prediction-model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Accident Hotspot Prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Road safety is of paramount importance for the Government and the entire mobility ecosystem. While the number of road accidents has reduced over the last couple of years, still the overall number of accidents that occurred in 2017 is alarming. It has been assessed that 90% of the accident fatalities occur due to human error. “Brasilia Declaration on Road Safety” envisages to halve road traffic deaths by the end of this decade. There are number of interventions being adopted to reduce the occurrence of the road accidents. A technological solution/platform which can rate driver behavior and predict accident hot spot is the need of the hour. This will enable enforcement agencies and the commuters to take preemptive actions to avoid accidents."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The prediction is carried out by taking the example of Bengaluru. We first take the JSON data of Bengaluru and convert it to XLSX using an online tool for better readability. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdf = pd.read_excel('bangalore-cas-alerts.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 207617 entries, 0 to 207616\n",
      "Data columns (total 7 columns):\n",
      "deviceCode_deviceCode                 207617 non-null int64\n",
      "deviceCode_location_latitude          207617 non-null float64\n",
      "deviceCode_location_longitude         207617 non-null float64\n",
      "deviceCode_location_wardName          207617 non-null object\n",
      "deviceCode_pyld_alarmType             207617 non-null object\n",
      "deviceCode_pyld_speed                 207617 non-null int64\n",
      "deviceCode_time_recordedTime_$date    207617 non-null object\n",
      "dtypes: float64(2), int64(2), object(3)\n",
      "memory usage: 11.1+ MB\n"
     ]
    }
   ],
   "source": [
    "bdf.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>deviceCode_deviceCode</th>\n",
       "      <th>deviceCode_location_latitude</th>\n",
       "      <th>deviceCode_location_longitude</th>\n",
       "      <th>deviceCode_location_wardName</th>\n",
       "      <th>deviceCode_pyld_alarmType</th>\n",
       "      <th>deviceCode_pyld_speed</th>\n",
       "      <th>deviceCode_time_recordedTime_$date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.984595</td>\n",
       "      <td>77.744087</td>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>PCW</td>\n",
       "      <td>32</td>\n",
       "      <td>2018-02-01T01:48:59.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.984595</td>\n",
       "      <td>77.744087</td>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>PCW</td>\n",
       "      <td>32</td>\n",
       "      <td>2018-02-01T01:48:59.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987233</td>\n",
       "      <td>77.741119</td>\n",
       "      <td>Garudachar Playa</td>\n",
       "      <td>FCW</td>\n",
       "      <td>41</td>\n",
       "      <td>2018-02-01T01:50:00.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987233</td>\n",
       "      <td>77.741119</td>\n",
       "      <td>Garudachar Playa</td>\n",
       "      <td>FCW</td>\n",
       "      <td>41</td>\n",
       "      <td>2018-02-01T01:50:00.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987503</td>\n",
       "      <td>77.740051</td>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>37</td>\n",
       "      <td>2018-02-01T01:50:11.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987503</td>\n",
       "      <td>77.740051</td>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>37</td>\n",
       "      <td>2018-02-01T01:50:11.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987523</td>\n",
       "      <td>77.736702</td>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>32</td>\n",
       "      <td>2018-02-01T01:50:50.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987523</td>\n",
       "      <td>77.736702</td>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>32</td>\n",
       "      <td>2018-02-01T01:50:50.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.988210</td>\n",
       "      <td>77.731369</td>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>27</td>\n",
       "      <td>2018-02-01T01:52:26.000Z</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.988210</td>\n",
       "      <td>77.731369</td>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>27</td>\n",
       "      <td>2018-02-01T01:52:26.000Z</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   deviceCode_deviceCode  deviceCode_location_latitude  \\\n",
       "0        864504031502210                     12.984595   \n",
       "1        864504031502210                     12.984595   \n",
       "2        864504031502210                     12.987233   \n",
       "3        864504031502210                     12.987233   \n",
       "4        864504031502210                     12.987503   \n",
       "5        864504031502210                     12.987503   \n",
       "6        864504031502210                     12.987523   \n",
       "7        864504031502210                     12.987523   \n",
       "8        864504031502210                     12.988210   \n",
       "9        864504031502210                     12.988210   \n",
       "\n",
       "   deviceCode_location_longitude deviceCode_location_wardName  \\\n",
       "0                      77.744087                     Kadugodi   \n",
       "1                      77.744087                     Kadugodi   \n",
       "2                      77.741119             Garudachar Playa   \n",
       "3                      77.741119             Garudachar Playa   \n",
       "4                      77.740051                         Hudi   \n",
       "5                      77.740051                         Hudi   \n",
       "6                      77.736702                     Kadugodi   \n",
       "7                      77.736702                     Kadugodi   \n",
       "8                      77.731369                         Hudi   \n",
       "9                      77.731369                         Hudi   \n",
       "\n",
       "  deviceCode_pyld_alarmType  deviceCode_pyld_speed  \\\n",
       "0                       PCW                     32   \n",
       "1                       PCW                     32   \n",
       "2                       FCW                     41   \n",
       "3                       FCW                     41   \n",
       "4                 Overspeed                     37   \n",
       "5                 Overspeed                     37   \n",
       "6                       HMW                     32   \n",
       "7                       HMW                     32   \n",
       "8                 Overspeed                     27   \n",
       "9                 Overspeed                     27   \n",
       "\n",
       "  deviceCode_time_recordedTime_$date  \n",
       "0           2018-02-01T01:48:59.000Z  \n",
       "1           2018-02-01T01:48:59.000Z  \n",
       "2           2018-02-01T01:50:00.000Z  \n",
       "3           2018-02-01T01:50:00.000Z  \n",
       "4           2018-02-01T01:50:11.000Z  \n",
       "5           2018-02-01T01:50:11.000Z  \n",
