Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "The goal of this notebook is to help adventure works predict if a customer wil buy a bike or not given customer information, average month spend and available bike buyer data. This could help adventure works target their marketing and improve their returns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# perform necessary import\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import numpy.random as nr\n",
    "import math\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# load datasets\n",
    "customer_info = pd.read_csv('data/AdvWorksCusts.csv')\n",
    "customer_spending = pd.read_csv('data/AW_AveMonthSpend.csv')\n",
    "customer_has_bike= pd.read_csv('data/AW_BikeBuyer.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we will clean up data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16519, 23)\n",
      "(16404,)\n"
     ]
    }
   ],
   "source": [
    "# remove duplicates\n",
    "print(customer_info.shape)\n",
    "print(customer_info.CustomerID.unique().shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16404, 23)\n",
      "(16404,)\n"
     ]
    }
   ],
   "source": [
    "customer_info.drop_duplicates(subset = 'CustomerID', keep = 'last', inplace = True)\n",
    "print(customer_info.shape)\n",
    "print(customer_info.CustomerID.unique().shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16519, 2)\n",
      "(16404,)\n"
     ]
    }
   ],
   "source": [
    "print(customer_spending.shape)\n",
    "print(customer_spending.CustomerID.unique().shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16404, 2)\n",
      "(16404,)\n"
     ]
    }
   ],
   "source": [
    "customer_spending.drop_duplicates(subset = 'CustomerID', keep = 'last', inplace = True)\n",
    "print(customer_spending.shape)\n",
    "print(customer_spending.CustomerID.unique().shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16519, 2)\n",
      "(16404,)\n"
     ]
    }
   ],
   "source": [
    "print(customer_has_bike.shape)\n",
    "print(customer_has_bike.CustomerID.unique().shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16404, 2)\n",
      "(16404,)\n"
     ]
    }
   ],
   "source": [
    "customer_has_bike.drop_duplicates(subset = 'CustomerID', keep = 'last', inplace = True)\n",
    "print(customer_has_bike.shape)\n",
    "print(customer_has_bike.CustomerID.unique().shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CustomerID    False\n",
       "BikeBuyer     False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check for missing values\n",
    "(customer_has_bike.astype(np.object).isnull()).any()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CustomerID       False\n",
       "AveMonthSpend    False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(customer_spending.astype(np.object).isnull()).any()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CustomerID              False\n",
       "Title                    True\n",
       "FirstName               False\n",
       "MiddleName               True\n",
       "LastName                False\n",
       "Suffix                   True\n",
       "AddressLine1            False\n",
       "AddressLine2             True\n",
       "City                    False\n",
       "StateProvinceName       False\n",
       "CountryRegionName       False\n",
       "PostalCode              False\n",
       "PhoneNumber             False\n",
       "BirthDate               False\n",
       "Education               False\n",
       "Occupation              False\n",
       "Gender                  False\n",
       "MaritalStatus           False\n",
       "HomeOwnerFlag           False\n",
       "NumberCarsOwned         False\n",
       "NumberChildrenAtHome    False\n",
       "TotalChildren           False\n",
       "YearlyIncome            False\n",
       "dtype: bool"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(customer_info.astype(np.object).isnull()).any()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Fine with the missing values check, no key features have missing values. Duplicates are now removed. Now on to checking the statistics of the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>HomeOwnerFlag</th>\n",
       "      <th>NumberCarsOwned</th>\n",
       "      <th>NumberChildrenAtHome</th>\n",
       "      <th>TotalChildren</th>\n",
       "      <th>YearlyIncome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>16404.000000</td>\n",
       "      <td>16404.000000</td>\n",
       "      <td>16404.000000</td>\n",
       "      <td>16404.000000</td>\n",
       "      <td>16404.000000</td>\n",
       "      <td>16404.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>20240.143136</td>\n",
       "      <td>0.674104</td>\n",
       "      <td>1.499756</td>\n",
       "      <td>0.991100</td>\n",
       "      <td>2.004511</td>\n",
       "      <td>78129.672397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>5340.370705</td>\n",
       "      <td>0.468723</td>\n",
       "      <td>1.139135</td>\n",
       "      <td>1.514278</td>\n",
       "      <td>1.682318</td>\n",
       "      <td>39728.383493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>11000.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9482.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>15617.750000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>47808.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>20231.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>76125.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>24862.250000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>105211.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>29482.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>196511.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         CustomerID  HomeOwnerFlag  NumberCarsOwned  NumberChildrenAtHome  \\\n",
       "count  16404.000000   16404.000000     16404.000000          16404.000000   \n",
       "mean   20240.143136       0.674104         1.499756              0.991100   \n",
       "std     5340.370705       0.468723         1.139135              1.514278   \n",
       "min    11000.000000       0.000000         0.000000              0.000000   \n",
       "25%    15617.750000       0.000000         1.000000              0.000000   \n",
       "50%    20231.500000       1.000000         2.000000              0.000000   \n",
       "75%    24862.250000       1.000000         2.000000              2.000000   \n",
       "max    29482.000000       1.000000         4.000000              5.000000   \n",
       "\n",
       "       TotalChildren   YearlyIncome  \n",
       "count   16404.000000   16404.000000  \n",
       "mean        2.004511   78129.672397  \n",
       "std         1.682318   39728.383493  \n",
       "min         0.000000    9482.000000  \n",
       "25%         0.000000   47808.750000  \n",
       "50%         2.000000   76125.000000  \n",
       "75%         3.000000  105211.750000  \n",
       "max         5.000000  196511.000000  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customer_info.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>AveMonthSpend</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>16404.000000</td>\n",
       "      <td>16404.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>20240.143136</td>\n",
       "      <td>72.391002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>5340.370705</td>\n",
       "      <td>27.269921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>11000.000000</td>\n",
       "      <td>22.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>15617.750000</td>\n",
       "      <td>52.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>20231.500000</td>\n",
       "      <td>68.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>24862.250000</td>\n",
       "      <td>84.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>29482.000000</td>\n",
       "      <td>176.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         CustomerID  AveMonthSpend\n",
       "count  16404.000000   16404.000000\n",
       "mean   20240.143136      72.391002\n",
       "std     5340.370705      27.269921\n",
       "min    11000.000000      22.000000\n",
       "25%    15617.750000      52.000000\n",
       "50%    20231.500000      68.000000\n",
       "75%    24862.250000      84.000000\n",
       "max    29482.000000     176.000000"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customer_spending.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>BikeBuyer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>16404.000000</td>\n",
       "      <td>16404.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>20240.143136</td>\n",
       "      <td>0.332541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>5340.370705</td>\n",
       "      <td>0.471138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>11000.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>15617.750000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>20231.500000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>24862.250000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>29482.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         CustomerID     BikeBuyer\n",
       "count  16404.000000  16404.000000\n",
       "mean   20240.143136      0.332541\n",
       "std     5340.370705      0.471138\n",
       "min    11000.000000      0.000000\n",
       "25%    15617.750000      0.000000\n",
       "50%    20231.500000      0.000000\n",
       "75%    24862.250000      1.000000\n",
       "max    29482.000000      1.000000"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customer_has_bike.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.667459\n",
       "1    0.332541\n",
       "Name: BikeBuyer, dtype: float64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customer_has_bike.BikeBuyer.value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will join the bike buyer column to customer info for one combined table. For test data challenge AveMonthSpend is not provided hence not a viable feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>Title</th>\n",
       "      <th>FirstName</th>\n",
       "      <th>MiddleName</th>\n",
       "      <th>LastName</th>\n",
       "      <th>Suffix</th>\n",
       "      <th>AddressLine1</th>\n",
       "      <th>AddressLine2</th>\n",
       "      <th>City</th>\n",
       "      <th>StateProvinceName</th>\n",
       "      <th>...</th>\n",
       "      <th>Education</th>\n",
       "      <th>Occupation</th>\n",
       "      <th>Gender</th>\n",
       "      <th>MaritalStatus</th>\n",
       "      <th>HomeOwnerFlag</th>\n",
       "      <th>NumberCarsOwned</th>\n",
       "      <th>NumberChildrenAtHome</th>\n",
       "      <th>TotalChildren</th>\n",
       "      <th>YearlyIncome</th>\n",
       "      <th>BikeBuyer</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>11000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Jon</td>\n",
       "      <td>V</td>\n",
       "      <td>Yang</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3761 N. 14th St</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Rockhampton</td>\n",
       "      <td>Queensland</td>\n",
       "      <td>...</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>Professional</td>\n",
       "      <td>M</td>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>137947</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11001</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Eugene</td>\n",
       "      <td>L</td>\n",
       "      <td>Huang</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2243 W St.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Seaford</td>\n",
       "      <td>Victoria</td>\n",
       "      <td>...</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>Professional</td>\n",
       "      <td>M</td>\n",
       "      <td>S</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>101141</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>11002</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Ruben</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Torres</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5844 Linden Land</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Hobart</td>\n",
       "      <td>Tasmania</td>\n",
       "      <td>...</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>Professional</td>\n",
       "      <td>M</td>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>91945</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>11003</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Christy</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Zhu</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1825 Village Pl.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>North Ryde</td>\n",
       "      <td>New South Wales</td>\n",
       "      <td>...</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>Professional</td>\n",
       "      <td>F</td>\n",
       "      <td>S</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>86688</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11004</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Elizabeth</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Johnson</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7553 Harness Circle</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Wollongong</td>\n",
       "      <td>New South Wales</td>\n",
       "      <td>...</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>Professional</td>\n",
       "      <td>F</td>\n",
       "      <td>S</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>92771</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 24 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   CustomerID Title  FirstName MiddleName LastName Suffix  \\\n",
       "0       11000   NaN        Jon          V     Yang    NaN   \n",
       "1       11001   NaN     Eugene          L    Huang    NaN   \n",
       "2       11002   NaN      Ruben        NaN   Torres    NaN   \n",
       "3       11003   NaN    Christy        NaN      Zhu    NaN   \n",
       "4       11004   NaN  Elizabeth        NaN  Johnson    NaN   \n",
       "\n",
       "          AddressLine1 AddressLine2         City StateProvinceName    ...      \\\n",
       "0      3761 N. 14th St          NaN  Rockhampton        Queensland    ...       \n",
       "1           2243 W St.          NaN      Seaford          Victoria    ...       \n",
       "2     5844 Linden Land          NaN       Hobart          Tasmania    ...       \n",
       "3     1825 Village Pl.          NaN   North Ryde   New South Wales    ...       \n",
       "4  7553 Harness Circle          NaN   Wollongong   New South Wales    ...       \n",
       "\n",
       "    Education    Occupation Gender MaritalStatus HomeOwnerFlag  \\\n",
       "0  Bachelors   Professional      M             M             1   \n",
       "1  Bachelors   Professional      M             S             0   \n",
       "2  Bachelors   Professional      M             M             1   \n",
       "3  Bachelors   Professional      F             S             0   \n",
       "4  Bachelors   Professional      F             S             1   \n",
       "\n",
       "  NumberCarsOwned NumberChildrenAtHome TotalChildren  YearlyIncome  BikeBuyer  \n",
       "0               0                    0             2        137947          0  \n",
       "1               1                    3             3        101141          1  \n",
       "2               1                    3             3         91945          0  \n",
       "3               1                    0             0         86688          0  \n",
       "4               4                    5             5         92771          1  \n",
       "\n",
       "[5 rows x 24 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "combined = customer_info.merge(customer_has_bike, on='CustomerID', how='left')\n",
    "combined.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Box plot is good to visualize classification numeric features. We will plot this and see if there is any significant differences in quartiles between bike buyers and none bike buyers to see if the numeric feature would be useful."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEFCAYAAAAmIwo/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBpfq2AP3agAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1187d4320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEFCAYAAADqujDUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRCrFAiAiUikWABGRSv0vZvKVeQ/ow4kAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11876c128>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEFCAYAAAAYKqc0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGD9JREFUeJzt3XtwjXfix/HPk4tIhJ6uVRajLh3abpURpnYz6rKy1EyE\n0kGJWTG2bK3F1q0bZEkRinFJa106XRFVLdPG7iyrZEdls6HpUNGEtrvrFtK6bNIjaSQ55/eHn1NR\nOR4n53GOx/v11zlPzvM8nxwnn3x981wMt9vtFgDAVkICHQAA4H+UOwDYEOUOADZEuQOADVHuAGBD\nYYEOcEN+fn6gIwDAfSkmJuYHy4Km3KXbBwQA1K2ugTHTMgBgQ5Q7ANgQ5Q4ANkS5A4ANUe42FB8f\nr/j4+EDHABBAlDsA2JClh0IOHTpU0dHRkqTWrVtr8eLFVu4OUq0Re3x8vHbt2hXANAACxbJyr6ys\nlNvtVkZGhlW7AGDSW2+9pZycnEDHkNPplCTPoC9QYmNjlZSUFNAMVrOs3IuKilRRUaGkpCRVV1dr\n+vTp6tq1q9d1OEvV/3hPIUklJSWqrKwMdAxVVFRIksLDwwOao6SkxPY/G5aVe8OGDTV+/Hi98MIL\n+u9//6sJEyZo9+7dCgure5ecoep/vKeQgudzMH78eEnSpk2bApzEPu75Gart2rXT4MGDZRiG2rVr\nJ4fDoW+++caq3eH/3TzHznw78OCyrNzff/99LVmyRNL1/wI5nU41a9bMqt0BAG5i2bTM8OHDNWfO\nHI0aNUqGYWjRokVep2TgP4zYAVjWtg0aNNDy5cut2jwAwAtOYgIAG6LcAcCGKHcAsCHKHQBsiHIH\nABui3AHAhih3ALAhyh0AbIhyBwAbotwBwIYodwCwIcodAGyIcgcAG6LcAcCGKHcAsCHKHQBsiHIH\nABui3AHAhih3ALAhyh0AbIhyBwAbotwBwIYodwCwIcodAGyIcgcAG6LcAcCGKHcAsCHKHQBsiHIH\nABui3AHAhih3G4qPj1d8fHygYwAIIEvL/dKlS+rdu7e++uorK3cDALiFZeVeVVWlefPmqWHDhlbt\nArdx84id0Tvw4AqzasNpaWkaOXKk1q9fb9UugPvCzJkzdenSpUDHCAoXL16UJI0fPz7ASYJD06ZN\ntXTpUku2bUm579y5Uz/60Y/Uq1evuyr3/Px8K+I80HhPA6+4uFilZaUKibRsLHXfcIW4JUkXnZcD\nnCTwXBXVqqystOxn9K4+baWlpXrooYfu+LodO3bIMAzl5uaqsLBQs2bN0ptvvqlmzZp5XS8mJuZu\n4sAE3tPAi4iIUEhkmB4e2CbQURBEruw+rYiIiHr/jNb1y8HUnHthYaEGDhyohIQElZSUKC4uTseP\nH6/z9ZmZmdqyZYsyMjL0xBNPKC0t7Y7FDv/YtWvXbR8DeLCYKvfU1FSlp6fL4XCoefPmSklJ0fz5\n863OBgDwkalyr6ioUIcOHTzPY2Njde3aNVM7yMjIqLUurLdr1y5G7cADzlS5OxwOFRUVyTAMSVJW\nVpapuXcAQGCY+oNqSkqKZs2apS+++ELdu3fXo48+qmXLllmdDQDgI1Pl3qZNG73zzjsqLy+Xy+VS\ndHS01bkAAPVgqtw/+eQT/fnPf1ZpaWmt5Zs3b7YkFACgfkyV++zZszV58mS1bNnS6jwAAD8wVe7N\nmzfXkCFDrM4CAPATU+WemJioV155RT179lRY2PerUPgAEJxMlfvWrVsl/fA0V8odAIKTqXL/5ptv\n9Le//c3qLAAAPzF1ElP37t2VnZ2t6upqq/MAAPzA1Mg9Oztb7733Xq1lhmGosLDQklAAgPoxVe4H\nDx60OgcAwI9MlXtFRYXWrl2r3Nxc1dTUqGfPnvrd736nqKgoq/MBAHxgas59wYIFqqio0KJFi5SW\nlqaqqiou+QsAQczUyP348ePKysryPJ83b54GDRpkWSgAQP2YGrm73W6VlZV5npeVlSk0NNSyUACA\n+jE1cv/Vr36l4cOHq1+/fnK73crOztavf/1rq7MBAHxkqtyHDRumzp076/Dhw3K5XFqzZo06depk\ndTYAgI+8lvvhw4drPe/YsaOk69Myhw8fVo8ePaxLBgDwmddyX716tefx8ePH9dRTT8ntdku6fhIT\n13MHgODktdwzMjI8j4cMGUKZA8B9wtTRMpI8N8cGAAQ/0+UOALh/eJ2WKS4u9jyuqqrS+fPnPXPu\nkrjtHgAEKa/lPmbMGBmG4Sn00aNHe75mGIb27dtnbToAgE+8lvv+/fvr/Nq1a9f8HgYA4B+m5txH\njBhR67nL5dKwYcMsCQQAqD+vI/exY8fq0KFDkqQnnnhC0vXrzISFhalfv37WpwMA+MRrud84rj01\nNVXJycm1vnbzH1YBAMHF1LTMzcVeUlKitWvXqm/fvpaFAgDUj6kLh0nSgQMHtG3bNh04cEDdunXj\nZh1BLD4+XpK0a9euACcBEChey/3SpUt67733tH37doWHh2vgwIE6fvw4lyEAgCDndVqmd+/eKioq\n0tq1a7Vnzx5NmzZNYWHmBvs1NTWaM2eORo4cqVGjRunkyZN+CQzvbozab30M4MHitdxnz56t06dP\n67e//a2WL1+uoqIi0xvOzs6WJG3btk1Tp07VypUr65cUAGDaHc9QHTNmjE6cOKGdO3cqKSlJZWVl\n2rRpk4YNGyaHw1Hnuv3791efPn0kXb+MQZMmTe4YJj8//+7S4454TwOvsrIy0BEQpCorKy37GTU1\nx9KpUyfNmTNHM2bMUHZ2tnbs2KH09HR9+umn3jceFqZZs2Zp7969ta4NX5eYmBhzqWEa72ngRURE\n6Nuqq4GOgSAUERFR75/Run453NVVIcPCwhQXF6d169Zp7969ptZJS0vTnj17NHfuXJWXl9/N7uCD\nm4+Q4WgZ4MFlauS+c+dOpaWlqaysTNL1E5gMw1BhYWGd63zwwQcqKSnRSy+9pMjISBmGoZAQrjAM\nAPeCqXJPT09XRkaG5x6qZvzyl7/UnDlzNHr0aFVXV+vVV19Vw4YNfQ4K8xixAzBV7s2bN7+rYpek\nqKgorVq1yqdQAID6MVXuP/3pTzVlyhTFxsYqIiLCs3zIkCGWBQMA+M5UuTudTjVq1EhHjhyptZxy\nB4DgZKrcFy9eLEkqLS3VQw89ZGkgAED9mTp8paioSAMHDlRCQoJKSkoUFxen48ePW50NAOAjU+W+\ncOFCpaeny+FwqHnz5kpJSeGqkAAQxEyVe0VFhTp06OB5Hhsbyz1UASCImSp3h8OhoqIiGYYhScrK\nymLuHQCCmKk/qKakpGjWrFn64osv1L17dz366KN6/fXXrc4GAPCRqXLPycnRO++8o/LycrlcLkVH\nR1udCwBQD6amZTIzMyVdP+uUYgeA4Gdq5N6iRQuNHTtWXbp0qXWG6uTJky0LBgDwnaly79q1q9U5\nAAB+ZKrcGaEDwP3Fa7k//vjjnsMfpes36wgJCdG1a9cUHR2tw4cPWx4QAHD3vJb7jRtiz58/X926\nddPgwYNlGIb27Nmjjz/++J4EBADcPVNHy3z22WdKSEjwjOIHDBigY8eOWRoMAOA7U+UeGRmpHTt2\nqLy8XE6nU5mZmXI4HFZnAwD4yFS5L1u2THv37lVsbKyeffZZ/etf/9LSpUutzgYA8JGpo2VatWql\ndevWWZ0FAOAnXsu9X79+tY6WudW+ffv8HggAUH9eyz0jI+Ne5QAA+JHXcj958qT69u2rDz744LZf\nb9WqlSWhAAD147Xcjx07pr59+yovL++2X+cG2QAQnLyW+5QpUyR9f4NsAMD9wdTRMv/4xz+Unp6u\nK1euyO12e5bzB1UACE6myv21117TH/7wBz322GNej54BAAQHU+XeuHFj9enTx+IoAAB/8VruN676\n2KFDB6WmpuoXv/iFwsK+X6VHjx7WpgMA+MRrua9evdrz+MKFCzpx4oTnuWEY2rx5s3XJAAA+4yQm\nALChO144bPXq1crNzfU8nzVrVq0RPQAg+Hgt91WrVqmwsFDt27f3LJs0aZI+//xzrV27ts71qqqq\nNGPGDL344osaPnw4h0wCwD3mtdw/+ugjrVq1Ss2bN/csa9u2rZYvX67du3fXuV5WVpYcDoe2bt2q\njRs3auHChf5LDAC4I69z7qGhoWrQoMEPljdq1KjWUTO3GjhwoAYMGCBJcrvdCg0NrWfM+8Nbb72l\nnJycQMeQ0+mUJEVHRwc0R2xsrJKSkgKaIRg4nU65Kqp1ZffpQEdBEHFVVMspp2Xb91rukZGROn36\ntNq0aVNr+alTpxQSUvegv1GjRpKuf6inTJmiqVOnmgqTn59v6nXBqqSkRJWVlYGOoYqKCklSeHh4\nQHOUlJTc9/+m/lBdXR3oCAhS1dXVlv2MeC33l156SUlJSZo8ebKefvppud1uFRQUKD09/Y6Fff78\neb388st68cUXFR8fbypMTEyM+eRBKFjyjx8/XpK0adOmACeBJDkcDl10XtbDA9vc+cV4YFzZfVqO\naEe9e6OuXw5ey71Pnz4KCQnRn/70J/3xj39USEiIOnfurLlz56pXr151rnfx4kUlJSVp3rx5+tnP\nflav4ACAu3fHyw88++yzOnfunDIzM01vdN26dSorK9Mbb7yhN954Q5K0YcMGNWzY0PekAADTTF1b\nJjMzU6NGjTK90eTkZCUnJ/scCgBQP6bKvUWLFho7dqy6dOmiiIgIz/LJkydbFgwA4DtT5d61a1er\ncwAA/MhUuU+ePFnl5eU6ffq0OnbsqO+++05RUVFWZwMA+OiO15aRpNzcXCUkJOg3v/mNLl68qH79\n+ungwYNWZwMA+MhUua9YsUJbt25VkyZN9Mgjj2jLli1aunSp1dkAAD4yVe4ul0vNmjXzPH/ssccs\nCwQAqD/TR8tkZ2fLMAyVlZUpMzNTLVu2tDobAMBHpkbuCxYs0K5du3T+/Hn1799fhYWFWrBggdXZ\nAAA+MjVyb9q0qVasWCGn06mwsDDONAWAIGeq3E+cOKHZs2eruLhYktS+fXulpaX94GqRAIDgYGpa\nZv78+Zo6dary8vKUl5enpKQkvfrqq1ZnAwD4yFS5V1ZWqnfv3p7ncXFxnhtCAACCj9dyLy4uVnFx\nsR5//HGtX79ely9fVmlpqbZs2aLu3bvfq4wAgLvkdc59zJgxMgxDbrdbeXl52rZtm+drhmFw5UcA\nCFJey33//v33KgcAwI9MHS3z73//W9u3b1dpaWmt5YsXL7YkFACgfkxfFXLQoEHq1KmT1XkAAH5g\nqtybNGnCjTkA4D5iqtyHDh2qlStXqmfPngoL+36VHj16WBYMAOA7U+V+6NAhHTt2TJ9++qlnmWEY\n2rx5s2XBAAC+M1XuBQUF+vvf/251FgCAn5g6Q7Vjx44qKiqyOgsAwE9MjdzPnDmjoUOHqlmzZgoP\nD5fb7ZZhGNq3b5/V+QAAPjBV7unp6VbnAAD4kalyP3z48G2Xt2rVyq9hAAD+Yarc8/LyPI+rqqqU\nn5+v7t27a8iQIZYFAwD4zlS533qZgf/973+aNm2aJYEAAPVn6miZW0VFRencuXP+zgIA8BNTI/fE\nxEQZhiFJcrvdOnPmTK2bdwAAgovXcr9xz9Tnn3/eU+6tW7eWw+FQVFSU9ekAAD4xfbOOGwzD0Ndf\nf63q6moVFhZaHhAAcPfu6mYdV69eVVpamg4ePKiFCxdaGgwA4DvTf1DNzc3V4MGDJUlZWVmKjY29\n4zpHjx5VYmKi7+kAAD654x9Uy8vLtWTJEs9o3UypS9KGDRuUlZWlyMjIeoc0Y+bMmbp06dI92Vew\nu3jxoiRp/PjxAU4SeE2bNtXSpUsDHQO457yWe25urpKTkxUbG6tdu3apUaNGpjfcpk0brVmzRjNn\nzjS9Tn5+vunX3qq4uFilpWUywu/NL5Ng5v7//5B9c8UZ4CSB5a6qUGVlZb0+V/5QWVkZ0P0jeFn5\n+fRa7uPGjVNYWJgOHjyonJwcz3IzFw4bMGCAzp49e1dhYmJi7ur1N4uIiJARHqnoxwb7vA3Yi/PL\nLEVERNTrc+UPERER+rbqakAzIDj54/NZ1y8Hr+XOVR8B4P7ktdy5MBgA3J98uvwAACC4WVrurVu3\n1vbt263cBQDgNhi5A4ANUe4AYEOUOwDYEOUOADZEuQOADVHuAGBDlDsA2BDlDgA2RLkDgA1R7gBg\nQ5Q7ANgQ5Q4ANkS5A4ANUe4AYEOUOwDYEOUOADZEuQOADVHuAGBDlDsA2BDlDgA2RLkDgA1R7gBg\nQ5Q7ANgQ5Q4ANkS5A4ANUe4AYEOUOwDYEOUOADZEuQOADVHuAGBDlDsA2FCYVRt2uVxKSUnRiRMn\n1KBBA6WmpurRRx+1ancAgJtYNnL/6KOPdO3aNb377rv6/e9/ryVLlli1KwDALSwbuefn56tXr16S\npK5du6qgoMCqXUmSnE6n3FXl+rbwXUv34507gPsOVkYA9+2W0xnA3d/EVVGtK7tPB27/12qkGj6f\nHqGGQhqEBjSCq6JairZu+5aVu9PpVHT098lDQ0NVXV2tsLC6d5mfn+/z/kJDQ2UYgSwSyc3Pzg8E\n9p/EUGhoaL0+V/4QERGhh5o8FNAMFdUVqqqpCmiGYBIeEq7I8MgAh7j+2bDq82lZuUdHR+vq1aue\n5y6Xy2uxS1JMTIzP+9u6davP6wJWqs/nGriTun45WDbn3q1bNx04cECSdOTIEXXs2NGqXQEAbmHZ\nyD0uLk45OTkaOXKk3G63Fi1aZNWuAAC3sKzcQ0JCtGDBAqs2DwDwgpOYAMCGKHcAsCHKHQBsiHIH\nABui3AHAhiw7WsYXgT6TEADswnC7OWkeAOyGaRkAsCHKHQBsiHIHABui3AHAhih3ALAhyh0AbIhy\ntxGXy6V58+ZpxIgRSkxM1KlTpwIdCajl6NGjSkxMDHSMB0JQncSE+rn5puRHjhzRkiVL9OabbwY6\nFiBJ2rBhg7KyshQZGeDb2z0gGLnbyL2+KTlwN9q0aaM1a9YEOsYDg3K3kbpuSg4EgwEDBtzxPsrw\nH8rdRny5KTkAe6LcbYSbkgO4gWGdjXBTcgA3cFVIALAhpmUAwIYodwCwIcodAGyIcgcAG6LcAcCG\nOBQS9728vDxNnDhRbdq0kdvtVlVVlQYPHqxJkyZpwoQJSk1NVU5Ojg4dOqQlS5aY2mZiYqIuXLig\nqKgouVwuNW7cWIsWLVLbtm2t/WYAP6HcYQtPPfWUMjIyJElXr17VoEGDFBcXpw0bNvi8zdTUVD3z\nzDOSpLffflsrV67UqlWr/JIXsBrTMrCd7777TqGhoWrcuLH69euns2fP1vr6okWLNH36dNXU1Oiz\nzz7TqFGjNHToUCUlJenMmTO33abT6dSPf/xjSdLOnTs1e/Zsz9cSExOVl5enGTNm6N133621/OjR\nozp16pTGjRunoUOHatSoUfr8888lSbNnz9bEiRP13HPPaf/+/f5+G/CAY+QOWygoKFBCQoJcLpdO\nnz6t5557To888sgPXrdmzRpduHBBK1asUE1NjZKTk7Vu3Tq1bNlSH3/8sebOnau3335bkpScnKyo\nqCh9++23Ki0t9fzPoC7Dhg3TmjVrNGLECJ07d06XL19Wly5dNHLkSM2bN09PPvmkvvzyS7388sva\ns2ePJMnhcGjdunV+fz8Ayh22cOu0zMSJE7V+/fparzlw4IAuX76s999/X2FhYTp58qTOnDmjSZMm\neV7jdDo9j2+elsnOzta4ceO0b9++OjM888wzmjt3rs6ePasPP/xQCQkJunr1qgoKCjRnzhzP68rL\ny3XlyhVJ0tNPP13/bx64DcodttOoUSP1799f//znP2stb9WqlaZNm6YFCxZo27Ztcrlcat26tT78\n8ENJUk1NjS5evHjbbfbt21cul0v/+c9/ZBiGbr5qR1VVlSTJMAwNGTJEf/3rX7V7925t3LhRLpdL\nDRo08OxDki5cuCCHwyFJatiwoV+/d+AG5txhOzU1NTp06JCefPLJWss7dOigF154QZGRkcrMzFT7\n9u1VWlqqTz75RJK0Y8cOvfLKK7fdZkFBgaqrq9WuXTs9/PDD+uqrr+R2u3XmzBmdOHHC87rnn39e\n27ZtU4sWLdS8eXM1btxYbdu29ZR7Tk6ORo8ebdF3DnyPkTts4cacuyRVVFSoc+fOmjBhQq0R8w0p\nKSkaNWqU4uLitGrVKr322muqrKxUdHS00tLSPK+7Mefudrvlcrn0+uuvKzo6Wj//+c+1Y8cODRw4\nUO3atVNMTIxnnZ/85Cdq0aKFhg4d6lm2bNkypaSkaOPGjQoPD9fKlStlGIaF7wbAVSEBv3G73fr6\n66+VmJiov/zlL2rQoEGgI+EBxrQM4Cd79uxRQkKCpk+fTrEj4Bi5A4ANMXIHABui3AHAhih3ALAh\nyh0AbIhyBwAb+j9FQfsq3EJmEAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x118490c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAEFCAYAAAAYKqc0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAFFRJREFUeJzt3W9wVNUdxvFn829JCBAiAqk0mEJpB22l5EWsjtaqAWSs\nGItIzKQzJpNOLGgVFAmFTIiACaAM0hYkwNACU3SCllRmYNDa0kYHpjtDbGyLkqlCoEQCEVwIm2z2\n9gXN1mg2GzZ7d5fD9/Mq3L33nF+Wk2dPTu4fh2VZlgAARomLdgEAgPAj3AHAQIQ7ABiIcAcAAxHu\nAGCghGgX0M3lckW7BAC4KmVnZ39lW8yEu9R7gQCAwAJNjFmWAQADEe4AYCDCHQAMRLgDgIEIdwAw\nEOEOAAay9VTIvLw8paamSpLGjBmjF154wc7uAAD/Y1u4ezweWZalbdu22dVFzNmyZYvq6+ujXYbc\nbrck+T9Yo+X2229XUVFRVGvAZYzNnq6FsWlbuP/rX/9Se3u7ioqK5PV6NW/ePE2aNKnPY672q1Rb\nWlrk8XiiXYba29slSYmJiVGto6Wl5ar/PzUFY7Ona2FsOux6WMeRI0fU0NCghx9+WB9//LFKSkq0\nd+9eJST0/nnicrm4QjVMiouLJUmbN2+OciVAT4zN8AuUnbbN3LOysjR27Fg5HA5lZWUpLS1Np0+f\nVkZGhl1dAgD+x7azZWpra1VVVSXp8q9Abrdb119/vV3dAQC+wLaZ+8yZM1VWVqb8/Hw5HA6tWLEi\n4JIMACC8bEvbpKQkvfjii3Y1DwDoAxcxAYCBCHcAMBDhDgAGItwBwECEOwAYiHAHAAMR7gBgIMId\nAAxEuAOAgQh3ADAQ4Q4ABiLcAcBAhDsAGIhwBwADEe4AYCDCHQAMRLgDgIEIdwAwEOEOAAYi3AHA\nQIQ7ABiIcAcAAxHuAGAgwh0ADES4A4CBCHcAMBDhDgAGItwBwECEOwAYiHAHAAMR7gBgIFvD/cyZ\nM/rBD36gpqYmO7sBAHyJbeHe2dmp8vJyDRo0yK4uAAABJNjVcHV1tWbPnq2NGzfa1QVwVViwYIHO\nnDkT7TJiQmtrqySpuLg4ypXEhuuuu04rV660pW1bwv31119Xenq67rjjjisKd5fLZUc51xyPxyOJ\n9zNWnDx5UufOn1Ncsm1zqauGL86SJLW6z0a5kujztXvl8Xhs+zm1ZbTt2rVLDodD7733nv75z3/q\nueee0/r163X99df3eVx2drYd5VxznE6nJN7PWOF0OhWXnKDh0zKjXQpiSNveY3I6nQP+OQ304WBL\nuO/YscP/dWFhoSoqKoIGOwAgfDgVEgAMZPsi4LZt2+zuAgDwJczcAcBAhDsAGIhwBwADEe4AYCDC\nHQAMRLgDgIEIdwAwEOEOAAYi3AHAQIQ7ABiIcAcAAxHuAGAgwh0ADES4A4CBCHcAMBDhDgAGItwB\nwECEOwAYiHAHAAMR7gBgIMIdAAxEuAOAgQh3ADAQ4Q4ABkroz04XL16U2+2WZVn+baNGjbKtKADA\nwAQN940bN2r9+vUaOnSof5vD4dCf/vQnO+sCAAxA0HB/7bXXtH//fo0YMSIS9QAAwiDomntGRobS\n09MjUQsAIEyCztyzsrJUUFCgW2+9VU6n07+9tLTU1sIAAKELGu7p6enKycmRZVm6dOlSJGoCAAxQ\n0HB/6qmn5PF4dPz4cY0bN04ej0eDBg2KRG0AgBAFXXM/dOiQ7r//fv30pz/V6dOnddddd+m9996L\nRG0AgBAFDffVq1dr+/btGjp0qEaOHKnf/OY3qqqqCtpwV1eXysrKNHv2bOXn5+vDDz8MS8EAgOCC\nLst0dXX1uGDpW9/6Vo+LmQJ55513JEk7d+7UwYMHtWbNGq1fv34ApfZtwYIFOnPmjG3tX01aW1sl\nScXFxVGuJPquu+46rVy5Mqo1uN1u+dq9att7LKp1ILb42r1yy21b+0HDfdSoUTpw4IAcDocuXryo\n7du3KyMjI2jD9957r+666y5J0smTJ3tcBBWIy+UKXnEAJ0+e1Llz5+VITA65DVNY//uF7HSbfQPn\namB1tsvj8QxoXIWD1+uNav+IXV6v17bxGTTcKysrVVlZqebmZt15553KyclRZWVl/xpPSNBzzz2n\n/fv36+WXXw66f3Z2dr/a7Y3T6ZQjMVmp4x8IuQ2YxX20Tk6nc0DjKhzS0tLU6j6r4dMyo1oHYkvb\n3mNKS00b8PgM9OEQNNx/97vf9SuYA6murtYzzzyjWbNmac+ePUpJSQm5LQBA/wT9g+r+/ftDavj3\nv/+9XnnlFUlScnKyHA6H4uK4CSUARELQmfvw4cM1ffp03XTTTT3Ob3/++ef7PG7KlCkqKytTQUGB\nvF6vFi1axPnxABAhQcP9/vvvD6nhlJQUrV27NqRjAQADEzDcW1paJEl33nlnxIoBAIRHwHCfNWuW\nHA6HOjo61NbWpoyMDMXHx+vEiRP6+te/rn379kWyTgDAFQgY7n/+858lSfPnz9esWbOUk5MjSTp8\n+LC2bt0akeIAAKEJevrKRx995A92SZo0aZKamppsLQoAMDBB/6A6cuRI/fKXv9T06dNlWZbq6uqU\nmcnFGAAQy/p147DTp09rzpw5euKJJ3ThwgVVV1dHojYAQIiCztzT0tK0dOnSSNQCAAiTgOF+0003\nyeFwyLIsORwO//bufzc2NkakQADAlQsY7u+//34k6wAAhFHAcH/jjTc0c+ZMbdiwodfXeUA2AMSu\ngOHu8XgkiYdiA8BVKGC4FxQUSLr8gGwAwNUl6Nkyu3bt0po1a/TZZ59J4g+qAHA1CBruv/rVr7Rp\n0yaNHz++x1kzAIDYFTTc09PT9e1vfzsStQAAwiRguP/hD3+QJI0ZM0ZPPPGE7rnnHsXHx/tf/9GP\nfmR/dQCAkAQM9wMHDkiSEhMTlZiYqPr6ev9rDoeDcAeAGBYw3FetWvWVbV1dXT1m7wCA2BTwxmEd\nHR1atGiR3nrrLf+2uXPnatGiRero6IhIcQCA0AQM95UrVyo+Pl633nqrf1v33SBXr15tf2UAgJAF\nXJY5ePCgdu/erbi4/+f/0KFDVVFRoYceeigixQEAQhNw5h4fH98j2LslJSUpISHoGZQAgCgKGO7D\nhg3TBx988JXtH3zwgZxOp61FAQAGJuAU/Oc//7lKS0tVUFCgW265RZZl6e9//7u2b9+uqqqqSNYI\nALhCAcN98uTJeuWVV7R582b/2vvNN9+smpoarlgFgBjX5+L5xIkT9eKLL0aqFgBAmAQM9ylTpvR6\no7Duu0Lu27fP1sIAAKELGO6bNm2KZB0AgDAKGO6ZmZmSLl+p+te//lUXL16UZVnq6upSc3Oz5s6d\nG7EiAQBXJugJ608++aTOnz+v5uZmfe9735PL5dLkyZMjURsAIEQBz3PvdvToUe3YsUNTpkxRaWmp\namtr9emnn0aiNgBAiIKG+4gRI+RwOJSVlaUjR45o9OjRQW8c1tnZqWeffVaPPvqoZs6cqbfffjts\nBQMAggu6LDNu3DgtX75cs2bN0oIFC3TmzBl1dnb2eUxdXZ3S0tK0atUqffbZZ3rwwQd1zz33hK1o\nAEDfgob70qVL5XK59M1vflM/+9nP9O677/Z6r/cvmjZtmqZOnSrp8qmT3AMeACIraLhXV1frF7/4\nhSQpNzdXubm5Kisr0wsvvBDwmMGDB0uS3G63nnzyST311FP9KsblcvVrv954PJ6Qj4W5PB7PgMZV\nuGoAemPn+AwY7kuWLNGJEyfU0NCgpqYm/3av16u2tragDf/nP//RnDlz9Oijj/b7kXzZ2dn92q83\nTqdTutj3chGuPU6nc0DjKlw1fN55Iao1IDaFY3wG+nAIGO4lJSVqbm7W8uXLVVJS4t8eHx+v8ePH\n99lZa2urioqKVF5eru9///shlgwACFXAs2UyMzN12223ac+ePRo5cqQ+/vhjNTU1adiwYUpPT++z\n0Q0bNuj8+fP69a9/rcLCQhUWFurSpUthLx4A0Luga+5vvvmmXnrpJd19992yLEtbtmzR3Llz+3wa\n0+LFi7V48eKwFgoA6L+g4V5TU6Pa2lr/bH3OnDn6yU9+wqP2ACCGBb2Iyefz9ViGSU9P7/VukQCA\n2BF05j5hwgRVV1dr5syZkqTa2lpNmDDB9sIAAKELOHN/4403JEnPP/+8LMvS/Pnz9fTTT8vn82np\n0qURKxAAcOUCztx/+9vfKi8vTykpKVq4cGEkawIADFDQNXcAwNUn4Mz9o48+6vVmX92P2eNOjwAQ\nuwKG+9ixY7Vx48ZI1gIACJOA4Z6YmKgbbrghkrUAAMIk4Jo7j9IDgKtXwHAvLy+PZB0AgDDibBkA\nMBDhDgAGItwBwECEOwAYiHAHAAMR7gBgIMIdAAxEuAOAgQh3ADAQ4Q4ABiLcAcBAhDsAGIhwBwAD\nEe4AYCDCHQAMRLgDgIEIdwAwEOEOAAYi3AHAQIQ7ABiIcAcAA9ka7g0NDSosLLSzCwBALxLsarim\npkZ1dXVKTk62q4se3G63rM52uY/WRaQ/xD6rs11ud7SruMzX7lXb3mPRLiPqfB1dkqS4pPgoVxJ9\nvnavlGpf+7aFe2ZmptatW6cFCxb0+xiXyxVyf16vN+RjYS6v1zugcRUOTqdTw4YOi2oNseJ8+3lJ\n0pDEwVGuJAYkXh4bdo1P28J96tSpam5uvqJjsrOzQ+4vLS1Np9vcSh3/QMhtwCzuo3VKS0sd0LgK\nh2j3H0uKi4slSZs3b45yJeYI9OHAH1QBwECEOwAYiHAHAAPZGu5jxozRa6+9ZmcXAIBeMHMHAAMR\n7gBgIMIdAAxEuAOAgQh3ADAQ4Q4ABiLcAcBAhDsAGIhwBwADEe4AYCDCHQAMRLgDgIEIdwAwEOEO\nAAYi3AHAQIQ7ABiIcAcAAxHuAGAgwh0ADES4A4CBCHcAMBDhDgAGItwBwECEOwAYiHAHAAMR7gBg\nIMIdAAxEuAOAgQh3ADAQ4Q4ABiLcAcBACXY17PP5VFFRoSNHjigpKUnLli3T2LFj7eoOAPAFts3c\n33rrLXV0dOjVV1/V/PnzVVVVZVdXAIAvsW3m7nK5dMcdd0iSJk2apMbGRru68rM62+U+Wmd7PwH7\n7+qQfF1R6z/mxMXLEZ8Ute6tznZJqVHrP5Zs2bJF9fX10S5Dra2tkqTi4uKo1nH77berqKgoqjXY\nzbZwd7vdSk39/w9WfHy8vF6vEhICd+lyuULuz+l0atiwoSEfHw7t7V51dhLu3RIT4pScnBjNCuR0\nOgc0rkzR0tIij8cT7TKUmHh5PES7lpaWFuPHhW3hnpqaqgsXLvj/7fP5+gx2ScrOzg65v4EcC5iO\nnw9zBfqQsm3NffLkyTpw4IAk6fDhw5owYYJdXQEAvsS2mXtubq7q6+s1e/ZsWZalFStW2NUVAOBL\nbAv3uLg4VVZW2tU8AKAPXMQEAAYi3AHAQIQ7ABiIcAcAAxHuAGAg286WCYXpV4wBQKQ4LMuyol0E\nACC8WJYBAAMR7gBgIMIdAAxEuAOAgQh3ADAQ4Q4ABiLcDeLz+VReXq5HHnlEhYWF+uSTT6JdEtBD\nQ0ODCgsLo13GNSGmLmLCwHzxoeSHDx9WVVWV1q9fH+2yAElSTU2N6urqlJycHO1SrgnM3A0SjYeS\nA/2VmZmpdevWRbuMawbhbpBADyUHYsHUqVODPkcZ4UO4GySUh5IDMBPhbhAeSg6gG9M6g/BQcgDd\nuCskABiIZRkAMBDhDgAGItwBwECEOwAYiHAHAANxKiSuegcPHlRpaakyMzNlWZY6Ozv1wAMP6PHH\nH1dJSYmWLVum+vp6HTp0SFVVVf1qs7CwUKdOnVJKSop8Pp+GDBmiFStW6MYbb7T3mwHChHCHEW6+\n+WZt27ZNknThwgVNnz5dubm5qqmpCbnNZcuWKScnR5K0detWrVmzRmvXrg1LvYDdWJaBcS5duqT4\n+HgNGTJEd999t5qbm3u8vmLFCs2bN09dXV16//33lZ+fr7y8PBUVFen48eO9tul2uzVixAhJ0uuv\nv66FCxf6XyssLNTBgwf17LPP6tVXX+2xvaGhQZ988okee+wx5eXlKT8/X//4xz8kSQsXLlRpaanu\nu+8+/fGPfwz324BrHDN3GKGxsVEzZsyQz+fTsWPHdN9992nkyJFf2W/dunU6deqUXnrpJXV1dWnx\n4sXasGGDvva1r+kvf/mLlixZoq1bt0qSFi9erJSUFH3++ec6d+6c/zeDQH784x9r3bp1euSRR3Ti\nxAmdPXtWt9xyi2bPnq3y8nJNnDhRR48e1Zw5c7Rv3z5JUlpamjZs2BD29wMg3GGELy/LlJaWauPG\njT32OXDggM6ePava2lolJCToww8/1PHjx/X444/793G73f6vv7gs88477+ixxx7T22+/HbCGnJwc\nLVmyRM3Nzdq9e7dmzJihCxcuqLGxUWVlZf79Ll68qLa2NknSd7/73YF/80AvCHcYZ/Dgwbr33nv1\n7rvv9th+ww036Omnn1ZlZaV27twpn8+nMWPGaPfu3ZKkrq4utba29trmD3/4Q/l8Pv373/+Ww+HQ\nF+/a0dnZKUlyOBx68MEHtWfPHu3du1ebNm2Sz+dTUlKSvw9JOnXqlNLS0iRJgwYNCuv3DnRjzR3G\n6erq0qFDhzRx4sQe28eNG6eHH35YycnJ2rFjh77xjW/o3Llz+tvf/iZJ2rVrl5555ple22xsbJTX\n61VWVpaGDx+upqYmWZal48eP68iRI/79HnroIe3cuVOjR4/WqFGjNGTIEN14443+cK+vr1dBQYFN\n3znwf8zcYYTuNXdJam9v13e+8x2VlJT0mDF3q6ioUH5+vnJzc7V27VotX75cHo9Hqampqq6u9u/X\nveZuWZZ8Pp9Wr16t1NRU3Xbbbdq1a5emTZumrKwsZWdn+4/JyMjQ6NGjlZeX59+2atUqVVRUaNOm\nTUpMTNSaNWvkcDhsfDcA7goJhI1lWfr0009VWFioN998U0lJSdEuCdcwlmWAMNm3b59mzJihefPm\nEeyIOmbuAGAgZu4AYCDCHQAMRLgDgIEIdwAwEOEOAAb6L/4jroTcb4FNAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1157d1eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# define function for plotting box plot\n",
    "def plot_box(combined, cols, col_x = 'BikeBuyer'):\n",
    "    for col in cols:\n",
    "        sns.set_style(\"whitegrid\")\n",
    "        sns.boxplot(col_x, col, data=combined)\n",
    "        plt.xlabel(col_x) # Set text for the x axis\n",
    "        plt.ylabel(col)# Set text for y axis\n",
    "        plt.show()\n",
    "\n",
    "cols = ['YearlyIncome','NumberCarsOwned', 'NumberChildrenAtHome', 'TotalChildren']       \n",
    "plot_box(combined, cols)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looks like they all have good difference in quartiles which signal usefulness."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Will now check categorical features to see if there are any great pointers to the BikeBuyer category, will be using barplots to see if there are differences in frequency for BikeBuyers and non BikeBuyers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5455, 24)\n",
      "(10949, 24)\n"
     ]
    }
   ],
   "source": [
    "is_bike_buyer= combined.BikeBuyer==1\n",
    "bike_buyers = combined[is_bike_buyer]\n",
    "print(bike_buyers.shape)\n",
    "is_non_bike_buyer= combined.BikeBuyer==0\n",
    "non_bike_buyers = combined[is_non_bike_buyer]\n",
    "print(non_bike_buyers.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# bar plot function for non numeric features\n",
    "# plot bar plot for bike buyer counts\n",
    "\n",
    "def plt_bar_plot(cat_cols):\n",
    "    combined['dummy'] = np.ones(shape = combined.shape[0])\n",
    "    for col in cat_cols:\n",
    "        print(col)\n",
    "        counts = combined[['dummy', 'BikeBuyer', col]].groupby(['BikeBuyer', col], as_index = False).count()\n",
    "        temp = counts[counts['BikeBuyer'] == 0][[col, 'dummy']]\n",
    "        temp = counts[counts['BikeBuyer'] == 0][[col, 'dummy']]\n",
    "        temp.plot.bar(x=col,y ='dummy')\n",
    "\n",
    "        plt.title('Counts for ' + col + '\\n non bike buyer')\n",
    "        plt.ylabel('count')\n",
    "        temp = counts[counts['BikeBuyer'] == 1][[col, 'dummy']]\n",
    "        temp.plot.bar(x=col,y ='dummy')\n",
    "\n",
    "        plt.title('Counts for ' + col + '\\n bike buyer')\n",
