1iwABs3bix3vE+fPli0aJHJMpX9Wz5+/Bjnzp2TpqpmZ2dj6NCh+PHHHxESElLxD0bV5rDP\nwOPj43HkyBFFYz5+/NhqnzhVKhX27NnD5E21qn79+li8eDFmzpwJDw8PNGzYEElJSQCA9evXIzo6\nGp6eniaPA6XjTh49elTh9YUQaNWqFebOnWu0p71KpcKAAQOwdetW3LlzBwBw7949XL9+HY8fP0aD\nBg3g6emJ27dvIzExUdbPdOTIEdy/fx9AaQv6/Pnz8PPzQ58+fbB9+3YUFBTg8ePH+Ne//iWV8fPz\nw4kTJwBAakUDpS1pfa/AunXrKhwgO2jQICxfvlz6QJGeno68vDxZ9a5IeHg4Dhw4gJycHJSUlODb\nb79F3759qyw3bdo0owFz+j8VJW+g8ha1l5cX7ty5g4yMDGRkZKBHjx5M3hbmsC3wKVOmoG3btorF\n0482X7lyJfr06aNY3OPHj6NDhw7w9vaGWu2wn8dIYYYtraCgILRv3x4JCQlYu3YtJkyYgLy8PLRr\n1w5r1qwBgAqPjxkzBhMmTICbmxuOHDli1EI1jBMfH1/uWJcuXTBnzhwMHDgQOp0Ozs7OWL58OSIi\nIhAcHIzOnTujZcuWiIqKqvBnMNVivHr1KiZOnAghBHQ6HV544QWp2/2ll15CYGAgmjZtivDwcClh\nTZ06Fb///e+xcuVKPP/889J133rrLYwcORLr1q3D4MGDjQZ5GcYeP348MjMzERISAiEEmjZtim3b\ntpWrm/7Rhd5zzz1Xrgei7LV9fHzw+eefo1+/fhBC4IUXXpCmetXm9NHevXvj0qVLyM3NRcuWLbF6\n9WrExMTgk08+QVhYmFnTy6h2OdRa6EVFRdBoNIonMsP/JGvWrMGYMWMUi/3JJ58gJiamwl9iRFQ9\ns2fPhru7O95//31rV4XqCLm5z6GabPPmzcPf//53RWOWXaRi7NixisafPXs2kzeRhXABJLJlDtUC\nf/LkCZydnaU5nZam7zY3nGuqRAs8KSkJOp1O0a56IiKyrDrXAtdqtcjOzgZQOipTqeQNlHbZv/vu\nu9LzPqW6zwsLC1FYWGjxOEREZLvsvgW+fft27N27F0uWLLFwrYiIiCxHbgvc7hM4ULqAgKOPwE5O\nTsbu3bvx4YcfWrsqRERkAXWiC12r1eLQoUPS946evAGgXbt2Zs3tJCKiusEuM9+1a9ewbt06RRdN\nycrKwsKFCxVfqEUfr3HjxtJSlURERHaZwDt06IBVq1YpNsVDP9q8sLBQ0WklGRkZGDx4MNcTJiKi\ncuwmgWu1WixZskTRvbwB6+3nDZQu37hkyRLORSUionLsJoGXlJTgwYMHiiZwayVv/T7eKpUKHTt2\nVCwuERHZD4cYhW4pQ4cORY8ePRRN3vn5+QgPD8eRI0fg6empWFwiIrIuh5pGJoTAu+++iw8++AAt\nWrRQvB65ublGmxM4elwiIrIeh5pGplKpMHDgQDRp0sQq8ZVMoleuXJH+4Zi8iYioKjadwAHgxRdf\nVHR5VGsQQmDy5Mm4ePGitatCRER2wma70Pfu3Yv+/fsrFnP06NHIzMxEgwYNsHHjRnh7eysWGyhN\n4hxtTkRUdznMM/DCwkJFdxVr164d8vPzAQBxcXFISEiweNyff/4Zbdu2RdOmTS0ei4iIbJvDPANX\nekvQli1bAgDCwsKwcuVKRWIfOHAAFy5cUCQWERE5FpttgStRLcN53m+99Rbi4+OxcuVKxbvPiYiI\nHKYL3dLVstYiLUlJScjNzcXgwYMVi0lERLZPbu5zsmBdbJpKpcKUKVMQHx+vaFyNRgONRqNoTCIi\ncjx1tgVORERkSxxmEJsjSUlJUXwjFCIicmxM4Aro0KEDnn/+eWtXg4iIHEidSOBZWVmYM2eO4t3y\n+nienp7o1auXorGJiMixOXwC1482V6vViq50dvPmTfTt2xc6nU6xmEREVHc49CA2a00V08vIyECb\nNm0Uj0tERPaH88D/y1rJ+9GjR9zHm4iIZGMC/6+4uDgEBwcrmrwLCwsRFhaGAwcOoFGjRorFJSIi\n+8cE/l/5+fmoX79+LdXI9uMSEZF9q9PzwOPj4xEdHY0hQ4agsLBQsbgXL16UBqsxeRMRkRIcKoGn\np6fjwIEDSExMVHSJ1I8//hhnz55VLB4REZFDrIWek5ODhg0bws3NDYCyW4ICwObNmxWdokZERGT3\nLfCsrCz06tULiYmJ2LhxI+Li4rBnzx6Lbwl69OhR3LhxAwCYvImISHF23QI3nCqmX6o0ISFBkdgp\nKSkoKChAy5YtFYlHRERkyG5HoVt7kRYiIqLaVCdGoVsreSclJWHr1q2KxSMiIqqIXXahu7q6Yvr0\n6Xj99dcVjevh4YGCggJFYxIREZlit13oREREjqROdKEr6dSpU5g0aZK1q0FERGSELfAq5OfnIy0t\nDZGRkdauChEROTCHa4FnZWVhxowZiif0kpISAKVLozJ5ExGRrbFKAt+1axc6d+6MDh064C9/+UuF\n5+lHm3t4eCi6WEp2djZ69uyJ4uJixWJW1/79+61dBbvA+2Q+3ivz8D6Zh/fJchRP4CUlJXjnnXew\na9cunD9/Hps2bcKFCxfKnWfNed7NmzfHP//5Tzg52f4gff7nMA/vk/l4r8zD+2Qe3ifLUTyBJycn\no3379vDz84OzszNefvllfP/99+XOs0byvnfvnvR1ixYtFItLREQkl+IJ/NatW0bLj/r6+uLWrVvl\nzlM6eRcVFaF///64c+eOYjGJiIiqS/FR6N999x127dqFVatWAQA2bNiAY8eOYcmSJdI57du3x9Wr\nV5WsFhERkVW1a9cOV65cMft8xR/ytmjRQtrFCwBu3LgBX19fo3Pk/ABERER1keJd6GFhYbh8+TIy\nMzOh1WqxefNmDB06VOlqEBER2TXFW+BOTk5YunQpBg0ahJKSEowbNw5dunRRuhpERER2zSZXYiMi\nIqLK2dxKbOYu8kKAn58fAgICEBwcjIiICGtXx2a8/vrraNasGbp37y4du3fvHmJiYtCxY0cMHDgQ\nDx48sGINbYOp+zRr1iz4+voiODgYwcHB2LVrlxVraBtu3LiBfv36wd/fH926dcPixYsB8D1lSkX3\niu8rYwUFBYiMjERQUBC6du2KGTNmAJD/nrKpFnhJSQk6deqEn376CS1atEB4eDg2bdrELvYKtGnT\nBikpKWjUqJG1q2JTDh06BHd3d/zxj3/EmTNnAADTp0/HU089henTp+Mvf/kL7t+/j88//9zKNbUu\nU/dp9uzZ8PDwwJQpU6xcO9uRnZ2N7OxsBAUFITc3F6Ghodi+fTvWrFnD91QZFd2rhIQEvq/KyMvL\ng5ubG4qLixEVFYUvvvgCP/zwg6z3lE21wM1d5IV+Y0Ofv2xG79690bBhQ6NjP/zwA0aPHg0AGD16\nNLZv326NqtkUU/cJ4HuqrObNmyMoKAgA4O7uji5duuDWrVt8T5lQ0b0C+L4qy83NDQCg1WpRUlKC\nhg0byn5P2VQCN3eRFyqlUqnw7LPPIiwsTJpXT6bdvn0bzZo1AwA0a9YMt2/ftnKNbNeSJUsQGBiI\ncePGsVu4jMzMTJw6dQqRkZF8T1VBf6969OgBgO+rsnQ6HYKCgtCsWTPpsYPc95RNJXAlNyxxBIcP\nH8apU6eQmJiIZcuW4dChQ9aukl1QqVR8r1Vg4sSJyMjIwOnTp+Hj44P333/f2lWyGbm5uRg5ciQW\nLVoEDw8Po9f4njKWm5uL3/3ud1i0aBHc3d35vjJBrVbj9OnTuHnzJg4ePIh9+/YZvW7Oe8qmErg5\ni7zQb3x8fAAATZo0wYgRI5CcnGzlGtmuZs2aITs7G0DpRjlNmza1co1sU9OmTaVfHOPHj+d76r+K\nioowcuRIvPbaaxg+fDgAvqcqor9Xr776qnSv+L6qmJeXF55//nmkpKTIfk/ZVALnIi/my8vLw+PH\njwEAT548we7du41GE5OxoUOHYu3atQCAtWvXSr9YyFhWVpb09bZt2/ieQumz23HjxqFr166YPHmy\ndJzvqfIquld8Xxm7e/eu9BghPz8fe/bsQXBwsPz3lLAxO3fuFB07dhTt2rUT8+bNs3Z1bNYvv/wi\nAgMDRWBgoPD39+e9MvDyyy8LHx8f4ezsLHx9fcXq1atFTk6OGDBggOjQoYOIiYkR9+/ft3Y1ra7s\nffrHP/4hXnvtNdG9e3cREBAghg0bJrKzs61dTas7dOiQUKlUIjAwUAQFBYmgoCCRmJjI95QJpu7V\nzp07+b4qIy0tTQQHB4vAwEDRvXt3MX/+fCGEkP2esqlpZERERGQem+pCJyIiIvMwgRMREdkhJnAi\nIiI7xARORERkh5jAiYiI7BATOBERkR1iAieqoZycHGmbRB8fH2nbxJCQEBQXF1u7epWaNWsWFi5c\naPE4hw4dgr+/P0JCQlBQUFDpuc8//zwePXpk8ToR2Tsna1eAyN41btwYp06dAmB6O86SkhJoNBpr\nVa9SNVm/W7+EhDnX+Oabb/A///M/GDVqVJXX27FjR7XrRFSXsAVOVMuEEBgzZgwmTJiAHj16YPr0\n6Zg9e7ZRS7dbt264fv06AGDDhg2IjIxEcHAwJkyYAJ1OV+6afn5+mDVrFkJDQxEQEIBLly4BKN+C\n1l83MzMTnTt3xtixY9GpUyeMGjUKu3fvRq9evdCxY0ccP35cKpOamoqePXuiY8eO+Oqrr6TjCxYs\nQEREBAIDAzFr1iwApTtMderUCaNHj0b37t1x8+ZNo3ru3bsXISEhCAgIwLhx46DVavHVV19hy5Yt\n+Oijj/Dqq68anV/2ejdu3ICfnx/u3buHzMxMdOnSBfHx8ejWrRsGDRoktd6PHz+OgIAABAcHY9q0