       "6           2018-02-01T01:50:50.000Z  \n",
       "7           2018-02-01T01:50:50.000Z  \n",
       "8           2018-02-01T01:52:26.000Z  \n",
       "9           2018-02-01T01:52:26.000Z  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bdf.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We rename the timestamp column for better readability."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdf = bdf.rename(columns = {'deviceCode_time_recordedTime_$date':'timestamp'})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We extract the date from the timestamp using the inbuilt functions of pandas and datetime libraries."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdf['timestamp'] = pd.to_datetime(bdf['timestamp'])\n",
    "bdf['eventDate'] = pd.to_datetime(bdf['timestamp'])\n",
    "bdf['eventDate'] = bdf['eventDate'].dt.strftime('%Y%m%d')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We take the hour during which the event occured and map them to different categories to later study if the vehicle was observed during peak hours, early hours, regular hours and so on."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdf['e_hour'] = pd.to_datetime(bdf['timestamp'], format = '%H:%M:%S').dt.hour\n",
    "bdf['ehourCat'] = 0\n",
    "bdf['ehourCat'] = np.where((bdf['e_hour'] >= 0) & (bdf['e_hour'] < 6), 1, bdf['ehourCat'])\n",
    "bdf['ehourCat'] = np.where((bdf['e_hour'] >= 6) & (bdf['e_hour'] < 10), 2, bdf['ehourCat'])\n",
    "bdf['ehourCat'] = np.where((bdf['e_hour'] >= 10) & (bdf['e_hour'] < 16), 3, bdf['ehourCat'])\n",
    "bdf['ehourCat'] = np.where((bdf['e_hour'] >= 16) & (bdf['e_hour'] < 21), 4, bdf['ehourCat'])\n",
    "bdf['ehourCat'] = np.where((bdf['e_hour'] >= 21) & (bdf['e_hour'] < 24), 5, bdf['ehourCat'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Currently, we have features such as speed, area, latitude, longitude, the time when the observation was made. In addition to these feature, we believe it would be beneficial to have the weather conditions when these events were observed. In order to do so, we scrape the weather data of Bangalore for the period of February until July 2018 from: https://www.timeanddate.com/weather/india/bangalore"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The follow block of code below scrapes the weather data and saves it to another XLSX file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finished scraping weather data!\n"
     ]
    }
   ],
   "source": [
    "from time import time\n",
    "from datetime import timedelta\n",
    "import datetime\n",
    "from urllib.request import urlopen\n",
    "import re\n",
    "from pandas import ExcelWriter\n",
    "\n",
    "def date_splitter(dateToSplit):\n",
    "    dateToSplit = dateToSplit.split('-', 1)\n",
    "    year = dateToSplit[0]\n",
    "    dateToSplit = dateToSplit[1].split('-', 1)\n",
    "    month = dateToSplit[0]\n",
    "    day = dateToSplit[1]\n",
    "    return year, month, day\n",
    "\n",
    "def extractTime(s):\n",
    "    time = re.findall('\\d+:\\d+', s)\n",
    "    return time\n",
    "\n",
    "def extractConditions(s):\n",
    "    x = re.findall('\"small\",h:\"', s)\n",
    "    condList = []\n",
    "    for i in range(len(x)):\n",
    "        s = s.split('\"small\",h:\"', 1)\n",
    "        s = s[1].split('.\"},{s', 1)\n",
    "        condList.append(s[0])\n",
    "        s = s[1]\n",
    "    return condList\n",
    "\n",
    "def extractVisibility(s):\n",
    "    s = s.replace('&nbsp;', '')\n",
    "    re1 = '(\\\\d+)'\n",
    "    re2 = ''\n",
    "    re3 = 'km'\n",
    "    visibilities = re.findall(re1 + re2 + re3, s)\n",
    "    return visibilities\n",
    "\n",
    "def extractTemperature(s):\n",
    "    s = s.replace('°', '')\n",
    "    s = s.replace('&nbsp;', '')\n",
    "    re1='\\d+'\n",
    "    re2='[A-Z]'\n",
    "    temperatures = re.findall(re1 + re2, s)\n",
    "    tempList = list()\n",
    "    for i in range(len(temperatures)):\n",
    "        x = re.findall(re1, temperatures[i])[0]\n",
    "        tempList.append(x)\n",
    "    return tempList\n",
    "\n",
    "finalRawDf = pd.DataFrame()\n",
    "\n",
    "# datetime.date.today equals 26th August, 2018. Change accordingly to make the script work for the duration of scraping\n",
    "\n",
    "for i in range(26, 193):\n",
    "    startDate = str(datetime.date(2018, 8, 26) + timedelta(days = -i))\n",
    "    d_year, d_month, d_day = date_splitter(startDate)\n",
    "    finalDate = d_year + d_month + d_day\n",
    "    if d_month == '02':\n",
    "        link = 'https://www.timeanddate.com/scripts/cityajax.php?n=india/bangalore&mode=historic&hd=' + finalDate + '&month=2&year=2018&json=1'\n",
    "    elif d_month == '03':\n",
    "        link = 'https://www.timeanddate.com/scripts/cityajax.php?n=india/bangalore&mode=historic&hd=' + finalDate + '&month=3&year=2018&json=1'\n",
    "    elif d_month == '04':\n",
    "        link = 'https://www.timeanddate.com/scripts/cityajax.php?n=india/bangalore&mode=historic&hd=' + finalDate + '&month=4&year=2018&json=1'\n",
    "    elif d_month == '05':\n",
    "        link = 'https://www.timeanddate.com/scripts/cityajax.php?n=india/bangalore&mode=historic&hd=' + finalDate + '&month=5&year=2018&json=1'\n",
    "    elif d_month == '06':\n",
    "        link = 'https://www.timeanddate.com/scripts/cityajax.php?n=india/bangalore&mode=historic&hd=' + finalDate + '&month=6&year=2018&json=1'\n",
    "    elif d_month == '07':\n",
    "        link = 'https://www.timeanddate.com/scripts/cityajax.php?n=india/bangalore&mode=historic&hd=' + finalDate + '&month=7&year=2018&json=1'\n",
    "\n",
    "    #print('Scraping:', link)\n",
    "    html = urlopen(link)\n",
    "    s = html.read().decode('utf-8')\n",
    "\n",
    "    times = extractTime(s)\n",
    "    visibility = extractVisibility(s)\n",
    "    temperature = extractTemperature(s)\n",
    "    condition = extractConditions(s)\n",
    "\n",
    "    if(len(visibility) < len(condition)):\n",
    "        remLength = len(condition) - len(visibility)\n",
    "        for i in range(remLength):\n",
    "            visibility.append(visibility[0])\n",
    "\n",
    "    tempDf = pd.DataFrame({\n",
    "            'weatherDate': finalDate,\n",
    "            'time': times,\n",
    "            'temperature': temperature,\n",
    "            'visibility': visibility,\n",
    "            'condition': condition\n",
    "            })\n",
    "\n",
    "    finalRawDf = finalRawDf.append(tempDf)\n",
    "\n",
    "print('Finished scraping weather data!')\n",
    "writer = ExcelWriter('bangalore-weather.xlsx')\n",
    "finalRawDf.to_excel(writer, index = False, sheet_name = 'Sheet1')\n",
    "writer.save()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We read the scraped data file now to get a glimpse of the contents."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "bwdf = pd.read_excel('bangalore-weather.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4636 entries, 0 to 4635\n",
      "Data columns (total 5 columns):\n",
      "condition      4636 non-null object\n",
      "temperature    4636 non-null int64\n",
      "time           4636 non-null object\n",
      "visibility     4636 non-null int64\n",
      "weatherDate    4636 non-null int64\n",
      "dtypes: int64(3), object(2)\n",
      "memory usage: 181.2+ KB\n"