    "        plt.ylabel('count')\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Occupation\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAFXCAYAAAC4K9ncAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/wMbAOrXe36IvQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11e5741d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAFXCAYAAAC4K9ncAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nICIiUSwIIiISxYIgIiJRLAgiIhL1/wF9ErRYEgHq1QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11e269c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Gender\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEeCAYAAACQfIJ4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11db9ef28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEeCAYAAACQfIJ4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11ca3d630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MaritalStatus\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEeCAYAAACQfIJ4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3X1cVGX+//HXcKvCEGLZ1sNsQ0VziwRZ/bomiamEgQIp\nMhTZelOa6eImonmDpaVkUJpZ3uW2KCKpq5l2o+iKeUM25ZoUfZXshjI03ZRBBZT5/eHX+UUeDYsR\nwffzL+aa65z5nDMH3lznzLnGZLfb7YiIiPyCS10XICIiVycFhIiIGFJAiIiIIQWEiIgYUkCIiIgh\nBYSIiBhSQMgVc/bsWZYsWUJsbCz9+vWjT58+zJo1i4qKCqe83t69e5kyZcplLfPBBx8QFhbGAw88\nwOnTp3/T665evZq2bdsye/bsau12u517772XyMjIy17nsGHDOHDgAACDBw/m2LFjl+yfn59f7XUW\nLlxIv3796Nu3L5GRkaSlpTn2e03302/Zn1K/KSDkipk6dSqffPIJb7zxBmvXrmXlypUcPHiQiRMn\nOuX1Dhw4QElJyWUts379egYMGMCqVato1KjRb37tm2++mXXr1lVr++ijj35z6CxcuJDWrVsDsH37\n9sta9p133mHTpk2sWLGCt956i1WrVvHll18yd+5coOb76bfsT6nfFBByRXz77besW7eO5557DrPZ\nDECTJk14+umn6dWrFwClpaWMHTuWyMhIoqKieP755zlz5gwAbdu2rfZf8/nH+fn5xMfHk5ycTHR0\nNH369GHXrl0cOnSIOXPm8NFHHzFhwgTKysoYPXo0/fr1IyYmhkmTJlFVVVWtxkWLFpGbm8vy5ctJ\nS0ujsrKSadOm0adPH6Kiopg4cSI2mw2AHj16kJSUREREBBs3brxgewMCAmjSpAkff/yxo+1f//oX\nffv2dTz+8ccfefzxxxk4cCA9evQgMTGRo0ePGq6/R48efPrpp0yYMAGAQYMGcejQIbZs2UJ8fDyx\nsbF0796dl1566YJajhw5wtmzZx3h5OnpyeTJk+nZs+cF+6mqqorp06czYMAA+vTpQ0REBFar9YJ+\nvxyh/PxxUVGRo6aYmBiWLVtWo2NErj4KCLkiPvvsM1q3bo23t3e19htuuIHevXsDMH36dHx9fVm3\nbh2rVq3iiy++4PXXX//Vde/du5fBgwezZs0a+vfvz9y5c7npppsYPXo0ISEhzJgxg40bN1JWVuYY\nucC50Pq5oUOH0qNHDx555BFSUlJ49dVXOXz4MGvXrmXt2rVUVVXx/PPPO/q3adOGd955xxFwvxQd\nHc3atWsBOHXqFFarlW7dujmeX79+PR06dGDFihXk5ubSqFEjR/+LrX/GjBkAvPHGG/zhD3/g9ddf\nZ+bMmaxevZoVK1awYMGCC04/xcTE4OPjw913383AgQOZOXMmhw4dIjAw8IL99J///IfDhw+zYsUK\nNmzYQExMDAsXLryg36UsXryYHj16sHr1ahYsWMBHH310QRhL/aCAkCvCxcXlV/9I5OXl8dBDD2Ey\nmfDw8CA+Pp68vLxfXffNN9/M7bffDkD79u05fvz4BX06duzIgQMHSExMZMGCBQwaNIhbb731V+uJ\nj4/H3d0dFxcXEhMT2bZtm+P5kJCQSy4fFRXFxo0bqaysdIwCXF1dHc8PGjSI4OBglixZwtSpU9m/\nfz8nT56s8fpNJhOvvfYaBQUFzJ07l5kzZ2K32zl16lS1fmazmddff5133nmH/v37c/ToUR599FFm\nzZp1wTqDgoJISkoiOzubtLQ03n33XcrKyi5Zxy/16tWLRYsW8cQTT/D+++8zadIkXFz0p6Y+0rsm\nV0RgYCBffvml4xTNeSUlJTz66KOcPn36ggCpqqpynGL6uV9e1P75tQKTyYTR9GK33HILGzdu5NFH\nH8Vms/HXv/6Vd99995I1G9VTWVnpeNykSZNLLn/DDTfQvn17tm7dypo1a4iJian2/KxZs5g9ezZN\nmzZl4MCBdO3atVrtv7b+kydPEhMTQ0FBAe3bt2fcuHG4ubldsP0LFy7k448/5pZbbmHAgAHMmjWL\nhQsXkpWVdcE6//3vf/PYY48BcO+992KxWAxf+5f7+ef7JSwsjPfee4+IiAg+//xzoqKi+Oabby65\nLXJ1UkDIFXHjjTcSFRXFU0895QgJm83G1KlT8fX1pVGjRtx9990sW7YMu91ORUUFOTk5/OUvfwHA\nz8+PTz/9FMDwnL8RV1dXR8BkZWUxYcIE7r77bpKTk7n77rvZv3//JZfv1q0b2dnZVFZWUlVVxbJl\ny+jatetlbXd0dDRLliyhtLSUgICAas998MEHDBo0iOjoaJo1a8aOHTs4e/Zsjbfr66+/xmazkZSU\nRI8ePfjwww+pqKi4INhOnz5Neno6P/30k6Pt4MGDtG/fvtr64NwF8LCwMBISErjzzjvZtGmTo6af\n9/Pz8+P777/n6NGj2O12Nm3a5Fj3k08+yYYNG7j//vtJTU3F29ubQ4cOXdZ+k6uDAkKumNTUVFq3\nbk18fDz9+vVjwIABtG7dmunTpwMwadIkjh07RlRUFFFRUdx2220MHz7c8dwzzzxDTEwMn332GTfc\ncMOvvl5QUBBffvklI0eOJDo6mrNnz9KnTx9iY2Ox2Ww8/PDDl1x+xIgRXH/99URHRxMREcGZM2cu\n+xNXPXv2pLCwkH79+l3w3MiRI3n++eeJjY3liSeeIDg4uEb/affq1YuEhARMJhPdu3cnIiKCmJgY\nNm/eTOvWrfn666+r9X/88cfp0qUL8fHxREREEB4ezq5duxwXtH++n+Lj49m9ezdRUVEMHDiQW265\nheLiYqqqqqr1O/8+PvDAA8TFxVV7Px5//HHWrVtH3759iYuLo2fPnnTq1Omy9ptcHUya7ltERIxo\nBCEiIoYUECIiYkgBISIihhQQIpdh/PjxLF682PC5fv36ceLECVavXu34qOhv8cu7lEXqiltdFyDS\nUPz8LmiRhkAjCGlwevTowcsvv0xCQgJhYWHVpsdYsWIFkZGR9O3bl8GDB3Pw4EHg3Mhg+vTpJCYm\n0qtXLx577LGL3kFstVqJi4ujT58+PPvssxedLwrg3XffpWfPnnz55ZcAvPnmm8TGxhIdHc0jjzxC\nUVGR4WucPHnSMXdUYmJitTp/PoI5/9hqtXLPPfc47oE4deoUXbp04ejRo5SUlDBy5EhiY2OJiori\ntddeA6C4uJh77rmHwYMHEx4ezuHDhy97X0vDpoCQBunkyZNkZWWRnZ3N0qVL+fbbb9m5cyeLFi3i\nn//8J2+99RaRkZGMHDnScUfwvn37WLx4MRs2bODw4cMXvdP6hx9+4B//+Adr1qyhsLCQnJwcw37r\n1q1j7ty5ZGZm4u/vz4cffsiaNWtYtmwZa9asYejQoYwaNcpw2UOHDvHII4+wdu1aIiMjGTdu3CW3\nt2PHjvj6+jqmAlm/fj1dunShWbNmJCcn88ADD7B69WpWrlzJjh072LBhg2NbHn/8cd577z2aN29e\no30r1w4FhDRI9957L3DuDu5mzZpx/Phxtm3bRp8+ffDz8wMgNjaWkpISiouLgXN3Tnt4eODu7k5A\nQIDhnE5w7lpDkyZN8PDwoG/fvuzYseOCPp9++ikpKSnEx8dz0003Aeemsfj6668dNwrOmjWL48eP\nV7vD+by2bdsSHBwMnJtsb9++fZSWll5ymx988EFHWK1YsQKLxcLJkyfZvXs3s2fPpl+/fsTFxXHo\n0CEKCwsBcHNzo0OHDr+6P+XapGsQ0iB5eno6fj4/b5DRPaF2u91xiqgmczoB1Sbcg3N/ZH/JbDaT\nnp5OUlIS3bt3p0WLFlRVVdGvXz+Sk5OBc3M7HT58mOuuu+6C5X85uZ3JZMLNze2ScyBFRUWRkZHB\nrl27OHnyJH/+85+x2WzY7Xays7Np3LgxAMeOHcPT05P//ve/eHh4GNYvAhpByDXk7rvvZsOGDY7r\nBKtWrcLX1/dXZ3X9pfXr11NRUUF5eTmrV68mNDT0gj5//OMf6dKlC4mJiaSkpFBVVUXXrl1Zv369\n41z/8uXLGTRokOFrfPHFF3z++efAudFAx44dady4MU2bNmXfvn3AuT/0H330kWOZxo0b07dvX556\n6ini4+MB8Pb2pkOHDixZsgSAEydOYLFYyM3NvaxtlmuT/nWQa0bXrl155JFHGDRoEFVVVfj5+TF/\n/vzLnoq6RYsWjtM3vXr1umCW1p8bPnw4mzdvZtGiRTz66KMMGzaMwYMHYzKZ8Pb2Zu7cuZhMpguW\n8/f3Z+7cuXz77bc0a9aMmTNnApCYmMjYsWMJDw+nRYsWF8xxFBsbS05ODtHR0Y62F154gWnTphEV\nFUVFRYXjIv35U2siF6O5mEQaCLvdzsKFC/nuu+94+umn67ocaQA0ghBpIO699178/Px49dVX67oU\naSCcNoJYvXo1//rXvwAoLy/n888/Jysri+eeew6TyUSbNm1ITU3FxcWFnJwcsrOzcXNzY8SIEYSF\nhXH69GmSk5M5evQoXl5epKWlOT59IiIizndFTjE9/fTTtGvXji1btvDXv/6Vzp07M2XKFLp160aH\nDh0YPHgwq1atory8nISEBFatWsWyZcuw2WyMGjWK9evX88knnzBp0iRnlyoiIv/H6Z9i+vTTTzlw\n4AADBw6koKDAcVEtNDSUHTt2sHfvXoKCgvDw8MBsNtOyZUsKCwurfcF7aGgoO3fudHapIiLyM06/\nBjF//nxGjhwJnLuIdv4TG15eXpSWlmKz2TCbzY7+Xl5e2Gy2au3n+/4aq9XqhC0QEWn4OnbseEGb\nUwPixIkTHDx4kP/5n/8Bqt/8U1ZWho+PD97e3tXmvCkrK8NsNldrP9+3Jow2Ui6f1WrVvpSrlo7P\n2nWxf66deopp9+7ddOnSxfG4ffv25OfnA5CXl0dISAiBgYFYrVbKy8spLS2lqKiIgIAAgoOD2bp1\nq6OvDgYRkSvLqSOIgwcP0qJFC8fjlJQUJk+eTEZGBv7+/oSHh+Pq6kpiYiIJCQnY7XbGjBmDp6cn\nFouFlJQULBYL7u7upKenO7NUERH5hQZ1o5yGnbVH+1KuZjo+a9fF9qfmYhIREUMKCBERMaSAEBER\nQwoIERExpMn6ROSaEfXk2lpd37r0frW6vquNAqIO1PZB6jRZV//3BTT0X1BpOMrLy4mIiGDz5s11\nXUqN6RSTiIgY0ghCRMRJysrKGDt2LCdOnKBly5bAuW8FnDp1Kq1atWL58uX8+OOPxMTEMGbMGG66\n6SaKi4u5//772b9/P5999hndu3fn73//O4mJibRt25b9+/fTpEkTQkJC+OCDDzhx4gSvv/46zzzz\nDFFRUXTv3p2ioiLS0tJYsGDB76pfIwgRESfJzs4mICCAZcuWOb4n/GK+/fZbnn32WebPn8/s2bMZ\nP348b775JitXrnT0CQwM5I033qCiooJGjRqxZMkSWrduze7duxkwYIDjO3hWrlxJ//79f3f9CggR\nESf56quvuPPOOwG46667cHOrftLm5xNZ3HLLLZjNZnx8fLj++uvx9fXF09Oz2neW/+lPfwLAx8eH\n1q1bO34uLy+nc+fOFBUVcezYMbZv305YWNjvrl8BISLiJK1atWLPnj0AfPbZZ5w5cwYPDw+OHDni\naDvv50HwW5hMJvr27cv06dPp2rUr7u7uv2t9oGsQInINudKferNYLIwbNw6LxYK/vz/u7u48/PDD\nPP3009x88800b968Vl8vNjaW7t27s3Zt7XxSUpP11YF68zHXekAfc7021Zff9SutpKSEcePG8cYb\nb1zWcpqsT0SkAXv//fcZOnQoo0ePrrV16hSTiEgD0Lt3b3r37l2r69QIQkREDCkgRETEkAJCREQM\nKSBERMSQAkJERAwpIERExJACQkREDCkgRETEkFNvlJs/fz6bN2+msrISi8VCp06dGD9+PCaTiTZt\n2pCamoqLiws5OTlkZ2fj5ubGiBEjCAsL4/Tp0yQnJ3P06FG8vLxIS0vDz8/PmeWKiMjPOG0EkZ+f\nzyeffMLy5cvJzMzkhx9+YMaMGSQlJZGVlYXdbic3N5cjR46QmZlJdnY2ixcvJiMjg4qKCpYvX05A\nQABZWVlER0czb948Z5UqIiIGnBYQH3zwAQEBAYwcOZLhw4fTvXt3CgoK6NSpEwChoaHs2LGDvXv3\nEhQUhIeHB2azmZYtW1JYWIjVaqVbt26Ovjt37nRWqSIiYsBpp5j++9//8v333/Paa69RXFzMiBEj\nsNvtjjnPvby8KC0txWazYTabHct5eXlhs9mqtZ/vWxNWq7X2N0auWnq/r116753PaQHh6+uLv78/\nHh4e+Pv74+npyQ8//OB4vqysDB8fH7y9vSkrK6vWbjabq7Wf71sT9WIK4Kziuq6gwagX77fUOk33\nXbsuFrZOO8XUsWNHtm3bht1up6SkhFOnTtGlSxfy8/MByMvLIyQkhMDAQKxWK+Xl5ZSWllJUVERA\nQADBwcFs3brV0VcHg4jIleW0EURYWBi7d++mf//+2O12pkyZQosWLZg8eTIZGRn4+/sTHh6Oq6sr\niYmJJCQkYLfbGTNmDJ6enlgsFlJSUrBYLLi7u5Oenu6sUkVExIC+Ua4O6Bvlao++Ue7aVF9+1+sL\nfaOciIhcFgWEiIgYUkCIiIghBYSIiBhSQIiIiCEFhIiIGFJAiIiIIQWEiIgYUkCIiIghBYSIiBhS\nQIiIiCEFhIiIGFJAiIiIIQWEiIgYUkCIiIghBYSIiBhSQIiIiCEFhIiIGFJAiIiIIQWEiIgYUkCI\niIghBYSIiBhSQIiIiCE3Z648JiYGb29vAFq0aMHw4cMZP348JpOJNm3akJqaiouLCzk5OWRnZ+Pm\n5saIESMICwvj9OnTJCcnc/ToUby8vEhLS8PPz8+Z5YqIyM84LSDKy8ux2+1kZmY62oYPH05SUhKd\nO3dmypQp5Obm0qFDBzIzM1m1ahXl5eUkJCTQtWtXli9fTkBAAKNGjWL9+vXMmzePSZMmOatcERH5\nBaedYiosLOTUqVMMHjyYhx9+mD179lBQUECnTp0ACA0NZceOHezdu5egoCA8PDwwm820bNmSwsJC\nrFYr3bp1c/TduXOns0oVEREDThtBNGrUiCFDhjBgwAC++uorhg0bht1ux2QyAeDl5UVpaSk2mw2z\n2exYzsvLC5vNVq39fN+asFqttb8xctXS+33t0nvvfE4LiNtuu41bb70Vk8nEbbfdhq+vLwUFBY7n\ny8rK8PHxwdvbm7KysmrtZrO5Wvv5vjXRsWPH2t0QZ8gqrusKGox68X5LrbNarXrva9HFwtZpp5hW\nrlzJzJkzASgpKcFms9G1a1fy8/MByMvLIyQkhMDAQKxWK+Xl5ZSWllJUVERAQADBwcFs3brV0VcH\ng4jIleW0EUT//v2ZMGECFosFk8nEc889R9OmTZk8eTIZGRn4+/sTHh6Oq6sriYmJJCQkYLfbGTNm\nDJ6enlgsFlJSUrBYLLi7u5Oenu6sUkVExIDJbrfb67qI2lJfhp1RT66t6xIajHXp/eq6BKkD9eV3\nvb642P7UjXIiImJIASEiIoYUECIiYkgBISIihhQQIiJiSAEhIiKGFBAiImJIASEiIoYUECIiYkgB\nISIihhQQIiJiSAEhIiKGFBAiImJIASEiIoYUECIiYkgBISIihhQQIiJiyGlfOSoi9U+9+rbDrOK6\nruBX1fdvPNQIQkREDCkgRETEkAJCREQMKSBERMSQAkJERAw5NSCOHj3KPffcQ1FREV9//TUWi4WE\nhARSU1OpqqoCICcnh9jYWOLi4tiyZQsAp0+fZtSoUSQkJDBs2DCOHTvmzDJFRMSA0wKisrKSKVOm\n0KhRIwBmzJhBUlISWVlZ2O12cnNzOXLkCJmZmWRnZ7N48WIyMjKoqKhg+fLlBAQEkJWVRXR0NPPm\nzXNWmSIichFOC4i0tDTi4+Np3rw5AAUFBXTq1AmA0NBQduzYwd69ewkKCsLDwwOz2UzLli0pLCzE\narXSrVs3R9+dO3c6q0wREbkIp9wot3r1avz8/OjWrRsLFiwAwG63YzKZAPDy8qK0tBSbzYbZbHYs\n5+Xlhc1mq9Z+vm9NWa3WWtwSudrp/ZarWX0/Pp0SEKtWrcJkMrFz504+//xzUlJSql1HKCsrw8fH\nB29vb8rKyqq1m83mau3n+9ZUx44da29DnKUe3AFaX9SL97s+0bFZq+rL8XmxIHPKKaZly5axdOlS\nMjMzuf3220lLSyM0NJT8/HwA8vLyCAkJITAwEKvVSnl5OaWlpRQVFREQEEBwcDBbt2519K0vO1lE\npCG5YnMxpaSkMHnyZDIyMvD39yc8PBxXV1cSExNJSEjAbrczZswYPD09sVgspKSkYLFYcHd3Jz09\n/UqVKSIi/8fpAZGZmen4eenSpRc8HxcXR1xcXLW2xo0bM2fOHGeXJiIil1CjU0zTpk27oC0lJaXW\nixERkavHJUcQEydO5Ntvv2Xfvn3s37/f0X7mzJnL+mSRiIjUP5cMiBEjRvDdd9/x7LPP8sQTTzja\nXV1dadWqldOLExGRunPJgGjRogUtWrTgrbfewmazUVpait1uB+DkyZP4+vpekSJFROTKq9FF6vnz\n5zN//vxqgWAymcjNzXVaYSIiUrdqFBBvvvkmmzZtws/Pz9n1iIjIVaJGn2K66aabuO6665xdi4iI\nXEVqNIL44x//SEJCAp07d8bDw8PR/vML1yIi0rDUKCBuvPFGbrzxRmfXIiIiV5EaBYRGCiIi154a\nBUS7du0cU3Wf17x5c8eEeiIi0vDUKCAKCwsdP1dWVrJp0yb27NnjtKJERKTuXfZ03+7u7kRERLBr\n1y5n1CMiIleJGo0g1qxZ4/jZbrezf/9+3N3dnVaUiIjUvRoFxPkv+jmvadOmvPjii04pSERErg41\nCogZM2ZQWVnJwYMHOXv2LG3atMHN7Yp915CIiNSBGv2V37dvH6NHj8bX15eqqip+/PFHXnnlFe66\n6y5n1yciInWkRgExffp0XnzxRUcg7Nmzh2nTprFy5UqnFiciInWnRp9iOnnyZLXRQocOHSgvL3da\nUSIiUvdqFBDXXXcdmzZtcjzetGmTvgtCRKSBq9EppmnTpvHYY48xceJER1t2drbTihIRkbpXoxFE\nXl4ejRs3ZsuWLbzxxhv4+fnx4YcfOrs2ERGpQzUKiJycHJYvX06TJk1o164dq1evZunSpc6uTURE\n6lCNTjFVVlZWu3O6JndRnz17lkmTJnHw4EFMJhNPP/00np6ejB8/HpPJRJs2bUhNTcXFxYWcnByy\ns7Nxc3NjxIgRhIWFcfr0aZKTkzl69CheXl6kpaXpG+1ERK6gGgVEz549GTRoEBEREQC8//773Hvv\nvZdcZsuWLcC5axX5+fm8+OKL2O12kpKS6Ny5M1OmTCE3N5cOHTqQmZnJqlWrKC8vJyEhga5du7J8\n+XICAgIYNWoU69evZ968eUyaNOl3bq6IiNRUjQIiOTmZd999l927d+Pm5sbDDz9Mz549L7lMz549\n6d69OwDff/89Pj4+7Nixg06dOgEQGhrK9u3bcXFxISgoCA8PDzw8PGjZsiWFhYVYrVaGDh3q6Dtv\n3rzfsZkiInK5ajxfxn333cd99913eSt3cyMlJYWNGzcyZ84ctm/f7vheCS8vL0pLS7HZbJjNZscy\nXl5e2Gy2au3n+9aE1Wq9rBqlftP7LVez+n58On1CpbS0NMaOHUtcXFy1m+vKysrw8fHB29ubsrKy\nau1ms7la+/m+NdGxY8fa3QBnyCqu6woajHrxftcnOjZrVX05Pi8WZJf9fRA1tWbNGubPnw9A48aN\nMZlM3HHHHY6ZYfPy8ggJCSEwMBCr1Up5eTmlpaUUFRUREBBAcHCw4xvr8vLy6s2OFhFpKJw2gujd\nuzcTJkzgwQcf5MyZMzz11FO0atWKyZMnk5GRgb+/P+Hh4bi6upKYmEhCQgJ2u50xY8bg6emJxWIh\nJSUFi8WCu7s76enpzipVREQMOC0gmjRpwuzZsy9oN7p/Ii4ujri4uGptjRs3Zs6cOc4qT0REfoXT\nTjGJiEj9poAQERFDCggRETGkgBAREUMKCBERMaSAEBERQwoIERExpIAQERFDCggRETGkgBAREUMK\nCBERMaSAEBERQwoIERExpIAQERFDCggRETGkgBAREUMKCBERMaSAEBERQwoIERExpIAQERFDCggR\nETGkgBAREUMKCBERMeTmjJVWVlby1FNP8d1331FRUcGIESNo3bo148ePx2Qy0aZNG1JTU3FxcSEn\nJ4fs7Gzc3NwYMWIEYWFhnD59muTkZI4ePYqXlxdpaWn4+fk5o1QREbkIp4wg3nrrLXx9fcnKymLR\nokVMmzaNGTNmkJSURFZWFna7ndzcXI4cOUJmZibZ2dksXryYjIwMKioqWL58OQEBAWRlZREdHc28\nefOcUaaIiFyCU0YQ9913H+Hh4QDY7XZcXV0pKCigU6dOAISGhrJ9+3ZcXFwICgrCw8MDDw8PWrZs\nSWFhIVarlaFDhzr6KiBERK48pwSEl5cXADabjdGjR5OUlERaWhomk8nxfGlpKTabDbPZXG05m81W\nrf1835qyWq21uCVytdP7LVez+n58OiUgAA4dOsTIkSNJSEggKiqKWbNmOZ4rKyvDx8cHb29vysrK\nqrWbzeZq7ef71lTHjh1rbyOcJau4ritoMOrF+12f6NisVfXl+LxYkDnlGsSPP/7I4MGDSU5Opn//\n/gC0b9+e/Px8APLy8ggJCSEwMBCr1Up5eTmlpaUUFRUREBBAcHAwW7dudfStLztZRKQhccoI4rXX\nXuPEiRPMmzfPcf1g4sSJTJ8+nYyMDPz9/QkPD8fV1ZXExEQSEhKw2+2MGTMGT09PLBYLKSkpWCwW\n3N3dSU9Pd0aZIiJyCSa73W6v6yJqi9VqrRejjagn19Z1CQ3GuvR+dV1Cg6Jjs3bVl+PzYn87daOc\niIgYUkCIiIghBYSIiBhSQIiIiCEFhIiIGFJAiIiIIQWEiIgYUkCIiIghBYSIiBhSQIiIiCEFhIiI\nGFJAiIjjJTngAAAKHklEQVSIIQWEiIgYUkCIiIghBYSIiBhSQIiIiCEFhIiIGFJAiIiIIQWEiIgY\nUkCIiIghBYSIiBhSQIiIiCEFhIiIGFJAiIiIIacGxH/+8x8SExMB+Prrr7FYLCQkJJCamkpVVRUA\nOTk5xMbGEhcXx5YtWwA4ffo0o0aNIiEhgWHDhnHs2DFnlikiIgacFhALFy5k0qRJlJeXAzBjxgyS\nkpLIysrCbreTm5vLkSNHyMzMJDs7m8WLF5ORkUFFRQXLly8nICCArKwsoqOjmTdvnrPKFBGRi3Bz\n1opbtmzJyy+/zLhx4wAoKCigU6dOAISGhrJ9+3ZcXFwICgrCw8MDDw8PWrZsSWFhIVarlaFDhzr6\nXk5AWK3W2t8YuWrp/ZarWX0/Pp0WEOHh4RQXFzse2+12TCYTAF5eXpSWlmKz2TCbzY4+Xl5e2Gy2\nau3n+9ZUx44da2kLnCir+Nf7SI3Ui/e7PtGxWavqy/F5sSC7YhepXVz+/0uVlZXh4+ODt7c3ZWVl\n1drNZnO19vN9RUTkyrpiAdG+fXvy8/MByMvLIyQkhMDAQKxWK+Xl5ZSWllJUVERAQADBwcFs3brV\n0be+pLCISEPitFNMv5SSksLkyZPJyMjA39+f8PBwXF1dSUxMJCEhAbvdzpgxY/D09MRisZCSkoLF\nYsHd3Z309PQrVaaIiPwfpwZEixYtyMnJAeC2225j6dKlF/SJi4sjLi6uWlvjxo2ZM2eOM0sTEZFf\noRvlRETEkAJCREQMKSBERMSQAkJERAwpIERExJACQkREDCkgRETEkAJCREQMKSBERMSQAkJERAwp\nIERExJACQkREDCkgRETEkAJCREQMKSBERMSQAkJERAwpIERExJACQkREDCkgRETEkAJCREQMKSBE\nRMSQAkJERAwpIERExJBbXRdwMVVVVUydOpUvvvgCDw8Ppk+fzq233lrXZYmIXDOu2hHEpk2bqKio\nYMWKFTz55JPMnDmzrksSEbmmmOx2u72uizAyY8YMAgMDuf/++wHo1q0b27Ztu+QyVqv1SpQmItLg\ndOzY8YK2q/YUk81mw9vb2/HY1dWVM2fO4OZ28ZKNNlBERH6bq/YUk7e3N2VlZY7HVVVVlwwHERGp\nXVdtQAQHB5OXlwfAnj17CAgIqOOKRESuLVftNYjzn2L63//9X+x2O8899xytWrWq67JERK4ZV21A\niIhI3bpqTzGJiEjdUkCIiIghBYSIiBhSQIhIvXLixAlsNltdl3FN0I0FAsCECRMu+tyMGTOuYCUi\n1RUUFDBx4kTefPNNtmzZQmpqKj4+PqSkpNCjR4+6Lq9BU0AIAPv27eP06dP07duXoKAg9OE2uVo8\n//zzzJw5E3d3d1566SUWLVrErbfeytChQxUQTqZTTALAunXreOWVVygvL2fBggXs2bOHli1b0q1b\nt7ouTa5xVVVVtGvXjpKSEk6dOsWf/vQnvL29cXHRny9n0whCHAICAhg7diwAu3fvJj09nR9++IGc\nnJw6rkyuZeen2Nm2bRtdunQBoLKystpUPOIcCgipxmazsXHjRt5++21OnTpF375967okucZ16dKF\n+Ph4fvjhB1599VW++eYbnnnmGfr06VPXpTV4upNaANiwYQMbNmzg+++/p3fv3kRGRtKiRYu6LksE\ngKKiIry9vbnxxhv55ptv+OKLL+jVq1ddl9XgKSAEgHbt2uHv70+7du0AMJlMjufS09PrqiwRqUM6\nxSQA/POf/6zrEkTkKqMRhIiIGNLnxERExJACQkREDCkg5JpRXFxM27ZtmTJlSrX2zz//nLZt27J6\n9eoaraekpIRhw4YBsHnzZpYsWXLJ/qtXr2b8+PEAfP/99wwfPpyoqCgiIyP529/+xtGjRwHYu3cv\ns2bNuuS6SktLefzxx2tUp8jvpYCQa4qvry/btm3j7NmzjrYNGzbg5+dX43XceOONLFy4EDg3T9Dl\nTBw3ZcoUIiMjWbduHW+//Tbt27cnNTUVgAMHDjjC4mKOHz9OYWFhjV9P5PdQQMg1xcvLi9tvv53d\nu3c72rZv385f/vIXAJYuXcqAAQOIjIwkKiqKoqIiAHr06EFSUhLh4eHs3buXHj16cODAAbKzs8nO\nzmbVqlWUlJQwZMgQ4uLiCAsL44UXXrjg9X/88UdOnTrlePzggw/y4IMPcuLECebMmcPmzZt59dVX\nsdlsjB49moEDBxIWFkZycjJ2u53p06dz+PBhRo4cSXFxcbW5iF5++WVefvllKisrSU5OJjo6mujo\naN0JL7+ZAkKuOREREbz33nvAudM6bdu2xd3dHZvNxqZNm8jMzOTtt9+mZ8+eZGVlOZYLDQ3lvffe\nc4w2WrduTXx8PPHx8TzwwAO8/fbbREZGkpOTw1tvvUVWVhbHjh2r9tp///vfeeGFFwgNDSUlJYWt\nW7fSuXNnfHx8GD16ND169GDEiBH8+9//5vbbb2fFihW899577Nmzh4KCAiZNmkTz5s155ZVXLrp9\nn3zyCcePH2fNmjUsWbKEjz/+2Al7Ua4Fug9CrjlhYWG89NJLVFVV8c477xAREcGGDRvw9vYmPT2d\n9evX89VXX7Ft2zZuv/12x3J33XXXJdc7ZMgQdu3axeLFi9m/fz+VlZXVRgtwLmTy8vLIz89n586d\nzJo1i/Xr1zNv3rxq/SIjI9m7dy//+Mc/+PLLL/npp584efIkvr6+v7p9bdq04eDBgwwZMoTQ0FDH\n/Foil0sjCLnmeHt7065dO6xWK7t27XKcXjp06BADBw6ktLSU0NBQYmJiqk177unpecn1zpw5k8zM\nTG6++WZGjBhB06ZNqy3/008/8dxzz+Hp6ekYQaxbt47t27dfMNLIzMzk+eefx8/Pj4ceeohWrVpd\nMAW7yWSq1nbmzBkAmjZtyvr163nooYc4ePAgMTExnDhx4rftLLmmKSDkmhQREUF6ejp33HGHY7bQ\nJk2acOutt/LII49w1113kZeXV+1ithFXV1fHH+bt27czZMgQIiIiOHToECUlJVRVVTn6ms1mNm/e\nzJo1axxt33zzDc2aNeO66667YF0DBw6kb9++mEwmCgsLqaqqws3NzdHHx8eH48ePc+zYMSoqKti2\nbRsAubm5jB07lu7duzNp0iSaNGnCoUOHam/nyTVDp5jkmhQWFsbEiRP529/+5mhzd3enqqqKPn36\n4OHhQWBgIPv377/kev785z+TkpLC9ddfz2OPPca4cePw8fGhWbNm3HHHHRQXFzv6urq6smDBAmbO\nnMns2bNp1KgRzZs357XXXsPV1ZXAwEDmzp3LCy+8wKBBg5g6dSqvv/46Xl5eBAUFUVxcTEhICDff\nfDOJiYlkZmYyZMgQ+vfvzx/+8AfuvPNO4P9fK7n//vvx9PSkd+/etG3b1jk7Uho0TbUhIiKGdIpJ\nREQMKSBERMSQAkJERAwpIERExJACQkREDCkgRETEkAJCREQM/T+CYvPQ5qmORAAAAABJRU5ErkJg\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11e274ac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEeCAYAAACQfIJ4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11da96438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols = ['Occupation', 'Gender','MaritalStatus']\n",
    "plt_bar_plot(cols)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looks like MaritalStatus, Gender and Occupation are good features as well"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hasChildAtHome</th>\n",
       "      <th>NumberChildrenAtHome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Y</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Y</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Y</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  hasChildAtHome  NumberChildrenAtHome\n",
       "0              N                     0\n",
       "1              Y                     3\n",
       "2              Y                     3\n",
       "3              N                     0\n",
       "4              Y                     5"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# try feature generation - hasChildren\n",
    "has_child_at_home = []\n",
    "def generate_has_child_at_home(customer_info, has_child_at_home):\n",
    "    for index, row in customer_info.iterrows():\n",
    "        if row.NumberChildrenAtHome>0:\n",
    "            has_child_at_home.append('Y')\n",
    "        else:\n",
    "            has_child_at_home.append('N')\n",
    "    return has_child_at_home\n",
    "        \n",
    "combined['hasChildAtHome'] = generate_has_child_at_home(customer_info, has_child_at_home)\n",
    "combined[['hasChildAtHome','NumberChildrenAtHome']].head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "hasChildAtHome\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEdCAYAAAAW6PDWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMfX/ADNRNDFHVODfAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11e573a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEdCAYAAAAW6PDWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XlYVfW+x/H3ZlQZUjN9MsVyQBOjHNKMNDESs1TUJMFo\nUNPM4WKFDOEUlppXPKFiVt6T1wKkNIfHboNDclKjc/bJSD2YGg1a11Ar2VvZTOv+4XWfyIViuUHk\n83oen8f927+113f99vBhrb3Wb1sMwzAQERH5HbfaLkBERK5MCggRETGlgBAREVMKCBERMaWAEBER\nUwoIERExpYCQP628vJy//vWvDB8+nKFDhzJo0CAWLlxISUmJS9aXl5fHzJkzL2mZTz75hNDQUEaM\nGEFxcfEfWu+SJUt4/vnn/9CyZi42bgkJCaxcudJ02aFDh3Lq1CnWrVvHhAkTTPs88MAD5ObmVmob\nPnw4gwYN4vdntycnJ7N3716g6u3s378/X3755SVvp9RdCgj502bPns3nn3/OqlWr2LBhA++88w4F\nBQU899xzLlnfoUOHOHbs2CUts3nzZkaOHMnatWtp0KCBS+q6VH9m3DZs2IC/v/8lrS8vL4+SkhI8\nPT3529/+Vum+Xbt2nRcaIh61XYDUbd9//z2bNm3ik08+wdfXF4BGjRoxZ84cPv/8cwCKioqYM2cO\n+fn5WCwW+vTpw9NPP42HhwcdO3Zk9+7dNG3aFMB5++DBgyxevJjWrVtz8OBBSkpKmDlzJm3atCEt\nLY2ioiISExNJTk4mMTGRb7/9Fjc3N4KCgnj++edxc/v33z6vv/46W7duxdvbm6KiIp5++mnmz5/P\n7t27cXd3Jzg4mMTERHx9fenfvz/BwcEcOHCAp59+mnvvvbfS9n799dfExMRQWFhIs2bNSE1NpXnz\n5mzfvp0VK1ZQUlLCyZMniYiIIDY2Frvdblrf0aNHLzpuAJ9//jmjRo3i+PHjdOjQgUWLFtGoUSPn\nOP3WoUOHSEpK4syZM7Rt25bTp09Xuj8jI4N+/frRpEkTVq1aRd++fQFYvHgxP/30E88++ywvvfRS\ntZ73LVu2sHTpUsrLy/H19SUxMZHg4GCWLFnCd999x/fff89PP/1EcHAwISEhrF+/niNHjhAXF8cD\nDzwAwPLly/nwww+pqKjghhtuYNasWbRo0aJa65caYoj8Ce+//74xYsSIC/aZPn26kZKSYlRUVBgO\nh8MYM2aMsWLFCsMwDCMwMNA4ceKEs++5259++qlx8803G/v37zcMwzBWrlxpjB492jAMw1i7dq0x\nfvx4wzAM49133zXGjBljGIZhlJWVGc8995zxzTffnFdDfHy88frrrxuGYRgvv/yyMXnyZKOkpMQo\nLy83EhISjBkzZhiGYRihoaHG0qVLTbcjLS3N6N+/v7PeiRMnGkuXLjUqKiqMhx9+2CgoKDAMwzD+\n93//17j55puNEydOVFlfdcYtPj7eePDBB43Tp08bZWVlxrBhw4x333230jj9diyGDh1qZGdnG4Zh\nGP/4xz+Mjh07Gp9++qlhGIbx888/G7fccotx4MAB49ixY0bnzp2NgwcPOtcVGhpq5OXlObezV69e\nxpAhQyr9CwoKMvLy8oxDhw4Zd955p/Hdd98ZhmEYu3btMkJCQoyioiIjLS3NCA0NNU6dOmWcOXPG\nuP3224158+YZhmEYH330kTFgwADn8xYbG2uUlpYahmEYWVlZxrhx4y44HlLzdIhJ/hQ3NzcqKiou\n2CcnJ4eHH34Yi8WCl5cXo0aNIicn56KP3bJlS26++WYAOnfuzK+//npen+7du3Po0CFiYmJ49dVX\nefTRR2nTps1F6xk1ahSenp64ubkRExNT6ZBLjx49qlw2JCTEubfTqVMnTp48icVi4ZVXXmHfvn0s\nXbqU+fPnYxgGZ86cqbK+6owbQFhYGA0bNsTd3Z0OHTpw8uRJ034///wzBw4cICIiwjkuHTp0cN6/\nbt062rdvT2BgIM2bN+fOO+9k1apVVa530KBBbNiwodK/5s2bA/Dpp59yxx130Lp1awB69+5N06ZN\nnd9h3Hnnnfj5+dGgQQOaN29Onz59AAgICOCXX34BYPv27XzxxReMGDGCoUOH8uabb1JQUHDR8ZCa\npYCQPyU4OJivv/4am81Wqf3YsWOMHz+e4uLi8z4IKyoqKCsrO++xfv+l9m+/K7BYLKbHyFu3bs1H\nH33E+PHjsdlsPP7447z//vsXrNmsntLSUuftRo0aVbmsh8e/j8qeq+n06dMMGzaMffv20blzZ6ZP\nn46HhweGYVRZX3XGrar1mbFYLACV7j+3rGEYZGVlcfToUfr370///v3Zu3cvGzdu5Oeff77gWJkx\nq8EwDOdz6uXlVem+327DORUVFYwbN84ZPmvXriUzM/OSaxHXUkDIn9KiRQsGDx5MUlKS88POZrMx\ne/ZsGjduTIMGDbjrrrt46623MAyDkpISsrOzufPOOwFo2rSp88yYjz76qFrrdHd3d34YZWRkkJiY\nyF133UVcXBx33XUXBw8evODyffr0ISsri9LSUioqKnjrrbcICQn5o0PAt99+i81mIzY2lv79+/PZ\nZ59RUlJCRUVFlfVVZ9wuRePGjQkKCuLtt98GYN++fXz11VcA7Ny5kxMnTrBlyxa2bdvGtm3b+Nvf\n/sZ1111HVlYWUHlML+aOO+5g586dfP/99wDs3r2bH3/8kVtvvbXa9d5111288847zm1/+eWXmT59\nerWXl5qhL6nlT5s1axbp6emMGjUKd3d3SkpKCAsLY8qUKcDZUyjnzp3L4MGDKS0tpU+fPjz55JPO\n+55//nn8/f258847ue666y66vq5du/KXv/yFSZMmsXDhQj777DMGDRpEw4YNadmyJY888sgFl584\ncSILFiwgIiKCsrIygoODmTFjxh/e/o4dO9KvXz/uu+8+/P39CQgIoH379nz77bdERERUWd/Fxu1S\npaamkpiYSFZWFgEBAbRt2xaAzMxMIiMj8fPzc/b18PBgwoQJpKWlMW7cOMLCwpg2bRpz58696Hra\nt2/PrFmzmDx5MuXl5TRo0IBXXnml0uNfzMiRIzl27BiRkZFYLBauv/565s+ff+kbLS5lMaraZxUR\nkXpNh5hERMSUAkJEREwpIERExJQCQuq9quYY+vLLL5k6dSpw4XmRquNyz+MkUhMUECJVuOWWW0hL\nS6vtMkRqjQJChLPXUwwbNoz777+fd955B4Dc3FznvEG/NW/ePB599FHsdjslJSW8+OKLDBs2jCFD\nhpCQkHDexW/nHD58mNGjR/PAAw8QFxfn7Pf7PZhzt5cvX84zzzzjbLdarc4rpf/5z38SHR3NsGHD\nGD58ONu3bwfOXjF9rj0mJubyDI7UW7oOQgTw9vbm3Xff5dixY0RERJhe9GUYBnPmzKGwsJDXXnsN\nLy8vli5diru7O+vWrcNisZCamsp//ud/Mnv27POW/+6771i7di1NmjQhLi6O5cuXExcXV2VNkZGR\nDBgwgF9++YXGjRuzZs0aRo0axa+//kpiYiIrV66kVatWzusJOnbsCJydtG/btm3OSQBF/igFhAgw\natQo4OyV4XfddRe7d+92fuCe88Ybb3DixAnWr1/vnE7i448/pqioiF27dgFQWlrKtddea7qOe++9\n1zmP04gRIy46c+q1115Lv3792LBhAxEREXzyySfMmjWLf/zjHxQWFjJp0iRnX4vFwoEDB4CzF+4p\nHORyUECIQKXpwQ3DMJ0/6Pbbb6dbt24kJiayZs0aPD09qaioICkpibvvvhsAu92Ow+EwXYe7u3uV\n6/jt9aq/nZNq9OjRzJ49Gw8PDwYMGICPjw/l5eW0a9fOOa0GnJ3DqWnTpmzatOmCc0mJXAp9ByEC\nvPvuuwD88MMP7Nq1i969e5/Xp0uXLjz88MP4+fmxdOlSAOc8U+fmXpoxYwapqamm69i2bRu//vor\n5eXlrFmzxvl7DL+dCXXPnj0UFhY6l+nWrRtubm6sXLmSqKgoAG677Ta+/fZb/v73vwPwr3/9i/Dw\ncH766afLNBoiZ2kPQgRwOBwMGzaM0tJSkpOTuemmm0w/cC0WCy+++CIRERHcfffdPPXUUyxYsIBh\nw4ZRXl7OzTffTEJCguk62rVrx4QJEzh16hTdu3dn/PjxADz77LPMnj2bNWvWEBQURFBQUKXlhg8f\nznvvvec85NW0aVPS0tJ46aWXcDgcGIbBSy+9xA033HCZR0XqO83FJHIFKysrY9KkSc7frBapSTrE\nJHKFOnToEL1798bX15eBAwfWdjlSD2kPQkRETGkPQkRETCkgRETE1FV1FpPVaq3tEkRE6qTu3buf\n13ZVBQSYb6RcOqvVqrGUK5Zen5dXVX9c6xCTiIiYUkCIiIgpBYSIiJhSQIiIiCkFhIiImFJAiIiI\nKQWEiIiYctl1EOXl5SQnJ1NQUIDFYmHOnDmUlZUxYcIEbrzxRgCioqIYNGgQ2dnZZGVl4eHhwcSJ\nEwkNDaW4uJi4uDhOnDiBj48PCxYscP4a158x+JkNf/oxfmvToqGX9fFERK4ULguIcz+inpWVRW5u\nLosXL6Z///48/vjjjBkzxtmvsLCQ1atXs3btWhwOB9HR0YSEhJCZmUlgYCBTpkxh8+bNpKenk5yc\n7Kpya4TD4eC+++5j27ZttV2KiKnL/QeUS2Ucqe0KLqqu/wHpsoAICwujX79+wNlf6fL392fv3r0U\nFBSwdetW2rRpQ1JSEnl5eXTt2hUvLy+8vLwICAggPz8fq9XKuHHjAOjbty/p6emuKlVEREy4dKoN\nDw8P4uPj+eijj0hLS+PYsWOMHDmSLl26sHz5cpYtW0anTp3w8/NzLuPj44PNZsNmsznbfXx8KCoq\nqtY6a3o+poutr7i4mGXLlmG322nRogUOh4OIiAjGjBnDDTfcwJYtW/jll1+4++67SUtL49prr6Ww\nsJDevXtz5MgRvvnmG2677TZGjRpFSkoKAQEBHDlyhAYNGtCxY0fy8vI4ffo0CQkJvPHGG4SEhNC1\na1eOHj3KW2+9xfTp0122bSJyYXX9PeTyuZgWLFjAs88+S2RkJFlZWbRo0QKAe++9l5SUFHr06IHd\nbnf2t9vt+Pn54evr62y32+34+/tXa30XnZ/lMu+WXmx9K1eu5Pbbb2fatGl88cUXTJs2DT8/P4KC\ngmjXrh1fffUVjRo1okuXLpw8eZI1a9ZQXFzMPffcQ05ODg0bNiQ0NJSFCxfi5+dHWFgYQ4YMYezY\nsbRv355Zs2YRHx9PaWkpEyZMIDMzk3HjxrFlyxbGjh37h+er0Vw39VQdOGxTl9SV91CNz8W0fv16\nVqxYAUDDhg2xWCxMnjyZvLw8AHbv3k1QUBDBwcFYrVYcDgdFRUUcPnyYwMBAunXrxo4dOwDIycmp\nMwP9e9988w233HILALfeeiseHpUz+be/19S6dWv8/Pzw9/enWbNmNG7cGG9vbywWi7PPud8r9vf3\np3379s7/OxwOevXqxeHDhzl58iQ7d+4kNDTU1ZsnIlcxl+1BDBgwgMTEREaPHk1ZWRlJSUlcf/31\npKSk4OnpSbNmzUhJScHX15eYmBiio6MxDINp06bh7e1NVFQU8fHxREVF4enpyaJFi1xVqku1a9eO\nPXv2EBYWxv79+ykrK8PLy4vCwkLatWvH/v37nXtVvw2CP8JisTBkyBDmzp1LSEgInp6el2MTRKSe\ncllANGrUiJdffvm89qysrPPaIiMjiYyMrNTWsGFD0tLSLntdNX1WQVRUFNOnTycqKoq2bdvi6enJ\nI488wpw5c2jZsiXNmze/rOsbPnw4/fr1Y8OGOnQ2iohcka6q36TWcXM4duwY06dPZ9WqVX/qcTSW\n9VOdOs21Dqgrp7lW9X7XldRXkQ8//JBx48YxderU2i5FRK4CV90vytVnAwYMYMCAAbVdhohcJbQH\nISIiphQQIiJiSgEhIiKmFBAiImJKASEiIqYUECIiYkoBISIiphQQIiJiSgEhIiKmFBAiImJKASEi\nIqYUECIiYkoBISIiphQQIiJiSgEhIiKmFBAiImJKASEiIqYUECIiYkoBISIiplz2m9Tl5eUkJydT\nUFCAxWJhzpw5eHt7k5CQgMVioUOHDsyaNQs3Nzeys7PJysrCw8ODiRMnEhoaSnFxMXFxcZw4cQIf\nHx8WLFhA06ZNXVWuiIj8jsv2ILZv3w5AVlYWsbGxLF68mHnz5hEbG0tGRgaGYbB161YKCwtZvXo1\nWVlZrFy5ktTUVEpKSsjMzCQwMJCMjAwiIiJIT093VakiImLCZXsQYWFh9OvXD4AffvgBf39/du3a\nRc+ePQHo27cvO3fuxM3Nja5du+Ll5YWXlxcBAQHk5+djtVoZN26cs291A8Jqtbpke+ojjaXIn1PX\n30MuCwgADw8P4uPj+eijj0hLS2Pnzp1YLBYAfHx8KCoqwmaz4efn51zGx8cHm81Wqf1c3+ro3r37\n5d+QeshqtWos66OMI7VdwVWlrryHqgoyl39JvWDBAj744ANmzJiBw+Fwttvtdvz9/fH19cVut1dq\n9/Pzq9R+rq+IiNQclwXE+vXrWbFiBQANGzbEYrHQpUsXcnNzAcjJyaFHjx4EBwdjtVpxOBwUFRVx\n+PBhAgMD6datGzt27HD2rStJLCJytXDZIaYBAwaQmJjI6NGjKSsrIykpiXbt2jFjxgxSU1Np27Yt\n4eHhuLu7ExMTQ3R0NIZhMG3aNLy9vYmKiiI+Pp6oqCg8PT1ZtGiRq0oVERETFsMwjNou4nLRcfPL\nR2NZPw1+ZkNtl3BV2bRoaG2XUC1Vvd91oZyIiJhSQIiIiCkFhIiImFJAiIiIKQWEiIiYUkCIiIgp\nBYSIiJhSQIiIiCkFhIiImFJAiIiIKQWEiIiYUkCIiIgpBYSIiJhSQIiIiCkFhIiImFJAiIiIKQWE\niIiYUkCIiIgpBYSIiJhSQIiIiCkFhIiImPJwxYOWlpaSlJTE0aNHKSkpYeLEiVx//fVMmDCBG2+8\nEYCoqCgGDRpEdnY2WVlZeHh4MHHiREJDQykuLiYuLo4TJ07g4+PDggULaNq0qStKFRGRKrgkIDZu\n3Ejjxo1ZuHAhv/zyCxEREUyaNInHH3+cMWPGOPsVFhayevVq1q5di8PhIDo6mpCQEDIzMwkMDGTK\nlCls3ryZ9PR0kpOTXVGqiIhUwSWHmAYOHMh//Md/AGAYBu7u7uzdu5ePP/6Y0aNHk5SUhM1mIy8v\nj65du+Ll5YWfnx8BAQHk5+djtVrp06cPAH379mX37t2uKFNERC7AJXsQPj4+ANhsNqZOnUpsbCwl\nJSWMHDmSLl26sHz5cpYtW0anTp3w8/OrtJzNZsNmsznbfXx8KCoqqva6rVbr5d2YekxjKfLn1PX3\nkEsCAuDHH39k0qRJREdHM3jwYE6dOoW/vz8A9957LykpKfTo0QO73e5cxm634+fnh6+vr7Pdbrc7\nl6uO7t27X94NqaesVqvGsj7KOFLbFVxV6sp7qKogc8khpuPHjzNmzBji4uJ48MEHARg7dix5eXkA\n7N69m6CgIIKDg7FarTgcDoqKijh8+DCBgYF069aNHTt2AJCTk1NnBllE5Grikj2IV155hVOnTpGe\nnk56ejoACQkJvPjii3h6etKsWTNSUlLw9fUlJiaG6OhoDMNg2rRpeHt7ExUVRXx8PFFRUXh6erJo\n0SJXlCkiIhdgMQzDqO0iLhcdFrl8NJb10+BnNtR2CVeVTYuG1nYJ1VLV+10XyomIiCkFhIiImFJA\niIiIKQWEiIiYUkCIiIgpBYSIiJhSQIiIiCkFhIiImFJAiIiIKQWEiIiYUkCIiIgpBYSIiJhSQIiI\niCkFhIiImFJAiIiIKQWEiIiYUkCIiIgpBYSIiJhSQIiIiCkFhIiImFJAiIiIKQ9XPGhpaSlJSUkc\nPXqUkpISJk6cSPv27UlISMBisdChQwdmzZqFm5sb2dnZZGVl4eHhwcSJEwkNDaW4uJi4uDhOnDiB\nj48PCxYsoGnTpq4oVUREqlCtPYiUlJTz2uLj46vsv3HjRho3bkxGRgavv/46KSkpzJs3j9jYWDIy\nMjAMg61bt1JYWMjq1avJyspi5cqVpKamUlJSQmZmJoGBgWRkZBAREUF6evof30IREflDLrgH8dxz\nz/H999+zd+9eDh486GwvKyujqKioyuUGDhxIeHg4AIZh4O7uzr59++jZsycAffv2ZefOnbi5udG1\na1e8vLzw8vIiICCA/Px8rFYr48aNc/ZVQIiI1LwLBsTEiRM5evQoL7zwApMnT3a2u7u7065duyqX\n8/HxAcBmszF16lRiY2NZsGABFovFeX9RURE2mw0/P79Ky9lstkrt5/pWl9VqrXZfuTCNpcifU9ff\nQxcMiFatWtGqVSs2btyIzWajqKgIwzAAOH36NI0bN65y2R9//JFJkyYRHR3N4MGDWbhwofM+u92O\nv78/vr6+2O32Su1+fn6V2s/1ra7u3btXu69UzWq1aizro4wjtV3BVaWuvIeqCrJqfUm9YsUKVqxY\nUSkQLBYLW7duNe1//PhxxowZw8yZM+nduzcAnTt3Jjc3l169epGTk8Mdd9xBcHAwf/nLX3A4HJSU\nlHD48GECAwPp1q0bO3bsIDg4mJycnDozyCIiV5NqBcTbb7/Nli1bqn0m0SuvvMKpU6dIT093fn/w\n3HPPMXfuXFJTU2nbti3h4eG4u7sTExNDdHQ0hmEwbdo0vL29iYqKIj4+nqioKDw9PVm0aNEf30IR\nEflDLMa5Y0YXEBMTwxtvvIG7u3tN1PSH1ZXDIoOf2VDbJVw1Ni0aWtslXFX02ry86srrs6rPzmrt\nQdx4441ER0fTq1cvvLy8nO2//eJaRESuLtUKiBYtWtCiRQtX1yIiIleQagWE9hREROqfagVEp06d\nnNcwnNO8eXN27NjhkqJERKT2VSsg8vPznf8vLS1ly5Yt7Nmzx2VFiYhI7bvk2Vw9PT257777+PTT\nT11Rj4iIXCGqtQexfv165/8Nw+DgwYN4enq6rCgREal91QqI3NzcSrebNGnC4sWLXVKQiIhcGaoV\nEPPmzaO0tJSCggLKy8vp0KEDHh4u+SkJERG5QlTrU37v3r1MnTqVxo0bU1FRwfHjx1m2bBm33nqr\nq+sTEZFaUq2AmDt3LosXL3YGwp49e0hJSeGdd95xaXEiIlJ7qnUW0+nTpyvtLdx22204HA6XFSUi\nIrWvWgFxzTXXsGXLFuftLVu2XPC3IEREpO6r1iGmlJQUJkyYwHPPPedsy8rKcllRIiJS+6q1B5GT\nk0PDhg3Zvn07q1atomnTpnz22Weurk1ERGpRtQIiOzubzMxMGjVqRKdOnVi3bh1vvvmmq2sTEZFa\nVK2AKC0trXTltK6iFhG5+lXrO4iwsDAeffRR7rvvPgA+/PBD7rnnHpcWJiIitataAREXF8f777/P\n3//+dzw8PHjkkUcICwtzdW0iIlKLqj1fxsCBAxk4cKAraxERkSvIJU/3LSIi9YMCQkRETLk0IL74\n4gtiYmIA2L9/P3369CEmJoaYmBjee+894OwptMOHDycyMpLt27cDUFxczJQpU4iOjuaJJ57g5MmT\nrixTRERMuGzO7tdee42NGzfSsGFDAPbt28fjjz/OmDFjnH0KCwtZvXo1a9euxeFwEB0dTUhICJmZ\nmQQGBjJlyhQ2b95Meno6ycnJripVRERMuCwgAgICWLJkCdOnTwfOThleUFDA1q1badOmDUlJSeTl\n5dG1a1e8vLzw8vIiICCA/Px8rFYr48aNA6Bv376kp6dXe71Wq9Ul2yNXJj3fciWr669PlwVEeHg4\nR44ccd4ODg5m5MiRdOnSheXLl7Ns2TI6deqEn5+fs4+Pjw82mw2bzeZs9/HxoaioqNrr7d69++Xb\nCFfJOHLxPlItdeL5rkv02rys6srrs6ogq7Evqe+99166dOni/P/+/fvx9fXFbrc7+9jtdvz8/Cq1\n2+12/P39a6pMERH5fzUWEGPHjiUvLw+A3bt3ExQURHBwMFarFYfDQVFREYcPHyYwMJBu3bqxY8cO\n4OxEgXUlhUVEriY19sPSs2fPJiUlBU9PT5o1a0ZKSgq+vr7ExMQQHR2NYRhMmzYNb29voqKiiI+P\nJyoqCk9PTxYtWlRTZYqIyP9zaUC0atWK7OxsAIKCgkx/QyIyMpLIyMhKbQ0bNiQtLc2VpYmIyEXo\nQjkRETGlgBAREVMKCBERMaWAEBERUwoIERExpYAQERFTCggRETGlgBAREVMKCBERMaWAEBERUwoI\nERExpYAQERFTCggRETGlgBAREVMKCBERMaWAEBERUwoIERExpYAQERFTCggRETGlgBAREVMKCBER\nMeXSgPjiiy+IiYkB4NtvvyUqKoro6GhmzZpFRUUFANnZ2QwfPpzIyEi2b98OQHFxMVOmTCE6Opon\nnniCkydPurJMEREx4bKAeO2110hOTsbhcAAwb948YmNjycjIwDAMtm7dSmFhIatXryYrK4uVK1eS\nmppKSUkJmZmZBAYGkpGRQUREBOnp6a4qU0REquCygAgICGDJkiXO2/v27aNnz54A9O3bl127dpGX\nl0fXrl3x8vLCz8+PgIAA8vPzsVqt9OnTx9l39+7dripTRESq4OGqBw4PD+fIkSPO24ZhYLFYAPDx\n8aGoqAibzYafn5+zj4+PDzabrVL7ub7VZbVaL9MWSF2g51uuZHX99emygPg9N7d/76zY7Xb8/f3x\n9fXFbrdXavfz86vUfq5vdXXv3v3yFe0qGUcu3keqpU4833WJXpuXVV15fVYVZDV2FlPnzp3Jzc0F\nICcnhx49ehAcHIzVasXhcFBUVMThw4cJDAykW7du7Nixw9m3rgyyiMjVpMb2IOLj45kxYwapqam0\nbduW8PBw3N3diYmJITo6GsMwmDZtGt7e3kRFRREfH09UVBSenp4sWrSopsoUEZH/59KAaNWqFdnZ\n2QDcdNNNvPnmm+f1iYyMJDIyslJbw4YNSUtLc2VpIiJyEbpQTkRETCkgRETElAJCRERMKSBERMSU\nAkJEREw7okdsAAAKKklEQVQpIERExJQCQkRETCkgRETElAJCRERMKSBERMSUAkJEREwpIERExJQC\nQkRETCkgRETElAJCRERMKSBERMSUAkJEREwpIERExJQCQkRETCkgRETElAJCRERMedT0CocNG4av\nry8ArVq14sknnyQhIQGLxUKHDh2YNWsWbm5uZGdnk5WVhYeHBxMnTiQ0NLSmSxURqddqNCAcDgeG\nYbB69Wpn25NPPklsbCy9evVi5syZbN26ldtuu43Vq1ezdu1aHA4H0dHRhISE4OXlVZPliojUazUa\nEPn5+Zw5c4YxY8ZQVlbG008/zb59++jZsycAffv2ZefOnbi5udG1a1e8vLzw8vIiICCA/Px8goOD\na7JcEZF6rUYDokGDBowdO5aRI0fyzTff8MQTT2AYBhaLBQAfHx+Kioqw2Wz4+fk5l/Px8cFms1Vr\nHVar1SW1y5VJz7dcyer667NGA+Kmm26iTZs2WCwWbrrpJho3bsy+ffuc99vtdvz9/fH19cVut1dq\n/21gXEj37t0ve92XXcaR2q7gqlEnnu+6RK/Ny6quvD6rCrIaPYvpnXfeYf78+QAcO3YMm81GSEgI\nubm5AOTk5NCjRw+Cg4OxWq04HA6Kioo4fPgwgYGBNVmqiEi9V6N7EA8++CCJiYlERUVhsVh48cUX\nadKkCTNmzCA1NZW2bdsSHh6Ou7s7MTExREdHYxgG06ZNw9vbuyZLFRGp92o0ILy8vFi0aNF57W++\n+eZ5bZGRkURGRtZEWSIiYkIXyomIiCkFhIiImFJAiIiIKQWEiIiYUkCIiIgpBYSIiJhSQIiIiCkF\nhIiImFJAiIiIKQWEiIiYUkCIiIgpBYSIiJhSQIiIiCkFhIiImFJAiIiIKQWEiIiYUkCIiIgpBYSI\niJhSQIiIiCkFhIiImFJAiIiIKQWEiIiY8qjtAqpSUVHB7NmzOXDgAF5eXsydO5c2bdrUdlkiIvXG\nFbsHsWXLFkpKSlizZg3PPPMM8+fPr+2SRETqlSs2IKxWK3369AHgtttuY+/evbVckYhI/XLFHmKy\n2Wz4+vo6b7u7u1NWVoaHx4VLtlqtri7tT5sd3aq2S7hq1IXnuy7Ra/Pyquuvzys2IHx9fbHb7c7b\nFRUVFw2H7t27u7osEZF644o9xNStWzdycnIA2LNnD4GBgbVckYhI/WIxDMOo7SLMnDuL6auvvsIw\nDF588UXatWtX22WJiNQbV2xAiIhI7bpiDzGJiEjtUkCIiIgpBYSIXLEOHDhQ2yXUa/oOQgBYv359\nlfdFRETUYCUi/zZw4EBGjRrFY489Vtul1EsKCAFg0aJFlW4bhsG6deto0KAB27Ztq6WqpL6z2+28\n9NJLHDlyhPnz53PdddfVdkn1igJCzvPdd98RHx/PTTfdRFJSUqUr2kVqw2effUZSUhK33nqrs+33\nf9TI5XfFXkktteOtt95i1apVJCYmEhoaWtvliHD48GFSU1Pp2bOnDnfWMAWEAHDs2DESExO55ppr\nePvtt7nmmmtquyQRXn31VbKyspg5cyb9+vWr7XLqHR1iEgB69OiBl5cXd9xxBxaLpdJ92pWX2jJ1\n6lTmzJlDkyZNaruUekkBIcDZY7xV6dmzZw1WIiJXCgWEiIiY0oVyIiJiSgEhIiKmFBBS5+Xm5hIT\nE/OnH2fPnj08+uijDBkyhAceeIDZs2dTXFwMQEJCAuvWrTtvmczMTDIzMwHo2LGj6eP279+fI0eO\nOG9v376djh07nvczuomJiRw9ehSAmJgYcnNzK91fVQ0irqKAEAHy8/OZPHkyTz/9NBs3bmT9+vUY\nhsGMGTMuuFxUVBRRUVGXtK5169YRHh5OVlZWpfbc3Fz0laBcSRQQclU4efIkTzzxBOHh4Tz55JOU\nlJSwePFiIiMjCQ8PZ9SoURQWFlJaWkpcXBwRERFERESQnZ0NwMqVK3nooYecV+p6eHgQFxdHWFiY\ncx0ff/wxDz74IKGhoaxZswaAJUuWsGTJkkq1/PLLLzzxxBMMHjyY2NhYHA5HpTp3797N9OnTef/9\n97HZbMDZ8/1/+uknxo8fz88//3zR7d2+fTtDhw5l8ODBPPXUUxw/fhw4u7eycOFC7r//foYMGcLH\nH3/MI488wt133817770HwPHjx3nqqacYPnw4I0aMYNeuXX902OUqp4CQq8IPP/zAzJkz+Z//+R+O\nHz9OZmYmX3/9NVlZWXzwwQcEBASwadMmPv/8c3799VfWr1/PX//6V/75z38C8K9//Yvg4OBKj+nr\n60t4eLjzdklJCW+//TYrVqxg8eLFVdaSlpZG586d2bRpE6NHj3Z+eANs2rSJkJAQWrVqRZcuXdiw\nYQMA48ePp3nz5rz66qvOc/6Tk5MZOnSo89+5ObFOnDjBzJkzWbZsGZs2baJbt248//zzznU0b96c\nzZs3ExQUxKuvvsp//dd/sXDhQl599VUAXnjhBUaMGMG6detYvnw5M2fOdAaVyG/pSmq5KnTq1InW\nrVsD0K5dO/z9/YmPj+ftt9+moKCAPXv2EBAQQIcOHSgoKGDs2LH07duXZ599FuC8iwPN3HPPPVgs\nFjp06HDBv/I/++wz58WFt99+u7MuOHt4afLkyQAMGjSIN998k9GjR5s+zty5c+nVq5fzdkJCAgB5\neXkEBwfTqlUrAB566CHnhz9A3759AWjZsiXNmzfHw8ODli1bcurUKQB27drF119/TVpaGgBlZWV8\n//333HzzzRcdA6lftAchVwUPj3//rWOxWPj5558ZO3YsFRUVhIeHExYWhmEYNGnShM2bN/Pwww9T\nUFDAsGHDOHXqFF26dOHLL7+s9Jg2m815uArA3d3d+fgXYrFYKn2XcG65/fv389VXX/HCCy/Qv39/\nli1bxsGDB/n8888vaVsrKioq3TYMg7KyMudtT09P5/9/Oy6/XX7VqlVs2LCBDRs2sGbNGgIDAy+p\nBqkfFBByVbJYLPTs2ZOoqCjat2/Pzp07KS8vZ+vWrTz77LP069eP5ORkGjVqxI8//shjjz1GZmYm\neXl5AJSWljJ//nx8fX3x8vK6pHX37t3beegoLy+P7777Dji79xAZGcnHH3/Mtm3b2LFjB0OHDnV+\nn+Hu7k55eflFH//WW2/liy++cJ4ZtWbNmkp7Ghdzxx13kJGRAcChQ4cYMmQIZ86cuaRtlPpBh5jk\nqlRcXEx+fj6DBw/G09OTjh07cuTIESZNmsQHH3zA/fffj7e3NwMGDHCenrpw4UJeeOEFzpw5Q1lZ\nGb179yY5OfmS1z116lQSEhK4//77adu2La1bt6a0tJRNmzbx3//935X6PvbYYzz00EMkJibSr18/\nxo8fz+uvv37Bx2/WrBnPP/88kydPprS0lJYtW/LCCy9Uu77k5GRmzpzJ4MGDAXjppZc0pbuY0lQb\nIiJiSoeYRETElAJCRERMKSBERMSUAkJEREwpIERExJQCQkRETCkgRETElAJCRERM/R/5TYG6Jf5K\n2AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1187d15f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols = ['hasChildAtHome']\n",
    "plt_bar_plot(cols)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "hasChildAtHome seems to be a good indicator of if a bike is bought or not."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>YearlyIncome</th>\n",
       "      <th>NumberCarsOwned</th>\n",
       "      <th>Occupation</th>\n",
       "      <th>Gender</th>\n",
       "      <th>MaritalStatus</th>\n",
       "      <th>hasChildAtHome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>137947</td>\n",
       "      <td>0</td>\n",
       "      <td>Professional</td>\n",
       "      <td>M</td>\n",
       "      <td>M</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>101141</td>\n",
       "      <td>1</td>\n",
       "      <td>Professional</td>\n",
       "      <td>M</td>\n",
       "      <td>S</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>91945</td>\n",
       "      <td>1</td>\n",
       "      <td>Professional</td>\n",
       "      <td>M</td>\n",
       "      <td>M</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>86688</td>\n",
       "      <td>1</td>\n",
       "      <td>Professional</td>\n",
       "      <td>F</td>\n",
       "      <td>S</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>92771</td>\n",
       "      <td>4</td>\n",
       "      <td>Professional</td>\n",
       "      <td>F</td>\n",
       "      <td>S</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   YearlyIncome  NumberCarsOwned    Occupation Gender MaritalStatus  \\\n",
       "0        137947                0  Professional      M             M   \n",
       "1        101141                1  Professional      M             S   \n",
       "2         91945                1  Professional      M             M   \n",
       "3         86688                1  Professional      F             S   \n",
       "4         92771                4  Professional      F             S   \n",
       "\n",
       "  hasChildAtHome  \n",
       "0              N  \n",
       "1              Y  \n",
       "2              Y  \n",
       "3              N  \n",
       "4              Y  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_chosen = ['YearlyIncome','NumberCarsOwned', 'Occupation', 'Gender','MaritalStatus','hasChildAtHome']\n",
    "features = combined[features_chosen]\n",
    "features.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Prepare data for sickit learn classification.\n",
    "1. Will encode categorical variables using one hot encoding.\n",
    "2. Will convert features and labels to numpy arrays."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 0 ..., 0 0 0]\n"
     ]
    }
   ],
   "source": [
    "labels = np.array(combined.BikeBuyer)\n",
    "print(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "import sklearn.model_selection as ms\n",
    "from sklearn import linear_model\n",
    "import sklearn.metrics as sklm\n",
    "\n",
    "def encode_string(cat_features):\n",
    "    enc = preprocessing.LabelEncoder()\n",
    "    enc.fit(cat_features)\n",
    "    enc_cat_features = enc.transform(cat_features)\n",