\nadLSnOfOnZPuZWBgIK5cuSL734/IXjCBE1mASqXCr7/+iqNHj5rsota3Wi9cuICEhAQcOXIEp06d\nglqtxjfffGPy/CZNmiAlJQUTJ07EF198YXSdstcFgKtXr2Lq1Km4ePEiLl26hM2bN+Pw4cP44osv\nMG/ePAClHzbS0tKwb98+HD16FJ9++imysrKwe/duXLlyBcnJyTh16hRSUlKk3e6uXLmCt99+G2fP\nnjXa/regoABjx45FQkIC0tLSUFxcjL///e8YP348hg4dii+++AIbNmwo97MZXq9Vq1ZGP8OVK1fw\nzjvv4OzZs/D29sZ3330HABg7dixWrVqFU6dOwcnJSSrz5ZdfYtKkSVKduRkSOTImcCILiYuLq7R7\nWQiBvXv3IiUlBWFhYQgODsa///1vZGRkmDw/NjYWABASEoLMzMwq47dp0wb+/v5QqVTw9/fHs88+\nC6C0la4vr1KpMHz4cLi4uKBx48bo168fkpOTsXv3buzevRvBwcEIDQ3FpUuXpNZs69atERERUS7e\npUuX0KZNG7Rv3x4AMHr0aBw8eNDo5zWlouvpf4aAgAAAQGhoKDIzM/Hw4UPk5uYiMjISAPDKK69I\n1+7ZsyfmzZuH+fPnIzMzE66urlXeJyJ7xWfgRBbi5uYmfe3k5GTUNW44kGv06NFSi7gyLi4uAACN\nRiMNjqvsuvrzgdK9h+vVqyd9XdngOv2HjhkzZiA+Pt7otczMTDRo0KDScnrmbrNQ0fUA459Bo9Eg\nPz+/3DmGcf7whz+gR48e+Ne//oUhQ4ZgxYoV6Nevn1n1ILI3bIETKcDPzw8nT54EAJw8eRIZGRlQ\nqVQYMGAAtm7dijt37gAA7t27Jz0br+515RBC4Pvvv0dhYSFycnKwf/9+REREYNCgQVi9ejWePHkC\nALh165ZUx4p07NgRmZmZuHr1KgBg/fr1iI6OllUfc3h5ecHDw0PaU/rbb7+VXvvll1/Qpk0b/OlP\nf8KwYcNw5syZWo9PZCvYAieyEMMW6ciRI7Fu3Tp069YNkZGR6NSpEwCgS5cumDNnDgYOHAidTgdn\nZ2csX74crVq1qvBaKpVK+r6i65YtY+oa+r8DAgLQr18/3L17Fx9//DGaN2+O5s2b48KFC3jmmWcA\nAB4eHtiwYYNR7LJcXV2xZs0axMXFobi4GBEREZgwYUKF9ZFTz7Lf/+Mf/8Abb7wBtVqNvn37wsvL\nCwCQkJCADRs2wNnZGT4+Ppg5c6bJmESOgNuJEpHdefLkidT1/vnnn+P27dv429/+ZuVaESmLLXAi\nsjs7duzAZ599huLiYvj5+eHrr7+2dpWIFMcWOBERkR3iIDYiIiI7xARORERkh5jAiYiI7BATOBER\nkR1iAiciIrJD/w+KsRB3+h96rQAAAABJRU5ErkJggg==\n", "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAH4CAYAAAC1/7+LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nuglzAAAAAElFTkSuQmCC\n", "text": [ "" ] } ], "prompt_number": 12 }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Plot learning curves" ] }, { "cell_type": "code", "collapsed": false, "input": [ "data_percentage_array = np.linspace(10, 100, 10)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 13 }, { "cell_type": "code", "collapsed": false, "input": [ "train_error = []\n", "test_error = []\n", "for data_percentage in data_percentage_array:\n", " model = DecisionTreeRegressor(max_depth=10)\n", " number_of_samples = int(data_percentage / 100. * len(train_y))\n", " model.fit(train_X[:number_of_samples,:], train_y[:number_of_samples])\n", "\n", " predicted_train_y = model.predict(train_X)\n", " predicted_test_y = model.predict(test_X)\n", "\n", " train_error.append((predicted_train_y - train_y).std())\n", " test_error.append((predicted_test_y - test_y).std())" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 