solution_telecom_churn.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# import libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "from IPython.core.interactiveshell import InteractiveShell\n",
    "InteractiveShell.ast_node_interactivity = \"all\"\n",
    "%matplotlib inline\n",
    "\n",
    "pd.set_option(\"display.max_columns\", 300)\n",
    "pd.set_option(\"display.max_rows\", 300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# read data\n",
    "churn = pd.read_csv(\"data/telecom_churn_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mobile_number</th>\n",
       "      <th>circle_id</th>\n",
       "      <th>loc_og_t2o_mou</th>\n",
       "      <th>std_og_t2o_mou</th>\n",
       "      <th>loc_ic_t2o_mou</th>\n",
       "      <th>last_date_of_month_6</th>\n",
       "      <th>last_date_of_month_7</th>\n",
       "      <th>last_date_of_month_8</th>\n",
       "      <th>last_date_of_month_9</th>\n",
       "      <th>arpu_6</th>\n",
       "      <th>arpu_7</th>\n",
       "      <th>arpu_8</th>\n",
       "      <th>arpu_9</th>\n",
       "      <th>onnet_mou_6</th>\n",
       "      <th>onnet_mou_7</th>\n",
       "      <th>onnet_mou_8</th>\n",
       "      <th>onnet_mou_9</th>\n",
       "      <th>offnet_mou_6</th>\n",
       "      <th>offnet_mou_7</th>\n",
       "      <th>offnet_mou_8</th>\n",
       "      <th>offnet_mou_9</th>\n",
       "      <th>roam_ic_mou_6</th>\n",
       "      <th>roam_ic_mou_7</th>\n",
       "      <th>roam_ic_mou_8</th>\n",
       "      <th>roam_ic_mou_9</th>\n",
       "      <th>roam_og_mou_6</th>\n",
       "      <th>roam_og_mou_7</th>\n",
       "      <th>roam_og_mou_8</th>\n",
       "      <th>roam_og_mou_9</th>\n",
       "      <th>loc_og_t2t_mou_6</th>\n",
       "      <th>loc_og_t2t_mou_7</th>\n",
       "      <th>loc_og_t2t_mou_8</th>\n",
       "      <th>loc_og_t2t_mou_9</th>\n",
       "      <th>loc_og_t2m_mou_6</th>\n",
       "      <th>loc_og_t2m_mou_7</th>\n",
       "      <th>loc_og_t2m_mou_8</th>\n",
       "      <th>loc_og_t2m_mou_9</th>\n",
       "      <th>loc_og_t2f_mou_6</th>\n",
       "      <th>loc_og_t2f_mou_7</th>\n",
       "      <th>loc_og_t2f_mou_8</th>\n",
       "      <th>loc_og_t2f_mou_9</th>\n",
       "      <th>loc_og_t2c_mou_6</th>\n",
       "      <th>loc_og_t2c_mou_7</th>\n",
       "      <th>loc_og_t2c_mou_8</th>\n",
       "      <th>loc_og_t2c_mou_9</th>\n",
       "      <th>loc_og_mou_6</th>\n",
       "      <th>loc_og_mou_7</th>\n",
       "      <th>loc_og_mou_8</th>\n",
       "      <th>loc_og_mou_9</th>\n",
       "      <th>std_og_t2t_mou_6</th>\n",
       "      <th>std_og_t2t_mou_7</th>\n",
       "      <th>std_og_t2t_mou_8</th>\n",
       "      <th>std_og_t2t_mou_9</th>\n",
       "      <th>std_og_t2m_mou_6</th>\n",
       "      <th>std_og_t2m_mou_7</th>\n",
       "      <th>std_og_t2m_mou_8</th>\n",
       "      <th>std_og_t2m_mou_9</th>\n",
       "      <th>std_og_t2f_mou_6</th>\n",
       "      <th>std_og_t2f_mou_7</th>\n",
       "      <th>std_og_t2f_mou_8</th>\n",
       "      <th>std_og_t2f_mou_9</th>\n",
       "      <th>std_og_t2c_mou_6</th>\n",
       "      <th>std_og_t2c_mou_7</th>\n",
       "      <th>std_og_t2c_mou_8</th>\n",
       "      <th>std_og_t2c_mou_9</th>\n",
       "      <th>std_og_mou_6</th>\n",
       "      <th>std_og_mou_7</th>\n",
       "      <th>std_og_mou_8</th>\n",
       "      <th>std_og_mou_9</th>\n",
       "      <th>isd_og_mou_6</th>\n",
       "      <th>isd_og_mou_7</th>\n",
       "      <th>isd_og_mou_8</th>\n",
       "      <th>isd_og_mou_9</th>\n",
       "      <th>spl_og_mou_6</th>\n",
       "      <th>spl_og_mou_7</th>\n",
       "      <th>spl_og_mou_8</th>\n",
       "      <th>spl_og_mou_9</th>\n",
       "      <th>og_others_6</th>\n",
       "      <th>og_others_7</th>\n",
       "      <th>og_others_8</th>\n",
       "      <th>og_others_9</th>\n",
       "      <th>total_og_mou_6</th>\n",
       "      <th>total_og_mou_7</th>\n",
       "      <th>total_og_mou_8</th>\n",
       "      <th>total_og_mou_9</th>\n",
       "      <th>loc_ic_t2t_mou_6</th>\n",
       "      <th>loc_ic_t2t_mou_7</th>\n",
       "      <th>loc_ic_t2t_mou_8</th>\n",
       "      <th>loc_ic_t2t_mou_9</th>\n",
       "      <th>loc_ic_t2m_mou_6</th>\n",
       "      <th>loc_ic_t2m_mou_7</th>\n",
       "      <th>loc_ic_t2m_mou_8</th>\n",
       "      <th>loc_ic_t2m_mou_9</th>\n",
       "      <th>loc_ic_t2f_mou_6</th>\n",
       "      <th>loc_ic_t2f_mou_7</th>\n",
       "      <th>loc_ic_t2f_mou_8</th>\n",
       "      <th>loc_ic_t2f_mou_9</th>\n",
       "      <th>loc_ic_mou_6</th>\n",
       "      <th>loc_ic_mou_7</th>\n",
       "      <th>loc_ic_mou_8</th>\n",
       "      <th>loc_ic_mou_9</th>\n",
       "      <th>std_ic_t2t_mou_6</th>\n",
       "      <th>std_ic_t2t_mou_7</th>\n",
       "      <th>std_ic_t2t_mou_8</th>\n",
       "      <th>std_ic_t2t_mou_9</th>\n",
       "      <th>std_ic_t2m_mou_6</th>\n",
       "      <th>std_ic_t2m_mou_7</th>\n",
       "      <th>std_ic_t2m_mou_8</th>\n",
       "      <th>std_ic_t2m_mou_9</th>\n",
       "      <th>std_ic_t2f_mou_6</th>\n",
       "      <th>std_ic_t2f_mou_7</th>\n",
       "      <th>std_ic_t2f_mou_8</th>\n",
       "      <th>std_ic_t2f_mou_9</th>\n",
       "      <th>std_ic_t2o_mou_6</th>\n",
       "      <th>std_ic_t2o_mou_7</th>\n",
       "      <th>std_ic_t2o_mou_8</th>\n",
       "      <th>std_ic_t2o_mou_9</th>\n",
       "      <th>std_ic_mou_6</th>\n",
       "      <th>std_ic_mou_7</th>\n",
       "      <th>std_ic_mou_8</th>\n",
       "      <th>std_ic_mou_9</th>\n",
       "      <th>total_ic_mou_6</th>\n",
       "      <th>total_ic_mou_7</th>\n",
       "      <th>total_ic_mou_8</th>\n",
       "      <th>total_ic_mou_9</th>\n",
       "      <th>spl_ic_mou_6</th>\n",
       "      <th>spl_ic_mou_7</th>\n",
       "      <th>spl_ic_mou_8</th>\n",
       "      <th>spl_ic_mou_9</th>\n",
       "      <th>isd_ic_mou_6</th>\n",
       "      <th>isd_ic_mou_7</th>\n",
       "      <th>isd_ic_mou_8</th>\n",
       "      <th>isd_ic_mou_9</th>\n",
       "      <th>ic_others_6</th>\n",
       "      <th>ic_others_7</th>\n",
       "      <th>ic_others_8</th>\n",
       "      <th>ic_others_9</th>\n",
       "      <th>total_rech_num_6</th>\n",
       "      <th>total_rech_num_7</th>\n",
       "      <th>total_rech_num_8</th>\n",
       "      <th>total_rech_num_9</th>\n",
       "      <th>total_rech_amt_6</th>\n",
       "      <th>total_rech_amt_7</th>\n",
       "      <th>total_rech_amt_8</th>\n",
       "      <th>total_rech_amt_9</th>\n",
       "      <th>max_rech_amt_6</th>\n",
       "      <th>max_rech_amt_7</th>\n",
       "      <th>max_rech_amt_8</th>\n",
       "      <th>max_rech_amt_9</th>\n",
       "      <th>date_of_last_rech_6</th>\n",
       "      <th>date_of_last_rech_7</th>\n",
       "      <th>date_of_last_rech_8</th>\n",
       "      <th>date_of_last_rech_9</th>\n",
       "      <th>last_day_rch_amt_6</th>\n",
       "      <th>last_day_rch_amt_7</th>\n",
       "      <th>last_day_rch_amt_8</th>\n",
       "      <th>last_day_rch_amt_9</th>\n",
       "      <th>date_of_last_rech_data_6</th>\n",
       "      <th>date_of_last_rech_data_7</th>\n",
       "      <th>date_of_last_rech_data_8</th>\n",
       "      <th>date_of_last_rech_data_9</th>\n",
       "      <th>total_rech_data_6</th>\n",
       "      <th>total_rech_data_7</th>\n",
       "      <th>total_rech_data_8</th>\n",
       "      <th>total_rech_data_9</th>\n",
       "      <th>max_rech_data_6</th>\n",
       "      <th>max_rech_data_7</th>\n",
       "      <th>max_rech_data_8</th>\n",
       "      <th>max_rech_data_9</th>\n",
       "      <th>count_rech_2g_6</th>\n",
       "      <th>count_rech_2g_7</th>\n",
       "      <th>count_rech_2g_8</th>\n",
       "      <th>count_rech_2g_9</th>\n",
       "      <th>count_rech_3g_6</th>\n",
       "      <th>count_rech_3g_7</th>\n",
       "      <th>count_rech_3g_8</th>\n",
       "      <th>count_rech_3g_9</th>\n",
       "      <th>av_rech_amt_data_6</th>\n",
       "      <th>av_rech_amt_data_7</th>\n",
       "      <th>av_rech_amt_data_8</th>\n",
       "      <th>av_rech_amt_data_9</th>\n",
       "      <th>vol_2g_mb_6</th>\n",
       "      <th>vol_2g_mb_7</th>\n",
       "      <th>vol_2g_mb_8</th>\n",
       "      <th>vol_2g_mb_9</th>\n",
       "      <th>vol_3g_mb_6</th>\n",
       "      <th>vol_3g_mb_7</th>\n",
       "      <th>vol_3g_mb_8</th>\n",
       "      <th>vol_3g_mb_9</th>\n",
       "      <th>arpu_3g_6</th>\n",
       "      <th>arpu_3g_7</th>\n",
       "      <th>arpu_3g_8</th>\n",
       "      <th>arpu_3g_9</th>\n",
       "      <th>arpu_2g_6</th>\n",
       "      <th>arpu_2g_7</th>\n",
       "      <th>arpu_2g_8</th>\n",
       "      <th>arpu_2g_9</th>\n",
       "      <th>night_pck_user_6</th>\n",
       "      <th>night_pck_user_7</th>\n",
       "      <th>night_pck_user_8</th>\n",
       "      <th>night_pck_user_9</th>\n",
       "      <th>monthly_2g_6</th>\n",
       "      <th>monthly_2g_7</th>\n",
       "      <th>monthly_2g_8</th>\n",
       "      <th>monthly_2g_9</th>\n",
       "      <th>sachet_2g_6</th>\n",
       "      <th>sachet_2g_7</th>\n",
       "      <th>sachet_2g_8</th>\n",
       "      <th>sachet_2g_9</th>\n",
       "      <th>monthly_3g_6</th>\n",
       "      <th>monthly_3g_7</th>\n",
       "      <th>monthly_3g_8</th>\n",
       "      <th>monthly_3g_9</th>\n",
       "      <th>sachet_3g_6</th>\n",
       "      <th>sachet_3g_7</th>\n",
       "      <th>sachet_3g_8</th>\n",
       "      <th>sachet_3g_9</th>\n",
       "      <th>fb_user_6</th>\n",
       "      <th>fb_user_7</th>\n",
       "      <th>fb_user_8</th>\n",
       "      <th>fb_user_9</th>\n",
       "      <th>aon</th>\n",
       "      <th>aug_vbc_3g</th>\n",
       "      <th>jul_vbc_3g</th>\n",
       "      <th>jun_vbc_3g</th>\n",
       "      <th>sep_vbc_3g</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7000842753</td>\n",
       "      <td>109</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>197.385</td>\n",
       "      <td>214.816</td>\n",
       "      <td>213.803</td>\n",
       "      <td>21.100</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.13</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.15</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.44</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>5.44</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>362</td>\n",
       "      <td>252</td>\n",
       "      <td>252</td>\n",
       "      <td>0</td>\n",
       "      <td>252</td>\n",
       "      <td>252</td>\n",
       "      <td>252</td>\n",
       "      <td>0</td>\n",
       "      <td>6/21/2014</td>\n",
       "      <td>7/16/2014</td>\n",
       "      <td>8/8/2014</td>\n",
       "      <td>9/28/2014</td>\n",
       "      <td>252</td>\n",
       "      <td>252</td>\n",
       "      <td>252</td>\n",
       "      <td>0</td>\n",
       "      <td>6/21/2014</td>\n",
       "      <td>7/16/2014</td>\n",
       "      <td>8/8/2014</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>252.0</td>\n",
       "      <td>252.0</td>\n",
       "      <td>252.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>252.0</td>\n",
       "      <td>252.0</td>\n",
       "      <td>252.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>30.13</td>\n",
       "      <td>1.32</td>\n",
       "      <td>5.75</td>\n",
       "      <td>0.0</td>\n",
       "      <td>83.57</td>\n",
       "      <td>150.76</td>\n",
       "      <td>109.61</td>\n",
       "      <td>0.00</td>\n",
       "      <td>212.17</td>\n",
       "      <td>212.17</td>\n",
       "      <td>212.17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>212.17</td>\n",
       "      <td>212.17</td>\n",
       "      <td>212.17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>968</td>\n",
       "      <td>30.4</td>\n",
       "      <td>0.0</td>\n",
       "      <td>101.20</td>\n",
       "      <td>3.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7001865778</td>\n",
       "      <td>109</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>34.047</td>\n",
       "      <td>355.074</td>\n",
       "      <td>268.321</td>\n",
       "      <td>86.285</td>\n",
       "      <td>24.11</td>\n",
       "      <td>78.68</td>\n",
       "      <td>7.68</td>\n",
       "      <td>18.34</td>\n",
       "      <td>15.74</td>\n",
       "      <td>99.84</td>\n",
       "      <td>304.76</td>\n",
       "      <td>53.76</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>23.88</td>\n",
       "      <td>74.56</td>\n",
       "      <td>7.68</td>\n",
       "      <td>18.34</td>\n",
       "      <td>11.51</td>\n",
       "      <td>75.94</td>\n",
       "      <td>291.86</td>\n",
       "      <td>53.76</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.91</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>35.39</td>\n",
       "      <td>150.51</td>\n",
       "      <td>299.54</td>\n",
       "      <td>72.11</td>\n",
       "      <td>0.23</td>\n",
       "      <td>4.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.46</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.23</td>\n",
       "      <td>4.58</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.68</td>\n",
       "      <td>23.43</td>\n",
       "      <td>12.76</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>40.31</td>\n",
       "      <td>178.53</td>\n",
       "      <td>312.44</td>\n",
       "      <td>72.11</td>\n",
       "      <td>1.61</td>\n",
       "      <td>29.91</td>\n",
       "      <td>29.23</td>\n",
       "      <td>116.09</td>\n",
       "      <td>17.48</td>\n",
       "      <td>65.38</td>\n",
       "      <td>375.58</td>\n",
       "      <td>56.93</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8.93</td>\n",
       "      <td>3.61</td>\n",
       "      <td>0.00</td>\n",
       "      <td>19.09</td>\n",
       "      <td>104.23</td>\n",
       "      <td>408.43</td>\n",
       "      <td>173.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2.35</td>\n",
       "      <td>0.00</td>\n",
       "      <td>5.90</td>\n",
       "      <td>0.00</td>\n",
       "      <td>12.49</td>\n",
       "      <td>15.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.90</td>\n",
       "      <td>0.00</td>\n",
       "      <td>14.84</td>\n",
       "      <td>15.01</td>\n",
       "      <td>26.83</td>\n",
       "      <td>104.23</td>\n",
       "      <td>423.28</td>\n",
       "      <td>188.04</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.83</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>4</td>\n",
       "      <td>9</td>\n",
       "      <td>11</td>\n",
       "      <td>5</td>\n",
       "      <td>74</td>\n",
       "      <td>384</td>\n",
       "      <td>283</td>\n",
       "      <td>121</td>\n",
       "      <td>44</td>\n",
       "      <td>154</td>\n",
       "      <td>65</td>\n",
       "      <td>50</td>\n",
       "      <td>6/29/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/28/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>44</td>\n",
       "      <td>23</td>\n",
       "      <td>30</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7/25/2014</td>\n",
       "      <td>8/10/2014</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>154.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>154.0</td>\n",
       "      <td>50.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>108.07</td>\n",
       "      <td>365.47</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28.61</td>\n",
       "      <td>7.60</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1006</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7001625959</td>\n",