    "    ohe = preprocessing.OneHotEncoder()\n",
    "    encoded = ohe.fit(enc_cat_features.reshape(-1,1))\n",
    "    return encoded.transform(enc_cat_features.reshape(-1,1)).toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def encode_cat_features(features):\n",
    "    cat_features = [ 'Gender','MaritalStatus','hasChildAtHome']\n",
    "    f = encode_string(features['Occupation'])\n",
    "    for cat in cat_features:\n",
    "        enc = encode_string(features[cat])\n",
    "        f = np.concatenate([f,enc],1)\n",
    "    return f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "numeric_features = np.array(combined[['YearlyIncome','NumberCarsOwned']])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "encoded_features = encode_cat_features(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "features = np.concatenate([encoded_features,numeric_features], 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00,\n",
       "         1.00000000e+00,   0.00000000e+00,   1.00000000e+00,\n",
       "         0.00000000e+00,   0.00000000e+00,   1.00000000e+00,\n",
       "         1.00000000e+00,   0.00000000e+00,   8.66880000e+04,\n",
       "         1.00000000e+00])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features.shape\n",
    "features[3,:13]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Conversion and encoding done now to split and train data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  1.00000000e+00,   0.00000000e+00,   0.00000000e+00,\n",
       "          0.00000000e+00,   0.00000000e+00,   1.00000000e+00,\n",
       "          0.00000000e+00,   0.00000000e+00,   1.00000000e+00,\n",
       "          1.00000000e+00,   0.00000000e+00,   3.97160000e+04,\n",
       "          1.00000000e+00],\n",
       "       [  0.00000000e+00,   1.00000000e+00,   0.00000000e+00,\n",
       "          0.00000000e+00,   0.00000000e+00,   0.00000000e+00,\n",
       "          1.00000000e+00,   0.00000000e+00,   1.00000000e+00,\n",
       "          1.00000000e+00,   0.00000000e+00,   1.72468000e+05,\n",
       "          2.00000000e+00]])"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Time to split and train\n",
    "nr.seed(9988)\n",
    "indx = range(features.shape[0])\n",
    "indx = ms.train_test_split(indx, test_size = 300)\n",
    "X_train = features[indx[0],:]\n",
    "y_train = np.ravel(labels[indx[0]])\n",
    "X_test = features[indx[1],:]\n",
    "y_test = np.ravel(labels[indx[1]])\n",
    "X_train[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.        ,  0.        ,  0.        ,  0.        ,  0.        ,\n",
       "         1.        ,  0.        ,  0.        ,  1.        ,  1.        ,\n",
       "         0.        , -0.96803832, -0.43876786],\n",
       "       [ 0.        ,  1.        ,  0.        ,  0.        ,  0.        ,\n",
       "         0.        ,  1.        ,  0.        ,  1.        ,  1.        ,\n",
       "         0.        ,  2.37320265,  0.43844103]])"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Time to scale, this would most especially be useful in yearly income\n",
    "scaler = preprocessing.StandardScaler().fit(X_train[:,11:])\n",
    "X_train[:,11:] = scaler.transform(X_train[:,11:])\n",
    "X_test[:,11:] = scaler.transform(X_test[:,11:])\n",
    "X_train[:2]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Features are now prepared, we will train with logistic regression\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Due to class imbalance for bike buyers and non bike buyers we wil be making use of class weight parameters\n",
    "logistic_mod = linear_model.LogisticRegression(class_weight=\"balanced\") "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.1"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# perform cross validation for the regularization parameter C\n",
    "nr.seed(123)\n",
    "inside = ms.KFold(n_splits=10, shuffle = True)\n",
    "nr.seed(321)\n",
    "outside = ms.KFold(n_splits=10, shuffle = True)\n",
    "nr.seed(3456)\n",
    "param_grid = {\"C\": [0.1, 1, 10, 100, 1000]}\n",
    "clf = ms.GridSearchCV(estimator = logistic_mod, param_grid = param_grid, \n",
    "                      cv = inside, # Use the inside folds\n",
    "                      scoring = 'roc_auc',\n",
    "                      return_train_score = True)\n",
    "clf.fit(features, labels)\n",
    "clf.best_estimator_.C"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean performance metric = 0.670\n",
      "SDT of the metric       = 0.010\n",
      "Outcomes by cv fold\n",
      "Fold  1    0.672\n",
      "Fold  2    0.652\n",
      "Fold  3    0.688\n",
      "Fold  4    0.664\n",
      "Fold  5    0.682\n",
      "Fold  6    0.665\n",
      "Fold  7    0.663\n",
      "Fold  8    0.663\n",
      "Fold  9    0.677\n",
      "Fold 10    0.678\n"
     ]
    }
   ],
   "source": [
    "nr.seed(498)\n",
    "cv_estimate = ms.cross_val_score(clf, features, labels, \n",
    "                                 cv = outside) # Use the outside folds\n",
    "print('Mean performance metric = %4.3f' % np.mean(cv_estimate))\n",
    "\n",
    "print('SDT of the metric       = %4.3f' % np.std(cv_estimate))\n",
    "print('Outcomes by cv fold')\n",
    "for i, x in enumerate(cv_estimate):\n",
    "    print('Fold %2d    %4.3f' % (i+1, x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.00230034]\n",
      "[[ 0.21402281 -0.8806718   0.46320162  0.00115858  0.20458913 -0.21770668\n",
      "   0.22000702 -0.73166213  0.73396247 -0.86942543  0.87172577  0.9021973\n",
      "  -0.00124658]]\n"
     ]
    }
   ],
   "source": [
    "# logistic regression with weight more on the bike buyer =1 case due to class imbalance {0:1,1:2}\n",
    "logistic_mod = linear_model.LogisticRegression(C=clf.best_estimator_.C, class_weight=\"balanced\") \n",
    "logistic_mod.fit( X_train, y_train)\n",
    "print(logistic_mod.intercept_)\n",
    "print(logistic_mod.coef_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.62415101  0.37584899]\n",
      " [ 0.38487313  0.61512687]\n",
      " [ 0.48483137  0.51516863]\n",
      " [ 0.33162968  0.66837032]\n",
      " [ 0.6888334   0.3111666 ]\n",
      " [ 0.90497289  0.09502711]\n",
      " [ 0.81824471  0.18175529]\n",
      " [ 0.84378025  0.15621975]\n",
      " [ 0.82469365  0.17530635]\n",
      " [ 0.47412803  0.52587197]\n",
      " [ 0.96465145  0.03534855]\n",
      " [ 0.29984697  0.70015303]\n",
      " [ 0.41667284  0.58332716]\n",
      " [ 0.52057317  0.47942683]\n",
      " [ 0.57184659  0.42815341]]\n"
     ]
    }
   ],
   "source": [
    "probabilities = logistic_mod.predict_proba(X_test)\n",
    "print(probabilities[:15,:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 1 1 0 0 0 0 0 1 0 1 1 0 0 1 0 0]\n",
      "[1 0 1 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0]\n"
     ]
    }
   ],
   "source": [
    "def score_model(probs, threshold):\n",
    "    return np.array([1 if x > threshold else 0 for x in probs[:,1]])\n",
    "threshold = 0.51\n",
    "scores = score_model(probabilities, threshold)\n",
    "print(np.array(scores[:18]))\n",
    "print(y_test[:18])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def print_metrics(labels, scores):\n",
    "    metrics = sklm.precision_recall_fscore_support(labels, scores)\n",
    "    conf = sklm.confusion_matrix(labels, scores)\n",
    "    print('                 Confusion matrix')\n",
    "    print('                 Score positive    Score negative')\n",
    "    print('Actual positive    %6d' % conf[0,0] + '             %5d' % conf[0,1])\n",
    "    print('Actual negative    %6d' % conf[1,0] + '             %5d' % conf[1,1])\n",
    "    print('')\n",
    "    print('Accuracy  %0.2f' % sklm.accuracy_score(labels, scores))\n",
    "    print(' ')\n",
    "    print('           Positive      Negative')\n",
    "    print('Num case   %6d' % metrics[3][0] + '        %6d' % metrics[3][1])\n",
    "    print('Precision  %6.2f' % metrics[0][0] + '        %6.2f' % metrics[0][1])\n",
    "    print('Recall     %6.2f' % metrics[1][0] + '        %6.2f' % metrics[1][1])\n",
    "    print('F1         %6.2f' % metrics[2][0] + '        %6.2f' % metrics[2][1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                 Confusion matrix\n",
      "                 Score positive    Score negative\n",
      "Actual positive       155                58\n",
      "Actual negative        25                62\n",
      "\n",
      "Accuracy  0.72\n",
      " \n",
      "           Positive      Negative\n",
      "Num case      213            87\n",
      "Precision    0.86          0.52\n",
      "Recall       0.73          0.71\n",
      "F1           0.79          0.60\n"
     ]
    }
   ],
   "source": [
    "print_metrics(y_test, scores)    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAETCAYAAADd6corAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1187ebdd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_auc(labels, probs):\n",
    "    ## Compute the false positive rate, true positive rate\n",
    "    ## and threshold along with the AUC\n",
    "    fpr, tpr, threshold = sklm.roc_curve(labels, probs[:,1])\n",
    "    auc = sklm.auc(fpr, tpr)\n",
    "    \n",
    "    ## Plot the result\n",
    "    plt.title('Receiver Operating Characteristic')\n",
    "    plt.plot(fpr, tpr, color = 'orange', label = 'AUC = %0.2f' % auc)\n",
    "    plt.legend(loc = 'lower right')\n",
    "    plt.plot([0, 1], [0, 1],'r--')\n",
    "    plt.xlim([0, 1])\n",
    "    plt.ylim([0, 1])\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.show()\n",
    "    \n",
    "plot_auc(y_test, probabilities)    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets try random forests with balanced arguement"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we do cross validation to choose optimal hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16404, 13)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3\n",
      "20\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "param_grid = {\"max_features\": [2, 3, 5, 10, 13], \"min_samples_leaf\":[3, 5, 10, 20]}\n",
    "nr.seed(3456)\n",
    "rf_clf = RandomForestClassifier(class_weight = \"balanced\")\n",
    "nr.seed(4455)\n",
    "rf_clf = ms.GridSearchCV(estimator = rf_clf, param_grid = param_grid, \n",
    "                      cv = inside, # Use the inside folds\n",
    "                      scoring = 'roc_auc',\n",
    "                      return_train_score = True)\n",
    "rf_clf.fit(features, labels)\n",
    "print(rf_clf.best_estimator_.max_features)\n",
    "print(rf_clf.best_estimator_.min_samples_leaf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                 Confusion matrix\n",
      "                 Score positive    Score negative\n",
      "Actual positive       166                47\n",
      "Actual negative        27                60\n",
      "\n",
      "Accuracy  0.75\n",
      " \n",
      "           Positive      Negative\n",
      "Num case      213            87\n",
      "Precision    0.86          0.56\n",
      "Recall       0.78          0.69\n",
      "F1           0.82          0.62\n"
     ]
    }
   ],
   "source": [
    "nr.seed(1115)\n",
    "rf_mod = RandomForestClassifier(class_weight = \"balanced\", \n",
    "                                max_features = rf_clf.best_estimator_.max_features, \n",
    "                                min_samples_leaf = rf_clf.best_estimator_.min_samples_leaf) \n",
    "rf_mod.fit(X_train, y_train)\n",
    "probabilities = rf_mod.predict_proba(X_test)\n",
    "scores = score_model(probabilities, 0.54)\n",
