07_data_prep.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Connecting Datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import datetime as dt\n",
    "from sqlalchemy import create_engine\n",
    "from sqlalchemy import inspect\n",
    "import geopandas\n",
    "import geopandas.tools\n",
    "from shapely.geometry import Point\n",
    "import pickle\n",
    "cnx = create_engine('')\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### Collect and preprocess data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Run SQL query to collect all 196,000 rows of data from the database on AWS, then tidy up the column names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "df = pd.read_sql_query('''SELECT * FROM incidents''', cnx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['incident_num', 'category', 'description', 'day_of_week',\n",
       "       'indicent_date', 'incident_time', 'pd_district', 'resolution',\n",
       "       'address', 'x', 'y', 'location', 'pd_id'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "df.rename(columns={'incident_num':'incident_num','category':'category',\n",
    "                   'description':'description','day_of_week':'day_of_week',\n",
    "                   'indicent_date':'incident_date','incident_time':'incident_time',\n",
    "                   'pd_district':'pd_district','resolution':'resolution',\n",
    "                   'address':'address','x':'longitude','y':'latitude',\n",
    "                   'location':'location','pd_id':'pd_id'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>incident_num</th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>incident_date</th>\n",
       "      <th>incident_time</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>location</th>\n",
       "      <th>pd_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>170007304.0</td>\n",
       "      <td>SUSPICIOUS OCC</td>\n",
       "      <td>SUSPICIOUS OCCURRENCE</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>2017-01-03</td>\n",
       "      <td>17:40:00</td>\n",
       "      <td>MISSION</td>\n",
       "      <td>NONE</td>\n",
       "      <td>3700 Block of 17TH ST</td>\n",
       "      <td>-122.429634</td>\n",
       "      <td>37.762765</td>\n",
       "      <td>(37.7627651977237,-122.429633638174)</td>\n",
       "      <td>1.700000e+13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>170007326.0</td>\n",
       "      <td>NON-CRIMINAL</td>\n",
       "      <td>DEATH REPORT, NATURAL CAUSES</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>2017-01-03</td>\n",
       "      <td>18:07:00</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>HARRISON ST / ESSEX ST</td>\n",
       "      <td>-122.394105</td>\n",
       "      <td>37.785151</td>\n",
       "      <td>(37.785150776928,-122.394104744923)</td>\n",
       "      <td>1.700000e+13</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   incident_num        category                   description day_of_week  \\\n",
       "0   170007304.0  SUSPICIOUS OCC         SUSPICIOUS OCCURRENCE     Tuesday   \n",
       "1   170007326.0    NON-CRIMINAL  DEATH REPORT, NATURAL CAUSES     Tuesday   \n",
       "\n",
       "  incident_date incident_time pd_district resolution                 address  \\\n",
       "0    2017-01-03      17:40:00     MISSION       NONE   3700 Block of 17TH ST   \n",
       "1    2017-01-03      18:07:00    SOUTHERN       NONE  HARRISON ST / ESSEX ST   \n",
       "\n",
       "    longitude   latitude                              location         pd_id  \n",
       "0 -122.429634  37.762765  (37.7627651977237,-122.429633638174)  1.700000e+13  \n",
       "1 -122.394105  37.785151   (37.785150776928,-122.394104744923)  1.700000e+13  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Combine the incident date and incident time columns into a combined date_time column, then create new columns for the hour of the day and month of the incident."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "df['date_time'] = df.apply(lambda r : pd.datetime.combine(r['incident_date'],\n",
    "                                                          r['incident_time']),1)\n",
    "df['hour_of_day'] = [r.hour for r in df['date_time']]\n",
    "df['month'] = df['date_time'].apply(lambda x: dt.datetime.strftime(x,'%b'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Finally, remove the unnecessary columns that I won't be using."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>60963</th>\n",
       "      <td>ASSAULT</td>\n",
       "      <td>THREATENING PHONE CALL(S)</td>\n",
       "      <td>Wednesday</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>1600 Block of MISSION ST</td>\n",
       "      <td>-122.419230</td>\n",
       "      <td>37.772033</td>\n",
       "      <td>2016-03-30 09:00:00</td>\n",
       "      <td>9</td>\n",
       "      <td>Mar</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55937</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Monday</td>\n",
       "      <td>NORTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>1300 Block of WEBSTER ST</td>\n",
       "      <td>-122.431046</td>\n",
       "      <td>37.783030</td>\n",
       "      <td>2016-03-28 16:40:00</td>\n",
       "      <td>16</td>\n",
       "      <td>Mar</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            category                   description day_of_week pd_district  \\\n",
       "60963        ASSAULT     THREATENING PHONE CALL(S)   Wednesday    SOUTHERN   \n",
       "55937  LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO      Monday    NORTHERN   \n",
       "\n",
       "      resolution                   address   longitude   latitude  \\\n",
       "60963       NONE  1600 Block of MISSION ST -122.419230  37.772033   \n",
       "55937       NONE  1300 Block of WEBSTER ST -122.431046  37.783030   \n",
       "\n",
       "                date_time  hour_of_day month  \n",
       "60963 2016-03-30 09:00:00            9   Mar  \n",
       "55937 2016-03-28 16:40:00           16   Mar  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.drop(['incident_num','incident_date','incident_time','location','pd_id'], axis=1, inplace=True)\n",
    "df.sample(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### Feature engineering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "There are too many individual categories, so I will bucket them into broader categories."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def rename_categories(input_val):\n",
    "    if input_val in ['SEX OFFENSES, FORCIBLE','SEX OFFENSES, NON FORCIBLE','ROBBERY',\n",
    "                     'ASSAULT','KIDNAPPING','SUICIDE','EXTORTION','BRIBERY','MISSING PERSON']:\n",
    "        return 'OFFENSES AGAINST THE PERSON'\n",
    "    elif input_val in ['BAD CHECKS','EMBEZZLEMENT','FORGERY/COUNTERFEITING','FRAUD']:\n",
    "        return 'WHITE COLLAR CRIME'\n",
    "#     elif input_val in ['LARCENY/THEFT','VEHICLE THEFT']:\n",
    "#         return 'THEFT'\n",
    "    elif input_val in ['BURGLARY','ARSON','VANDALISM','STOLEN PROPERTY','TRESPASS']:\n",
    "        return 'PROPERTY CRIME'\n",
    "    elif input_val in ['DRUG/NARCOTIC','PROSTITUTION','GAMBLING','PORNOGRAPHY/OBSCENE MAT',\n",
    "                       'DISORDERLY CONDUCT','DRUNKENNESS','DRIVING UNDER THE INFLUENCE',\n",
    "                       'LIQUOR LAWS']:\n",
    "        return 'SOCIAL ORDER OFFENSES'\n",
    "    elif input_val == 'SECONDARY CODES':\n",
    "        return 'DOMESTIC VIOLENCE AND JUVENILE CRIME'\n",
    "    elif input_val in ['SUSPICIOUS OCC','LOITERING','TREA']:\n",
    "        return 'SUSPICIOUS BEHAVIOR'\n",
    "    elif input_val in ['RUNAWAY','FAMILY OFFENSES','RECOVERED VEHICLE','WEAPON LAWS',\n",
    "                       'WARRANTS']:\n",
    "        return 'OTHER OFFENSES'\n",
    "    else:\n",
    "        return input_val\n",
    "\n",
    "df['broad_category'] = df['category'].apply(rename_categories)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Remove some categories that are not useful geography-wise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "df = df[df['broad_category'] != 'SUSPICIOUS BEHAVIOR']\n",
    "df = df[df['broad_category'] != 'OTHER OFFENSES']\n",
    "df = df[df['broad_category'] != 'NON-CRIMINAL']\n",
    "df = df[df['broad_category'] != 'WHITE COLLAR CRIME']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "I will be doing a lot of geospatial data processing, so I will convert the latitude and longitude columns into Points using Shapely. I will also create an arbitrary 'id' column so that I have a 'primary key' to be able to work with, should I ever need it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>month</th>\n",
       "      <th>broad_category</th>\n",
       "      <th>geometry</th>\n",
       "      <th>id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>57913</th>\n",
       "      <td>VANDALISM</td>\n",
       "      <td>MALICIOUS MISCHIEF, VANDALISM OF VEHICLES</td>\n",
       "      <td>Saturday</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>800 Block of BRYANT ST</td>\n",
       "      <td>-122.403405</td>\n",
       "      <td>37.775421</td>\n",
       "      <td>2016-04-02 02:00:00</td>\n",
       "      <td>2</td>\n",
       "      <td>Apr</td>\n",
       "      <td>PROPERTY CRIME</td>\n",
       "      <td>POINT (-122.4034048 37.77542071)</td>\n",
       "      <td>57913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32960</th>\n",
       "      <td>ARSON</td>\n",
       "      <td>ARSON OF AN INHABITED DWELLING</td>\n",
       "      <td>Monday</td>\n",
       "      <td>TARAVAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>100 Block of CAMBON DR</td>\n",
       "      <td>-122.474171</td>\n",
       "      <td>37.716657</td>\n",
       "      <td>2017-04-17 19:01:00</td>\n",
       "      <td>19</td>\n",
       "      <td>Apr</td>\n",
       "      <td>PROPERTY CRIME</td>\n",
       "      <td>POINT (-122.474171 37.71665697)</td>\n",