     ]
    }
   ],
   "source": [
    "bwdf.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>condition</th>\n",
       "      <th>temperature</th>\n",
       "      <th>time</th>\n",
       "      <th>visibility</th>\n",
       "      <th>weatherDate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Partly cloudy</td>\n",
       "      <td>20</td>\n",
       "      <td>05:30</td>\n",
       "      <td>8</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Partly cloudy</td>\n",
       "      <td>20</td>\n",
       "      <td>06:00</td>\n",
       "      <td>6</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Broken clouds</td>\n",
       "      <td>21</td>\n",
       "      <td>07:00</td>\n",
       "      <td>6</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Broken clouds</td>\n",
       "      <td>22</td>\n",
       "      <td>07:30</td>\n",
       "      <td>6</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Broken clouds</td>\n",
       "      <td>22</td>\n",
       "      <td>08:00</td>\n",
       "      <td>6</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Broken clouds</td>\n",
       "      <td>22</td>\n",
       "      <td>08:30</td>\n",
       "      <td>6</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Broken clouds</td>\n",
       "      <td>23</td>\n",
       "      <td>09:00</td>\n",
       "      <td>6</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Broken clouds</td>\n",
       "      <td>24</td>\n",
       "      <td>09:30</td>\n",
       "      <td>6</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Scattered clouds</td>\n",
       "      <td>24</td>\n",
       "      <td>10:00</td>\n",
       "      <td>8</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Scattered clouds</td>\n",
       "      <td>24</td>\n",
       "      <td>10:30</td>\n",
       "      <td>8</td>\n",
       "      <td>20180731</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          condition  temperature   time  visibility  weatherDate\n",
       "0     Partly cloudy           20  05:30           8     20180731\n",
       "1     Partly cloudy           20  06:00           6     20180731\n",
       "2     Broken clouds           21  07:00           6     20180731\n",
       "3     Broken clouds           22  07:30           6     20180731\n",
       "4     Broken clouds           22  08:00           6     20180731\n",
       "5     Broken clouds           22  08:30           6     20180731\n",
       "6     Broken clouds           23  09:00           6     20180731\n",
       "7     Broken clouds           24  09:30           6     20180731\n",
       "8  Scattered clouds           24  10:00           8     20180731\n",
       "9  Scattered clouds           24  10:30           8     20180731"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bwdf.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We follow similar strategy here to get the hour to equate it to the hour from the previous dataframe of events recorded by the Collision Avoidance System."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "bwdf['w_hour'] = pd.to_datetime(bwdf['time'], format= '%H:%M').dt.hour\n",
    "bwdf['hourCat'] = 0\n",
    "bwdf['hourCat'] = np.where((bwdf['w_hour'] >= 0) & (bwdf['w_hour'] < 6), 1, bwdf['hourCat'])\n",
    "bwdf['hourCat'] = np.where((bwdf['w_hour'] >= 6) & (bwdf['w_hour'] < 10), 2, bwdf['hourCat'])\n",
    "bwdf['hourCat'] = np.where((bwdf['w_hour'] >= 10) & (bwdf['w_hour'] < 16), 3, bwdf['hourCat'])\n",
    "bwdf['hourCat'] = np.where((bwdf['w_hour'] >= 16) & (bwdf['w_hour'] < 21), 4, bwdf['hourCat'])\n",
    "bwdf['hourCat'] = np.where((bwdf['w_hour'] >= 21) & (bwdf['w_hour'] < 24), 5, bwdf['hourCat'])\n",
    "bwdf = bwdf.drop_duplicates(subset = ['weatherDate', 'hourCat'], keep = 'first')\n",
    "bwdf['ehourCat'] = bwdf['hourCat']\n",
    "bwdf['weatherDate'] = bwdf['weatherDate'].astype(str)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To merge the two dataframes on a mutual column, we create an extra column with the same name weatherDate in our previous dataframe bdf."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "bdf['weatherDate'] = bdf['eventDate']\n",
    "bdf['weatherDate'] = bdf['weatherDate'].astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "b1 = pd.merge(bdf, bwdf, on = ['weatherDate', 'ehourCat'], how = 'left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our merged dataset now looks like shown below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 207617 entries, 0 to 207616\n",
      "Data columns (total 17 columns):\n",
      "deviceCode_deviceCode            207617 non-null int64\n",
      "deviceCode_location_latitude     207617 non-null float64\n",
      "deviceCode_location_longitude    207617 non-null float64\n",
      "deviceCode_location_wardName     207617 non-null object\n",
      "deviceCode_pyld_alarmType        207617 non-null object\n",
      "deviceCode_pyld_speed            207617 non-null int64\n",
      "timestamp                        207617 non-null datetime64[ns]\n",
      "eventDate                        207617 non-null object\n",
      "e_hour                           207617 non-null int64\n",
      "ehourCat                         207617 non-null int64\n",
      "weatherDate                      207617 non-null object\n",
      "condition                        190275 non-null object\n",
      "temperature                      190275 non-null float64\n",
      "time                             190275 non-null object\n",
      "visibility                       190275 non-null float64\n",
      "w_hour                           190275 non-null float64\n",
      "hourCat                          190275 non-null float64\n",
      "dtypes: datetime64[ns](1), float64(6), int64(4), object(6)\n",
      "memory usage: 28.5+ MB\n"
     ]
    }
   ],
   "source": [
    "b1.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>deviceCode_deviceCode</th>\n",
       "      <th>deviceCode_location_latitude</th>\n",
       "      <th>deviceCode_location_longitude</th>\n",
       "      <th>deviceCode_location_wardName</th>\n",
       "      <th>deviceCode_pyld_alarmType</th>\n",
       "      <th>deviceCode_pyld_speed</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>eventDate</th>\n",
       "      <th>e_hour</th>\n",
       "      <th>ehourCat</th>\n",
       "      <th>weatherDate</th>\n",
       "      <th>condition</th>\n",
       "      <th>temperature</th>\n",
       "      <th>time</th>\n",
       "      <th>visibility</th>\n",
       "      <th>w_hour</th>\n",
       "      <th>hourCat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.984595</td>\n",
       "      <td>77.744087</td>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>PCW</td>\n",
       "      <td>32</td>\n",
       "      <td>2018-02-01 01:48:59</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.984595</td>\n",
       "      <td>77.744087</td>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>PCW</td>\n",
       "      <td>32</td>\n",
       "      <td>2018-02-01 01:48:59</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987233</td>\n",
       "      <td>77.741119</td>\n",
       "      <td>Garudachar Playa</td>\n",
       "      <td>FCW</td>\n",
       "      <td>41</td>\n",
       "      <td>2018-02-01 01:50:00</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987233</td>\n",