       "      <td>109</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>167.690</td>\n",
       "      <td>189.058</td>\n",
       "      <td>210.226</td>\n",
       "      <td>290.714</td>\n",
       "      <td>11.54</td>\n",
       "      <td>55.24</td>\n",
       "      <td>37.26</td>\n",
       "      <td>74.81</td>\n",
       "      <td>143.33</td>\n",
       "      <td>220.59</td>\n",
       "      <td>208.36</td>\n",
       "      <td>118.91</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>38.49</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>70.94</td>\n",
       "      <td>7.19</td>\n",
       "      <td>28.74</td>\n",
       "      <td>13.58</td>\n",
       "      <td>14.39</td>\n",
       "      <td>29.34</td>\n",
       "      <td>16.86</td>\n",
       "      <td>38.46</td>\n",
       "      <td>28.16</td>\n",
       "      <td>24.11</td>\n",
       "      <td>21.79</td>\n",
       "      <td>15.61</td>\n",
       "      <td>22.24</td>\n",
       "      <td>0.0</td>\n",
       "      <td>135.54</td>\n",
       "      <td>45.76</td>\n",
       "      <td>0.48</td>\n",
       "      <td>60.66</td>\n",
       "      <td>67.41</td>\n",
       "      <td>67.66</td>\n",
       "      <td>64.81</td>\n",
       "      <td>4.34</td>\n",
       "      <td>26.49</td>\n",
       "      <td>22.58</td>\n",
       "      <td>8.76</td>\n",
       "      <td>41.81</td>\n",
       "      <td>67.41</td>\n",
       "      <td>75.53</td>\n",
       "      <td>9.28</td>\n",
       "      <td>1.48</td>\n",
       "      <td>14.76</td>\n",
       "      <td>22.83</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>47.64</td>\n",
       "      <td>108.68</td>\n",
       "      <td>120.94</td>\n",
       "      <td>18.04</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>46.56</td>\n",
       "      <td>236.84</td>\n",
       "      <td>96.84</td>\n",
       "      <td>42.08</td>\n",
       "      <td>0.45</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>155.33</td>\n",
       "      <td>412.94</td>\n",
       "      <td>285.46</td>\n",
       "      <td>124.94</td>\n",
       "      <td>115.69</td>\n",
       "      <td>71.11</td>\n",
       "      <td>67.46</td>\n",
       "      <td>148.23</td>\n",
       "      <td>14.38</td>\n",
       "      <td>15.44</td>\n",
       "      <td>38.89</td>\n",
       "      <td>38.98</td>\n",
       "      <td>99.48</td>\n",
       "      <td>122.29</td>\n",
       "      <td>49.63</td>\n",
       "      <td>158.19</td>\n",
       "      <td>229.56</td>\n",
       "      <td>208.86</td>\n",
       "      <td>155.99</td>\n",
       "      <td>345.41</td>\n",
       "      <td>72.41</td>\n",
       "      <td>71.29</td>\n",
       "      <td>28.69</td>\n",
       "      <td>49.44</td>\n",
       "      <td>45.18</td>\n",
       "      <td>177.01</td>\n",
       "      <td>167.09</td>\n",
       "      <td>118.18</td>\n",
       "      <td>21.73</td>\n",
       "      <td>58.34</td>\n",
       "      <td>43.23</td>\n",
       "      <td>3.86</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>139.33</td>\n",
       "      <td>306.66</td>\n",
       "      <td>239.03</td>\n",
       "      <td>171.49</td>\n",
       "      <td>370.04</td>\n",
       "      <td>519.53</td>\n",
       "      <td>395.03</td>\n",
       "      <td>517.74</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.45</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.93</td>\n",
       "      <td>3.14</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.36</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>168</td>\n",
       "      <td>315</td>\n",
       "      <td>116</td>\n",
       "      <td>358</td>\n",
       "      <td>86</td>\n",
       "      <td>200</td>\n",
       "      <td>86</td>\n",
       "      <td>100</td>\n",
       "      <td>6/17/2014</td>\n",
       "      <td>7/24/2014</td>\n",
       "      <td>8/14/2014</td>\n",
       "      <td>9/29/2014</td>\n",
       "      <td>0</td>\n",
       "      <td>200</td>\n",
       "      <td>86</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9/17/2014</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>46.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>46.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>8.42</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.84</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1103</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.17</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mobile_number  circle_id  loc_og_t2o_mou  std_og_t2o_mou  loc_ic_t2o_mou  \\\n",
       "0     7000842753        109             0.0             0.0             0.0   \n",
       "1     7001865778        109             0.0             0.0             0.0   \n",
       "2     7001625959        109             0.0             0.0             0.0   \n",
       "\n",
       "  last_date_of_month_6 last_date_of_month_7 last_date_of_month_8  \\\n",
       "0            6/30/2014            7/31/2014            8/31/2014   \n",
       "1            6/30/2014            7/31/2014            8/31/2014   \n",
       "2            6/30/2014            7/31/2014            8/31/2014   \n",
       "\n",
       "  last_date_of_month_9   arpu_6   arpu_7   arpu_8   arpu_9  onnet_mou_6  \\\n",
       "0            9/30/2014  197.385  214.816  213.803   21.100          NaN   \n",
       "1            9/30/2014   34.047  355.074  268.321   86.285        24.11   \n",
       "2            9/30/2014  167.690  189.058  210.226  290.714        11.54   \n",
       "\n",
       "   onnet_mou_7  onnet_mou_8  onnet_mou_9  offnet_mou_6  offnet_mou_7  \\\n",
       "0          NaN         0.00          NaN           NaN           NaN   \n",
       "1        78.68         7.68        18.34         15.74         99.84   \n",
       "2        55.24        37.26        74.81        143.33        220.59   \n",
       "\n",
       "   offnet_mou_8  offnet_mou_9  roam_ic_mou_6  roam_ic_mou_7  roam_ic_mou_8  \\\n",
       "0          0.00           NaN            NaN            NaN            0.0   \n",
       "1        304.76         53.76            0.0            0.0            0.0   \n",
       "2        208.36        118.91            0.0            0.0            0.0   \n",
       "\n",
       "   roam_ic_mou_9  roam_og_mou_6  roam_og_mou_7  roam_og_mou_8  roam_og_mou_9  \\\n",
       "0            NaN            NaN            NaN            0.0            NaN   \n",
       "1           0.00            0.0            0.0            0.0           0.00   \n",
       "2          38.49            0.0            0.0            0.0          70.94   \n",
       "\n",
       "   loc_og_t2t_mou_6  loc_og_t2t_mou_7  loc_og_t2t_mou_8  loc_og_t2t_mou_9  \\\n",
       "0               NaN               NaN              0.00               NaN   \n",
       "1             23.88             74.56              7.68             18.34   \n",
       "2              7.19             28.74             13.58             14.39   \n",
       "\n",
       "   loc_og_t2m_mou_6  loc_og_t2m_mou_7  loc_og_t2m_mou_8  loc_og_t2m_mou_9  \\\n",
       "0               NaN               NaN              0.00               NaN   \n",
       "1             11.51             75.94            291.86             53.76   \n",
       "2             29.34             16.86             38.46             28.16   \n",
       "\n",
       "   loc_og_t2f_mou_6  loc_og_t2f_mou_7  loc_og_t2f_mou_8  loc_og_t2f_mou_9  \\\n",
       "0               NaN               NaN              0.00               NaN   \n",
       "1              0.00              0.00              0.00              0.00   \n",
       "2             24.11             21.79             15.61             22.24   \n",
       "\n",
       "   loc_og_t2c_mou_6  loc_og_t2c_mou_7  loc_og_t2c_mou_8  loc_og_t2c_mou_9  \\\n",
       "0               NaN               NaN              0.00               NaN   \n",
       "1               0.0              2.91              0.00              0.00   \n",
       "2               0.0            135.54             45.76              0.48   \n",
       "\n",
       "   loc_og_mou_6  loc_og_mou_7  loc_og_mou_8  loc_og_mou_9  std_og_t2t_mou_6  \\\n",
       "0           NaN           NaN          0.00           NaN               NaN   \n",
       "1         35.39        150.51        299.54         72.11              0.23   \n",
       "2         60.66         67.41         67.66         64.81              4.34   \n",
       "\n",
       "   std_og_t2t_mou_7  std_og_t2t_mou_8  std_og_t2t_mou_9  std_og_t2m_mou_6  \\\n",
       "0               NaN              0.00               NaN               NaN   \n",
       "1              4.11              0.00              0.00              0.00   \n",
       "2             26.49             22.58              8.76             41.81   \n",
       "\n",
       "   std_og_t2m_mou_7  std_og_t2m_mou_8  std_og_t2m_mou_9  std_og_t2f_mou_6  \\\n",
       "0               NaN              0.00               NaN               NaN   \n",
       "1              0.46              0.13              0.00              0.00   \n",
       "2             67.41             75.53              9.28              1.48   \n",
       "\n",
       "   std_og_t2f_mou_7  std_og_t2f_mou_8  std_og_t2f_mou_9  std_og_t2c_mou_6  \\\n",
       "0               NaN              0.00               NaN               NaN   \n",
       "1              0.00              0.00               0.0               0.0   \n",
       "2             14.76             22.83               0.0               0.0   \n",
       "\n",
       "   std_og_t2c_mou_7  std_og_t2c_mou_8  std_og_t2c_mou_9  std_og_mou_6  \\\n",
       "0               NaN               0.0               NaN           NaN   \n",
       "1               0.0               0.0               0.0          0.23   \n",
       "2               0.0               0.0               0.0         47.64   \n",
       "\n",
       "   std_og_mou_7  std_og_mou_8  std_og_mou_9  isd_og_mou_6  isd_og_mou_7  \\\n",
       "0           NaN          0.00           NaN           NaN           NaN   \n",
       "1          4.58          0.13          0.00           0.0           0.0   \n",
       "2        108.68        120.94         18.04           0.0           0.0   \n",
       "\n",
       "   isd_og_mou_8  isd_og_mou_9  spl_og_mou_6  spl_og_mou_7  spl_og_mou_8  \\\n",
       "0           0.0           NaN           NaN           NaN          0.00   \n",
       "1           0.0           0.0          4.68         23.43         12.76   \n",
       "2           0.0           0.0         46.56        236.84         96.84   \n",
       "\n",
       "   spl_og_mou_9  og_others_6  og_others_7  og_others_8  og_others_9  \\\n",
       "0           NaN          NaN          NaN          0.0          NaN   \n",
       "1          0.00         0.00          0.0          0.0          0.0   \n",
       "2         42.08         0.45          0.0          0.0          0.0   \n",
       "\n",
       "   total_og_mou_6  total_og_mou_7  total_og_mou_8  total_og_mou_9  \\\n",
       "0            0.00            0.00            0.00            0.00   \n",
       "1           40.31          178.53          312.44           72.11   \n",
       "2          155.33          412.94          285.46          124.94   \n",
       "\n",
       "   loc_ic_t2t_mou_6  loc_ic_t2t_mou_7  loc_ic_t2t_mou_8  loc_ic_t2t_mou_9  \\\n",
       "0               NaN               NaN              0.16               NaN   \n",
       "1              1.61             29.91             29.23            116.09   \n",
       "2            115.69             71.11             67.46            148.23   \n",
       "\n",
       "   loc_ic_t2m_mou_6  loc_ic_t2m_mou_7  loc_ic_t2m_mou_8  loc_ic_t2m_mou_9  \\\n",
       "0               NaN               NaN              4.13               NaN   \n",
       "1             17.48             65.38            375.58             56.93   \n",
       "2             14.38             15.44             38.89             38.98   \n",
       "\n",
       "   loc_ic_t2f_mou_6  loc_ic_t2f_mou_7  loc_ic_t2f_mou_8  loc_ic_t2f_mou_9  \\\n",
       "0               NaN               NaN              1.15               NaN   \n",
       "1              0.00              8.93              3.61              0.00   \n",
       "2             99.48            122.29             49.63            158.19   \n",
       "\n",
       "   loc_ic_mou_6  loc_ic_mou_7  loc_ic_mou_8  loc_ic_mou_9  std_ic_t2t_mou_6  \\\n",
       "0           NaN           NaN          5.44           NaN               NaN   \n",
       "1         19.09        104.23        408.43        173.03              0.00   \n",
       "2        229.56        208.86        155.99        345.41             72.41   \n",
       "\n",
       "   std_ic_t2t_mou_7  std_ic_t2t_mou_8  std_ic_t2t_mou_9  std_ic_t2m_mou_6  \\\n",
       "0               NaN              0.00               NaN               NaN   \n",
       "1              0.00              2.35              0.00              5.90   \n",
       "2             71.29             28.69             49.44             45.18   \n",
       "\n",
       "   std_ic_t2m_mou_7  std_ic_t2m_mou_8  std_ic_t2m_mou_9  std_ic_t2f_mou_6  \\\n",
       "0               NaN              0.00               NaN               NaN   \n",
       "1              0.00             12.49             15.01              0.00   \n",
       "2            177.01            167.09            118.18             21.73   \n",
       "\n",
       "   std_ic_t2f_mou_7  std_ic_t2f_mou_8  std_ic_t2f_mou_9  std_ic_t2o_mou_6  \\\n",
       "0               NaN              0.00               NaN               NaN   \n",
       "1              0.00              0.00              0.00               0.0   \n",
       "2             58.34             43.23              3.86               0.0   \n",
       "\n",
       "   std_ic_t2o_mou_7  std_ic_t2o_mou_8  std_ic_t2o_mou_9  std_ic_mou_6  \\\n",
       "0               NaN               0.0               NaN           NaN   \n",
       "1               0.0               0.0               0.0          5.90   \n",
       "2               0.0               0.0               0.0        139.33   \n",
       "\n",
       "   std_ic_mou_7  std_ic_mou_8  std_ic_mou_9  total_ic_mou_6  total_ic_mou_7  \\\n",
       "0           NaN          0.00           NaN            0.00            0.00   \n",
       "1          0.00         14.84         15.01           26.83          104.23   \n",
       "2        306.66        239.03        171.49          370.04          519.53   \n",
       "\n",
       "   total_ic_mou_8  total_ic_mou_9  spl_ic_mou_6  spl_ic_mou_7  spl_ic_mou_8  \\\n",
       "0            5.44            0.00           NaN           NaN           0.0   \n",
       "1          423.28          188.04          0.00           0.0           0.0   \n",
       "2          395.03          517.74          0.21           0.0           0.0   \n",
       "\n",
       "   spl_ic_mou_9  isd_ic_mou_6  isd_ic_mou_7  isd_ic_mou_8  isd_ic_mou_9  \\\n",
       "0           NaN           NaN           NaN           0.0           NaN   \n",
       "1          0.00          1.83          0.00           0.0          0.00   \n",
       "2          0.45          0.00          0.85           0.0          0.01   \n",
       "\n",
       "   ic_others_6  ic_others_7  ic_others_8  ic_others_9  total_rech_num_6  \\\n",
       "0          NaN          NaN          0.0          NaN                 4   \n",
       "1         0.00         0.00          0.0         0.00                 4   \n",
       "2         0.93         3.14          0.0         0.36                 5   \n",
       "\n",