       "      <td>32960</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        category                                description day_of_week  \\\n",
       "57913  VANDALISM  MALICIOUS MISCHIEF, VANDALISM OF VEHICLES    Saturday   \n",
       "32960      ARSON             ARSON OF AN INHABITED DWELLING      Monday   \n",
       "\n",
       "      pd_district resolution                 address   longitude   latitude  \\\n",
       "57913    SOUTHERN       NONE  800 Block of BRYANT ST -122.403405  37.775421   \n",
       "32960     TARAVAL       NONE  100 Block of CAMBON DR -122.474171  37.716657   \n",
       "\n",
       "                date_time  hour_of_day month  broad_category  \\\n",
       "57913 2016-04-02 02:00:00            2   Apr  PROPERTY CRIME   \n",
       "32960 2017-04-17 19:01:00           19   Apr  PROPERTY CRIME   \n",
       "\n",
       "                               geometry     id  \n",
       "57913  POINT (-122.4034048 37.77542071)  57913  \n",
       "32960   POINT (-122.474171 37.71665697)  32960  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['geometry'] = df.apply(lambda row: Point(row['longitude'], row['latitude']), axis=1)\n",
    "df['id'] = df.index\n",
    "df.sample(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Geospatial data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "First up is to connect the zip code + population geospatial data that I downloaded from the SF open data website. I will be using geopandas to work with all of my geospatial data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "zip_df = geopandas.read_file('zip.geojson')\n",
    "zip_df.drop(['id','multigeom','objectid','po_name','st_area_sh','st_length_','state',\n",
    "             'zip'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>geometry</th>\n",
       "      <th>pop10_sqmi</th>\n",
       "      <th>pop2010</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>POLYGON ((-122.4555040090645 37.76091139152248...</td>\n",
       "      <td>11153.0</td>\n",
       "      <td>24202</td>\n",
       "      <td>2.17</td>\n",
       "      <td>94131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>POLYGON ((-122.3917009776518 37.79411300167845...</td>\n",
       "      <td>46226.7</td>\n",
       "      <td>20802</td>\n",
       "      <td>0.45</td>\n",
       "      <td>94105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>POLYGON ((-122.3908201190462 37.7766660063356,...</td>\n",
       "      <td>9522.0</td>\n",
       "      <td>4761</td>\n",
       "      <td>0.5</td>\n",
       "      <td>94158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>POLYGON ((-122.4244074100221 37.81072189731741...</td>\n",
       "      <td>19925.4</td>\n",
       "      <td>22715</td>\n",
       "      <td>1.14</td>\n",
       "      <td>94123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>POLYGON ((-122.486279012768 37.73682097594768,...</td>\n",
       "      <td>6440.6</td>\n",
       "      <td>26986</td>\n",
       "      <td>4.19</td>\n",
       "      <td>94132</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             geometry pop10_sqmi pop2010  \\\n",
       "2   POLYGON ((-122.4555040090645 37.76091139152248...    11153.0   24202   \n",
       "13  POLYGON ((-122.3917009776518 37.79411300167845...    46226.7   20802   \n",
       "7   POLYGON ((-122.3908201190462 37.7766660063356,...     9522.0    4761   \n",
       "29  POLYGON ((-122.4244074100221 37.81072189731741...    19925.4   22715   \n",
       "3   POLYGON ((-122.486279012768 37.73682097594768,...     6440.6   26986   \n",
       "\n",
       "    sqmi zip_code  \n",
       "2   2.17    94131  \n",
       "13  0.45    94105  \n",
       "7    0.5    94158  \n",
       "29  1.14    94123  \n",
       "3   4.19    94132  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zip_df.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1130c0b70>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAUwAAAD8CAYAAAAc052eAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113d959b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "zip_df.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>geometry</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>146322</th>\n",
       "      <td>146322</td>\n",
       "      <td>POINT (-122.4815173 37.78117781)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>163259</th>\n",
       "      <td>163259</td>\n",
       "      <td>POINT (-122.4039949 37.79461182)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            id                          geometry\n",
       "146322  146322  POINT (-122.4815173 37.78117781)\n",
       "163259  163259  POINT (-122.4039949 37.79461182)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create a GeoDataFrame of the geometry data from the original data-set\n",
    "gdf = geopandas.GeoDataFrame(df[['id','geometry']], geometry='geometry')\n",
    "gdf.crs = {\"init\": \"epsg:4326\"}\n",
    "gdf.sample(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>pop10_sqmi</th>\n",
       "      <th>pop2010</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>61875</th>\n",
       "      <td>61875</td>\n",
       "      <td>22830.1</td>\n",
       "      <td>30364</td>\n",
       "      <td>1.33</td>\n",
       "      <td>94103</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10752</th>\n",
       "      <td>10752</td>\n",
       "      <td>6440.6</td>\n",
       "      <td>26986</td>\n",
       "      <td>4.19</td>\n",
       "      <td>94132</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          id pop10_sqmi pop2010  sqmi zip_code\n",
       "61875  61875    22830.1   30364  1.33    94103\n",
       "10752  10752     6440.6   26986  4.19    94132"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# do a spatial join on the two GeoDataFrames\n",
    "result = geopandas.tools.sjoin(gdf, zip_df, how='left')\n",
    "result.drop(['geometry','index_right'], axis=1, inplace=True)\n",
    "result.sample(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>month</th>\n",
       "      <th>broad_category</th>\n",
       "      <th>geometry</th>\n",
       "      <th>id</th>\n",
       "      <th>pop10_sqmi</th>\n",
       "      <th>pop2010</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ASSAULT</td>\n",
       "      <td>BATTERY</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>TARAVAL</td>\n",
       "      <td>ARREST, BOOKED</td>\n",
       "      <td>3200 Block of 20TH AV</td>\n",
       "      <td>-122.475647</td>\n",
       "      <td>37.728528</td>\n",
       "      <td>2017-01-03 17:43:00</td>\n",
       "      <td>17</td>\n",
       "      <td>Jan</td>\n",
       "      <td>OFFENSES AGAINST THE PERSON</td>\n",
       "      <td>POINT (-122.4756475 37.72852806)</td>\n",
       "      <td>2</td>\n",
       "      <td>6440.6</td>\n",
       "      <td>26986</td>\n",
       "      <td>4.19</td>\n",
       "      <td>94132</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>PETTY THEFT SHOPLIFTING</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>TARAVAL</td>\n",
       "      <td>ARREST, BOOKED</td>\n",
       "      <td>3200 Block of 20TH AV</td>\n",
       "      <td>-122.475647</td>\n",
       "      <td>37.728528</td>\n",
       "      <td>2017-01-03 17:43:00</td>\n",
       "      <td>17</td>\n",
       "      <td>Jan</td>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>POINT (-122.4756475 37.72852806)</td>\n",
       "      <td>3</td>\n",
       "      <td>6440.6</td>\n",
       "      <td>26986</td>\n",
       "      <td>4.19</td>\n",
       "      <td>94132</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        category              description day_of_week pd_district  \\\n",
       "0        ASSAULT                  BATTERY     Tuesday     TARAVAL   \n",
       "1  LARCENY/THEFT  PETTY THEFT SHOPLIFTING     Tuesday     TARAVAL   \n",
       "\n",
       "       resolution                address   longitude   latitude  \\\n",
       "0  ARREST, BOOKED  3200 Block of 20TH AV -122.475647  37.728528   \n",
       "1  ARREST, BOOKED  3200 Block of 20TH AV -122.475647  37.728528   \n",
       "\n",
       "            date_time  hour_of_day month               broad_category  \\\n",
       "0 2017-01-03 17:43:00           17   Jan  OFFENSES AGAINST THE PERSON   \n",
       "1 2017-01-03 17:43:00           17   Jan                LARCENY/THEFT   \n",
       "\n",
       "                           geometry  id pop10_sqmi pop2010  sqmi zip_code  \n",
       "0  POINT (-122.4756475 37.72852806)   2     6440.6   26986  4.19    94132  \n",
       "1  POINT (-122.4756475 37.72852806)   3     6440.6   26986  4.19    94132  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# reconnect with original data-set\n",
    "df = df.merge(result, left_on='id', right_on='id')\n",
    "df['zip_code'] = df['zip_code'].apply(lambda x: str(x))\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Income data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Next task is to connect the median income by zip code data that I downloaded from another source."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>zip_code</th>\n",
       "      <th>median_income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>94127</td>\n",
       "      <td>95313.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>94105</td>\n",
       "      <td>88976.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  zip_code  median_income\n",
       "0    94127        95313.0\n",
       "1    94105        88976.0"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "income_df = pd.read_csv('median_income_data.csv')\n",
    "income_df.drop(['latitude','longitude','population'], axis=1, inplace=True)\n",
    "income_df['median_income'] = income_df['median_income'].apply(lambda x: float(str(x).replace(',','')))\n",
    "income_df['zip_code'] = income_df['zip_code'].apply(lambda x: str(x))\n",
    "income_df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>month</th>\n",
       "      <th>broad_category</th>\n",
       "      <th>geometry</th>\n",
       "      <th>id</th>\n",
       "      <th>pop10_sqmi</th>\n",
       "      <th>pop2010</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "      <th>median_income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ASSAULT</td>\n",
       "      <td>BATTERY</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>TARAVAL</td>\n",
       "      <td>ARREST, BOOKED</td>\n",
       "      <td>3200 Block of 20TH AV</td>\n",
       "      <td>-122.475647</td>\n",
       "      <td>37.728528</td>\n",
       "      <td>2017-01-03 17:43:00</td>\n",
       "      <td>17</td>\n",
       "      <td>Jan</td>\n",
       "      <td>OFFENSES AGAINST THE PERSON</td>\n",
       "      <td>POINT (-122.4756475 37.72852806)</td>\n",
       "      <td>2</td>\n",
       "      <td>6440.6</td>\n",
       "      <td>26986</td>\n",
       "      <td>4.19</td>\n",
       "      <td>94132</td>\n",
       "      <td>55000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>PETTY THEFT SHOPLIFTING</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>TARAVAL</td>\n",
       "      <td>ARREST, BOOKED</td>\n",
       "      <td>3200 Block of 20TH AV</td>\n",
       "      <td>-122.475647</td>\n",
       "      <td>37.728528</td>\n",
       "      <td>2017-01-03 17:43:00</td>\n",
       "      <td>17</td>\n",
       "      <td>Jan</td>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>POINT (-122.4756475 37.72852806)</td>\n",
       "      <td>3</td>\n",
       "      <td>6440.6</td>\n",
       "      <td>26986</td>\n",
       "      <td>4.19</td>\n",
       "      <td>94132</td>\n",
       "      <td>55000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        category              description day_of_week pd_district  \\\n",
       "0        ASSAULT                  BATTERY     Tuesday     TARAVAL   \n",
       "1  LARCENY/THEFT  PETTY THEFT SHOPLIFTING     Tuesday     TARAVAL   \n",
       "\n",
       "       resolution                address   longitude   latitude  \\\n",
       "0  ARREST, BOOKED  3200 Block of 20TH AV -122.475647  37.728528   \n",
       "1  ARREST, BOOKED  3200 Block of 20TH AV -122.475647  37.728528   \n",
       "\n",
       "            date_time  hour_of_day month               broad_category  \\\n",
       "0 2017-01-03 17:43:00           17   Jan  OFFENSES AGAINST THE PERSON   \n",
       "1 2017-01-03 17:43:00           17   Jan                LARCENY/THEFT   \n",
       "\n",
       "                           geometry  id pop10_sqmi pop2010  sqmi zip_code  \\\n",
       "0  POINT (-122.4756475 37.72852806)   2     6440.6   26986  4.19    94132   \n",
       "1  POINT (-122.4756475 37.72852806)   3     6440.6   26986  4.19    94132   \n",
       "\n",
       "   median_income  \n",
       "0        55000.0  \n",
       "1        55000.0  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.merge(income_df, left_on='zip_code', right_on='zip_code', how='left')\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "I also have average house price (Z index) data from retrieved from the Zillow API.."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Zillow API Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>zip_code</th>\n",
       "      <th>z_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>94102</td>\n",
       "      <td>703800.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>94103</td>\n",
       "      <td>913350.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>94105</td>\n",
       "      <td>1193700.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  zip_code    z_index\n",
       "0    94102   703800.0\n",
       "1    94103   913350.0\n",
       "2    94105  1193700.0"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zillow_df = pickle.load(open('z_index_data.pkl', 'rb'))\n",
    "zillow_df['zip_code'] = zillow_df['zip_code'].apply(lambda x: str(x))\n",
    "zillow_df.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>month</th>\n",
       "      <th>broad_category</th>\n",
       "      <th>geometry</th>\n",
       "      <th>id</th>\n",
       "      <th>pop10_sqmi</th>\n",
       "      <th>pop2010</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "      <th>median_income</th>\n",
       "      <th>z_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>25547</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Friday</td>\n",
       "      <td>CENTRAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>100 Block of NORTHPOINT ST</td>\n",
       "      <td>-122.409724</td>\n",
       "      <td>37.807115</td>\n",
       "      <td>2017-03-10 20:00:00</td>\n",
       "      <td>20</td>\n",
       "      <td>Mar</td>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>POINT (-122.4097236 37.80711542)</td>\n",
       "      <td>40328</td>\n",
       "      <td>34200.0</td>\n",
       "      <td>24966</td>\n",
       "      <td>0.73</td>\n",
       "      <td>94133</td>\n",
       "      <td>40990.0</td>\n",
       "      <td>1.036800e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27355</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Wednesday</td>\n",
       "      <td>RICHMOND</td>\n",
       "      <td>NONE</td>\n",
       "      <td>3000 Block of PACIFIC AV</td>\n",
       "      <td>-122.445504</td>\n",
       "      <td>37.792035</td>\n",
       "      <td>2017-04-12 22:00:00</td>\n",
       "      <td>22</td>\n",
       "      <td>Apr</td>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>POINT (-122.4455045 37.79203536)</td>\n",
       "      <td>42310</td>\n",
       "      <td>25896.4</td>\n",
       "      <td>28486</td>\n",
       "      <td>1.1</td>\n",
       "      <td>94115</td>\n",
       "      <td>54879.0</td>\n",
       "      <td>1.160933e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            category                   description day_of_week pd_district  \\\n",
       "25547  LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO      Friday     CENTRAL   \n",
       "27355  LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO   Wednesday    RICHMOND   \n",
       "\n",
       "      resolution                     address   longitude   latitude  \\\n",
       "25547       NONE  100 Block of NORTHPOINT ST -122.409724  37.807115   \n",
       "27355       NONE    3000 Block of PACIFIC AV -122.445504  37.792035   \n",
       "\n",
       "                date_time  hour_of_day month broad_category  \\\n",
       "25547 2017-03-10 20:00:00           20   Mar  LARCENY/THEFT   \n",
       "27355 2017-04-12 22:00:00           22   Apr  LARCENY/THEFT   \n",
       "\n",
       "                               geometry     id pop10_sqmi pop2010  sqmi  \\\n",
       "25547  POINT (-122.4097236 37.80711542)  40328    34200.0   24966  0.73   \n",
       "27355  POINT (-122.4455045 37.79203536)  42310    25896.4   28486   1.1   \n",
       "\n",
       "      zip_code  median_income       z_index  \n",
       "25547    94133        40990.0  1.036800e+06  \n",
       "27355    94115        54879.0  1.160933e+06  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.merge(zillow_df, left_on='zip_code', right_on='zip_code', how='left')\n",
    "df.sample(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Now to calculate the distance (distance = euclidean distance) of each police incident to the nearest BART or Caltrain station..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### BART / Caltrain stations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "station_locations = pickle.load(open('station_locations.pkl', 'rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>month</th>\n",
       "      <th>broad_category</th>\n",
       "      <th>geometry</th>\n",
       "      <th>id</th>\n",
       "      <th>pop10_sqmi</th>\n",
       "      <th>pop2010</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "      <th>median_income</th>\n",
       "      <th>z_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>117964</th>\n",
       "      <td>PROSTITUTION</td>\n",
       "      <td>SOLICITS FOR ACT OF PROSTITUTION</td>\n",
       "      <td>Friday</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>100 Block of 3RD ST</td>\n",
       "      <td>-122.400916</td>\n",
       "      <td>37.785457</td>\n",
       "      <td>2016-01-29 20:35:00</td>\n",
       "      <td>20</td>\n",
       "      <td>Jan</td>\n",
       "      <td>SOCIAL ORDER OFFENSES</td>\n",
       "      <td>POINT (-122.400916 37.78545749)</td>\n",
       "      <td>184597</td>\n",
       "      <td>22830.1</td>\n",
       "      <td>30364</td>\n",
       "      <td>1.33</td>\n",
       "      <td>94103</td>\n",
       "      <td>31131.0</td>\n",
       "      <td>913350.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115193</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Wednesday</td>\n",
       "      <td>TARAVAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>2300 Block of 24TH AV</td>\n",
       "      <td>-122.481183</td>\n",
       "      <td>37.743727</td>\n",
       "      <td>2016-03-16 14:00:00</td>\n",
       "      <td>14</td>\n",
       "      <td>Mar</td>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>POINT (-122.4811832 37.74372682)</td>\n",
       "      <td>179791</td>\n",
       "      <td>16360.7</td>\n",
       "      <td>42865</td>\n",
       "      <td>2.62</td>\n",
       "      <td>94116</td>\n",
       "      <td>66627.0</td>\n",
       "      <td>1153700.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             category                       description day_of_week  \\\n",
       "117964   PROSTITUTION  SOLICITS FOR ACT OF PROSTITUTION      Friday   \n",
       "115193  LARCENY/THEFT      GRAND THEFT FROM LOCKED AUTO   Wednesday   \n",
       "\n",
       "       pd_district resolution                address   longitude   latitude  \\\n",
       "117964    SOUTHERN       NONE    100 Block of 3RD ST -122.400916  37.785457   \n",
       "115193     TARAVAL       NONE  2300 Block of 24TH AV -122.481183  37.743727   \n",
       "\n",
       "                 date_time  hour_of_day month         broad_category  \\\n",
       "117964 2016-01-29 20:35:00           20   Jan  SOCIAL ORDER OFFENSES   \n",