       "      <td>77.741119</td>\n",
       "      <td>Garudachar Playa</td>\n",
       "      <td>FCW</td>\n",
       "      <td>41</td>\n",
       "      <td>2018-02-01 01:50:00</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987503</td>\n",
       "      <td>77.740051</td>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>37</td>\n",
       "      <td>2018-02-01 01:50:11</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987503</td>\n",
       "      <td>77.740051</td>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>37</td>\n",
       "      <td>2018-02-01 01:50:11</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987523</td>\n",
       "      <td>77.736702</td>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>32</td>\n",
       "      <td>2018-02-01 01:50:50</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.987523</td>\n",
       "      <td>77.736702</td>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>32</td>\n",
       "      <td>2018-02-01 01:50:50</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.988210</td>\n",
       "      <td>77.731369</td>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>27</td>\n",
       "      <td>2018-02-01 01:52:26</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>864504031502210</td>\n",
       "      <td>12.988210</td>\n",
       "      <td>77.731369</td>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>27</td>\n",
       "      <td>2018-02-01 01:52:26</td>\n",
       "      <td>20180201</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>20180201</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   deviceCode_deviceCode  deviceCode_location_latitude  \\\n",
       "0        864504031502210                     12.984595   \n",
       "1        864504031502210                     12.984595   \n",
       "2        864504031502210                     12.987233   \n",
       "3        864504031502210                     12.987233   \n",
       "4        864504031502210                     12.987503   \n",
       "5        864504031502210                     12.987503   \n",
       "6        864504031502210                     12.987523   \n",
       "7        864504031502210                     12.987523   \n",
       "8        864504031502210                     12.988210   \n",
       "9        864504031502210                     12.988210   \n",
       "\n",
       "   deviceCode_location_longitude deviceCode_location_wardName  \\\n",
       "0                      77.744087                     Kadugodi   \n",
       "1                      77.744087                     Kadugodi   \n",
       "2                      77.741119             Garudachar Playa   \n",
       "3                      77.741119             Garudachar Playa   \n",
       "4                      77.740051                         Hudi   \n",
       "5                      77.740051                         Hudi   \n",
       "6                      77.736702                     Kadugodi   \n",
       "7                      77.736702                     Kadugodi   \n",
       "8                      77.731369                         Hudi   \n",
       "9                      77.731369                         Hudi   \n",
       "\n",
       "  deviceCode_pyld_alarmType  deviceCode_pyld_speed           timestamp  \\\n",
       "0                       PCW                     32 2018-02-01 01:48:59   \n",
       "1                       PCW                     32 2018-02-01 01:48:59   \n",
       "2                       FCW                     41 2018-02-01 01:50:00   \n",
       "3                       FCW                     41 2018-02-01 01:50:00   \n",
       "4                 Overspeed                     37 2018-02-01 01:50:11   \n",
       "5                 Overspeed                     37 2018-02-01 01:50:11   \n",
       "6                       HMW                     32 2018-02-01 01:50:50   \n",
       "7                       HMW                     32 2018-02-01 01:50:50   \n",
       "8                 Overspeed                     27 2018-02-01 01:52:26   \n",
       "9                 Overspeed                     27 2018-02-01 01:52:26   \n",
       "\n",
       "  eventDate  e_hour  ehourCat weatherDate condition  temperature time  \\\n",
       "0  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "1  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "2  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "3  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "4  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "5  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "6  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "7  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "8  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "9  20180201       1         1    20180201       NaN          NaN  NaN   \n",
       "\n",
       "   visibility  w_hour  hourCat  \n",
       "0         NaN     NaN      NaN  \n",
       "1         NaN     NaN      NaN  \n",
       "2         NaN     NaN      NaN  \n",
       "3         NaN     NaN      NaN  \n",
       "4         NaN     NaN      NaN  \n",
       "5         NaN     NaN      NaN  \n",
       "6         NaN     NaN      NaN  \n",
       "7         NaN     NaN      NaN  \n",
       "8         NaN     NaN      NaN  \n",
       "9         NaN     NaN      NaN  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b1.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We rename the warname column to Area."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "b1 = b1.rename(columns = {'deviceCode_location_wardName':'Area'})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The dataset we have currently does not mention about the accident prone zones of Bengaluru. This information was obtained from: http://bengaluru.citizenmatters.in/top-10-places-where-most-accidents-in-bengaluru-take-place-7260"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We mapped the areas which we have in our dataset to the nearest accident prone zone marked in the link based on the number of fatalities reported by the traffic police. We performed this task manually. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Similary, we also obtained the pothole data of Bengaluru from this link and repeated the task of mapping the pothole centric areas to the areas in our initial dataset. The pothole data was obtained manually from: https://www.potholeraja.com/live-potholeview"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "badf = pd.read_excel('bangalore-accident-zones.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have a look at the data of accident prone zones and pothole centric zones of Bengaluru. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 50 entries, 0 to 49\n",
      "Data columns (total 4 columns):\n",
      "Area                 50 non-null object\n",
      "Mapped_Location      50 non-null object\n",
      "Accident_Severity    50 non-null object\n",
      "Pothole_Severity     50 non-null object\n",
      "dtypes: object(4)\n",
      "memory usage: 1.6+ KB\n"
     ]
    }
   ],
   "source": [
    "badf.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Area</th>\n",
       "      <th>Mapped_Location</th>\n",
       "      <th>Accident_Severity</th>\n",