       "   total_rech_num_7  total_rech_num_8  total_rech_num_9  total_rech_amt_6  \\\n",
       "0                 3                 2                 6               362   \n",
       "1                 9                11                 5                74   \n",
       "2                 4                 2                 7               168   \n",
       "\n",
       "   total_rech_amt_7  total_rech_amt_8  total_rech_amt_9  max_rech_amt_6  \\\n",
       "0               252               252                 0             252   \n",
       "1               384               283               121              44   \n",
       "2               315               116               358              86   \n",
       "\n",
       "   max_rech_amt_7  max_rech_amt_8  max_rech_amt_9 date_of_last_rech_6  \\\n",
       "0             252             252               0           6/21/2014   \n",
       "1             154              65              50           6/29/2014   \n",
       "2             200              86             100           6/17/2014   \n",
       "\n",
       "  date_of_last_rech_7 date_of_last_rech_8 date_of_last_rech_9  \\\n",
       "0           7/16/2014            8/8/2014           9/28/2014   \n",
       "1           7/31/2014           8/28/2014           9/30/2014   \n",
       "2           7/24/2014           8/14/2014           9/29/2014   \n",
       "\n",
       "   last_day_rch_amt_6  last_day_rch_amt_7  last_day_rch_amt_8  \\\n",
       "0                 252                 252                 252   \n",
       "1                  44                  23                  30   \n",
       "2                   0                 200                  86   \n",
       "\n",
       "   last_day_rch_amt_9 date_of_last_rech_data_6 date_of_last_rech_data_7  \\\n",
       "0                   0                6/21/2014                7/16/2014   \n",
       "1                   0                      NaN                7/25/2014   \n",
       "2                   0                      NaN                      NaN   \n",
       "\n",
       "  date_of_last_rech_data_8 date_of_last_rech_data_9  total_rech_data_6  \\\n",
       "0                 8/8/2014                      NaN                1.0   \n",
       "1                8/10/2014                      NaN                NaN   \n",
       "2                      NaN                9/17/2014                NaN   \n",
       "\n",
       "   total_rech_data_7  total_rech_data_8  total_rech_data_9  max_rech_data_6  \\\n",
       "0                1.0                1.0                NaN            252.0   \n",
       "1                1.0                2.0                NaN              NaN   \n",
       "2                NaN                NaN                1.0              NaN   \n",
       "\n",
       "   max_rech_data_7  max_rech_data_8  max_rech_data_9  count_rech_2g_6  \\\n",
       "0            252.0            252.0              NaN              0.0   \n",
       "1            154.0             25.0              NaN              NaN   \n",
       "2              NaN              NaN             46.0              NaN   \n",
       "\n",
       "   count_rech_2g_7  count_rech_2g_8  count_rech_2g_9  count_rech_3g_6  \\\n",
       "0              0.0              0.0              NaN              1.0   \n",
       "1              1.0              2.0              NaN              NaN   \n",
       "2              NaN              NaN              1.0              NaN   \n",
       "\n",
       "   count_rech_3g_7  count_rech_3g_8  count_rech_3g_9  av_rech_amt_data_6  \\\n",
       "0              1.0              1.0              NaN               252.0   \n",
       "1              0.0              0.0              NaN                 NaN   \n",
       "2              NaN              NaN              0.0                 NaN   \n",
       "\n",
       "   av_rech_amt_data_7  av_rech_amt_data_8  av_rech_amt_data_9  vol_2g_mb_6  \\\n",
       "0               252.0               252.0                 NaN        30.13   \n",
       "1               154.0                50.0                 NaN         0.00   \n",
       "2                 NaN                 NaN                46.0         0.00   \n",
       "\n",
       "   vol_2g_mb_7  vol_2g_mb_8  vol_2g_mb_9  vol_3g_mb_6  vol_3g_mb_7  \\\n",
       "0         1.32         5.75          0.0        83.57       150.76   \n",
       "1       108.07       365.47          0.0         0.00         0.00   \n",
       "2         0.00         0.00          0.0         0.00         0.00   \n",
       "\n",
       "   vol_3g_mb_8  vol_3g_mb_9  arpu_3g_6  arpu_3g_7  arpu_3g_8  arpu_3g_9  \\\n",
       "0       109.61         0.00     212.17     212.17     212.17        NaN   \n",
       "1         0.00         0.00        NaN       0.00       0.00        NaN   \n",
       "2         0.00         8.42        NaN        NaN        NaN       2.84   \n",
       "\n",
       "   arpu_2g_6  arpu_2g_7  arpu_2g_8  arpu_2g_9  night_pck_user_6  \\\n",
       "0     212.17     212.17     212.17        NaN               0.0   \n",
       "1        NaN      28.61       7.60        NaN               NaN   \n",
       "2        NaN        NaN        NaN        0.0               NaN   \n",
       "\n",
       "   night_pck_user_7  night_pck_user_8  night_pck_user_9  monthly_2g_6  \\\n",
       "0               0.0               0.0               NaN             0   \n",
       "1               0.0               0.0               NaN             0   \n",
       "2               NaN               NaN               0.0             0   \n",
       "\n",
       "   monthly_2g_7  monthly_2g_8  monthly_2g_9  sachet_2g_6  sachet_2g_7  \\\n",
       "0             0             0             0            0            0   \n",
       "1             1             0             0            0            0   \n",
       "2             0             0             0            0            0   \n",
       "\n",
       "   sachet_2g_8  sachet_2g_9  monthly_3g_6  monthly_3g_7  monthly_3g_8  \\\n",
       "0            0            0             1             1             1   \n",
       "1            2            0             0             0             0   \n",
       "2            0            1             0             0             0   \n",
       "\n",
       "   monthly_3g_9  sachet_3g_6  sachet_3g_7  sachet_3g_8  sachet_3g_9  \\\n",
       "0             0            0            0            0            0   \n",
       "1             0            0            0            0            0   \n",
       "2             0            0            0            0            0   \n",
       "\n",
       "   fb_user_6  fb_user_7  fb_user_8  fb_user_9   aon  aug_vbc_3g  jul_vbc_3g  \\\n",
       "0        1.0        1.0        1.0        NaN   968        30.4         0.0   \n",
       "1        NaN        1.0        1.0        NaN  1006         0.0         0.0   \n",
       "2        NaN        NaN        NaN        1.0  1103         0.0         0.0   \n",
       "\n",
       "   jun_vbc_3g  sep_vbc_3g  \n",
       "0      101.20        3.58  \n",
       "1        0.00        0.00  \n",
       "2        4.17        0.00  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# look at initial rows of the data\n",
    "churn.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 99999 entries, 0 to 99998\n",
      "Columns: 226 entries, mobile_number to sep_vbc_3g\n",
      "dtypes: float64(179), int64(35), object(12)\n",
      "memory usage: 172.4+ MB\n"
     ]
    }
   ],
   "source": [
    "# feature type summary\n",
    "churn.info(verbose=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are 99999 rows and 226 columns in the data. Lot of the columns are numeric type, but we need to inspect which are the categorical columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mobile_number</th>\n",
       "      <th>circle_id</th>\n",
       "      <th>loc_og_t2o_mou</th>\n",
       "      <th>std_og_t2o_mou</th>\n",
       "      <th>loc_ic_t2o_mou</th>\n",
       "      <th>last_date_of_month_6</th>\n",
       "      <th>last_date_of_month_7</th>\n",
       "      <th>last_date_of_month_8</th>\n",
       "      <th>last_date_of_month_9</th>\n",
       "      <th>arpu_6</th>\n",
       "      <th>arpu_7</th>\n",
       "      <th>arpu_8</th>\n",
       "      <th>arpu_9</th>\n",
       "      <th>onnet_mou_6</th>\n",
       "      <th>onnet_mou_7</th>\n",
       "      <th>onnet_mou_8</th>\n",
       "      <th>onnet_mou_9</th>\n",
       "      <th>offnet_mou_6</th>\n",
       "      <th>offnet_mou_7</th>\n",
       "      <th>offnet_mou_8</th>\n",
       "      <th>offnet_mou_9</th>\n",
       "      <th>roam_ic_mou_6</th>\n",
       "      <th>roam_ic_mou_7</th>\n",
       "      <th>roam_ic_mou_8</th>\n",
       "      <th>roam_ic_mou_9</th>\n",
       "      <th>roam_og_mou_6</th>\n",
       "      <th>roam_og_mou_7</th>\n",
       "      <th>roam_og_mou_8</th>\n",
       "      <th>roam_og_mou_9</th>\n",
       "      <th>loc_og_t2t_mou_6</th>\n",
       "      <th>loc_og_t2t_mou_7</th>\n",
       "      <th>loc_og_t2t_mou_8</th>\n",
       "      <th>loc_og_t2t_mou_9</th>\n",
       "      <th>loc_og_t2m_mou_6</th>\n",
       "      <th>loc_og_t2m_mou_7</th>\n",
       "      <th>loc_og_t2m_mou_8</th>\n",
       "      <th>loc_og_t2m_mou_9</th>\n",
       "      <th>loc_og_t2f_mou_6</th>\n",
       "      <th>loc_og_t2f_mou_7</th>\n",
       "      <th>loc_og_t2f_mou_8</th>\n",
       "      <th>loc_og_t2f_mou_9</th>\n",
       "      <th>loc_og_t2c_mou_6</th>\n",
       "      <th>loc_og_t2c_mou_7</th>\n",
       "      <th>loc_og_t2c_mou_8</th>\n",
       "      <th>loc_og_t2c_mou_9</th>\n",
       "      <th>loc_og_mou_6</th>\n",
       "      <th>loc_og_mou_7</th>\n",
       "      <th>loc_og_mou_8</th>\n",
       "      <th>loc_og_mou_9</th>\n",
       "      <th>std_og_t2t_mou_6</th>\n",
       "      <th>std_og_t2t_mou_7</th>\n",
       "      <th>std_og_t2t_mou_8</th>\n",
       "      <th>std_og_t2t_mou_9</th>\n",
       "      <th>std_og_t2m_mou_6</th>\n",
       "      <th>std_og_t2m_mou_7</th>\n",
       "      <th>std_og_t2m_mou_8</th>\n",
       "      <th>std_og_t2m_mou_9</th>\n",
       "      <th>std_og_t2f_mou_6</th>\n",
       "      <th>std_og_t2f_mou_7</th>\n",
       "      <th>std_og_t2f_mou_8</th>\n",
       "      <th>std_og_t2f_mou_9</th>\n",
       "      <th>std_og_t2c_mou_6</th>\n",
       "      <th>std_og_t2c_mou_7</th>\n",
       "      <th>std_og_t2c_mou_8</th>\n",
       "      <th>std_og_t2c_mou_9</th>\n",
       "      <th>std_og_mou_6</th>\n",
       "      <th>std_og_mou_7</th>\n",
       "      <th>std_og_mou_8</th>\n",
       "      <th>std_og_mou_9</th>\n",
       "      <th>isd_og_mou_6</th>\n",
       "      <th>isd_og_mou_7</th>\n",
       "      <th>isd_og_mou_8</th>\n",
       "      <th>isd_og_mou_9</th>\n",
       "      <th>spl_og_mou_6</th>\n",
       "      <th>spl_og_mou_7</th>\n",
       "      <th>spl_og_mou_8</th>\n",
       "      <th>spl_og_mou_9</th>\n",
       "      <th>og_others_6</th>\n",
       "      <th>og_others_7</th>\n",
       "      <th>og_others_8</th>\n",
       "      <th>og_others_9</th>\n",
       "      <th>total_og_mou_6</th>\n",
       "      <th>total_og_mou_7</th>\n",
       "      <th>total_og_mou_8</th>\n",
       "      <th>total_og_mou_9</th>\n",
       "      <th>loc_ic_t2t_mou_6</th>\n",
       "      <th>loc_ic_t2t_mou_7</th>\n",
       "      <th>loc_ic_t2t_mou_8</th>\n",
       "      <th>loc_ic_t2t_mou_9</th>\n",
       "      <th>loc_ic_t2m_mou_6</th>\n",
       "      <th>loc_ic_t2m_mou_7</th>\n",
       "      <th>loc_ic_t2m_mou_8</th>\n",
       "      <th>loc_ic_t2m_mou_9</th>\n",
       "      <th>loc_ic_t2f_mou_6</th>\n",
       "      <th>loc_ic_t2f_mou_7</th>\n",
       "      <th>loc_ic_t2f_mou_8</th>\n",
       "      <th>loc_ic_t2f_mou_9</th>\n",
       "      <th>loc_ic_mou_6</th>\n",
       "      <th>loc_ic_mou_7</th>\n",
       "      <th>loc_ic_mou_8</th>\n",
       "      <th>loc_ic_mou_9</th>\n",
       "      <th>std_ic_t2t_mou_6</th>\n",
       "      <th>std_ic_t2t_mou_7</th>\n",
       "      <th>std_ic_t2t_mou_8</th>\n",
       "      <th>std_ic_t2t_mou_9</th>\n",
       "      <th>std_ic_t2m_mou_6</th>\n",
       "      <th>std_ic_t2m_mou_7</th>\n",
       "      <th>std_ic_t2m_mou_8</th>\n",
       "      <th>std_ic_t2m_mou_9</th>\n",
       "      <th>std_ic_t2f_mou_6</th>\n",
       "      <th>std_ic_t2f_mou_7</th>\n",
       "      <th>std_ic_t2f_mou_8</th>\n",
       "      <th>std_ic_t2f_mou_9</th>\n",
       "      <th>std_ic_t2o_mou_6</th>\n",
       "      <th>std_ic_t2o_mou_7</th>\n",
       "      <th>std_ic_t2o_mou_8</th>\n",
       "      <th>std_ic_t2o_mou_9</th>\n",
       "      <th>std_ic_mou_6</th>\n",
       "      <th>std_ic_mou_7</th>\n",
       "      <th>std_ic_mou_8</th>\n",
       "      <th>std_ic_mou_9</th>\n",
       "      <th>total_ic_mou_6</th>\n",
       "      <th>total_ic_mou_7</th>\n",
       "      <th>total_ic_mou_8</th>\n",
       "      <th>total_ic_mou_9</th>\n",
       "      <th>spl_ic_mou_6</th>\n",
       "      <th>spl_ic_mou_7</th>\n",
       "      <th>spl_ic_mou_8</th>\n",
       "      <th>spl_ic_mou_9</th>\n",
       "      <th>isd_ic_mou_6</th>\n",
       "      <th>isd_ic_mou_7</th>\n",
       "      <th>isd_ic_mou_8</th>\n",
       "      <th>isd_ic_mou_9</th>\n",
       "      <th>ic_others_6</th>\n",
       "      <th>ic_others_7</th>\n",
       "      <th>ic_others_8</th>\n",
       "      <th>ic_others_9</th>\n",
       "      <th>total_rech_num_6</th>\n",
       "      <th>total_rech_num_7</th>\n",
       "      <th>total_rech_num_8</th>\n",
       "      <th>total_rech_num_9</th>\n",
       "      <th>total_rech_amt_6</th>\n",
       "      <th>total_rech_amt_7</th>\n",
       "      <th>total_rech_amt_8</th>\n",
       "      <th>total_rech_amt_9</th>\n",
       "      <th>max_rech_amt_6</th>\n",
       "      <th>max_rech_amt_7</th>\n",
       "      <th>max_rech_amt_8</th>\n",
       "      <th>max_rech_amt_9</th>\n",
       "      <th>date_of_last_rech_6</th>\n",
       "      <th>date_of_last_rech_7</th>\n",
       "      <th>date_of_last_rech_8</th>\n",
       "      <th>date_of_last_rech_9</th>\n",
       "      <th>last_day_rch_amt_6</th>\n",
       "      <th>last_day_rch_amt_7</th>\n",
       "      <th>last_day_rch_amt_8</th>\n",
       "      <th>last_day_rch_amt_9</th>\n",
       "      <th>date_of_last_rech_data_6</th>\n",
       "      <th>date_of_last_rech_data_7</th>\n",
       "      <th>date_of_last_rech_data_8</th>\n",
       "      <th>date_of_last_rech_data_9</th>\n",
       "      <th>total_rech_data_6</th>\n",
       "      <th>total_rech_data_7</th>\n",
       "      <th>total_rech_data_8</th>\n",
       "      <th>total_rech_data_9</th>\n",
       "      <th>max_rech_data_6</th>\n",
       "      <th>max_rech_data_7</th>\n",
       "      <th>max_rech_data_8</th>\n",
       "      <th>max_rech_data_9</th>\n",
       "      <th>count_rech_2g_6</th>\n",
       "      <th>count_rech_2g_7</th>\n",
       "      <th>count_rech_2g_8</th>\n",
       "      <th>count_rech_2g_9</th>\n",
       "      <th>count_rech_3g_6</th>\n",
       "      <th>count_rech_3g_7</th>\n",
       "      <th>count_rech_3g_8</th>\n",
       "      <th>count_rech_3g_9</th>\n",
       "      <th>av_rech_amt_data_6</th>\n",
       "      <th>av_rech_amt_data_7</th>\n",
       "      <th>av_rech_amt_data_8</th>\n",
       "      <th>av_rech_amt_data_9</th>\n",
       "      <th>vol_2g_mb_6</th>\n",
       "      <th>vol_2g_mb_7</th>\n",
       "      <th>vol_2g_mb_8</th>\n",
       "      <th>vol_2g_mb_9</th>\n",
       "      <th>vol_3g_mb_6</th>\n",
       "      <th>vol_3g_mb_7</th>\n",
       "      <th>vol_3g_mb_8</th>\n",