       "115193 2016-03-16 14:00:00           14   Mar          LARCENY/THEFT   \n",
       "\n",
       "                                geometry      id pop10_sqmi pop2010  sqmi  \\\n",
       "117964   POINT (-122.400916 37.78545749)  184597    22830.1   30364  1.33   \n",
       "115193  POINT (-122.4811832 37.74372682)  179791    16360.7   42865  2.62   \n",
       "\n",
       "       zip_code  median_income    z_index  \n",
       "117964    94103        31131.0   913350.0  \n",
       "115193    94116        66627.0  1153700.0  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def min_distance_to_station(point):\n",
    "    return station_locations.distance(point).min()\n",
    "\n",
    "def num_close_stations(point):\n",
    "    distances = station_locations.distance(point)\n",
    "    counter = 0\n",
    "    for d in distances:\n",
    "        if d < 0.01:\n",
    "            counter += 1\n",
    "    return counter\n",
    "\n",
    "df['dist_to_train_station'] = df['geometry'].apply(min_distance_to_station)\n",
    "df['num_close_train_stations'] = df['geometry'].apply(num_close_stations)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>...</th>\n",
       "      <th>geometry</th>\n",
       "      <th>id</th>\n",
       "      <th>pop10_sqmi</th>\n",
       "      <th>pop2010</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "      <th>median_income</th>\n",
       "      <th>z_index</th>\n",
       "      <th>dist_to_train_station</th>\n",
       "      <th>num_close_train_stations</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15758</th>\n",
       "      <td>BURGLARY</td>\n",
       "      <td>BURGLARY OF STORE, UNLAWFUL ENTRY</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>RICHMOND</td>\n",
       "      <td>ARREST, BOOKED</td>\n",
       "      <td>800 Block of LA PLAYA ST</td>\n",
       "      <td>-122.509895</td>\n",
       "      <td>37.772313</td>\n",
       "      <td>2017-03-30 16:35:00</td>\n",
       "      <td>16</td>\n",
       "      <td>...</td>\n",
       "      <td>POINT (-122.5098954 37.7723132)</td>\n",
       "      <td>26939</td>\n",
       "      <td>18809.6</td>\n",
       "      <td>41193</td>\n",
       "      <td>2.19</td>\n",
       "      <td>94121</td>\n",
       "      <td>61776.0</td>\n",
       "      <td>2085900.0</td>\n",
       "      <td>0.080410</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108578</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM UNLOCKED AUTO</td>\n",
       "      <td>Saturday</td>\n",
       "      <td>MISSION</td>\n",
       "      <td>NONE</td>\n",
       "      <td>22ND ST / HAMPSHIRE ST</td>\n",
       "      <td>-122.407605</td>\n",
       "      <td>37.756110</td>\n",
       "      <td>2016-11-05 22:06:00</td>\n",
       "      <td>22</td>\n",
       "      <td>...</td>\n",
       "      <td>POINT (-122.4076048 37.75610979)</td>\n",
       "      <td>170474</td>\n",
       "      <td>27982.9</td>\n",
       "      <td>67159</td>\n",
       "      <td>2.4</td>\n",
       "      <td>94110</td>\n",
       "      <td>53795.0</td>\n",
       "      <td>1095100.0</td>\n",
       "      <td>0.011149</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110358</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>800 Block of BRYANT ST</td>\n",
       "      <td>-122.403405</td>\n",
       "      <td>37.775421</td>\n",
       "      <td>2016-11-27 20:40:00</td>\n",
       "      <td>20</td>\n",
       "      <td>...</td>\n",
       "      <td>POINT (-122.4034048 37.77542071)</td>\n",
       "      <td>172476</td>\n",
       "      <td>22830.1</td>\n",
       "      <td>30364</td>\n",
       "      <td>1.33</td>\n",
       "      <td>94103</td>\n",
       "      <td>31131.0</td>\n",
       "      <td>913350.0</td>\n",
       "      <td>0.009013</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85895</th>\n",
       "      <td>VEHICLE THEFT</td>\n",
       "      <td>STOLEN AND RECOVERED VEHICLE</td>\n",
       "      <td>Wednesday</td>\n",
       "      <td>BAYVIEW</td>\n",
       "      <td>NONE</td>\n",
       "      <td>2900 Block of HAWES ST</td>\n",
       "      <td>-122.391431</td>\n",
       "      <td>37.717810</td>\n",
       "      <td>2016-12-07 22:30:00</td>\n",
       "      <td>22</td>\n",
       "      <td>...</td>\n",
       "      <td>POINT (-122.3914312 37.71780992)</td>\n",
       "      <td>142016</td>\n",
       "      <td>7645.4</td>\n",
       "      <td>37233</td>\n",
       "      <td>4.87</td>\n",
       "      <td>94124</td>\n",
       "      <td>37146.0</td>\n",
       "      <td>612900.0</td>\n",
       "      <td>0.014725</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66647</th>\n",
       "      <td>VANDALISM</td>\n",
       "      <td>MALICIOUS MISCHIEF, VANDALISM</td>\n",
       "      <td>Monday</td>\n",
       "      <td>INGLESIDE</td>\n",
       "      <td>NONE</td>\n",
       "      <td>5800 Block of MISSION ST</td>\n",
       "      <td>-122.451819</td>\n",
       "      <td>37.709154</td>\n",
       "      <td>2016-09-05 13:15:00</td>\n",
       "      <td>13</td>\n",
       "      <td>...</td>\n",
       "      <td>POINT (-122.4518185 37.70915421)</td>\n",
       "      <td>109557</td>\n",
       "      <td>25086.1</td>\n",
       "      <td>82784</td>\n",
       "      <td>3.3</td>\n",
       "      <td>94112</td>\n",
       "      <td>57629.0</td>\n",
       "      <td>796450.0</td>\n",
       "      <td>0.013158</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             category                        description day_of_week  \\\n",
       "15758        BURGLARY  BURGLARY OF STORE, UNLAWFUL ENTRY    Thursday   \n",
       "108578  LARCENY/THEFT     GRAND THEFT FROM UNLOCKED AUTO    Saturday   \n",
       "110358  LARCENY/THEFT       GRAND THEFT FROM LOCKED AUTO      Sunday   \n",
       "85895   VEHICLE THEFT       STOLEN AND RECOVERED VEHICLE   Wednesday   \n",
       "66647       VANDALISM      MALICIOUS MISCHIEF, VANDALISM      Monday   \n",
       "\n",
       "       pd_district      resolution                   address   longitude  \\\n",
       "15758     RICHMOND  ARREST, BOOKED  800 Block of LA PLAYA ST -122.509895   \n",
       "108578     MISSION            NONE    22ND ST / HAMPSHIRE ST -122.407605   \n",
       "110358    SOUTHERN            NONE    800 Block of BRYANT ST -122.403405   \n",
       "85895      BAYVIEW            NONE    2900 Block of HAWES ST -122.391431   \n",
       "66647    INGLESIDE            NONE  5800 Block of MISSION ST -122.451819   \n",
       "\n",
       "         latitude           date_time  hour_of_day            ...             \\\n",
       "15758   37.772313 2017-03-30 16:35:00           16            ...              \n",
       "108578  37.756110 2016-11-05 22:06:00           22            ...              \n",
       "110358  37.775421 2016-11-27 20:40:00           20            ...              \n",
       "85895   37.717810 2016-12-07 22:30:00           22            ...              \n",
       "66647   37.709154 2016-09-05 13:15:00           13            ...              \n",
       "\n",
       "                                geometry      id pop10_sqmi  pop2010  sqmi  \\\n",
       "15758    POINT (-122.5098954 37.7723132)   26939    18809.6    41193  2.19   \n",
       "108578  POINT (-122.4076048 37.75610979)  170474    27982.9    67159   2.4   \n",
       "110358  POINT (-122.4034048 37.77542071)  172476    22830.1    30364  1.33   \n",
       "85895   POINT (-122.3914312 37.71780992)  142016     7645.4    37233  4.87   \n",
       "66647   POINT (-122.4518185 37.70915421)  109557    25086.1    82784   3.3   \n",
       "\n",
       "       zip_code median_income    z_index  dist_to_train_station  \\\n",
       "15758     94121       61776.0  2085900.0               0.080410   \n",
       "108578    94110       53795.0  1095100.0               0.011149   \n",
       "110358    94103       31131.0   913350.0               0.009013   \n",
       "85895     94124       37146.0   612900.0               0.014725   \n",
       "66647     94112       57629.0   796450.0               0.013158   \n",
       "\n",
       "        num_close_train_stations  \n",
       "15758                          0  \n",
       "108578                         0  \n",
       "110358                         1  \n",
       "85895                          0  \n",
       "66647                          0  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Distance to the nearest police station"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>geometry</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>POINT (-122.4099186 37.7987322)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>POINT (-122.3894121 37.7723802)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>POINT (-122.3979806 37.7297322)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>POINT (-122.422005 37.762849)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>POINT (-122.4324665 37.7801858)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>POINT (-122.455287 37.767797)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>POINT (-122.4644668 37.7799276)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>POINT (-122.4462154 37.7246756)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>POINT (-122.4815003 37.7437335)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>POINT (-122.4128962 37.783669)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          geometry\n",
       "0  POINT (-122.4099186 37.7987322)\n",
       "1  POINT (-122.3894121 37.7723802)\n",
       "2  POINT (-122.3979806 37.7297322)\n",
       "3    POINT (-122.422005 37.762849)\n",
       "4  POINT (-122.4324665 37.7801858)\n",
       "5    POINT (-122.455287 37.767797)\n",
       "6  POINT (-122.4644668 37.7799276)\n",
       "7  POINT (-122.4462154 37.7246756)\n",
       "8  POINT (-122.4815003 37.7437335)\n",
       "9   POINT (-122.4128962 37.783669)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "police_locations = pickle.load(open('police_locations.pkl', 'rb'))\n",