       "      <th>Pothole_Severity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A Narayanapura</td>\n",
       "      <td>Electronic City</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Agaram</td>\n",
       "      <td>Electronic City</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Banasavadi</td>\n",
       "      <td>Banasavadi</td>\n",
       "      <td>Low</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Basavanapura</td>\n",
       "      <td>K R Puram</td>\n",
       "      <td>Medium</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Bellanduru</td>\n",
       "      <td>Electronic City</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Benniganahalli</td>\n",
       "      <td>Banasavadi</td>\n",
       "      <td>Low</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Bharathi Nagar</td>\n",
       "      <td>K R Puram</td>\n",
       "      <td>Medium</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>BTM Layout</td>\n",
       "      <td>Madivala</td>\n",
       "      <td>High</td>\n",
       "      <td>Medium</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>C V Raman Nagar</td>\n",
       "      <td>Banasavadi</td>\n",
       "      <td>Low</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Chickpete</td>\n",
       "      <td>Madivala</td>\n",
       "      <td>High</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              Area  Mapped_Location Accident_Severity Pothole_Severity\n",
       "0   A Narayanapura  Electronic City              High             High\n",
       "1           Agaram  Electronic City              High             High\n",
       "2       Banasavadi       Banasavadi               Low              Low\n",
       "3     Basavanapura        K R Puram            Medium             High\n",
       "4       Bellanduru  Electronic City              High             High\n",
       "5   Benniganahalli       Banasavadi               Low              Low\n",
       "6   Bharathi Nagar        K R Puram            Medium             High\n",
       "7       BTM Layout         Madivala              High           Medium\n",
       "8  C V Raman Nagar       Banasavadi               Low              Low\n",
       "9        Chickpete         Madivala              High              Low"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "badf.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We merge this data to our initial dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "b = pd.merge(b1, badf, on = ['Area'], how = 'left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We rename the Alarm type and Plying speed columns for better readability."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "b = b.rename(columns = {'deviceCode_pyld_alarmType':'Alarm_Type'})\n",
    "b = b.rename(columns = {'deviceCode_pyld_speed':'Plying_Speed'})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The city speed limit of Bengaluru is 50kmph but given that some of the roads may connect to highways, we have given extra buffer of 10kmph to mark it as an event of Overpseeding. The source of this information can be found here: http://www.bangaloretrafficpolice.gov.in/faqs.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "b['hasOversped'] = np.where(b.Plying_Speed > 60, 1, 0)\n",
    "b['hasOversped'] = np.where(b.Alarm_Type == 'Overspeed', 1, b['hasOversped'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We inspect the current dataset we have to see how it looks so far. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 207617 entries, 0 to 207616\n",
      "Data columns (total 21 columns):\n",
      "deviceCode_deviceCode            207617 non-null int64\n",
      "deviceCode_location_latitude     207617 non-null float64\n",
      "deviceCode_location_longitude    207617 non-null float64\n",
      "Area                             207617 non-null object\n",
      "Alarm_Type                       207617 non-null object\n",
      "Plying_Speed                     207617 non-null int64\n",
      "timestamp                        207617 non-null datetime64[ns]\n",
      "eventDate                        207617 non-null object\n",
      "e_hour                           207617 non-null int64\n",
      "ehourCat                         207617 non-null int64\n",
      "weatherDate                      207617 non-null object\n",
      "condition                        190275 non-null object\n",
      "temperature                      190275 non-null float64\n",
      "time                             190275 non-null object\n",
      "visibility                       190275 non-null float64\n",
      "w_hour                           190275 non-null float64\n",
      "hourCat                          190275 non-null float64\n",
      "Mapped_Location                  207617 non-null object\n",
      "Accident_Severity                207617 non-null object\n",
      "Pothole_Severity                 207617 non-null object\n",
      "hasOversped                      207617 non-null int64\n",
      "dtypes: datetime64[ns](1), float64(6), int64(5), object(9)\n",
      "memory usage: 34.8+ MB\n"
     ]
    }
   ],
   "source": [
    "b.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Some of the columns do not have enough entries compared to our initial entries. This is due to the fact that the weather scraper could not find information for all the hours when the events were recorded by the Collision Avoidance System. Hence, we replace the NaN data for which we could not find information with the Mode values, that is the most frequently occuring values of certain columns which we will use for our final prediction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "for column in ['temperature', 'visibility', 'condition']:\n",
    "    b[column].fillna(b[column].mode()[0], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 207617 entries, 0 to 207616\n",
      "Data columns (total 21 columns):\n",
      "deviceCode_deviceCode            207617 non-null int64\n",
      "deviceCode_location_latitude     207617 non-null float64\n",
      "deviceCode_location_longitude    207617 non-null float64\n",
      "Area                             207617 non-null object\n",
      "Alarm_Type                       207617 non-null object\n",
      "Plying_Speed                     207617 non-null int64\n",
      "timestamp                        207617 non-null datetime64[ns]\n",
      "eventDate                        207617 non-null object\n",
      "e_hour                           207617 non-null int64\n",
      "ehourCat                         207617 non-null int64\n",
      "weatherDate                      207617 non-null object\n",
      "condition                        207617 non-null object\n",
      "temperature                      207617 non-null float64\n",
      "time                             190275 non-null object\n",
      "visibility                       207617 non-null float64\n",
      "w_hour                           190275 non-null float64\n",
      "hourCat                          190275 non-null float64\n",
      "Mapped_Location                  207617 non-null object\n",
      "Accident_Severity                207617 non-null object\n",
      "Pothole_Severity                 207617 non-null object\n",
      "hasOversped                      207617 non-null int64\n",
      "dtypes: datetime64[ns](1), float64(6), int64(5), object(9)\n",