       "      <th>vol_3g_mb_9</th>\n",
       "      <th>arpu_3g_6</th>\n",
       "      <th>arpu_3g_7</th>\n",
       "      <th>arpu_3g_8</th>\n",
       "      <th>arpu_3g_9</th>\n",
       "      <th>arpu_2g_6</th>\n",
       "      <th>arpu_2g_7</th>\n",
       "      <th>arpu_2g_8</th>\n",
       "      <th>arpu_2g_9</th>\n",
       "      <th>night_pck_user_6</th>\n",
       "      <th>night_pck_user_7</th>\n",
       "      <th>night_pck_user_8</th>\n",
       "      <th>night_pck_user_9</th>\n",
       "      <th>monthly_2g_6</th>\n",
       "      <th>monthly_2g_7</th>\n",
       "      <th>monthly_2g_8</th>\n",
       "      <th>monthly_2g_9</th>\n",
       "      <th>sachet_2g_6</th>\n",
       "      <th>sachet_2g_7</th>\n",
       "      <th>sachet_2g_8</th>\n",
       "      <th>sachet_2g_9</th>\n",
       "      <th>monthly_3g_6</th>\n",
       "      <th>monthly_3g_7</th>\n",
       "      <th>monthly_3g_8</th>\n",
       "      <th>monthly_3g_9</th>\n",
       "      <th>sachet_3g_6</th>\n",
       "      <th>sachet_3g_7</th>\n",
       "      <th>sachet_3g_8</th>\n",
       "      <th>sachet_3g_9</th>\n",
       "      <th>fb_user_6</th>\n",
       "      <th>fb_user_7</th>\n",
       "      <th>fb_user_8</th>\n",
       "      <th>fb_user_9</th>\n",
       "      <th>aon</th>\n",
       "      <th>aug_vbc_3g</th>\n",
       "      <th>jul_vbc_3g</th>\n",
       "      <th>jun_vbc_3g</th>\n",
       "      <th>sep_vbc_3g</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>9.999900e+04</td>\n",
       "      <td>99999.0</td>\n",
       "      <td>98981.0</td>\n",
       "      <td>98981.0</td>\n",
       "      <td>98981.0</td>\n",
       "      <td>99999</td>\n",
       "      <td>99398</td>\n",
       "      <td>98899</td>\n",
       "      <td>98340</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.0</td>\n",
       "      <td>96140.0</td>\n",
       "      <td>94621.0</td>\n",
       "      <td>92254.0</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.0</td>\n",
       "      <td>96140.0</td>\n",
       "      <td>94621.0</td>\n",
       "      <td>92254.0</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>96062.000000</td>\n",
       "      <td>96140.000000</td>\n",
       "      <td>94621.000000</td>\n",
       "      <td>92254.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>98392</td>\n",
       "      <td>98232</td>\n",
       "      <td>96377</td>\n",
       "      <td>95239</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>25153</td>\n",
       "      <td>25571</td>\n",
       "      <td>26339</td>\n",
       "      <td>25922</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.00000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "      <td>99999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>30</td>\n",
       "      <td>31</td>\n",
       "      <td>31</td>\n",
       "      <td>30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>30</td>\n",
       "      <td>31</td>\n",
       "      <td>31</td>\n",
       "      <td>30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>9/29/2014</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>9/29/2014</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>99999</td>\n",
       "      <td>99398</td>\n",
       "      <td>98899</td>\n",
       "      <td>98340</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16960</td>\n",
       "      <td>17288</td>\n",
       "      <td>14706</td>\n",
       "      <td>22623</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1888</td>\n",
       "      <td>1813</td>\n",
       "      <td>1998</td>\n",
       "      <td>2329</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>7.001207e+09</td>\n",
       "      <td>109.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>282.987358</td>\n",
       "      <td>278.536648</td>\n",
       "      <td>279.154731</td>\n",
       "      <td>261.645069</td>\n",
       "      <td>132.395875</td>\n",
       "      <td>133.670805</td>\n",
       "      <td>133.018098</td>\n",
       "      <td>130.302327</td>\n",
       "      <td>197.935577</td>\n",
       "      <td>197.045133</td>\n",
       "      <td>196.574803</td>\n",
       "      <td>190.337222</td>\n",
       "      <td>9.950013</td>\n",
       "      <td>7.149898</td>\n",
       "      <td>7.292981</td>\n",
       "      <td>6.343841</td>\n",
       "      <td>13.911337</td>\n",
       "      <td>9.818732</td>\n",
       "      <td>9.971890</td>\n",
       "      <td>8.555519</td>\n",
       "      <td>47.100763</td>\n",
       "      <td>46.473010</td>\n",
       "      <td>45.887806</td>\n",
       "      <td>44.584446</td>\n",
       "      <td>93.342088</td>\n",
       "      <td>91.397131</td>\n",
       "      <td>91.755128</td>\n",
       "      <td>90.463192</td>\n",
       "      <td>3.751013</td>\n",
       "      <td>3.792985</td>\n",
       "      <td>3.677991</td>\n",
       "      <td>3.655123</td>\n",
       "      <td>1.123056</td>\n",
       "      <td>1.368500</td>\n",
       "      <td>1.433821</td>\n",
       "      <td>1.232726</td>\n",
       "      <td>144.201175</td>\n",
       "      <td>141.670476</td>\n",
       "      <td>141.328209</td>\n",
       "      <td>138.709970</td>\n",
       "      <td>79.829870</td>\n",
       "      <td>83.299598</td>\n",
       "      <td>83.282673</td>\n",
       "      <td>82.342919</td>\n",
       "      <td>87.299624</td>\n",
       "      <td>90.804137</td>\n",
       "      <td>89.838390</td>\n",
       "      <td>86.276622</td>\n",
       "      <td>1.129011</td>\n",
       "      <td>1.115010</td>\n",
       "      <td>1.067792</td>\n",
       "      <td>1.042362</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>168.261218</td>\n",
       "      <td>175.221436</td>\n",
       "      <td>174.191498</td>\n",
       "      <td>169.664466</td>\n",
       "      <td>0.798277</td>\n",
       "      <td>0.776572</td>\n",
       "      <td>0.791247</td>\n",
       "      <td>0.723892</td>\n",
       "      <td>3.916811</td>\n",
       "      <td>4.978279</td>\n",
       "      <td>5.053769</td>\n",
       "      <td>4.412767</td>\n",
       "      <td>0.454157</td>\n",
       "      <td>0.030235</td>\n",
       "      <td>0.033372</td>\n",
       "      <td>0.047456</td>\n",
       "      <td>305.133424</td>\n",
       "      <td>310.231175</td>\n",
       "      <td>304.119513</td>\n",
       "      <td>289.279198</td>\n",
       "      <td>47.922365</td>\n",
       "      <td>47.990520</td>\n",
       "      <td>47.211362</td>\n",
       "      <td>46.281794</td>\n",
       "      <td>107.475650</td>\n",
       "      <td>107.120493</td>\n",
       "      <td>108.460515</td>\n",
       "      <td>106.155471</td>\n",
       "      <td>12.084305</td>\n",
       "      <td>12.599697</td>\n",
       "      <td>11.751834</td>\n",
       "      <td>12.173105</td>\n",
       "      <td>167.491059</td>\n",
       "      <td>167.719540</td>\n",
       "      <td>167.432575</td>\n",
       "      <td>164.619293</td>\n",
       "      <td>9.575993</td>\n",
       "      <td>10.011904</td>\n",
       "      <td>9.883921</td>\n",
       "      <td>9.432479</td>\n",
       "      <td>20.722240</td>\n",
       "      <td>21.656415</td>\n",
       "      <td>21.183211</td>\n",
       "      <td>19.620913</td>\n",
       "      <td>2.156397</td>\n",
       "      <td>2.216923</td>\n",
       "      <td>2.085004</td>\n",
       "      <td>2.173419</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>32.457179</td>\n",
       "      <td>33.887833</td>\n",
       "      <td>33.154735</td>\n",
       "      <td>31.229344</td>\n",
       "      <td>200.130037</td>\n",
       "      <td>202.853055</td>\n",
       "      <td>198.750783</td>\n",
       "      <td>189.214260</td>\n",
       "      <td>0.061557</td>\n",
       "      <td>0.033585</td>\n",
       "      <td>0.040361</td>\n",
       "      <td>0.163137</td>\n",
       "      <td>7.460608</td>\n",
       "      <td>8.334936</td>\n",
       "      <td>8.442001</td>\n",
       "      <td>8.063003</td>\n",
       "      <td>0.854656</td>\n",
       "      <td>1.012960</td>\n",
       "      <td>0.970800</td>\n",
       "      <td>1.017162</td>\n",
       "      <td>7.558806</td>\n",
       "      <td>7.700367</td>\n",
       "      <td>7.212912</td>\n",
       "      <td>6.893019</td>\n",
       "      <td>327.514615</td>\n",
       "      <td>322.962970</td>\n",
       "      <td>324.157122</td>\n",
       "      <td>303.345673</td>\n",
       "      <td>104.637486</td>\n",
       "      <td>104.752398</td>\n",
       "      <td>107.728207</td>\n",
       "      <td>101.943889</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>63.156252</td>\n",
       "      <td>59.385804</td>\n",
       "      <td>62.641716</td>\n",
       "      <td>43.901249</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.463802</td>\n",
       "      <td>2.666419</td>\n",
       "      <td>2.651999</td>\n",
       "      <td>2.441170</td>\n",
       "      <td>126.393392</td>\n",
       "      <td>126.729459</td>\n",
       "      <td>125.717301</td>\n",
       "      <td>124.94144</td>\n",
       "      <td>1.864668</td>\n",
       "      <td>2.044699</td>\n",
       "      <td>2.016288</td>\n",
       "      <td>1.781807</td>\n",
       "      <td>0.599133</td>\n",
       "      <td>0.621720</td>\n",
       "      <td>0.635711</td>\n",
       "      <td>0.659363</td>\n",
       "      <td>192.600982</td>\n",
       "      <td>200.981292</td>\n",
       "      <td>197.526489</td>\n",
       "      <td>192.734315</td>\n",
       "      <td>51.904956</td>\n",
       "      <td>51.229937</td>\n",
       "      <td>50.170154</td>\n",
       "      <td>44.719701</td>\n",
       "      <td>121.396219</td>\n",
       "      <td>128.995847</td>\n",
       "      <td>135.410689</td>\n",
       "      <td>136.056613</td>\n",
       "      <td>89.555057</td>\n",
       "      <td>89.384120</td>\n",
       "      <td>91.173849</td>\n",
       "      <td>100.264116</td>\n",
       "      <td>86.398003</td>\n",
       "      <td>85.914450</td>\n",
       "      <td>86.599478</td>\n",
       "      <td>93.712026</td>\n",
       "      <td>0.025086</td>\n",
       "      <td>0.023034</td>\n",
       "      <td>0.020844</td>\n",
       "      <td>0.015971</td>\n",
       "      <td>0.079641</td>\n",
       "      <td>0.083221</td>\n",
       "      <td>0.081001</td>\n",
       "      <td>0.068781</td>\n",
       "      <td>0.389384</td>\n",
       "      <td>0.439634</td>\n",
       "      <td>0.450075</td>\n",
       "      <td>0.393104</td>\n",
       "      <td>0.075921</td>\n",
       "      <td>0.078581</td>\n",
       "      <td>0.082941</td>\n",
       "      <td>0.086341</td>\n",
       "      <td>0.074781</td>\n",
       "      <td>0.080401</td>\n",
       "      <td>0.084501</td>\n",
       "      <td>0.084581</td>\n",
       "      <td>0.914404</td>\n",
       "      <td>0.908764</td>\n",
       "      <td>0.890808</td>\n",
       "      <td>0.860968</td>\n",
       "      <td>1219.854749</td>\n",
       "      <td>68.170248</td>\n",
       "      <td>66.839062</td>\n",
       "      <td>60.021204</td>\n",
       "      <td>3.299373</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.956694e+05</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>328.439770</td>\n",
       "      <td>338.156291</td>\n",
       "      <td>344.474791</td>\n",
       "      <td>341.998630</td>\n",
       "      <td>297.207406</td>\n",
       "      <td>308.794148</td>\n",
       "      <td>308.951589</td>\n",
       "      <td>308.477668</td>\n",
       "      <td>316.851613</td>\n",
       "      <td>325.862803</td>\n",
       "      <td>327.170662</td>\n",
       "      <td>319.396092</td>\n",
       "      <td>72.825411</td>\n",
       "      <td>73.447948</td>\n",
       "      <td>68.402466</td>\n",
       "      <td>57.137537</td>\n",
       "      <td>71.443196</td>\n",
       "      <td>58.455762</td>\n",
       "      <td>64.713221</td>\n",
       "      <td>58.438186</td>\n",
       "      <td>150.856393</td>\n",
       "      <td>155.318705</td>\n",
       "      <td>151.184830</td>\n",
       "      <td>147.995390</td>\n",
       "      <td>162.780544</td>\n",
       "      <td>157.492308</td>\n",
       "      <td>156.537048</td>\n",
       "      <td>158.681454</td>\n",
       "      <td>14.230438</td>\n",
       "      <td>14.264986</td>\n",
       "      <td>13.270996</td>\n",
       "      <td>13.457549</td>\n",
       "      <td>5.448946</td>\n",
       "      <td>7.533445</td>\n",
       "      <td>6.783335</td>\n",
       "      <td>5.619021</td>\n",
       "      <td>251.751489</td>\n",
       "      <td>248.731086</td>\n",
       "      <td>245.914311</td>\n",
       "      <td>245.934517</td>\n",
       "      <td>252.476533</td>\n",
       "      <td>263.631042</td>\n",
       "      <td>265.486090</td>\n",
       "      <td>267.184991</td>\n",
       "      <td>255.617850</td>\n",
       "      <td>269.347911</td>\n",
       "      <td>271.757783</td>\n",
       "      <td>261.407396</td>\n",
       "      <td>7.984970</td>\n",
       "      <td>8.599406</td>\n",
       "      <td>7.905971</td>\n",
       "      <td>8.261770</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>389.948499</td>\n",
       "      <td>408.922934</td>\n",
       "      <td>411.633049</td>\n",
       "      <td>405.138658</td>\n",
       "      <td>25.765248</td>\n",
       "      <td>25.603052</td>\n",
       "      <td>25.544471</td>\n",
       "      <td>21.310751</td>\n",
       "      <td>14.936449</td>\n",
       "      <td>20.661570</td>\n",
       "      <td>17.855111</td>\n",
       "      <td>16.328227</td>\n",
       "      <td>4.125911</td>\n",
       "      <td>2.161717</td>\n",
       "      <td>2.323464</td>\n",
       "      <td>3.635466</td>\n",
       "      <td>463.419481</td>\n",
       "      <td>480.031178</td>\n",
       "      <td>478.150031</td>\n",
       "      <td>468.980002</td>\n",
       "      <td>140.258485</td>\n",
       "      <td>145.795055</td>\n",
       "      <td>137.239552</td>\n",
       "      <td>140.130610</td>\n",
       "      <td>171.713903</td>\n",
       "      <td>169.423620</td>\n",
       "      <td>169.723759</td>\n",
       "      <td>165.492803</td>\n",
       "      <td>40.140895</td>\n",
       "      <td>42.977442</td>\n",
       "      <td>39.125379</td>\n",
       "      <td>43.840776</td>\n",
       "      <td>254.124029</td>\n",
       "      <td>256.242707</td>\n",
       "      <td>250.025523</td>\n",
       "      <td>249.845070</td>\n",
       "      <td>54.330607</td>\n",
       "      <td>57.411971</td>\n",
       "      <td>55.073186</td>\n",
       "      <td>53.376273</td>\n",
       "      <td>80.793414</td>\n",
       "      <td>86.521393</td>\n",
       "      <td>83.683565</td>\n",
       "      <td>74.913050</td>\n",
       "      <td>16.495594</td>\n",
       "      <td>16.454061</td>\n",
       "      <td>15.812580</td>\n",
       "      <td>15.978601</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>106.283386</td>\n",
       "      <td>113.720168</td>\n",
       "      <td>110.127008</td>\n",
       "      <td>101.982303</td>\n",
       "      <td>291.651671</td>\n",
       "      <td>298.124954</td>\n",
       "      <td>289.321094</td>\n",
       "      <td>284.823024</td>\n",
       "      <td>0.160920</td>\n",
       "      <td>0.155725</td>\n",
       "      <td>0.146147</td>\n",
       "      <td>0.527860</td>\n",
       "      <td>59.722948</td>\n",
       "      <td>65.219829</td>\n",
       "      <td>63.813098</td>\n",
       "      <td>63.505379</td>\n",
       "      <td>11.955164</td>\n",
       "      <td>12.673099</td>\n",