    "police_locations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def min_distance_to_police(point):\n",
    "    return police_locations.distance(point).min()\n",
    "\n",
    "def num_close_police_stations(point):\n",
    "    distances = police_locations.distance(point)\n",
    "    counter = 0\n",
    "    for d in distances:\n",
    "        if d < 0.01:\n",
    "            counter += 1\n",
    "    return counter\n",
    "\n",
    "df['dist_to_police_station'] = df['geometry'].apply(min_distance_to_police)\n",
    "df['num_close_police_stations'] = df['geometry'].apply(num_close_police_stations)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>...</th>\n",
       "      <th>pop10_sqmi</th>\n",
       "      <th>pop2010</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "      <th>median_income</th>\n",
       "      <th>z_index</th>\n",
       "      <th>dist_to_train_station</th>\n",
       "      <th>num_close_train_stations</th>\n",
       "      <th>dist_to_police_station</th>\n",
       "      <th>num_close_police_stations</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>74522</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>RICHMOND</td>\n",
       "      <td>NONE</td>\n",
       "      <td>1000 Block of POINTLOBOS AV</td>\n",
       "      <td>-122.513642</td>\n",
       "      <td>37.778469</td>\n",
       "      <td>2016-10-13 17:00:00</td>\n",
       "      <td>17</td>\n",
       "      <td>...</td>\n",
       "      <td>18809.6</td>\n",
       "      <td>41193</td>\n",
       "      <td>2.19</td>\n",
       "      <td>94121</td>\n",
       "      <td>61776.0</td>\n",
       "      <td>2085900.0</td>\n",
       "      <td>0.087234</td>\n",
       "      <td>0</td>\n",
       "      <td>0.047325</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24998</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Saturday</td>\n",
       "      <td>CENTRAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>2300 Block of STOCKTON ST</td>\n",
       "      <td>-122.410525</td>\n",
       "      <td>37.807445</td>\n",
       "      <td>2017-03-04 13:00:00</td>\n",
       "      <td>13</td>\n",
       "      <td>...</td>\n",
       "      <td>34200.0</td>\n",
       "      <td>24966</td>\n",
       "      <td>0.73</td>\n",
       "      <td>94133</td>\n",
       "      <td>40990.0</td>\n",
       "      <td>1036800.0</td>\n",
       "      <td>0.019867</td>\n",
       "      <td>0</td>\n",
       "      <td>0.008734</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            category                   description day_of_week pd_district  \\\n",
       "74522  LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO    Thursday    RICHMOND   \n",
       "24998  LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO    Saturday     CENTRAL   \n",
       "\n",
       "      resolution                      address   longitude   latitude  \\\n",
       "74522       NONE  1000 Block of POINTLOBOS AV -122.513642  37.778469   \n",
       "24998       NONE    2300 Block of STOCKTON ST -122.410525  37.807445   \n",
       "\n",
       "                date_time  hour_of_day            ...             pop10_sqmi  \\\n",
       "74522 2016-10-13 17:00:00           17            ...                18809.6   \n",
       "24998 2017-03-04 13:00:00           13            ...                34200.0   \n",
       "\n",
       "      pop2010  sqmi  zip_code median_income    z_index dist_to_train_station  \\\n",
       "74522   41193  2.19     94121       61776.0  2085900.0              0.087234   \n",
       "24998   24966  0.73     94133       40990.0  1036800.0              0.019867   \n",
       "\n",
       "      num_close_train_stations  dist_to_police_station  \\\n",
       "74522                        0                0.047325   \n",
       "24998                        0                0.008734   \n",
       "\n",
       "       num_close_police_stations  \n",
       "74522                          0  \n",
       "24998                          1  \n",
       "\n",
       "[2 rows x 24 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Dispensary location data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>geometry</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>POINT (-122.4194132797827 37.78662965327916)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>POINT (-122.4298778642076 37.77189194436477)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>POINT (-122.4385234945558 37.77606702194213)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>POINT (-122.4711317429206 37.7804004970458)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>POINT (-122.4635424309064 37.72602894819491)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       geometry\n",
       "0  POINT (-122.4194132797827 37.78662965327916)\n",
       "1  POINT (-122.4298778642076 37.77189194436477)\n",
       "2  POINT (-122.4385234945558 37.77606702194213)\n",
       "3   POINT (-122.4711317429206 37.7804004970458)\n",
       "4  POINT (-122.4635424309064 37.72602894819491)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dispensary_locations = pickle.load(open('dispensary_locations.pkl', 'rb'))\n",
    "dispensary_locations.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def min_distance_to_dispensary(point):\n",
    "    return dispensary_locations.distance(point).min()\n",
    "\n",
    "def num_close_dispensaries(point):\n",
    "    distances = dispensary_locations.distance(point)\n",
    "    counter = 0\n",
    "    for d in distances:\n",
    "        if d < 0.01:\n",
    "            counter += 1\n",
    "    return counter\n",
    "\n",
    "df['dist_to_dispensary'] = df['geometry'].apply(min_distance_to_dispensary)\n",
    "df['num_close_dispensaries'] = df['geometry'].apply(num_close_dispensaries)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>...</th>\n",
       "      <th>sqmi</th>\n",
       "      <th>zip_code</th>\n",
       "      <th>median_income</th>\n",
       "      <th>z_index</th>\n",
       "      <th>dist_to_train_station</th>\n",
       "      <th>num_close_train_stations</th>\n",
       "      <th>dist_to_police_station</th>\n",
       "      <th>num_close_police_stations</th>\n",
       "      <th>dist_to_dispensary</th>\n",
       "      <th>num_close_dispensaries</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>77694</th>\n",
       "      <td>ASSAULT</td>\n",
       "      <td>BATTERY</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>3RD ST / BRANNAN ST</td>\n",
       "      <td>-122.394305</td>\n",
       "      <td>37.780091</td>\n",
       "      <td>2016-10-20 13:40:00</td>\n",
       "      <td>13</td>\n",
       "      <td>...</td>\n",
       "      <td>1.86</td>\n",
       "      <td>94107</td>\n",
       "      <td>61362.0</td>\n",
       "      <td>1059450.0</td>\n",
       "      <td>0.003704</td>\n",
       "      <td>1</td>\n",
       "      <td>0.009132</td>\n",
       "      <td>1</td>\n",
       "      <td>0.007937</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76835</th>\n",
       "      <td>VANDALISM</td>\n",
       "      <td>MALICIOUS MISCHIEF, VANDALISM</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>CENTRAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>1400 Block of WASHINGTON ST</td>\n",
       "      <td>-122.417119</td>\n",
       "      <td>37.793732</td>\n",
       "      <td>2016-10-25 19:00:00</td>\n",
       "      <td>19</td>\n",
       "      <td>...</td>\n",
       "      <td>1.03</td>\n",
       "      <td>94109</td>\n",
       "      <td>43444.0</td>\n",
       "      <td>1279800.0</td>\n",
       "      <td>0.013015</td>\n",
       "      <td>0</td>\n",
       "      <td>0.008766</td>\n",
       "      <td>1</td>\n",
       "      <td>0.007464</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        category                    description day_of_week pd_district  \\\n",
       "77694    ASSAULT                        BATTERY    Thursday    SOUTHERN   \n",
       "76835  VANDALISM  MALICIOUS MISCHIEF, VANDALISM     Tuesday     CENTRAL   \n",
       "\n",
       "      resolution                      address   longitude   latitude  \\\n",
       "77694       NONE          3RD ST / BRANNAN ST -122.394305  37.780091   \n",
       "76835       NONE  1400 Block of WASHINGTON ST -122.417119  37.793732   \n",
       "\n",
       "                date_time  hour_of_day           ...            sqmi zip_code  \\\n",
       "77694 2016-10-20 13:40:00           13           ...            1.86    94107   \n",
       "76835 2016-10-25 19:00:00           19           ...            1.03    94109   \n",
       "\n",
       "      median_income    z_index dist_to_train_station num_close_train_stations  \\\n",
       "77694       61362.0  1059450.0              0.003704                        1   \n",
       "76835       43444.0  1279800.0              0.013015                        0   \n",
       "\n",
       "      dist_to_police_station num_close_police_stations  dist_to_dispensary  \\\n",
       "77694               0.009132                         1            0.007937   \n",
       "76835               0.008766                         1            0.007464   \n",
       "\n",
       "       num_close_dispensaries  \n",
       "77694                       2  \n",
       "76835                       1  \n",
       "\n",
       "[2 rows x 26 columns]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Health care facility locations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>geometry</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>POINT (-122.4535495 37.78740232)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>POINT (-122.45591166 37.78614406)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>POINT (-122.4358177 37.76905419)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>POINT (-122.43103755 37.79142444)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>POINT (-122.42124885 37.74804296)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            geometry\n",