      "memory usage: 34.8+ MB\n"
     ]
    }
   ],
   "source": [
    "b.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The temperature, condition and visibility columns have the same number of records since we replaced them with the Mode values of each of those columns."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We create a flag for low and high visibility. Anything great than 10 would be equated to high visibility whereas the other case would be low visibility. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "b['visibility'] = np.where(b['visibility'] < 10, 0, 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now map all the columns we need to its numerical equivalent in the form of flags. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = b.copy()\n",
    "df['hasOversped'] = np.where(b.hasOversped == 1, 'Yes', 'No')\n",
    "df['visibility'] = np.where(b.visibility == 0, 'Low', 'High')\n",
    "df['ehourCat'] = b['ehourCat'].map({1: 'Early', 2: 'PeakM', 3: 'RegularM'})\n",
    "b['Accident_Severity'] = b['Accident_Severity'].map({'High': 3, 'Medium': 2, 'Low': 1})\n",
    "b['Pothole_Severity'] = b['Pothole_Severity'].map({'High': 3, 'Medium': 2, 'Low': 1})\n",
    "b['Alarm_Type'] = b['Alarm_Type'].map({'PCW': 1, 'FCW': 2, 'Overspeed': 3, 'HMW': 4, 'UFCW': 5, 'LDWL': 6, 'LDWR': 7})\n",
    "b['condition'] = b['condition'].map({'Clear': 1, 'Sunny': 2, 'Passing clouds': 3,\n",
    "       'Broken clouds': 4, 'Scattered clouds': 5, 'Fog': 6, 'Haze': 7, 'Partly cloudy': 8,\n",
    "       'Mild': 9, 'Drizzle. Broken clouds': 10})\n",
    "b['Area'] = b['Area'].map({'Kadugodi': 1, 'Garudachar Playa': 2, 'Hudi': 3, 'Other': 4, 'Devasandra': 5,\n",
    "       'Hagadur': 6, 'Bellanduru': 7, 'Marathahalli': 8, 'Dodda Nekkundi': 9, 'Varthuru': 10,\n",
    "       'HAL Airport': 11, 'Vijnana Nagar': 12, 'Konena Agrahara': 13, 'A Narayanapura': 14,\n",
    "       'C V Raman Nagar': 15, 'Jeevanbhima Nagar': 16, 'HSR Layout': 17, 'Domlur': 18, 'Jogupalya': 19,\n",
    "       'Hoysala Nagar': 20, 'New Tippasandara': 21, 'Benniganahalli': 22, 'Singasandra': 23,\n",
    "       'Basavanapura': 24, 'Halsoor': 25, 'Agaram': 26, 'Shantala Nagar': 27, 'Sampangiram Nagar': 28,\n",
    "       'Sudham Nagara': 29, 'Dharmaraya Swamy Temple': 30, 'Chickpete': 31, 'Banasavadi': 32,\n",
    "       'Horamavu': 33, 'Kacharkanahalli': 34, 'Kammanahalli': 35, 'Vijnanapura': 36, 'Ramamurthy Nagar': 37,\n",
    "       'K R Puram': 38, 'BTM Layout': 39, 'Madivala': 40, 'Gurappanapalya': 41, 'J P Nagar': 42, 'Sarakki': 43,\n",
    "       'Jaraganahalli': 44, 'Vasanthpura': 45, 'Hemmigepura': 46, 'Yelchenahalli': 47,\n",
    "       'Jayanagar East': 48, 'Bharathi Nagar': 49, 'other': 4})\n",
    "\n",
    "writer = ExcelWriter('bangalore-consolidated-data.xlsx')\n",
    "b.to_excel(writer, index = False, sheet_name = 'Sheet1')\n",
    "df.to_excel(writer, index = False, sheet_name = 'Sheet2')\n",
    "writer.save()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We get rid of the columns that we do not need and feel has no impact on the chance of accident likelihood. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "del b['deviceCode_deviceCode'], b['deviceCode_location_latitude'], b['deviceCode_location_longitude']\n",
    "del b['w_hour'], b['Mapped_Location'], b['timestamp'], b['e_hour'], b['weatherDate']\n",
    "del b['hourCat'], b['time'], b['temperature'], b['eventDate'], b['Plying_Speed']\n",
    "\n",
    "del df['deviceCode_deviceCode'], df['deviceCode_location_latitude'], df['deviceCode_location_longitude']\n",
    "del df['w_hour'], df['Mapped_Location'], df['timestamp'], df['e_hour'], df['weatherDate']\n",
    "del df['hourCat'], df['time'], df['temperature'], df['eventDate'], df['Plying_Speed']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now have two datasets: one that has all the data in words for understandability and the other in numerical format to fit to our prediction model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 207617 entries, 0 to 207616\n",
      "Data columns (total 8 columns):\n",
      "Area                 207617 non-null int64\n",
      "Alarm_Type           207617 non-null int64\n",
      "ehourCat             207617 non-null int64\n",
      "condition            207617 non-null int64\n",
      "visibility           207617 non-null int64\n",
      "Accident_Severity    207617 non-null int64\n",
      "Pothole_Severity     207617 non-null int64\n",
      "hasOversped          207617 non-null int64\n",
      "dtypes: int64(8)\n",
      "memory usage: 14.3 MB\n"
     ]
    }
   ],
   "source": [
    "b.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Area</th>\n",
       "      <th>Alarm_Type</th>\n",
       "      <th>ehourCat</th>\n",
       "      <th>condition</th>\n",
       "      <th>visibility</th>\n",
       "      <th>Accident_Severity</th>\n",
       "      <th>Pothole_Severity</th>\n",
       "      <th>hasOversped</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Area  Alarm_Type  ehourCat  condition  visibility  Accident_Severity  \\\n",
       "0     1           1         1          3           0                  2   \n",
       "1     1           1         1          3           0                  2   \n",
       "2     2           2         1          3           0                  2   \n",
       "3     2           2         1          3           0                  2   \n",
       "4     3           3         1          3           0                  2   \n",
       "5     3           3         1          3           0                  2   \n",
       "6     1           4         1          3           0                  2   \n",
       "7     1           4         1          3           0                  2   \n",
       "8     3           3         1          3           0                  2   \n",
       "9     3           3         1          3           0                  2   \n",
       "\n",
       "   Pothole_Severity  hasOversped  \n",
       "0                 2            0  \n",
       "1                 2            0  \n",
       "2                 2            0  \n",
       "3                 2            0  \n",
       "4                 2            1  \n",
       "5                 2            1  \n",
       "6                 2            0  \n",
       "7                 2            0  \n",
       "8                 2            1  \n",
       "9                 2            1  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The equivalent dataset for this which can be in human understandable format is as below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 207617 entries, 0 to 207616\n",
      "Data columns (total 8 columns):\n",
      "Area                 207617 non-null object\n",
      "Alarm_Type           207617 non-null object\n",
      "ehourCat             207617 non-null object\n",
      "condition            207617 non-null object\n",
      "visibility           207617 non-null object\n",
      "Accident_Severity    207617 non-null object\n",
      "Pothole_Severity     207617 non-null object\n",
      "hasOversped          207617 non-null object\n",
      "dtypes: object(8)\n",
      "memory usage: 14.3+ MB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Area</th>\n",
       "      <th>Alarm_Type</th>\n",