    "print_metrics(y_test, scores)     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAETCAYAAADd6corAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1187eb048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "  plot_auc(y_test, probabilities)    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Random Forest claassifier does a better job of predicting on the different test sets, will make use of this model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(500, 23)\n",
      "(500,)\n"
     ]
    }
   ],
   "source": [
    "# import final test data\n",
    "final = pd.read_csv('data/AW_test.csv')\n",
    "print(final.shape)\n",
    "print(final.CustomerID.unique().shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>CustomerID</th>\n",
       "      <th>Title</th>\n",
       "      <th>FirstName</th>\n",
       "      <th>MiddleName</th>\n",
       "      <th>LastName</th>\n",
       "      <th>Suffix</th>\n",
       "      <th>AddressLine1</th>\n",
       "      <th>AddressLine2</th>\n",
       "      <th>City</th>\n",
       "      <th>StateProvinceName</th>\n",
       "      <th>...</th>\n",
       "      <th>BirthDate</th>\n",
       "      <th>Education</th>\n",
       "      <th>Occupation</th>\n",
       "      <th>Gender</th>\n",
       "      <th>MaritalStatus</th>\n",
       "      <th>HomeOwnerFlag</th>\n",
       "      <th>NumberCarsOwned</th>\n",
       "      <th>NumberChildrenAtHome</th>\n",
       "      <th>TotalChildren</th>\n",
       "      <th>YearlyIncome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>18988</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Courtney</td>\n",
       "      <td>A</td>\n",
       "      <td>Baker</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8727 Buena Vista Ave.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Fremont</td>\n",
       "      <td>California</td>\n",
       "      <td>...</td>\n",
       "      <td>1/5/1945</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>Management</td>\n",
       "      <td>F</td>\n",
       "      <td>S</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>86931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>29135</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Adam</td>\n",
       "      <td>C</td>\n",
       "      <td>Allen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3491 Cook Street</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Haney</td>\n",
       "      <td>British Columbia</td>\n",
       "      <td>...</td>\n",
       "      <td>10/4/1964</td>\n",
       "      <td>Bachelors</td>\n",
       "      <td>Skilled Manual</td>\n",
       "      <td>M</td>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>100125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12156</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Bonnie</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Raji</td>\n",
       "      <td>NaN</td>\n",
       "      <td>359 Pleasant Hill Rd</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Burbank</td>\n",
       "      <td>California</td>\n",
       "      <td>...</td>\n",
       "      <td>1/12/1934</td>\n",
       "      <td>Graduate Degree</td>\n",
       "      <td>Management</td>\n",
       "      <td>F</td>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>103985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>13749</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Julio</td>\n",
       "      <td>C</td>\n",
       "      <td>Alonso</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8945 Euclid Ave.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Burlingame</td>\n",
       "      <td>California</td>\n",
       "      <td>...</td>\n",
       "      <td>9/22/1958</td>\n",
       "      <td>Graduate Degree</td>\n",
       "      <td>Skilled Manual</td>\n",
       "      <td>M</td>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>127161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>27780</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Christy</td>\n",
       "      <td>A</td>\n",
       "      <td>Andersen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>42, boulevard Tremblay</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Dunkerque</td>\n",
       "      <td>Nord</td>\n",
       "      <td>...</td>\n",
       "      <td>3/19/1965</td>\n",
       "      <td>High School</td>\n",
       "      <td>Manual</td>\n",
       "      <td>F</td>\n",
       "      <td>M</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>21876</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   CustomerID Title FirstName MiddleName  LastName Suffix  \\\n",
       "0       18988   NaN  Courtney          A     Baker    NaN   \n",
       "1       29135   NaN      Adam          C     Allen    NaN   \n",
       "2       12156   NaN    Bonnie        NaN      Raji    NaN   \n",
       "3       13749   NaN     Julio          C    Alonso    NaN   \n",
       "4       27780   NaN   Christy          A  Andersen    NaN   \n",
       "\n",
       "             AddressLine1 AddressLine2        City StateProvinceName  \\\n",
       "0   8727 Buena Vista Ave.          NaN     Fremont        California   \n",
       "1        3491 Cook Street          NaN       Haney  British Columbia   \n",
       "2    359 Pleasant Hill Rd          NaN     Burbank        California   \n",
       "3        8945 Euclid Ave.          NaN  Burlingame        California   \n",
       "4  42, boulevard Tremblay          NaN   Dunkerque              Nord   \n",
       "\n",
       "       ...       BirthDate        Education      Occupation Gender  \\\n",
       "0      ...        1/5/1945       Bachelors       Management      F   \n",
       "1      ...       10/4/1964       Bachelors   Skilled Manual      M   \n",
       "2      ...       1/12/1934  Graduate Degree      Management      F   \n",
       "3      ...       9/22/1958  Graduate Degree  Skilled Manual      M   \n",
       "4      ...       3/19/1965      High School          Manual      F   \n",
       "\n",
       "  MaritalStatus HomeOwnerFlag NumberCarsOwned NumberChildrenAtHome  \\\n",
       "0             S             0               2                    0   \n",
       "1             M             1               2                    2   \n",
       "2             M             1               2                    0   \n",
       "3             M             1               0                    0   \n",
       "4             M             1               1                    2   \n",
       "\n",
       "   TotalChildren  YearlyIncome  \n",
       "0              5         86931  \n",
       "1              4        100125  \n",
       "2              4        103985  \n",
       "3              4        127161  \n",
       "4              2         21876  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "final['hasChildAtHome'] = generate_has_child_at_home(final,[])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>YearlyIncome</th>\n",
       "      <th>NumberCarsOwned</th>\n",
       "      <th>Occupation</th>\n",
       "      <th>Gender</th>\n",
       "      <th>MaritalStatus</th>\n",
       "      <th>hasChildAtHome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>86931</td>\n",
       "      <td>2</td>\n",
       "      <td>Management</td>\n",
       "      <td>F</td>\n",
       "      <td>S</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100125</td>\n",
       "      <td>2</td>\n",
       "      <td>Skilled Manual</td>\n",
       "      <td>M</td>\n",
       "      <td>M</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>103985</td>\n",
       "      <td>2</td>\n",
       "      <td>Management</td>\n",
       "      <td>F</td>\n",
       "      <td>M</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>127161</td>\n",
       "      <td>0</td>\n",
       "      <td>Skilled Manual</td>\n",
       "      <td>M</td>\n",
       "      <td>M</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>21876</td>\n",
       "      <td>1</td>\n",
       "      <td>Manual</td>\n",
       "      <td>F</td>\n",
       "      <td>M</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   YearlyIncome  NumberCarsOwned      Occupation Gender MaritalStatus  \\\n",
       "0         86931                2      Management      F             S   \n",
       "1        100125                2  Skilled Manual      M             M   \n",
       "2        103985                2      Management      F             M   \n",
       "3        127161                0  Skilled Manual      M             M   \n",
       "4         21876                1          Manual      F             M   \n",
       "\n",
       "  hasChildAtHome  \n",
       "0              N  \n",
       "1              Y  \n",
       "2              N  \n",
       "3              N  \n",
       "4              Y  "
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_features = final[features_chosen]\n",
    "final_features.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00,\n",
       "         0.00000000e+00,   1.00000000e+00,   0.00000000e+00,\n",
       "         1.00000000e+00,   1.00000000e+00,   0.00000000e+00,\n",
       "         1.00000000e+00,   0.00000000e+00,   1.27161000e+05,\n",
       "         0.00000000e+00])"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numeric_final_features = np.array(final_features[['YearlyIncome','NumberCarsOwned']])\n",
    "encoded_final_features = encode_cat_features(final_features)\n",
    "final_features = np.concatenate([encoded_final_features,numeric_final_features], 1)\n",
    "final_features[3,:13]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.        ,  1.        ,  0.        ,  0.        ,  0.        ,\n",
       "         1.        ,  0.        ,  0.        ,  1.        ,  1.        ,\n",
       "         0.        ,  0.2203181 ,  0.43844103],\n",
       "       [ 0.        ,  0.        ,  0.        ,  0.        ,  1.        ,\n",
       "         0.        ,  1.        ,  1.        ,  0.        ,  0.        ,\n",
       "         1.        ,  0.55239847,  0.43844103]])"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_features[:,11:] = scaler.transform(final_features[:,11:])\n",
    "final_features[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "probabilities = rf_mod.predict_proba(final_features)\n",
    "scores = score_model(probabilities, 0.54)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 0 0 0 0 1 1 1 0 1 0 0 0 0 0 1 0 1 1 0 0 1 0 0 0 0 1 1 0 1 1 0 0 1 1 1\n",
      " 0 0 0 1 1 0 0 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 1 0 0 0 1 0 1 1 1 0 1 0 0 1 0\n",
      " 0 1 1 1 0 0 1 1 0 1 0 0 0 1 1 0 0 1 0 1 0 1 0 0 0 0 1 0 0 1 0 1 0 0 0 0 1\n",
      " 0 0 0 0 1 0 0 0 1 1 1 1 0 0 0 0 0 1 1 0 0 1 0 0 1 1 0 1 0 0 0 1 0 1 0 0 0\n",
      " 0 0 0 0 0 0 1 0 1 0 1 1 0 1 1 1 1 0 1 1 0 0 0 1 0 0 1 0 1 0 0 1 0 0 1 0 0\n",
      " 0 1 0 0 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0 1 1 0 1 0 1 1 0 1 1 0 0 1 1 1 1\n",
      " 1 1 0 0 0 1 1 0 1 0 0 0 1 0 1 0 0 0 1 1 0 0 1 0 0 0 1 1 1 0 1 0 1 1 0 0 1\n",
      " 0 0 1 1 0 1 1 0 0 1 0 1 1 0 0 1 0 1 0 1 0 0 1 0 0 0 0 0 0 0 1 0 0 0 1 0 1\n",
      " 1 0 0 0 1 0 1 1 0 0 1 1 0 1 0 1 1 0 0 0 0 1 1 0 0 0 0 0 1 0 0 1 0 0 1 0 1\n",
      " 0 1 1 1 0 1 0 0 0 0 1 0 0 1 0 1 1 1 1 0 1 0 1 1 1 0 0 1 1 0 1 0 0 1 0 0 0\n",
      " 1 0 0 1 0 0 1 0 0 1 1 1 0 0 0 1 1 1 1 0 0 0 0 0 0 0 1 0 0 0 0 1 1 1 1 0 1\n",
      " 0 1 0 1 0 0 0 1 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 1 1 0 1 0 1 0 1\n",
      " 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 1 0 0 1 0 1 1 1 1 1 0 0 0 1 0 0 0 0 1 1 1 1\n",
      " 1 1 1 0 0 1 0 0 1 1 1 1 1 0 1 1 1 0 0]\n"
     ]
    }
   ],
   "source": [
    "print(scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "np.savetxt('final_answers.csv', scores, delimiter=',',fmt='%i')"
   ]
  }
 ],
 "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}