       "0   POINT (-122.4535495 37.78740232)\n",
       "1  POINT (-122.45591166 37.78614406)\n",
       "2   POINT (-122.4358177 37.76905419)\n",
       "3  POINT (-122.43103755 37.79142444)\n",
       "4  POINT (-122.42124885 37.74804296)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "care_locations = pickle.load(open('care_locations.pkl', 'rb'))\n",
    "care_locations.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def min_distance_to_health_facility(point):\n",
    "    return care_locations.distance(point).min()\n",
    "\n",
    "\n",
    "def num_close_health_facilities(point):\n",
    "    distances = care_locations.distance(point)\n",
    "    counter = 0\n",
    "    for d in distances:\n",
    "        if d < 0.01:\n",
    "            counter += 1\n",
    "    return counter\n",
    "\n",
    "df['dist_to_health_facility'] = df['geometry'].apply(min_distance_to_health_facility)\n",
    "df['num_close_health_facilities'] = df['geometry'].apply(num_close_health_facilities)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>...</th>\n",
       "      <th>median_income</th>\n",
       "      <th>z_index</th>\n",
       "      <th>dist_to_train_station</th>\n",
       "      <th>num_close_train_stations</th>\n",
       "      <th>dist_to_police_station</th>\n",
       "      <th>num_close_police_stations</th>\n",
       "      <th>dist_to_dispensary</th>\n",
       "      <th>num_close_dispensaries</th>\n",
       "      <th>dist_to_health_facility</th>\n",
       "      <th>num_close_health_facilities</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>35559</th>\n",
       "      <td>VANDALISM</td>\n",
       "      <td>MALICIOUS MISCHIEF, STREET CARS/BUSES</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>INGLESIDE</td>\n",
       "      <td>NONE</td>\n",
       "      <td>CORTLAND AV / MISSION ST</td>\n",
       "      <td>-122.422842</td>\n",
       "      <td>37.741029</td>\n",
       "      <td>2016-03-27 21:20:00</td>\n",
       "      <td>21</td>\n",
       "      <td>...</td>\n",
       "      <td>53795.0</td>\n",
       "      <td>1095100.0</td>\n",
       "      <td>0.012369</td>\n",
       "      <td>0</td>\n",
       "      <td>0.021836</td>\n",
       "      <td>0</td>\n",
       "      <td>0.003047</td>\n",
       "      <td>2</td>\n",
       "      <td>0.002529</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48318</th>\n",
       "      <td>ROBBERY</td>\n",
       "      <td>ROBBERY, BODILY FORCE</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>RICHMOND</td>\n",
       "      <td>NONE</td>\n",
       "      <td>30TH AV / FULTON ST</td>\n",
       "      <td>-122.489539</td>\n",
       "      <td>37.772325</td>\n",
       "      <td>2016-05-31 17:00:00</td>\n",
       "      <td>17</td>\n",
       "      <td>...</td>\n",
       "      <td>61776.0</td>\n",
       "      <td>2085900.0</td>\n",
       "      <td>0.065888</td>\n",
       "      <td>0</td>\n",
       "      <td>0.026199</td>\n",
       "      <td>0</td>\n",
       "      <td>0.020101</td>\n",
       "      <td>0</td>\n",
       "      <td>0.008775</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        category                            description day_of_week  \\\n",
       "35559  VANDALISM  MALICIOUS MISCHIEF, STREET CARS/BUSES      Sunday   \n",
       "48318    ROBBERY                  ROBBERY, BODILY FORCE     Tuesday   \n",
       "\n",
       "      pd_district resolution                   address   longitude   latitude  \\\n",
       "35559   INGLESIDE       NONE  CORTLAND AV / MISSION ST -122.422842  37.741029   \n",
       "48318    RICHMOND       NONE       30TH AV / FULTON ST -122.489539  37.772325   \n",
       "\n",
       "                date_time  hour_of_day             ...               \\\n",
       "35559 2016-03-27 21:20:00           21             ...                \n",
       "48318 2016-05-31 17:00:00           17             ...                \n",
       "\n",
       "      median_income    z_index dist_to_train_station  \\\n",
       "35559       53795.0  1095100.0              0.012369   \n",
       "48318       61776.0  2085900.0              0.065888   \n",
       "\n",
       "       num_close_train_stations dist_to_police_station  \\\n",
       "35559                         0               0.021836   \n",
       "48318                         0               0.026199   \n",
       "\n",
       "      num_close_police_stations dist_to_dispensary num_close_dispensaries  \\\n",
       "35559                         0           0.003047                      2   \n",
       "48318                         0           0.020101                      0   \n",
       "\n",
       "       dist_to_health_facility  num_close_health_facilities  \n",
       "35559                 0.002529                            7  \n",
       "48318                 0.008775                            1  \n",
       "\n",
       "[2 rows x 28 columns]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['category', 'description', 'day_of_week', 'pd_district', 'resolution',\n",
       "       'address', 'longitude', 'latitude', 'date_time', 'hour_of_day', 'month',\n",
       "       'broad_category', 'geometry', 'id', 'pop10_sqmi', 'pop2010', 'sqmi',\n",
       "       'zip_code', 'median_income', 'z_index', 'dist_to_train_station',\n",
       "       'num_close_train_stations', 'dist_to_police_station',\n",
       "       'num_close_police_stations', 'dist_to_dispensary',\n",
       "       'num_close_dispensaries', 'dist_to_health_facility',\n",
       "       'num_close_health_facilities'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Shelter locations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>geometry</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>POINT (-122.4375855 37.781571)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>POINT (-122.3933527 37.7910388)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>POINT (-122.4072079 37.7765976)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>POINT (-122.3990467 37.7265131)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>POINT (-122.4141134 37.7764266)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          geometry\n",
       "1   POINT (-122.4375855 37.781571)\n",
       "2  POINT (-122.3933527 37.7910388)\n",
       "3  POINT (-122.4072079 37.7765976)\n",
       "4  POINT (-122.3990467 37.7265131)\n",
       "6  POINT (-122.4141134 37.7764266)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "shelter_locations = pickle.load(open('shelter_locations.pkl', 'rb'))\n",
    "shelter_locations.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def min_distance_to_shelters(point):\n",
    "    return shelter_locations.distance(point).min()\n",
    "\n",
    "def num_close_shelters(point):\n",
    "    distances = shelter_locations.distance(point)\n",
    "    counter = 0\n",
    "    for d in distances:\n",
    "        if d < 0.01:\n",
    "            counter += 1\n",
    "    return counter\n",
    "\n",
    "df['dist_to_shelter'] = df['geometry'].apply(min_distance_to_shelters)\n",
    "df['num_close_shelters'] = df['geometry'].apply(num_close_shelters)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>...</th>\n",
       "      <th>dist_to_train_station</th>\n",
       "      <th>num_close_train_stations</th>\n",
       "      <th>dist_to_police_station</th>\n",
       "      <th>num_close_police_stations</th>\n",
       "      <th>dist_to_dispensary</th>\n",
       "      <th>num_close_dispensaries</th>\n",
       "      <th>dist_to_health_facility</th>\n",
       "      <th>num_close_health_facilities</th>\n",
       "      <th>dist_to_shelter</th>\n",
       "      <th>num_close_shelters</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>38996</th>\n",
       "      <td>VEHICLE THEFT</td>\n",
       "      <td>STOLEN TRUCK</td>\n",
       "      <td>Wednesday</td>\n",
       "      <td>CENTRAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>CALIFORNIA ST / BATTERY ST</td>\n",
       "      <td>-122.399963</td>\n",
       "      <td>37.793209</td>\n",
       "      <td>2016-04-13 10:00:00</td>\n",
       "      <td>10</td>\n",
       "      <td>...</td>\n",
       "      <td>0.002962</td>\n",
       "      <td>2</td>\n",
       "      <td>0.011385</td>\n",
       "      <td>0</td>\n",
       "      <td>0.005006</td>\n",
       "      <td>2</td>\n",
       "      <td>0.006542</td>\n",
       "      <td>3</td>\n",
       "      <td>0.001134</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35865</th>\n",
       "      <td>VEHICLE THEFT</td>\n",
       "      <td>STOLEN AUTOMOBILE</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>INGLESIDE</td>\n",
       "      <td>NONE</td>\n",
       "      <td>MONTEREY BL / JOOST AV</td>\n",
       "      <td>-122.434992</td>\n",
       "      <td>37.732295</td>\n",
       "      <td>2016-02-18 20:30:00</td>\n",
       "      <td>20</td>\n",
       "      <td>...</td>\n",
       "      <td>0.001404</td>\n",
       "      <td>1</td>\n",
       "      <td>0.013566</td>\n",
       "      <td>0</td>\n",
       "      <td>0.005025</td>\n",
       "      <td>1</td>\n",
       "      <td>0.005809</td>\n",
       "      <td>2</td>\n",
       "      <td>0.008457</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95382</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT OF PROPERTY</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>NORTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>LARKIN ST / MCALLISTER ST</td>\n",
       "      <td>-122.416949</td>\n",
       "      <td>37.780469</td>\n",
       "      <td>2016-04-03 07:30:00</td>\n",
       "      <td>7</td>\n",
       "      <td>...</td>\n",
       "      <td>0.002920</td>\n",
       "      <td>2</td>\n",
       "      <td>0.005164</td>\n",
       "      <td>1</td>\n",
       "      <td>0.004441</td>\n",
       "      <td>6</td>\n",
       "      <td>0.002083</td>\n",
       "      <td>20</td>\n",
       "      <td>0.001437</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22092</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM UNLOCKED AUTO</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>800 Block of BRYANT ST</td>\n",