       "      <th>ehourCat</th>\n",
       "      <th>condition</th>\n",
       "      <th>visibility</th>\n",
       "      <th>Accident_Severity</th>\n",
       "      <th>Pothole_Severity</th>\n",
       "      <th>hasOversped</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>PCW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>PCW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Garudachar Playa</td>\n",
       "      <td>FCW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Garudachar Playa</td>\n",
       "      <td>FCW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Yes</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Area Alarm_Type ehourCat       condition visibility  \\\n",
       "0          Kadugodi        PCW    Early  Passing clouds        Low   \n",
       "1          Kadugodi        PCW    Early  Passing clouds        Low   \n",
       "2  Garudachar Playa        FCW    Early  Passing clouds        Low   \n",
       "3  Garudachar Playa        FCW    Early  Passing clouds        Low   \n",
       "4              Hudi  Overspeed    Early  Passing clouds        Low   \n",
       "5              Hudi  Overspeed    Early  Passing clouds        Low   \n",
       "6          Kadugodi        HMW    Early  Passing clouds        Low   \n",
       "7          Kadugodi        HMW    Early  Passing clouds        Low   \n",
       "8              Hudi  Overspeed    Early  Passing clouds        Low   \n",
       "9              Hudi  Overspeed    Early  Passing clouds        Low   \n",
       "\n",
       "  Accident_Severity Pothole_Severity hasOversped  \n",
       "0            Medium           Medium          No  \n",
       "1            Medium           Medium          No  \n",
       "2            Medium           Medium          No  \n",
       "3            Medium           Medium          No  \n",
       "4            Medium           Medium         Yes  \n",
       "5            Medium           Medium         Yes  \n",
       "6            Medium           Medium          No  \n",
       "7            Medium           Medium          No  \n",
       "8            Medium           Medium         Yes  \n",
       "9            Medium           Medium         Yes  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since the data is unlabeled, we opt for unsupervised learning approach using clustering techniques. We have tried different clustering algorithms but we present the ones which gave us the most suitable labels as per our intuition."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finished clustering using K-Means\n"
     ]
    }
   ],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "X = b.values.astype(np.float)\n",
    "kmeans = KMeans(n_clusters = 2, max_iter = 2000, algorithm = 'full').fit(X)\n",
    "kmf2labels = kmeans.labels_\n",
    "kmf2labels = kmf2labels.tolist()\n",
    "print('Finished clustering using K-Means')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "While different variation of K-Means exists, we used the Expectation Maximization method. In order to get better convergence of our datapoints to centroid, we also increased the number of iterations from a default of 300 to 2000. The Expectation Maximization algorithm can be found here: https://en.wikipedia.org/wiki/Expectation–maximization_algorithm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We create two sets of clusters here: one for low likelihood accident zones whereas the other one for high likelihood accident zones. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "207617"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(kmf2labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So, we now have labels for all the data points from our dataset containing features: area, accident severity, pothole severity, speed, alarm type, overspeeding flag, hour category, visibility and sky conditions, which can be crucial towards understanding an accident would be taking place or not."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let us join this list of labels to the dataframe to study them carefully."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "b['labels'] = kmf2labels\n",
    "df['labels'] = kmf2labels\n",
    "df['labels'] = df['labels'].map({0: 'High', 1: 'Low'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Area</th>\n",
       "      <th>Alarm_Type</th>\n",
       "      <th>ehourCat</th>\n",
       "      <th>condition</th>\n",
       "      <th>visibility</th>\n",
       "      <th>Accident_Severity</th>\n",
       "      <th>Pothole_Severity</th>\n",
       "      <th>hasOversped</th>\n",
       "      <th>labels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Area  Alarm_Type  ehourCat  condition  visibility  Accident_Severity  \\\n",
       "0     1           1         1          3           0                  2   \n",
       "1     1           1         1          3           0                  2   \n",
       "2     2           2         1          3           0                  2   \n",
       "3     2           2         1          3           0                  2   \n",
       "4     3           3         1          3           0                  2   \n",
       "5     3           3         1          3           0                  2   \n",
       "6     1           4         1          3           0                  2   \n",
       "7     1           4         1          3           0                  2   \n",
       "8     3           3         1          3           0                  2   \n",
       "9     3           3         1          3           0                  2   \n",
       "\n",
       "   Pothole_Severity  hasOversped  labels  \n",
       "0                 2            0       0  \n",
       "1                 2            0       0  \n",
       "2                 2            0       0  \n",
       "3                 2            0       0  \n",
       "4                 2            1       0  \n",
       "5                 2            1       0  \n",
       "6                 2            0       0  \n",
       "7                 2            0       0  \n",
       "8                 2            1       0  \n",
       "9                 2            1       0  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Area</th>\n",
       "      <th>Alarm_Type</th>\n",
       "      <th>ehourCat</th>\n",
       "      <th>condition</th>\n",
       "      <th>visibility</th>\n",
       "      <th>Accident_Severity</th>\n",
       "      <th>Pothole_Severity</th>\n",
       "      <th>hasOversped</th>\n",
       "      <th>labels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>PCW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>PCW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Garudachar Playa</td>\n",
       "      <td>FCW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Garudachar Playa</td>\n",
       "      <td>FCW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Yes</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Yes</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Yes</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Hudi</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>Early</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Yes</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Area Alarm_Type ehourCat       condition visibility  \\\n",