       "      <td>13.284348</td>\n",
       "      <td>12.381172</td>\n",
       "      <td>7.078405</td>\n",
       "      <td>7.070422</td>\n",
       "      <td>7.203753</td>\n",
       "      <td>7.096261</td>\n",
       "      <td>398.019701</td>\n",
       "      <td>408.114237</td>\n",
       "      <td>416.540455</td>\n",
       "      <td>404.588583</td>\n",
       "      <td>120.614894</td>\n",
       "      <td>124.523970</td>\n",
       "      <td>126.902505</td>\n",
       "      <td>125.375109</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>97.356649</td>\n",
       "      <td>95.915385</td>\n",
       "      <td>104.431816</td>\n",
       "      <td>90.809712</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.789128</td>\n",
       "      <td>3.031593</td>\n",
       "      <td>3.074987</td>\n",
       "      <td>2.516339</td>\n",
       "      <td>108.477235</td>\n",
       "      <td>109.765267</td>\n",
       "      <td>109.437851</td>\n",
       "      <td>111.36376</td>\n",
       "      <td>2.570254</td>\n",
       "      <td>2.768332</td>\n",
       "      <td>2.720132</td>\n",
       "      <td>2.214701</td>\n",
       "      <td>1.274428</td>\n",
       "      <td>1.394524</td>\n",
       "      <td>1.422827</td>\n",
       "      <td>1.411513</td>\n",
       "      <td>192.646318</td>\n",
       "      <td>196.791224</td>\n",
       "      <td>191.301305</td>\n",
       "      <td>188.400286</td>\n",
       "      <td>213.356445</td>\n",
       "      <td>212.302217</td>\n",
       "      <td>212.347892</td>\n",
       "      <td>198.653570</td>\n",
       "      <td>544.247227</td>\n",
       "      <td>541.494013</td>\n",
       "      <td>558.775335</td>\n",
       "      <td>577.394194</td>\n",
       "      <td>193.124653</td>\n",
       "      <td>195.893924</td>\n",
       "      <td>188.180936</td>\n",
       "      <td>216.291992</td>\n",
       "      <td>172.767523</td>\n",
       "      <td>176.379871</td>\n",
       "      <td>168.247852</td>\n",
       "      <td>171.384224</td>\n",
       "      <td>0.156391</td>\n",
       "      <td>0.150014</td>\n",
       "      <td>0.142863</td>\n",
       "      <td>0.125366</td>\n",
       "      <td>0.295058</td>\n",
       "      <td>0.304395</td>\n",
       "      <td>0.299568</td>\n",
       "      <td>0.278120</td>\n",
       "      <td>1.497320</td>\n",
       "      <td>1.636230</td>\n",
       "      <td>1.630263</td>\n",
       "      <td>1.347140</td>\n",
       "      <td>0.363371</td>\n",
       "      <td>0.387231</td>\n",
       "      <td>0.384947</td>\n",
       "      <td>0.384978</td>\n",
       "      <td>0.568344</td>\n",
       "      <td>0.628334</td>\n",
       "      <td>0.660234</td>\n",
       "      <td>0.650457</td>\n",
       "      <td>0.279772</td>\n",
       "      <td>0.287950</td>\n",
       "      <td>0.311885</td>\n",
       "      <td>0.345987</td>\n",
       "      <td>954.733842</td>\n",
       "      <td>267.580450</td>\n",
       "      <td>271.201856</td>\n",
       "      <td>253.938223</td>\n",
       "      <td>32.408353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>7.000000e+09</td>\n",
       "      <td>109.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-2258.709000</td>\n",
       "      <td>-2014.045000</td>\n",
       "      <td>-945.808000</td>\n",
       "      <td>-1899.505000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-30.820000</td>\n",
       "      <td>-26.040000</td>\n",
       "      <td>-24.490000</td>\n",
       "      <td>-71.090000</td>\n",
       "      <td>-35.830000</td>\n",
       "      <td>-15.480000</td>\n",
       "      <td>-55.830000</td>\n",
       "      <td>-45.740000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>180.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>7.000606e+09</td>\n",
       "      <td>109.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>93.411500</td>\n",
       "      <td>86.980500</td>\n",
       "      <td>84.126000</td>\n",
       "      <td>62.685000</td>\n",
       "      <td>7.380000</td>\n",
       "      <td>6.660000</td>\n",
       "      <td>6.460000</td>\n",
       "      <td>5.330000</td>\n",
       "      <td>34.730000</td>\n",
       "      <td>32.190000</td>\n",
       "      <td>31.630000</td>\n",
       "      <td>27.130000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.660000</td>\n",
       "      <td>1.630000</td>\n",
       "      <td>1.600000</td>\n",
       "      <td>1.360000</td>\n",
       "      <td>9.880000</td>\n",
       "      <td>10.025000</td>\n",
       "      <td>9.810000</td>\n",
       "      <td>8.810000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>17.110000</td>\n",
       "      <td>17.480000</td>\n",
       "      <td>17.110000</td>\n",
       "      <td>15.560000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>44.740000</td>\n",
       "      <td>43.010000</td>\n",
       "      <td>38.580000</td>\n",
       "      <td>25.510000</td>\n",
       "      <td>2.990000</td>\n",
       "      <td>3.230000</td>\n",
       "      <td>3.280000</td>\n",
       "      <td>3.290000</td>\n",
       "      <td>17.290000</td>\n",
       "      <td>18.590000</td>\n",
       "      <td>18.930000</td>\n",
       "      <td>18.560000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>30.390000</td>\n",
       "      <td>32.460000</td>\n",
       "      <td>32.740000</td>\n",
       "      <td>32.290000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>38.530000</td>\n",
       "      <td>41.190000</td>\n",
       "      <td>38.290000</td>\n",
       "      <td>32.370000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>109.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>52.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>82.000000</td>\n",
       "      <td>92.000000</td>\n",
       "      <td>87.000000</td>\n",
       "      <td>69.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>467.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>7.001205e+09</td>\n",
       "      <td>109.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>197.704000</td>\n",
       "      <td>191.640000</td>\n",
       "      <td>192.080000</td>\n",
       "      <td>176.849000</td>\n",
       "      <td>34.310000</td>\n",
       "      <td>32.330000</td>\n",
       "      <td>32.360000</td>\n",
       "      <td>29.840000</td>\n",
       "      <td>96.310000</td>\n",
       "      <td>91.735000</td>\n",
       "      <td>92.140000</td>\n",
       "      <td>87.290000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>11.910000</td>\n",
       "      <td>11.610000</td>\n",
       "      <td>11.730000</td>\n",
       "      <td>11.260000</td>\n",
       "      <td>41.030000</td>\n",
       "      <td>40.430000</td>\n",
       "      <td>40.360000</td>\n",
       "      <td>39.120000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>65.110000</td>\n",
       "      <td>63.685000</td>\n",
       "      <td>63.730000</td>\n",
       "      <td>61.840000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.950000</td>\n",
       "      <td>3.635000</td>\n",
       "      <td>3.310000</td>\n",
       "      <td>2.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11.640000</td>\n",
       "      <td>11.090000</td>\n",
       "      <td>10.410000</td>\n",
       "      <td>8.410000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>145.140000</td>\n",
       "      <td>141.530000</td>\n",
       "      <td>138.610000</td>\n",
       "      <td>125.460000</td>\n",
       "      <td>15.690000</td>\n",
       "      <td>15.740000</td>\n",
       "      <td>16.030000</td>\n",
       "      <td>15.660000</td>\n",
       "      <td>56.490000</td>\n",
       "      <td>57.080000</td>\n",
       "      <td>58.240000</td>\n",
       "      <td>56.610000</td>\n",
       "      <td>0.880000</td>\n",
       "      <td>0.930000</td>\n",
       "      <td>0.930000</td>\n",
       "      <td>0.960000</td>\n",
       "      <td>92.160000</td>\n",
       "      <td>92.550000</td>\n",
       "      <td>93.830000</td>\n",
       "      <td>91.640000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.030000</td>\n",
       "      <td>2.040000</td>\n",
       "      <td>2.030000</td>\n",
       "      <td>1.740000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.890000</td>\n",
       "      <td>5.960000</td>\n",
       "      <td>5.880000</td>\n",
       "      <td>5.380000</td>\n",
       "      <td>114.740000</td>\n",
       "      <td>116.340000</td>\n",
       "      <td>114.660000</td>\n",
       "      <td>105.890000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>230.000000</td>\n",
       "      <td>220.000000</td>\n",
       "      <td>225.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>110.000000</td>\n",
       "      <td>110.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>61.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>30.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>145.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>164.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.480000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.880000</td>\n",
       "      <td>2.605000</td>\n",
       "      <td>10.830000</td>\n",
       "      <td>8.810000</td>\n",
       "      <td>9.270000</td>\n",
       "      <td>14.800000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>863.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.001812e+09</td>\n",
       "      <td>109.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>371.060000</td>\n",
       "      <td>365.344500</td>\n",
       "      <td>369.370500</td>\n",
       "      <td>353.466500</td>\n",
       "      <td>118.740000</td>\n",
       "      <td>115.595000</td>\n",
       "      <td>115.860000</td>\n",
       "      <td>112.130000</td>\n",
       "      <td>231.860000</td>\n",
       "      <td>226.815000</td>\n",
       "      <td>228.260000</td>\n",
       "      <td>220.505000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.960000</td>\n",
       "      <td>39.910000</td>\n",
       "      <td>40.110000</td>\n",
       "      <td>39.280000</td>\n",
       "      <td>110.390000</td>\n",
       "      <td>107.560000</td>\n",
       "      <td>109.090000</td>\n",
       "      <td>106.810000</td>\n",
       "      <td>2.080000</td>\n",
       "      <td>2.090000</td>\n",
       "      <td>2.040000</td>\n",
       "      <td>1.940000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>168.270000</td>\n",
       "      <td>164.382500</td>\n",
       "      <td>166.110000</td>\n",
       "      <td>162.225000</td>\n",
       "      <td>30.807500</td>\n",
       "      <td>31.132500</td>\n",
       "      <td>30.580000</td>\n",
       "      <td>28.230000</td>\n",
       "      <td>53.290000</td>\n",
       "      <td>54.040000</td>\n",
       "      <td>52.490000</td>\n",
       "      <td>48.560000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>144.837500</td>\n",
       "      <td>150.615000</td>\n",
       "      <td>147.940000</td>\n",
       "      <td>142.105000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.430000</td>\n",
       "      <td>3.710000</td>\n",
       "      <td>3.990000</td>\n",
       "      <td>3.230000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>372.860000</td>\n",
       "      <td>378.570000</td>\n",
       "      <td>369.900000</td>\n",
       "      <td>353.480000</td>\n",
       "      <td>46.840000</td>\n",
       "      <td>45.810000</td>\n",
       "      <td>46.290000</td>\n",
       "      <td>45.180000</td>\n",
       "      <td>132.387500</td>\n",
       "      <td>130.960000</td>\n",
       "      <td>133.930000</td>\n",
       "      <td>130.490000</td>\n",
       "      <td>8.140000</td>\n",
       "      <td>8.282500</td>\n",
       "      <td>8.110000</td>\n",
       "      <td>8.140000</td>\n",
       "      <td>208.075000</td>\n",
       "      <td>205.837500</td>\n",
       "      <td>207.280000</td>\n",
       "      <td>202.737500</td>\n",
       "      <td>4.060000</td>\n",
       "      <td>4.230000</td>\n",
       "      <td>4.080000</td>\n",
       "      <td>3.510000</td>\n",
       "      <td>15.030000</td>\n",
       "      <td>15.740000</td>\n",
       "      <td>15.360000</td>\n",
       "      <td>14.260000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>26.930000</td>\n",
       "      <td>28.310000</td>\n",
       "      <td>27.710000</td>\n",
       "      <td>25.690000</td>\n",
       "      <td>251.670000</td>\n",
       "      <td>250.660000</td>\n",
       "      <td>248.990000</td>\n",
       "      <td>236.320000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.060000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>437.500000</td>\n",
       "      <td>428.000000</td>\n",
       "      <td>434.500000</td>\n",
       "      <td>415.000000</td>\n",
       "      <td>120.000000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>144.000000</td>\n",
       "      <td>144.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>110.000000</td>\n",
       "      <td>110.000000</td>\n",
       "      <td>130.000000</td>\n",
       "      <td>50.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>177.000000</td>\n",
       "      <td>177.000000</td>\n",
       "      <td>179.000000</td>\n",
       "      <td>179.00000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>252.000000</td>\n",
       "      <td>252.000000</td>\n",
       "      <td>252.000000</td>\n",
       "      <td>252.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>122.070000</td>\n",
       "      <td>119.560000</td>\n",
       "      <td>122.070000</td>\n",
       "      <td>140.010000</td>\n",
       "      <td>122.070000</td>\n",
       "      <td>122.070000</td>\n",
       "      <td>122.070000</td>\n",
       "      <td>140.010000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1807.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7.002411e+09</td>\n",
       "      <td>109.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>27731.088000</td>\n",
       "      <td>35145.834000</td>\n",
       "      <td>33543.624000</td>\n",
       "      <td>38805.617000</td>\n",
       "      <td>7376.710000</td>\n",
       "      <td>8157.780000</td>\n",
       "      <td>10752.560000</td>\n",
       "      <td>10427.460000</td>\n",
       "      <td>8362.360000</td>\n",
       "      <td>9667.130000</td>\n",
       "      <td>14007.340000</td>\n",
       "      <td>10310.760000</td>\n",
       "      <td>13724.380000</td>\n",
       "      <td>15371.040000</td>\n",
       "      <td>13095.360000</td>\n",
       "      <td>8464.030000</td>\n",
       "      <td>3775.110000</td>\n",
       "      <td>2812.040000</td>\n",
       "      <td>5337.040000</td>\n",
       "      <td>4428.460000</td>\n",
       "      <td>6431.330000</td>\n",
       "      <td>7400.660000</td>\n",
       "      <td>10752.560000</td>\n",
       "      <td>10389.240000</td>\n",
       "      <td>4729.740000</td>\n",
       "      <td>4557.140000</td>\n",
       "      <td>4961.330000</td>\n",
       "      <td>4429.880000</td>\n",
       "      <td>1466.030000</td>\n",
       "      <td>1196.430000</td>\n",
       "      <td>928.490000</td>\n",
       "      <td>927.410000</td>\n",
       "      <td>342.860000</td>\n",
       "      <td>916.240000</td>\n",
       "      <td>502.090000</td>\n",
       "      <td>339.840000</td>\n",
       "      <td>10643.380000</td>\n",
       "      <td>7674.780000</td>\n",
       "      <td>11039.910000</td>\n",
       "      <td>11099.260000</td>\n",
       "      <td>7366.580000</td>\n",
       "      <td>8133.660000</td>\n",
       "      <td>8014.430000</td>\n",
       "      <td>9382.580000</td>\n",
       "      <td>8314.760000</td>\n",