       "      <td>-122.403405</td>\n",
       "      <td>37.775421</td>\n",
       "      <td>2017-01-24 09:00:00</td>\n",
       "      <td>9</td>\n",
       "      <td>...</td>\n",
       "      <td>0.009013</td>\n",
       "      <td>1</td>\n",
       "      <td>0.012575</td>\n",
       "      <td>0</td>\n",
       "      <td>0.006862</td>\n",
       "      <td>3</td>\n",
       "      <td>0.003678</td>\n",
       "      <td>3</td>\n",
       "      <td>0.000950</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97861</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>800 Block of BRYANT ST</td>\n",
       "      <td>-122.403405</td>\n",
       "      <td>37.775421</td>\n",
       "      <td>2016-05-19 13:30:00</td>\n",
       "      <td>13</td>\n",
       "      <td>...</td>\n",
       "      <td>0.009013</td>\n",
       "      <td>1</td>\n",
       "      <td>0.012575</td>\n",
       "      <td>0</td>\n",
       "      <td>0.006862</td>\n",
       "      <td>3</td>\n",
       "      <td>0.003678</td>\n",
       "      <td>3</td>\n",
       "      <td>0.000950</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            category                     description day_of_week pd_district  \\\n",
       "38996  VEHICLE THEFT                    STOLEN TRUCK   Wednesday     CENTRAL   \n",
       "35865  VEHICLE THEFT               STOLEN AUTOMOBILE    Thursday   INGLESIDE   \n",
       "95382  LARCENY/THEFT         GRAND THEFT OF PROPERTY      Sunday    NORTHERN   \n",
       "22092  LARCENY/THEFT  GRAND THEFT FROM UNLOCKED AUTO     Tuesday    SOUTHERN   \n",
       "97861  LARCENY/THEFT    GRAND THEFT FROM LOCKED AUTO    Thursday    SOUTHERN   \n",
       "\n",
       "      resolution                     address   longitude   latitude  \\\n",
       "38996       NONE  CALIFORNIA ST / BATTERY ST -122.399963  37.793209   \n",
       "35865       NONE      MONTEREY BL / JOOST AV -122.434992  37.732295   \n",
       "95382       NONE   LARKIN ST / MCALLISTER ST -122.416949  37.780469   \n",
       "22092       NONE      800 Block of BRYANT ST -122.403405  37.775421   \n",
       "97861       NONE      800 Block of BRYANT ST -122.403405  37.775421   \n",
       "\n",
       "                date_time  hour_of_day         ...          \\\n",
       "38996 2016-04-13 10:00:00           10         ...           \n",
       "35865 2016-02-18 20:30:00           20         ...           \n",
       "95382 2016-04-03 07:30:00            7         ...           \n",
       "22092 2017-01-24 09:00:00            9         ...           \n",
       "97861 2016-05-19 13:30:00           13         ...           \n",
       "\n",
       "      dist_to_train_station num_close_train_stations dist_to_police_station  \\\n",
       "38996              0.002962                        2               0.011385   \n",
       "35865              0.001404                        1               0.013566   \n",
       "95382              0.002920                        2               0.005164   \n",
       "22092              0.009013                        1               0.012575   \n",
       "97861              0.009013                        1               0.012575   \n",
       "\n",
       "       num_close_police_stations dist_to_dispensary num_close_dispensaries  \\\n",
       "38996                          0           0.005006                      2   \n",
       "35865                          0           0.005025                      1   \n",
       "95382                          1           0.004441                      6   \n",
       "22092                          0           0.006862                      3   \n",
       "97861                          0           0.006862                      3   \n",
       "\n",
       "      dist_to_health_facility num_close_health_facilities  dist_to_shelter  \\\n",
       "38996                0.006542                           3         0.001134   \n",
       "35865                0.005809                           2         0.008457   \n",
       "95382                0.002083                          20         0.001437   \n",
       "22092                0.003678                           3         0.000950   \n",
       "97861                0.003678                           3         0.000950   \n",
       "\n",
       "       num_close_shelters  \n",
       "38996                   8  \n",
       "35865                   1  \n",
       "95382                  40  \n",
       "22092                  28  \n",
       "97861                  28  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Distance to Union Sq"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Adding a distance to union square metric."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def distance_to_union_sq(point):\n",
    "    return union_sq.distance(point).min()\n",
    "\n",
    "u_sq = {'lng': [-122.4075],\n",
    "        'lat': [37.788056]}\n",
    "\n",
    "union_sq = geopandas.GeoDataFrame(u_sq)\n",
    "union_sq['geometry'] = union_sq.apply(lambda row: Point(row['lng'], row['lat']), axis=1)\n",
    "union_sq.drop(['lat','lng'], axis=1, inplace=True)\n",
    "df['dist_to_union_sq'] = df['geometry'].apply(distance_to_union_sq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>category</th>\n",
       "      <th>description</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>pd_district</th>\n",
       "      <th>resolution</th>\n",
       "      <th>address</th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>date_time</th>\n",
       "      <th>hour_of_day</th>\n",
       "      <th>...</th>\n",
       "      <th>num_close_train_stations</th>\n",
       "      <th>dist_to_police_station</th>\n",
       "      <th>num_close_police_stations</th>\n",
       "      <th>dist_to_dispensary</th>\n",
       "      <th>num_close_dispensaries</th>\n",
       "      <th>dist_to_health_facility</th>\n",
       "      <th>num_close_health_facilities</th>\n",
       "      <th>dist_to_shelter</th>\n",
       "      <th>num_close_shelters</th>\n",
       "      <th>dist_to_union_sq</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17939</th>\n",
       "      <td>ASSAULT</td>\n",
       "      <td>FALSE IMPRISONMENT</td>\n",
       "      <td>Monday</td>\n",
       "      <td>BAYVIEW</td>\n",
       "      <td>NONE</td>\n",
       "      <td>MINNESOTA ST / MARIPOSA ST</td>\n",
       "      <td>-122.390820</td>\n",
       "      <td>37.764201</td>\n",
       "      <td>2017-04-10 16:00:00</td>\n",
       "      <td>16</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0.008300</td>\n",
       "      <td>1</td>\n",
       "      <td>0.022100</td>\n",
       "      <td>0</td>\n",
       "      <td>0.012699</td>\n",
       "      <td>0</td>\n",
       "      <td>0.010131</td>\n",
       "      <td>0</td>\n",
       "      <td>0.029109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116545</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>PETTY THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>CENTRAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>2200 Block of STOCKTON ST</td>\n",
       "      <td>-122.410474</td>\n",
       "      <td>37.806496</td>\n",
       "      <td>2016-01-24 20:18:00</td>\n",
       "      <td>20</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.007783</td>\n",
       "      <td>1</td>\n",
       "      <td>0.020842</td>\n",
       "      <td>0</td>\n",
       "      <td>0.006756</td>\n",
       "      <td>3</td>\n",
       "      <td>0.011593</td>\n",
       "      <td>0</td>\n",
       "      <td>0.018678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74599</th>\n",
       "      <td>VEHICLE THEFT</td>\n",
       "      <td>STOLEN TRUCK</td>\n",
       "      <td>Friday</td>\n",
       "      <td>INGLESIDE</td>\n",
       "      <td>NONE</td>\n",
       "      <td>0 Block of PROSPECT AV</td>\n",
       "      <td>-122.417503</td>\n",
       "      <td>37.745189</td>\n",
       "      <td>2016-01-22 21:00:00</td>\n",
       "      <td>21</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0.018225</td>\n",
       "      <td>0</td>\n",
       "      <td>0.002143</td>\n",
       "      <td>2</td>\n",
       "      <td>0.002780</td>\n",
       "      <td>8</td>\n",
       "      <td>0.002318</td>\n",
       "      <td>2</td>\n",
       "      <td>0.044018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97850</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>PARK</td>\n",
       "      <td>NONE</td>\n",
       "      <td>100 Block of PARNASSUS AV</td>\n",
       "      <td>-122.451880</td>\n",
       "      <td>37.764542</td>\n",
       "      <td>2016-05-19 19:20:00</td>\n",
       "      <td>19</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.004712</td>\n",
       "      <td>1</td>\n",
       "      <td>0.017642</td>\n",
       "      <td>0</td>\n",
       "      <td>0.005655</td>\n",
       "      <td>5</td>\n",
       "      <td>0.002953</td>\n",
       "      <td>5</td>\n",
       "      <td>0.050225</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90864</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Monday</td>\n",
       "      <td>SOUTHERN</td>\n",
       "      <td>NONE</td>\n",
       "      <td>HARRISON ST / 7TH ST</td>\n",
       "      <td>-122.406300</td>\n",
       "      <td>37.775513</td>\n",
       "      <td>2016-01-04 13:25:00</td>\n",
       "      <td>13</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>0.010489</td>\n",
       "      <td>0</td>\n",
       "      <td>0.006726</td>\n",
       "      <td>5</td>\n",
       "      <td>0.000902</td>\n",
       "      <td>6</td>\n",
       "      <td>0.001414</td>\n",
       "      <td>35</td>\n",
       "      <td>0.012600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94355</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Monday</td>\n",
       "      <td>PARK</td>\n",
       "      <td>NONE</td>\n",
       "      <td>SCOTT ST / EDDY ST</td>\n",