       "0          Kadugodi        PCW    Early  Passing clouds        Low   \n",
       "1          Kadugodi        PCW    Early  Passing clouds        Low   \n",
       "2  Garudachar Playa        FCW    Early  Passing clouds        Low   \n",
       "3  Garudachar Playa        FCW    Early  Passing clouds        Low   \n",
       "4              Hudi  Overspeed    Early  Passing clouds        Low   \n",
       "5              Hudi  Overspeed    Early  Passing clouds        Low   \n",
       "6          Kadugodi        HMW    Early  Passing clouds        Low   \n",
       "7          Kadugodi        HMW    Early  Passing clouds        Low   \n",
       "8              Hudi  Overspeed    Early  Passing clouds        Low   \n",
       "9              Hudi  Overspeed    Early  Passing clouds        Low   \n",
       "\n",
       "  Accident_Severity Pothole_Severity hasOversped labels  \n",
       "0            Medium           Medium          No   High  \n",
       "1            Medium           Medium          No   High  \n",
       "2            Medium           Medium          No   High  \n",
       "3            Medium           Medium          No   High  \n",
       "4            Medium           Medium         Yes   High  \n",
       "5            Medium           Medium         Yes   High  \n",
       "6            Medium           Medium          No   High  \n",
       "7            Medium           Medium          No   High  \n",
       "8            Medium           Medium         Yes   High  \n",
       "9            Medium           Medium         Yes   High  "
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It can be seen that in cases of medium accident prone zones and pothole severities, overspeeding with low visibility during early hours of the data can have high chances of accident as per the given area and alarm type."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "writer = ExcelWriter('bangalore-labeled-data.xlsx')\n",
    "b.to_excel(writer, index = False, sheet_name = 'Sheet1')\n",
    "df.to_excel(writer, index = False, sheet_name = 'Sheet2')\n",
    "writer.save()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Area</th>\n",
       "      <th>Alarm_Type</th>\n",
       "      <th>ehourCat</th>\n",
       "      <th>condition</th>\n",
       "      <th>visibility</th>\n",
       "      <th>Accident_Severity</th>\n",
       "      <th>Pothole_Severity</th>\n",
       "      <th>hasOversped</th>\n",
       "      <th>labels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3251</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3252</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3253</th>\n",
       "      <td>Kadugodi</td>\n",
       "      <td>HMW</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>Medium</td>\n",
       "      <td>Medium</td>\n",
       "      <td>No</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3254</th>\n",
       "      <td>Bellanduru</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3255</th>\n",
       "      <td>Bellanduru</td>\n",
       "      <td>Overspeed</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "      <td>Yes</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3256</th>\n",
       "      <td>Bellanduru</td>\n",
       "      <td>HMW</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "      <td>No</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3257</th>\n",
       "      <td>Bellanduru</td>\n",
       "      <td>HMW</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "      <td>No</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3258</th>\n",
       "      <td>Bellanduru</td>\n",
       "      <td>HMW</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "      <td>No</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3259</th>\n",
       "      <td>Bellanduru</td>\n",
       "      <td>HMW</td>\n",
       "      <td>PeakM</td>\n",
       "      <td>Passing clouds</td>\n",
       "      <td>Low</td>\n",
       "      <td>High</td>\n",
       "      <td>High</td>\n",
       "      <td>No</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Area Alarm_Type ehourCat       condition visibility  \\\n",
       "3251    Kadugodi        HMW    PeakM  Passing clouds        Low   \n",
       "3252    Kadugodi        HMW    PeakM  Passing clouds        Low   \n",
       "3253    Kadugodi        HMW    PeakM  Passing clouds        Low   \n",
       "3254  Bellanduru  Overspeed    PeakM  Passing clouds        Low   \n",
       "3255  Bellanduru  Overspeed    PeakM  Passing clouds        Low   \n",
       "3256  Bellanduru        HMW    PeakM  Passing clouds        Low   \n",
       "3257  Bellanduru        HMW    PeakM  Passing clouds        Low   \n",
       "3258  Bellanduru        HMW    PeakM  Passing clouds        Low   \n",
       "3259  Bellanduru        HMW    PeakM  Passing clouds        Low   \n",
       "\n",
       "     Accident_Severity Pothole_Severity hasOversped labels  \n",
       "3251            Medium           Medium          No   High  \n",
       "3252            Medium           Medium          No   High  \n",
       "3253            Medium           Medium          No   High  \n",
       "3254              High             High         Yes    Low  \n",
       "3255              High             High         Yes    Low  \n",
       "3256              High             High          No    Low  \n",
       "3257              High             High          No    Low  \n",
       "3258              High             High          No    Low  \n",
       "3259              High             High          No    Low  "
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[3251:3260]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above results, we can see areas with high accident prone zones and high potholes are more vulnerable to accidents no matter what the conditions are during the morning hours. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x144cf3c88>"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x144dfeac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.violinplot(x = 'labels', y = 'Accident_Severity', data = b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x14572c128>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x145726b00>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.violinplot(x = 'labels', y = 'hasOversped', data = b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1456e4d68>"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x145714128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.violinplot(x = 'labels', y = 'Alarm_Type', data = b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x14582c5c0>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1456d7160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.violinplot(x = 'labels', y = 'ehourCat', data = b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x14524eb38>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x145851d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.violinplot(x = 'labels', y = 'Pothole_Severity', data = b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.PairGrid at 0x115e8bc88>"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10eb00898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.pairplot(b, hue = 'labels')"
   ]
  }
 ],
 "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}