       "      <td>9284.740000</td>\n",
       "      <td>13950.040000</td>\n",
       "      <td>10223.430000</td>\n",
       "      <td>628.560000</td>\n",
       "      <td>544.630000</td>\n",
       "      <td>516.910000</td>\n",
       "      <td>808.490000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8432.990000</td>\n",
       "      <td>10936.730000</td>\n",
       "      <td>13980.060000</td>\n",
       "      <td>11495.310000</td>\n",
       "      <td>5900.660000</td>\n",
       "      <td>5490.280000</td>\n",
       "      <td>5681.540000</td>\n",
       "      <td>4244.530000</td>\n",
       "      <td>1023.210000</td>\n",
       "      <td>2372.510000</td>\n",
       "      <td>1390.880000</td>\n",
       "      <td>1635.710000</td>\n",
       "      <td>800.890000</td>\n",
       "      <td>370.130000</td>\n",
       "      <td>394.930000</td>\n",
       "      <td>787.790000</td>\n",
       "      <td>10674.030000</td>\n",
       "      <td>11365.310000</td>\n",
       "      <td>14043.060000</td>\n",
       "      <td>11517.730000</td>\n",
       "      <td>6626.930000</td>\n",
       "      <td>9324.660000</td>\n",
       "      <td>10696.230000</td>\n",
       "      <td>10598.830000</td>\n",
       "      <td>4693.860000</td>\n",
       "      <td>4455.830000</td>\n",
       "      <td>6274.190000</td>\n",
       "      <td>5463.780000</td>\n",
       "      <td>1872.340000</td>\n",
       "      <td>1983.010000</td>\n",
       "      <td>2433.060000</td>\n",
       "      <td>4318.280000</td>\n",
       "      <td>7454.630000</td>\n",
       "      <td>9669.910000</td>\n",
       "      <td>10830.160000</td>\n",
       "      <td>10796.290000</td>\n",
       "      <td>5459.560000</td>\n",
       "      <td>5800.930000</td>\n",
       "      <td>4309.290000</td>\n",
       "      <td>3819.830000</td>\n",
       "      <td>5647.160000</td>\n",
       "      <td>6141.880000</td>\n",
       "      <td>5645.860000</td>\n",
       "      <td>5689.760000</td>\n",
       "      <td>1351.110000</td>\n",
       "      <td>1136.080000</td>\n",
       "      <td>1394.890000</td>\n",
       "      <td>1431.960000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5712.110000</td>\n",
       "      <td>6745.760000</td>\n",
       "      <td>5957.140000</td>\n",
       "      <td>5956.660000</td>\n",
       "      <td>7716.140000</td>\n",
       "      <td>9699.010000</td>\n",
       "      <td>10830.380000</td>\n",
       "      <td>10796.590000</td>\n",
       "      <td>19.760000</td>\n",
       "      <td>21.330000</td>\n",
       "      <td>16.860000</td>\n",
       "      <td>62.380000</td>\n",
       "      <td>6789.410000</td>\n",
       "      <td>5289.540000</td>\n",
       "      <td>4127.010000</td>\n",
       "      <td>5057.740000</td>\n",
       "      <td>1362.940000</td>\n",
       "      <td>1495.940000</td>\n",
       "      <td>2327.510000</td>\n",
       "      <td>1005.230000</td>\n",
       "      <td>307.000000</td>\n",
       "      <td>138.000000</td>\n",
       "      <td>196.000000</td>\n",
       "      <td>131.000000</td>\n",
       "      <td>35190.000000</td>\n",
       "      <td>40335.000000</td>\n",
       "      <td>45320.000000</td>\n",
       "      <td>37235.000000</td>\n",
       "      <td>4010.000000</td>\n",
       "      <td>4010.000000</td>\n",
       "      <td>4449.000000</td>\n",
       "      <td>3399.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4010.000000</td>\n",
       "      <td>4010.000000</td>\n",
       "      <td>4449.000000</td>\n",
       "      <td>3399.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>61.000000</td>\n",
       "      <td>54.000000</td>\n",
       "      <td>60.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>1555.000000</td>\n",
       "      <td>1555.000000</td>\n",
       "      <td>1555.000000</td>\n",
       "      <td>1555.00000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>49.000000</td>\n",
       "      <td>7546.000000</td>\n",
       "      <td>4365.000000</td>\n",
       "      <td>4076.000000</td>\n",
       "      <td>4061.000000</td>\n",
       "      <td>10285.900000</td>\n",
       "      <td>7873.550000</td>\n",
       "      <td>11117.610000</td>\n",
       "      <td>8993.950000</td>\n",
       "      <td>45735.400000</td>\n",
       "      <td>28144.120000</td>\n",
       "      <td>30036.060000</td>\n",
       "      <td>39221.270000</td>\n",
       "      <td>6362.280000</td>\n",
       "      <td>4980.900000</td>\n",
       "      <td>3716.900000</td>\n",
       "      <td>13884.310000</td>\n",
       "      <td>6433.760000</td>\n",
       "      <td>4809.360000</td>\n",
       "      <td>3483.170000</td>\n",
       "      <td>3467.170000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>41.000000</td>\n",
       "      <td>49.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>4337.000000</td>\n",
       "      <td>12916.220000</td>\n",
       "      <td>9165.600000</td>\n",
       "      <td>11166.210000</td>\n",
       "      <td>2618.570000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        mobile_number  circle_id  loc_og_t2o_mou  std_og_t2o_mou  \\\n",
       "count    9.999900e+04    99999.0         98981.0         98981.0   \n",
       "unique            NaN        NaN             NaN             NaN   \n",
       "top               NaN        NaN             NaN             NaN   \n",
       "freq              NaN        NaN             NaN             NaN   \n",
       "mean     7.001207e+09      109.0             0.0             0.0   \n",
       "std      6.956694e+05        0.0             0.0             0.0   \n",
       "min      7.000000e+09      109.0             0.0             0.0   \n",
       "25%      7.000606e+09      109.0             0.0             0.0   \n",
       "50%      7.001205e+09      109.0             0.0             0.0   \n",
       "75%      7.001812e+09      109.0             0.0             0.0   \n",
       "max      7.002411e+09      109.0             0.0             0.0   \n",
       "\n",
       "        loc_ic_t2o_mou last_date_of_month_6 last_date_of_month_7  \\\n",
       "count          98981.0                99999                99398   \n",
       "unique             NaN                    1                    1   \n",
       "top                NaN            6/30/2014            7/31/2014   \n",
       "freq               NaN                99999                99398   \n",
       "mean               0.0                  NaN                  NaN   \n",
       "std                0.0                  NaN                  NaN   \n",
       "min                0.0                  NaN                  NaN   \n",
       "25%                0.0                  NaN                  NaN   \n",
       "50%                0.0                  NaN                  NaN   \n",
       "75%                0.0                  NaN                  NaN   \n",
       "max                0.0                  NaN                  NaN   \n",
       "\n",
       "       last_date_of_month_8 last_date_of_month_9        arpu_6        arpu_7  \\\n",
       "count                 98899                98340  99999.000000  99999.000000   \n",
       "unique                    1                    1           NaN           NaN   \n",
       "top               8/31/2014            9/30/2014           NaN           NaN   \n",
       "freq                  98899                98340           NaN           NaN   \n",
       "mean                    NaN                  NaN    282.987358    278.536648   \n",
       "std                     NaN                  NaN    328.439770    338.156291   \n",
       "min                     NaN                  NaN  -2258.709000  -2014.045000   \n",
       "25%                     NaN                  NaN     93.411500     86.980500   \n",
       "50%                     NaN                  NaN    197.704000    191.640000   \n",
       "75%                     NaN                  NaN    371.060000    365.344500   \n",
       "max                     NaN                  NaN  27731.088000  35145.834000   \n",
       "\n",
       "              arpu_8        arpu_9   onnet_mou_6   onnet_mou_7   onnet_mou_8  \\\n",
       "count   99999.000000  99999.000000  96062.000000  96140.000000  94621.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean      279.154731    261.645069    132.395875    133.670805    133.018098   \n",
       "std       344.474791    341.998630    297.207406    308.794148    308.951589   \n",
       "min      -945.808000  -1899.505000      0.000000      0.000000      0.000000   \n",
       "25%        84.126000     62.685000      7.380000      6.660000      6.460000   \n",
       "50%       192.080000    176.849000     34.310000     32.330000     32.360000   \n",
       "75%       369.370500    353.466500    118.740000    115.595000    115.860000   \n",
       "max     33543.624000  38805.617000   7376.710000   8157.780000  10752.560000   \n",
       "\n",
       "         onnet_mou_9  offnet_mou_6  offnet_mou_7  offnet_mou_8  offnet_mou_9  \\\n",
       "count   92254.000000  96062.000000  96140.000000  94621.000000  92254.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean      130.302327    197.935577    197.045133    196.574803    190.337222   \n",
       "std       308.477668    316.851613    325.862803    327.170662    319.396092   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         5.330000     34.730000     32.190000     31.630000     27.130000   \n",
       "50%        29.840000     96.310000     91.735000     92.140000     87.290000   \n",
       "75%       112.130000    231.860000    226.815000    228.260000    220.505000   \n",
       "max     10427.460000   8362.360000   9667.130000  14007.340000  10310.760000   \n",
       "\n",
       "        roam_ic_mou_6  roam_ic_mou_7  roam_ic_mou_8  roam_ic_mou_9  \\\n",
       "count    96062.000000   96140.000000   94621.000000   92254.000000   \n",
       "unique            NaN            NaN            NaN            NaN   \n",
       "top               NaN            NaN            NaN            NaN   \n",
       "freq              NaN            NaN            NaN            NaN   \n",
       "mean         9.950013       7.149898       7.292981       6.343841   \n",
       "std         72.825411      73.447948      68.402466      57.137537   \n",
       "min          0.000000       0.000000       0.000000       0.000000   \n",
       "25%          0.000000       0.000000       0.000000       0.000000   \n",
       "50%          0.000000       0.000000       0.000000       0.000000   \n",
       "75%          0.000000       0.000000       0.000000       0.000000   \n",
       "max      13724.380000   15371.040000   13095.360000    8464.030000   \n",
       "\n",
       "        roam_og_mou_6  roam_og_mou_7  roam_og_mou_8  roam_og_mou_9  \\\n",
       "count    96062.000000   96140.000000   94621.000000   92254.000000   \n",
       "unique            NaN            NaN            NaN            NaN   \n",
       "top               NaN            NaN            NaN            NaN   \n",
       "freq              NaN            NaN            NaN            NaN   \n",
       "mean        13.911337       9.818732       9.971890       8.555519   \n",
       "std         71.443196      58.455762      64.713221      58.438186   \n",
       "min          0.000000       0.000000       0.000000       0.000000   \n",
       "25%          0.000000       0.000000       0.000000       0.000000   \n",
       "50%          0.000000       0.000000       0.000000       0.000000   \n",
       "75%          0.000000       0.000000       0.000000       0.000000   \n",
       "max       3775.110000    2812.040000    5337.040000    4428.460000   \n",
       "\n",
       "        loc_og_t2t_mou_6  loc_og_t2t_mou_7  loc_og_t2t_mou_8  \\\n",
       "count       96062.000000      96140.000000      94621.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean           47.100763         46.473010         45.887806   \n",
       "std           150.856393        155.318705        151.184830   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             1.660000          1.630000          1.600000   \n",
       "50%            11.910000         11.610000         11.730000   \n",
       "75%            40.960000         39.910000         40.110000   \n",
       "max          6431.330000       7400.660000      10752.560000   \n",
       "\n",
       "        loc_og_t2t_mou_9  loc_og_t2m_mou_6  loc_og_t2m_mou_7  \\\n",
       "count       92254.000000      96062.000000      96140.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean           44.584446         93.342088         91.397131   \n",
       "std           147.995390        162.780544        157.492308   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             1.360000          9.880000         10.025000   \n",
       "50%            11.260000         41.030000         40.430000   \n",
       "75%            39.280000        110.390000        107.560000   \n",
       "max         10389.240000       4729.740000       4557.140000   \n",
       "\n",
       "        loc_og_t2m_mou_8  loc_og_t2m_mou_9  loc_og_t2f_mou_6  \\\n",
       "count       94621.000000      92254.000000      96062.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean           91.755128         90.463192          3.751013   \n",
       "std           156.537048        158.681454         14.230438   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             9.810000          8.810000          0.000000   \n",
       "50%            40.360000         39.120000          0.000000   \n",
       "75%           109.090000        106.810000          2.080000   \n",
       "max          4961.330000       4429.880000       1466.030000   \n",
       "\n",
       "        loc_og_t2f_mou_7  loc_og_t2f_mou_8  loc_og_t2f_mou_9  \\\n",
       "count       96140.000000      94621.000000      92254.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean            3.792985          3.677991          3.655123   \n",
       "std            14.264986         13.270996         13.457549   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             0.000000          0.000000          0.000000   \n",
       "50%             0.000000          0.000000          0.000000   \n",
       "75%             2.090000          2.040000          1.940000   \n",
       "max          1196.430000        928.490000        927.410000   \n",
       "\n",
       "        loc_og_t2c_mou_6  loc_og_t2c_mou_7  loc_og_t2c_mou_8  \\\n",
       "count       96062.000000      96140.000000      94621.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean            1.123056          1.368500          1.433821   \n",
       "std             5.448946          7.533445          6.783335   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             0.000000          0.000000          0.000000   \n",
       "50%             0.000000          0.000000          0.000000   \n",
       "75%             0.000000          0.000000          0.000000   \n",
       "max           342.860000        916.240000        502.090000   \n",
       "\n",
       "        loc_og_t2c_mou_9  loc_og_mou_6  loc_og_mou_7  loc_og_mou_8  \\\n",
       "count       92254.000000  96062.000000  96140.000000  94621.000000   \n",
       "unique               NaN           NaN           NaN           NaN   \n",
       "top                  NaN           NaN           NaN           NaN   \n",
       "freq                 NaN           NaN           NaN           NaN   \n",
       "mean            1.232726    144.201175    141.670476    141.328209   \n",
       "std             5.619021    251.751489    248.731086    245.914311   \n",
