{
 "metadata": {
  "name": "",
  "signature": "sha256:eba89a5303a3bfeb8cd3c5e2ffb4d81a96cbed610317fe5c18d05898ff4e3a35"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import pandas as pd\n",
      "import numpy as np\n",
      "import matplotlib.pyplot as plt\n",
      "\n",
      "import folium as fm\n",
      "import geopy\n",
      "\n",
      "from IPython.display import HTML\n",
      "\n",
      "from sklearn import linear_model, naive_bayes, feature_selection, metrics, tree\n",
      "from sklearn.cross_validation import train_test_split\n",
      "\n",
      "%matplotlib inline"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "tips_adj_df = pd.read_pickle('./dumps/tips_complete_features.pkl')\n",
      "zipcode_dummies = pd.get_dummies(tips_adj_df['ZIPCODE'].astype(int))\n",
      "tips_adj_df = tips_adj_df.join(zipcode_dummies)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 34
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "# Predicting Grade \"A\" Restaurants by zipcode"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "zip_score_list_A = []\n",
      "\n",
      "for zipcode in zipcode_dummies.columns.values:\n",
      "    tip_zip_subset = tips_adj_df[tips_adj_df[zipcode] == 1]\n",
      "    \n",
      "    X_adjs = tip_zip_subset.ix[:, 34:289]\n",
      "    y = tip_zip_subset['grade_A']\n",
      "    \n",
      "    clf = naive_bayes.MultinomialNB()\n",
      "    clf = clf.fit(X_adjs, y)\n",
      "    score = clf.score(X_adjs, y)\n",
      "    \n",
      "    zip_score_list_A.append({'zipcode': zipcode, 'score': score, 'tip_count': len(tip_zip_subset)})"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 35
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "tip_by_zip_df_A = pd.DataFrame(zip_score_list_A)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 36
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "map = fm.Map(location=[40.714623, -74.006605], zoom_start=11, tiles='Stamen Toner')\n",
      "map.geo_json(geo_path='./nyc-zip-code-tabulation-areas-polygons.geojson',\n",
      "             data=tip_by_zip_df_A, columns=['zipcode', 'score'],\n",
      "             key_on='feature.properties.postalCode',\n",
      "             threshold_scale=[0.50, 0.8, 0.85, 0.90, 0.95, 0.99],\n",
      "             fill_color='RdPu', fill_opacity=0.65, line_opacity=0.5)\n",
      "map.create_map(path='foursquare_zips_all_A.html')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 37
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "HTML('<iframe src=foursquare_zips_all_A.html width=1000 height = 500></iframe>')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<iframe src=foursquare_zips_all_A.html width=1000 height = 500></iframe>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 38,
       "text": [
        "<IPython.core.display.HTML at 0x10dbaeed0>"
       ]
      }
     ],
     "prompt_number": 38
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "tip_threshold_by_zip_A_df = tip_by_zip_df_A[tip_by_zip_df_A['tip_count'] >= 50]\n",
      "\n",
      "map = fm.Map(location=[40.714623, -74.006605], zoom_start=11, tiles='Stamen Toner')\n",
      "map.geo_json(geo_path='./nyc-zip-code-tabulation-areas-polygons.geojson',\n",
      "             data=tip_threshold_by_zip_A_df, columns=['zipcode', 'score'],\n",
      "             key_on='feature.properties.postalCode',\n",
      "             threshold_scale=[0.50, 0.8, 0.85, 0.90, 0.95, 0.99],\n",
      "             fill_color='RdPu', fill_opacity=0.65, line_opacity=0.5)\n",
      "map.create_map(path='foursquare_zips_threshold_A.html')\n",
      "\n",
      "HTML('<iframe src=foursquare_zips_threshold_A.html width=1000 height = 500></iframe>')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<iframe src=foursquare_zips_threshold_A.html width=1000 height = 500></iframe>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 39,
       "text": [
        "<IPython.core.display.HTML at 0x11365ee50>"
       ]
      }
     ],
     "prompt_number": 39
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "# Predicting Grade \"C\" Restaurants by zipcode"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "zip_score_list_C = []\n",
      "\n",
      "for zipcode in zipcode_dummies.columns.values:\n",
      "    tip_zip_subset = tips_adj_df[tips_adj_df[zipcode] == 1]\n",
      "    \n",
      "    X_adjs = tip_zip_subset.ix[:, 34:289]\n",
      "    y = tip_zip_subset['grade_C']\n",
      "    \n",
      "    clf = naive_bayes.MultinomialNB()\n",
      "    clf = clf.fit(X_adjs, y)\n",
      "    score = clf.score(X_adjs, y)\n",
      "    \n",
      "    zip_score_list_C.append({'zipcode': zipcode, 'score': score, 'tip_count': len(tip_zip_subset)})"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 40
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "tip_by_zip_C_df = pd.DataFrame(zip_score_list_C)\n",
      "\n",
      "map = fm.Map(location=[40.714623, -74.006605], zoom_start=11, tiles='Stamen Toner')\n",
      "map.geo_json(geo_path='./nyc-zip-code-tabulation-areas-polygons.geojson',\n",
      "             data=tip_by_zip_C_df, columns=['zipcode', 'score'],\n",
      "             key_on='feature.properties.postalCode',\n",
      "             threshold_scale=[0.50, 0.8, 0.85, 0.90, 0.95, 0.99],\n",
      "             fill_color='BuPu', fill_opacity=0.65, line_opacity=0.5)\n",
      "map.create_map(path='foursquare_zips_all_C.html')\n",
      "\n",
      "HTML('<iframe src=foursquare_zips_all_C.html width=1000 height = 500></iframe>')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<iframe src=foursquare_zips_all_C.html width=1000 height = 500></iframe>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 41,
       "text": [
        "<IPython.core.display.HTML at 0x11353c150>"
       ]
      }
     ],
     "prompt_number": 41
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "tip_threshold_by_zip_C_df = tip_by_zip_C_df[tip_by_zip_C_df['tip_count'] >= 50]\n",
      "\n",
      "map = fm.Map(location=[40.714623, -74.006605], zoom_start=11, tiles='Stamen Toner')\n",
      "map.geo_json(geo_path='./nyc-zip-code-tabulation-areas-polygons.geojson',\n",
      "             data=tip_threshold_by_zip_C_df, columns=['zipcode', 'score'],\n",
      "             key_on='feature.properties.postalCode',\n",
      "             threshold_scale=[0.50, 0.8, 0.85, 0.90, 0.95, 0.99],\n",
      "             fill_color='BuPu', fill_opacity=0.65, line_opacity=0.5)\n",
      "map.create_map(path='foursquare_zips_threshold_C.html')\n",
      "\n",
      "HTML('<iframe src=foursquare_zips_threshold_C.html width=1000 height = 500></iframe>')"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "html": [
        "<iframe src=foursquare_zips_threshold_C.html width=1000 height = 500></iframe>"
       ],
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 42,
       "text": [
        "<IPython.core.display.HTML at 0x113667190>"
       ]
      }
     ],
     "prompt_number": 42
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}