       "      <td>-122.437263</td>\n",
       "      <td>37.780802</td>\n",
       "      <td>2016-03-14 08:30:00</td>\n",
       "      <td>8</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.004835</td>\n",
       "      <td>1</td>\n",
       "      <td>0.004900</td>\n",
       "      <td>1</td>\n",
       "      <td>0.001798</td>\n",
       "      <td>4</td>\n",
       "      <td>0.000834</td>\n",
       "      <td>9</td>\n",
       "      <td>0.030634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40001</th>\n",
       "      <td>ASSAULT</td>\n",
       "      <td>FALSE IMPRISONMENT</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>TENDERLOIN</td>\n",
       "      <td>ARREST, BOOKED</td>\n",
       "      <td>300 Block of TURK ST</td>\n",
       "      <td>-122.414879</td>\n",
       "      <td>37.782634</td>\n",
       "      <td>2016-04-19 13:33:00</td>\n",
       "      <td>13</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>0.002237</td>\n",
       "      <td>1</td>\n",
       "      <td>0.005710</td>\n",
       "      <td>5</td>\n",
       "      <td>0.000299</td>\n",
       "      <td>19</td>\n",
       "      <td>0.000641</td>\n",
       "      <td>43</td>\n",
       "      <td>0.009157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91269</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>PETTY THEFT OF PROPERTY</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>CENTRAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>300 Block of KEARNY ST</td>\n",
       "      <td>-122.404195</td>\n",
       "      <td>37.791226</td>\n",
       "      <td>2016-01-14 15:30:00</td>\n",
       "      <td>15</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>0.009440</td>\n",
       "      <td>1</td>\n",
       "      <td>0.004804</td>\n",
       "      <td>4</td>\n",
       "      <td>0.004049</td>\n",
       "      <td>7</td>\n",
       "      <td>0.002232</td>\n",
       "      <td>15</td>\n",
       "      <td>0.004579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103390</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT OF PROPERTY</td>\n",
       "      <td>Monday</td>\n",
       "      <td>CENTRAL</td>\n",
       "      <td>NONE</td>\n",
       "      <td>0 Block of HOTALING PL</td>\n",
       "      <td>-122.402762</td>\n",
       "      <td>37.796052</td>\n",
       "      <td>2016-08-08 09:00:00</td>\n",
       "      <td>9</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>0.007642</td>\n",
       "      <td>1</td>\n",
       "      <td>0.008162</td>\n",
       "      <td>1</td>\n",
       "      <td>0.002651</td>\n",
       "      <td>6</td>\n",
       "      <td>0.004377</td>\n",
       "      <td>6</td>\n",
       "      <td>0.009295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47442</th>\n",
       "      <td>LARCENY/THEFT</td>\n",
       "      <td>GRAND THEFT FROM LOCKED AUTO</td>\n",
       "      <td>Saturday</td>\n",
       "      <td>RICHMOND</td>\n",
       "      <td>NONE</td>\n",
       "      <td>1100 Block of 9TH AV</td>\n",
       "      <td>-122.466504</td>\n",
       "      <td>37.766201</td>\n",
       "      <td>2016-05-28 09:52:00</td>\n",
       "      <td>9</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.011330</td>\n",
       "      <td>0</td>\n",
       "      <td>0.014934</td>\n",
       "      <td>0</td>\n",
       "      <td>0.008428</td>\n",
       "      <td>1</td>\n",
       "      <td>0.001650</td>\n",
       "      <td>3</td>\n",
       "      <td>0.062922</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             category                   description day_of_week pd_district  \\\n",
       "17939         ASSAULT            FALSE IMPRISONMENT      Monday     BAYVIEW   \n",
       "116545  LARCENY/THEFT  PETTY THEFT FROM LOCKED AUTO      Sunday     CENTRAL   \n",
       "74599   VEHICLE THEFT                  STOLEN TRUCK      Friday   INGLESIDE   \n",
       "97850   LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO    Thursday        PARK   \n",
       "90864   LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO      Monday    SOUTHERN   \n",
       "94355   LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO      Monday        PARK   \n",
       "40001         ASSAULT            FALSE IMPRISONMENT     Tuesday  TENDERLOIN   \n",
       "91269   LARCENY/THEFT       PETTY THEFT OF PROPERTY    Thursday     CENTRAL   \n",
       "103390  LARCENY/THEFT       GRAND THEFT OF PROPERTY      Monday     CENTRAL   \n",
       "47442   LARCENY/THEFT  GRAND THEFT FROM LOCKED AUTO    Saturday    RICHMOND   \n",
       "\n",
       "            resolution                     address   longitude   latitude  \\\n",
       "17939             NONE  MINNESOTA ST / MARIPOSA ST -122.390820  37.764201   \n",
       "116545            NONE   2200 Block of STOCKTON ST -122.410474  37.806496   \n",
       "74599             NONE      0 Block of PROSPECT AV -122.417503  37.745189   \n",
       "97850             NONE   100 Block of PARNASSUS AV -122.451880  37.764542   \n",
       "90864             NONE        HARRISON ST / 7TH ST -122.406300  37.775513   \n",
       "94355             NONE          SCOTT ST / EDDY ST -122.437263  37.780802   \n",
       "40001   ARREST, BOOKED        300 Block of TURK ST -122.414879  37.782634   \n",
       "91269             NONE      300 Block of KEARNY ST -122.404195  37.791226   \n",
       "103390            NONE      0 Block of HOTALING PL -122.402762  37.796052   \n",
       "47442             NONE        1100 Block of 9TH AV -122.466504  37.766201   \n",
       "\n",
       "                 date_time  hour_of_day        ...         \\\n",
       "17939  2017-04-10 16:00:00           16        ...          \n",
       "116545 2016-01-24 20:18:00           20        ...          \n",
       "74599  2016-01-22 21:00:00           21        ...          \n",
       "97850  2016-05-19 19:20:00           19        ...          \n",
       "90864  2016-01-04 13:25:00           13        ...          \n",
       "94355  2016-03-14 08:30:00            8        ...          \n",
       "40001  2016-04-19 13:33:00           13        ...          \n",
       "91269  2016-01-14 15:30:00           15        ...          \n",
       "103390 2016-08-08 09:00:00            9        ...          \n",
       "47442  2016-05-28 09:52:00            9        ...          \n",
       "\n",
       "       num_close_train_stations dist_to_police_station  \\\n",
       "17939                         1               0.008300   \n",
       "116545                        0               0.007783   \n",
       "74599                         1               0.018225   \n",
       "97850                         0               0.004712   \n",
       "90864                         2               0.010489   \n",
       "94355                         0               0.004835   \n",
       "40001                         2               0.002237   \n",
       "91269                         3               0.009440   \n",
       "103390                        2               0.007642   \n",
       "47442                         0               0.011330   \n",
       "\n",
       "       num_close_police_stations  dist_to_dispensary num_close_dispensaries  \\\n",
       "17939                          1            0.022100                      0   \n",
       "116545                         1            0.020842                      0   \n",
       "74599                          0            0.002143                      2   \n",
       "97850                          1            0.017642                      0   \n",
       "90864                          0            0.006726                      5   \n",
       "94355                          1            0.004900                      1   \n",
       "40001                          1            0.005710                      5   \n",
       "91269                          1            0.004804                      4   \n",
       "103390                         1            0.008162                      1   \n",
       "47442                          0            0.014934                      0   \n",
       "\n",
       "       dist_to_health_facility num_close_health_facilities dist_to_shelter  \\\n",
       "17939                 0.012699                           0        0.010131   \n",
       "116545                0.006756                           3        0.011593   \n",
       "74599                 0.002780                           8        0.002318   \n",
       "97850                 0.005655                           5        0.002953   \n",
       "90864                 0.000902                           6        0.001414   \n",
       "94355                 0.001798                           4        0.000834   \n",
       "40001                 0.000299                          19        0.000641   \n",
       "91269                 0.004049                           7        0.002232   \n",
       "103390                0.002651                           6        0.004377   \n",
       "47442                 0.008428                           1        0.001650   \n",
       "\n",
       "        num_close_shelters  dist_to_union_sq  \n",
       "17939                    0          0.029109  \n",
       "116545                   0          0.018678  \n",
       "74599                    2          0.044018  \n",
       "97850                    5          0.050225  \n",
       "90864                   35          0.012600  \n",
       "94355                    9          0.030634  \n",
       "40001                   43          0.009157  \n",
       "91269                   15          0.004579  \n",
       "103390                   6          0.009295  \n",
       "47442                    3          0.062922  \n",
       "\n",
       "[10 rows x 31 columns]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### Save dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "pickle.dump(df, open('complete_data_set.pkl', 'wb'))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}