14 }, { "cell_type": "code", "collapsed": false, "input": [ "plt.plot(data_percentage_array, train_error, label='training')\n", "plt.plot(data_percentage_array, test_error, label='validation')\n", "plt.grid()\n", "plt.legend(loc=3)\n", "plt.xlabel('Data percentage')\n", "plt.ylabel('Root mean square error');" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAEPCAYAAACp/QjLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKg6EEEJkUHEghBAi4/8B8Dkc9mY9JuEAAAAASUVORK5CYII=\n", "text": [ "" ] } ], "prompt_number": 15 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Inspecting the learning curve, we see that the training error decreases as the `DecisionTree` fits better the data, but the validation error shows a significant gap compared to training error.\n", "That typically means the model is not fitting the data well." ] }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Fit a Random Forest estimator to the data" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In this case it is useful to test a more sophisticated model, for example Random Forests, i.e. a method that trains several Decision Trees and averages them.\n", "\n", "`sklearn` has a very consistent interface, for any estimator, call `fit` for training then `predict` to estimate." ] }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn.ensemble import RandomForestRegressor\n", "model = RandomForestRegressor(n_estimators=5)\n", "model.fit(train_X, train_y)\n", "predicted_test_y = model.predict(test_X)\n", "rms_random_forest = scatter_y(test_y, predicted_test_y)" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAHuCAYAAABgcF4NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIrJB/w+7M/CgTl+mogAAAABJRU5ErkJggg==\n", "text": [ "" ] } ], "prompt_number": 16 }, { "cell_type": "code", "collapsed": false, "input": [ "print \"Root Mean Square error decreases from %.2g to %.2g.\" % (rms_decision_tree, rms_random_forest)" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "Root Mean Square error decreases from 2.7 to 2.5.\n" ] } ], "prompt_number": 17 }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Optimize model parameters" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "However we have not yet performed any optimization of the Random Forest parameters:\n", "\n", " * `max_depth`: the maximum depth of the Decision Trees\n", " * `max_features`: the max number of features to consider for each split" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Given a list (or distributions) of parameters, `RandomizedSearchCV` performs a randomized search (less expensive than an exhaustive grid search) for the best parameters.