       "min             0.000000      0.000000      0.000000      0.000000   \n",
       "25%             0.000000     17.110000     17.480000     17.110000   \n",
       "50%             0.000000     65.110000     63.685000     63.730000   \n",
       "75%             0.000000    168.270000    164.382500    166.110000   \n",
       "max           339.840000  10643.380000   7674.780000  11039.910000   \n",
       "\n",
       "        loc_og_mou_9  std_og_t2t_mou_6  std_og_t2t_mou_7  std_og_t2t_mou_8  \\\n",
       "count   92254.000000      96062.000000      96140.000000      94621.000000   \n",
       "unique           NaN               NaN               NaN               NaN   \n",
       "top              NaN               NaN               NaN               NaN   \n",
       "freq             NaN               NaN               NaN               NaN   \n",
       "mean      138.709970         79.829870         83.299598         83.282673   \n",
       "std       245.934517        252.476533        263.631042        265.486090   \n",
       "min         0.000000          0.000000          0.000000          0.000000   \n",
       "25%        15.560000          0.000000          0.000000          0.000000   \n",
       "50%        61.840000          0.000000          0.000000          0.000000   \n",
       "75%       162.225000         30.807500         31.132500         30.580000   \n",
       "max     11099.260000       7366.580000       8133.660000       8014.430000   \n",
       "\n",
       "        std_og_t2t_mou_9  std_og_t2m_mou_6  std_og_t2m_mou_7  \\\n",
       "count       92254.000000      96062.000000      96140.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean           82.342919         87.299624         90.804137   \n",
       "std           267.184991        255.617850        269.347911   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             0.000000          0.000000          0.000000   \n",
       "50%             0.000000          3.950000          3.635000   \n",
       "75%            28.230000         53.290000         54.040000   \n",
       "max          9382.580000       8314.760000       9284.740000   \n",
       "\n",
       "        std_og_t2m_mou_8  std_og_t2m_mou_9  std_og_t2f_mou_6  \\\n",
       "count       94621.000000      92254.000000      96062.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean           89.838390         86.276622          1.129011   \n",
       "std           271.757783        261.407396          7.984970   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             0.000000          0.000000          0.000000   \n",
       "50%             3.310000          2.500000          0.000000   \n",
       "75%            52.490000         48.560000          0.000000   \n",
       "max         13950.040000      10223.430000        628.560000   \n",
       "\n",
       "        std_og_t2f_mou_7  std_og_t2f_mou_8  std_og_t2f_mou_9  \\\n",
       "count       96140.000000      94621.000000      92254.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean            1.115010          1.067792          1.042362   \n",
       "std             8.599406          7.905971          8.261770   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             0.000000          0.000000          0.000000   \n",
       "50%             0.000000          0.000000          0.000000   \n",
       "75%             0.000000          0.000000          0.000000   \n",
       "max           544.630000        516.910000        808.490000   \n",
       "\n",
       "        std_og_t2c_mou_6  std_og_t2c_mou_7  std_og_t2c_mou_8  \\\n",
       "count            96062.0           96140.0           94621.0   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean                 0.0               0.0               0.0   \n",
       "std                  0.0               0.0               0.0   \n",
       "min                  0.0               0.0               0.0   \n",
       "25%                  0.0               0.0               0.0   \n",
       "50%                  0.0               0.0               0.0   \n",
       "75%                  0.0               0.0               0.0   \n",
       "max                  0.0               0.0               0.0   \n",
       "\n",
       "        std_og_t2c_mou_9  std_og_mou_6  std_og_mou_7  std_og_mou_8  \\\n",
       "count            92254.0  96062.000000  96140.000000  94621.000000   \n",
       "unique               NaN           NaN           NaN           NaN   \n",
       "top                  NaN           NaN           NaN           NaN   \n",
       "freq                 NaN           NaN           NaN           NaN   \n",
       "mean                 0.0    168.261218    175.221436    174.191498   \n",
       "std                  0.0    389.948499    408.922934    411.633049   \n",
       "min                  0.0      0.000000      0.000000      0.000000   \n",
       "25%                  0.0      0.000000      0.000000      0.000000   \n",
       "50%                  0.0     11.640000     11.090000     10.410000   \n",
       "75%                  0.0    144.837500    150.615000    147.940000   \n",
       "max                  0.0   8432.990000  10936.730000  13980.060000   \n",
       "\n",
       "        std_og_mou_9  isd_og_mou_6  isd_og_mou_7  isd_og_mou_8  isd_og_mou_9  \\\n",
       "count   92254.000000  96062.000000  96140.000000  94621.000000  92254.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean      169.664466      0.798277      0.776572      0.791247      0.723892   \n",
       "std       405.138658     25.765248     25.603052     25.544471     21.310751   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         8.410000      0.000000      0.000000      0.000000      0.000000   \n",
       "75%       142.105000      0.000000      0.000000      0.000000      0.000000   \n",
       "max     11495.310000   5900.660000   5490.280000   5681.540000   4244.530000   \n",
       "\n",
       "        spl_og_mou_6  spl_og_mou_7  spl_og_mou_8  spl_og_mou_9   og_others_6  \\\n",
       "count   96062.000000  96140.000000  94621.000000  92254.000000  96062.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean        3.916811      4.978279      5.053769      4.412767      0.454157   \n",
       "std        14.936449     20.661570     17.855111     16.328227      4.125911   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "75%         2.430000      3.710000      3.990000      3.230000      0.000000   \n",
       "max      1023.210000   2372.510000   1390.880000   1635.710000    800.890000   \n",
       "\n",
       "         og_others_7   og_others_8   og_others_9  total_og_mou_6  \\\n",
       "count   96140.000000  94621.000000  92254.000000    99999.000000   \n",
       "unique           NaN           NaN           NaN             NaN   \n",
       "top              NaN           NaN           NaN             NaN   \n",
       "freq             NaN           NaN           NaN             NaN   \n",
       "mean        0.030235      0.033372      0.047456      305.133424   \n",
       "std         2.161717      2.323464      3.635466      463.419481   \n",
       "min         0.000000      0.000000      0.000000        0.000000   \n",
       "25%         0.000000      0.000000      0.000000       44.740000   \n",
       "50%         0.000000      0.000000      0.000000      145.140000   \n",
       "75%         0.000000      0.000000      0.000000      372.860000   \n",
       "max       370.130000    394.930000    787.790000    10674.030000   \n",
       "\n",
       "        total_og_mou_7  total_og_mou_8  total_og_mou_9  loc_ic_t2t_mou_6  \\\n",
       "count     99999.000000    99999.000000    99999.000000      96062.000000   \n",
       "unique             NaN             NaN             NaN               NaN   \n",
       "top                NaN             NaN             NaN               NaN   \n",
       "freq               NaN             NaN             NaN               NaN   \n",
       "mean        310.231175      304.119513      289.279198         47.922365   \n",
       "std         480.031178      478.150031      468.980002        140.258485   \n",
       "min           0.000000        0.000000        0.000000          0.000000   \n",
       "25%          43.010000       38.580000       25.510000          2.990000   \n",
       "50%         141.530000      138.610000      125.460000         15.690000   \n",
       "75%         378.570000      369.900000      353.480000         46.840000   \n",
       "max       11365.310000    14043.060000    11517.730000       6626.930000   \n",
       "\n",
       "        loc_ic_t2t_mou_7  loc_ic_t2t_mou_8  loc_ic_t2t_mou_9  \\\n",
       "count       96140.000000      94621.000000      92254.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean           47.990520         47.211362         46.281794   \n",
       "std           145.795055        137.239552        140.130610   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             3.230000          3.280000          3.290000   \n",
       "50%            15.740000         16.030000         15.660000   \n",
       "75%            45.810000         46.290000         45.180000   \n",
       "max          9324.660000      10696.230000      10598.830000   \n",
       "\n",
       "        loc_ic_t2m_mou_6  loc_ic_t2m_mou_7  loc_ic_t2m_mou_8  \\\n",
       "count       96062.000000      96140.000000      94621.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean          107.475650        107.120493        108.460515   \n",
       "std           171.713903        169.423620        169.723759   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%            17.290000         18.590000         18.930000   \n",
       "50%            56.490000         57.080000         58.240000   \n",
       "75%           132.387500        130.960000        133.930000   \n",
       "max          4693.860000       4455.830000       6274.190000   \n",
       "\n",
       "        loc_ic_t2m_mou_9  loc_ic_t2f_mou_6  loc_ic_t2f_mou_7  \\\n",
       "count       92254.000000      96062.000000      96140.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean          106.155471         12.084305         12.599697   \n",
       "std           165.492803         40.140895         42.977442   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%            18.560000          0.000000          0.000000   \n",
       "50%            56.610000          0.880000          0.930000   \n",
       "75%           130.490000          8.140000          8.282500   \n",
       "max          5463.780000       1872.340000       1983.010000   \n",
       "\n",
       "        loc_ic_t2f_mou_8  loc_ic_t2f_mou_9  loc_ic_mou_6  loc_ic_mou_7  \\\n",
       "count       94621.000000      92254.000000  96062.000000  96140.000000   \n",
       "unique               NaN               NaN           NaN           NaN   \n",
       "top                  NaN               NaN           NaN           NaN   \n",
       "freq                 NaN               NaN           NaN           NaN   \n",
       "mean           11.751834         12.173105    167.491059    167.719540   \n",
       "std            39.125379         43.840776    254.124029    256.242707   \n",
       "min             0.000000          0.000000      0.000000      0.000000   \n",
       "25%             0.000000          0.000000     30.390000     32.460000   \n",
       "50%             0.930000          0.960000     92.160000     92.550000   \n",
       "75%             8.110000          8.140000    208.075000    205.837500   \n",
       "max          2433.060000       4318.280000   7454.630000   9669.910000   \n",
       "\n",
       "        loc_ic_mou_8  loc_ic_mou_9  std_ic_t2t_mou_6  std_ic_t2t_mou_7  \\\n",
       "count   94621.000000  92254.000000      96062.000000      96140.000000   \n",
       "unique           NaN           NaN               NaN               NaN   \n",
       "top              NaN           NaN               NaN               NaN   \n",
       "freq             NaN           NaN               NaN               NaN   \n",
       "mean      167.432575    164.619293          9.575993         10.011904   \n",
       "std       250.025523    249.845070         54.330607         57.411971   \n",
       "min         0.000000      0.000000          0.000000          0.000000   \n",
       "25%        32.740000     32.290000          0.000000          0.000000   \n",
       "50%        93.830000     91.640000          0.000000          0.000000   \n",
       "75%       207.280000    202.737500          4.060000          4.230000   \n",
       "max     10830.160000  10796.290000       5459.560000       5800.930000   \n",
       "\n",
       "        std_ic_t2t_mou_8  std_ic_t2t_mou_9  std_ic_t2m_mou_6  \\\n",
       "count       94621.000000      92254.000000      96062.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean            9.883921          9.432479         20.722240   \n",
       "std            55.073186         53.376273         80.793414   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             0.000000          0.000000          0.000000   \n",
       "50%             0.000000          0.000000          2.030000   \n",
       "75%             4.080000          3.510000         15.030000   \n",
       "max          4309.290000       3819.830000       5647.160000   \n",
       "\n",
       "        std_ic_t2m_mou_7  std_ic_t2m_mou_8  std_ic_t2m_mou_9  \\\n",
       "count       96140.000000      94621.000000      92254.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean           21.656415         21.183211         19.620913   \n",
       "std            86.521393         83.683565         74.913050   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             0.000000          0.000000          0.000000   \n",
       "50%             2.040000          2.030000          1.740000   \n",
       "75%            15.740000         15.360000         14.260000   \n",
       "max          6141.880000       5645.860000       5689.760000   \n",
       "\n",
       "        std_ic_t2f_mou_6  std_ic_t2f_mou_7  std_ic_t2f_mou_8  \\\n",
       "count       96062.000000      96140.000000      94621.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean            2.156397          2.216923          2.085004   \n",
       "std            16.495594         16.454061         15.812580   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             0.000000          0.000000          0.000000   \n",
       "50%             0.000000          0.000000          0.000000   \n",
       "75%             0.000000          0.000000          0.000000   \n",
       "max          1351.110000       1136.080000       1394.890000   \n",
       "\n",
       "        std_ic_t2f_mou_9  std_ic_t2o_mou_6  std_ic_t2o_mou_7  \\\n",
       "count       92254.000000           96062.0           96140.0   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean            2.173419               0.0               0.0   \n",
       "std            15.978601               0.0               0.0   \n",
       "min             0.000000               0.0               0.0   \n",
       "25%             0.000000               0.0               0.0   \n",
       "50%             0.000000               0.0               0.0   \n",
       "75%             0.000000               0.0               0.0   \n",
       "max          1431.960000               0.0               0.0   \n",
       "\n",
       "        std_ic_t2o_mou_8  std_ic_t2o_mou_9  std_ic_mou_6  std_ic_mou_7  \\\n",
       "count            94621.0           92254.0  96062.000000  96140.000000   \n",