\n", "\n", "`RandomizedSearchCV` uses cross validation, i.e. randomly splits the input data in order to estimate the performance of each model more robustly." ] }, { "cell_type": "code", "collapsed": false, "input": [ "model = RandomForestRegressor(n_estimators=100)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 18 }, { "cell_type": "code", "collapsed": false, "input": [ "n_features = X.shape[1]" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 19 }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn.grid_search import RandomizedSearchCV\n", "grid = RandomizedSearchCV(model, n_iter=20, \n", " param_distributions=dict(\n", " max_depth=np.arange(5,20+1), \n", " max_features=np.arange(1, n_features+1)\n", " )\n", " )\n", "grid.fit(X, y)\n", "print grid.best_params_" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "{'max_features': 5, 'max_depth': 9}\n" ] } ], "prompt_number": 20 }, { "cell_type": "code", "collapsed": false, "input": [ "model = RandomForestRegressor(max_features=grid.best_params_[\"max_features\"],\n", " max_depth=grid.best_params_[\"max_depth\"])\n", "model.fit(train_X, train_y)\n", "predicted_test_y = model.predict(test_X)\n", "rms_optimized_random_forest = scatter_y(test_y, predicted_test_y)" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAfAAAAHuCAYAAABgcF4NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niAGciIjIDjGAExER2SEGcCIiIjv0/6vw1GUxB48jAAAAAElFTkSuQmCC\n", "text": [ "" ] } ], "prompt_number": 21 }, { "cell_type": "code", "collapsed": false, "input": [ "print \"Root Mean Square error decreases from %.2g to %.2g.\" % (rms_random_forest, rms_optimized_random_forest)" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "Root Mean Square error decreases from 2.5 to 2.3.\n" ] } ], "prompt_number": 22 }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Conclusion" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In this tutorial we learnt about:\n", "\n", " * Prepare `y` vector of outputs and `X` matrix of features for `sklearn`\n", " * Split train and test sets: `cross_validation.train_test_split(X, y)`\n", " * Train an estimator: `model.fit(train_X, train_y)`\n", " * Use a trained estimator to predict unknown output values: `model.predict(test_X)`\n", " * Use `RandomizedSearchCV` to optimize model parameters" ] } ], "metadata": {} } ] }