       "unique               NaN               NaN           NaN           NaN   \n",
       "top                  NaN               NaN           NaN           NaN   \n",
       "freq                 NaN               NaN           NaN           NaN   \n",
       "mean                 0.0               0.0     32.457179     33.887833   \n",
       "std                  0.0               0.0    106.283386    113.720168   \n",
       "min                  0.0               0.0      0.000000      0.000000   \n",
       "25%                  0.0               0.0      0.000000      0.000000   \n",
       "50%                  0.0               0.0      5.890000      5.960000   \n",
       "75%                  0.0               0.0     26.930000     28.310000   \n",
       "max                  0.0               0.0   5712.110000   6745.760000   \n",
       "\n",
       "        std_ic_mou_8  std_ic_mou_9  total_ic_mou_6  total_ic_mou_7  \\\n",
       "count   94621.000000  92254.000000    99999.000000    99999.000000   \n",
       "unique           NaN           NaN             NaN             NaN   \n",
       "top              NaN           NaN             NaN             NaN   \n",
       "freq             NaN           NaN             NaN             NaN   \n",
       "mean       33.154735     31.229344      200.130037      202.853055   \n",
       "std       110.127008    101.982303      291.651671      298.124954   \n",
       "min         0.000000      0.000000        0.000000        0.000000   \n",
       "25%         0.010000      0.000000       38.530000       41.190000   \n",
       "50%         5.880000      5.380000      114.740000      116.340000   \n",
       "75%        27.710000     25.690000      251.670000      250.660000   \n",
       "max      5957.140000   5956.660000     7716.140000     9699.010000   \n",
       "\n",
       "        total_ic_mou_8  total_ic_mou_9  spl_ic_mou_6  spl_ic_mou_7  \\\n",
       "count     99999.000000    99999.000000  96062.000000  96140.000000   \n",
       "unique             NaN             NaN           NaN           NaN   \n",
       "top                NaN             NaN           NaN           NaN   \n",
       "freq               NaN             NaN           NaN           NaN   \n",
       "mean        198.750783      189.214260      0.061557      0.033585   \n",
       "std         289.321094      284.823024      0.160920      0.155725   \n",
       "min           0.000000        0.000000      0.000000      0.000000   \n",
       "25%          38.290000       32.370000      0.000000      0.000000   \n",
       "50%         114.660000      105.890000      0.000000      0.000000   \n",
       "75%         248.990000      236.320000      0.000000      0.000000   \n",
       "max       10830.380000    10796.590000     19.760000     21.330000   \n",
       "\n",
       "        spl_ic_mou_8  spl_ic_mou_9  isd_ic_mou_6  isd_ic_mou_7  isd_ic_mou_8  \\\n",
       "count   94621.000000  92254.000000  96062.000000  96140.000000  94621.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean        0.040361      0.163137      7.460608      8.334936      8.442001   \n",
       "std         0.146147      0.527860     59.722948     65.219829     63.813098   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "75%         0.000000      0.060000      0.000000      0.000000      0.000000   \n",
       "max        16.860000     62.380000   6789.410000   5289.540000   4127.010000   \n",
       "\n",
       "        isd_ic_mou_9   ic_others_6   ic_others_7   ic_others_8   ic_others_9  \\\n",
       "count   92254.000000  96062.000000  96140.000000  94621.000000  92254.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean        8.063003      0.854656      1.012960      0.970800      1.017162   \n",
       "std        63.505379     11.955164     12.673099     13.284348     12.381172   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "75%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "max      5057.740000   1362.940000   1495.940000   2327.510000   1005.230000   \n",
       "\n",
       "        total_rech_num_6  total_rech_num_7  total_rech_num_8  \\\n",
       "count       99999.000000      99999.000000      99999.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean            7.558806          7.700367          7.212912   \n",
       "std             7.078405          7.070422          7.203753   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             3.000000          3.000000          3.000000   \n",
       "50%             6.000000          6.000000          5.000000   \n",
       "75%             9.000000         10.000000          9.000000   \n",
       "max           307.000000        138.000000        196.000000   \n",
       "\n",
       "        total_rech_num_9  total_rech_amt_6  total_rech_amt_7  \\\n",
       "count       99999.000000      99999.000000      99999.000000   \n",
       "unique               NaN               NaN               NaN   \n",
       "top                  NaN               NaN               NaN   \n",
       "freq                 NaN               NaN               NaN   \n",
       "mean            6.893019        327.514615        322.962970   \n",
       "std             7.096261        398.019701        408.114237   \n",
       "min             0.000000          0.000000          0.000000   \n",
       "25%             3.000000        109.000000        100.000000   \n",
       "50%             5.000000        230.000000        220.000000   \n",
       "75%             9.000000        437.500000        428.000000   \n",
       "max           131.000000      35190.000000      40335.000000   \n",
       "\n",
       "        total_rech_amt_8  total_rech_amt_9  max_rech_amt_6  max_rech_amt_7  \\\n",
       "count       99999.000000      99999.000000    99999.000000    99999.000000   \n",
       "unique               NaN               NaN             NaN             NaN   \n",
       "top                  NaN               NaN             NaN             NaN   \n",
       "freq                 NaN               NaN             NaN             NaN   \n",
       "mean          324.157122        303.345673      104.637486      104.752398   \n",
       "std           416.540455        404.588583      120.614894      124.523970   \n",
       "min             0.000000          0.000000        0.000000        0.000000   \n",
       "25%            90.000000         52.000000       30.000000       30.000000   \n",
       "50%           225.000000        200.000000      110.000000      110.000000   \n",
       "75%           434.500000        415.000000      120.000000      128.000000   \n",
       "max         45320.000000      37235.000000     4010.000000     4010.000000   \n",
       "\n",
       "        max_rech_amt_8  max_rech_amt_9 date_of_last_rech_6  \\\n",
       "count     99999.000000    99999.000000               98392   \n",
       "unique             NaN             NaN                  30   \n",
       "top                NaN             NaN           6/30/2014   \n",
       "freq               NaN             NaN               16960   \n",
       "mean        107.728207      101.943889                 NaN   \n",
       "std         126.902505      125.375109                 NaN   \n",
       "min           0.000000        0.000000                 NaN   \n",
       "25%          30.000000       28.000000                 NaN   \n",
       "50%          98.000000       61.000000                 NaN   \n",
       "75%         144.000000      144.000000                 NaN   \n",
       "max        4449.000000     3399.000000                 NaN   \n",
       "\n",
       "       date_of_last_rech_7 date_of_last_rech_8 date_of_last_rech_9  \\\n",
       "count                98232               96377               95239   \n",
       "unique                  31                  31                  30   \n",
       "top              7/31/2014           8/31/2014           9/29/2014   \n",
       "freq                 17288               14706               22623   \n",
       "mean                   NaN                 NaN                 NaN   \n",
       "std                    NaN                 NaN                 NaN   \n",
       "min                    NaN                 NaN                 NaN   \n",
       "25%                    NaN                 NaN                 NaN   \n",
       "50%                    NaN                 NaN                 NaN   \n",
       "75%                    NaN                 NaN                 NaN   \n",
       "max                    NaN                 NaN                 NaN   \n",
       "\n",
       "        last_day_rch_amt_6  last_day_rch_amt_7  last_day_rch_amt_8  \\\n",
       "count         99999.000000        99999.000000        99999.000000   \n",
       "unique                 NaN                 NaN                 NaN   \n",
       "top                    NaN                 NaN                 NaN   \n",
       "freq                   NaN                 NaN                 NaN   \n",
       "mean             63.156252           59.385804           62.641716   \n",
       "std              97.356649           95.915385          104.431816   \n",
       "min               0.000000            0.000000            0.000000   \n",
       "25%               0.000000            0.000000            0.000000   \n",
       "50%              30.000000           30.000000           30.000000   \n",
       "75%             110.000000          110.000000          130.000000   \n",
       "max            4010.000000         4010.000000         4449.000000   \n",
       "\n",
       "        last_day_rch_amt_9 date_of_last_rech_data_6 date_of_last_rech_data_7  \\\n",
       "count         99999.000000                    25153                    25571   \n",
       "unique                 NaN                       30                       31   \n",
       "top                    NaN                6/30/2014                7/31/2014   \n",
       "freq                   NaN                     1888                     1813   \n",
       "mean             43.901249                      NaN                      NaN   \n",
       "std              90.809712                      NaN                      NaN   \n",
       "min               0.000000                      NaN                      NaN   \n",
       "25%               0.000000                      NaN                      NaN   \n",
       "50%               0.000000                      NaN                      NaN   \n",
       "75%              50.000000                      NaN                      NaN   \n",
       "max            3399.000000                      NaN                      NaN   \n",
       "\n",
       "       date_of_last_rech_data_8 date_of_last_rech_data_9  total_rech_data_6  \\\n",
       "count                     26339                    25922       25153.000000   \n",
       "unique                       31                       30                NaN   \n",
       "top                   8/31/2014                9/29/2014                NaN   \n",
       "freq                       1998                     2329                NaN   \n",
       "mean                        NaN                      NaN           2.463802   \n",
       "std                         NaN                      NaN           2.789128   \n",
       "min                         NaN                      NaN           1.000000   \n",
       "25%                         NaN                      NaN           1.000000   \n",
       "50%                         NaN                      NaN           1.000000   \n",
       "75%                         NaN                      NaN           3.000000   \n",
       "max                         NaN                      NaN          61.000000   \n",
       "\n",
       "        total_rech_data_7  total_rech_data_8  total_rech_data_9  \\\n",
       "count        25571.000000       26339.000000       25922.000000   \n",
       "unique                NaN                NaN                NaN   \n",
       "top                   NaN                NaN                NaN   \n",
       "freq                  NaN                NaN                NaN   \n",
       "mean             2.666419           2.651999           2.441170   \n",
       "std              3.031593           3.074987           2.516339   \n",
       "min              1.000000           1.000000           1.000000   \n",
       "25%              1.000000           1.000000           1.000000   \n",
       "50%              1.000000           1.000000           2.000000   \n",
       "75%              3.000000           3.000000           3.000000   \n",
       "max             54.000000          60.000000          84.000000   \n",
       "\n",
       "        max_rech_data_6  max_rech_data_7  max_rech_data_8  max_rech_data_9  \\\n",
       "count      25153.000000     25571.000000     26339.000000      25922.00000   \n",
       "unique              NaN              NaN              NaN              NaN   \n",
       "top                 NaN              NaN              NaN              NaN   \n",
       "freq                NaN              NaN              NaN              NaN   \n",
       "mean         126.393392       126.729459       125.717301        124.94144   \n",
       "std          108.477235       109.765267       109.437851        111.36376   \n",
       "min            1.000000         1.000000         1.000000          1.00000   \n",
       "25%           25.000000        25.000000        25.000000         25.00000   \n",
       "50%          145.000000       145.000000       145.000000        145.00000   \n",
       "75%          177.000000       177.000000       179.000000        179.00000   \n",
       "max         1555.000000      1555.000000      1555.000000       1555.00000   \n",
       "\n",
       "        count_rech_2g_6  count_rech_2g_7  count_rech_2g_8  count_rech_2g_9  \\\n",
       "count      25153.000000     25571.000000     26339.000000     25922.000000   \n",
       "unique              NaN              NaN              NaN              NaN   \n",
       "top                 NaN              NaN              NaN              NaN   \n",
       "freq                NaN              NaN              NaN              NaN   \n",
       "mean           1.864668         2.044699         2.016288         1.781807   \n",
       "std            2.570254         2.768332         2.720132         2.214701   \n",
       "min            0.000000         0.000000         0.000000         0.000000   \n",
       "25%            1.000000         1.000000         1.000000         1.000000   \n",
       "50%            1.000000         1.000000         1.000000         1.000000   \n",
       "75%            2.000000         2.000000         2.000000         2.000000   \n",
       "max           42.000000        48.000000        44.000000        40.000000   \n",
       "\n",
       "        count_rech_3g_6  count_rech_3g_7  count_rech_3g_8  count_rech_3g_9  \\\n",
       "count      25153.000000     25571.000000     26339.000000     25922.000000   \n",
       "unique              NaN              NaN              NaN              NaN   \n",
       "top                 NaN              NaN              NaN              NaN   \n",
       "freq                NaN              NaN              NaN              NaN   \n",
       "mean           0.599133         0.621720         0.635711         0.659363   \n",
       "std            1.274428         1.394524         1.422827         1.411513   \n",
       "min            0.000000         0.000000         0.000000         0.000000   \n",
       "25%            0.000000         0.000000         0.000000         0.000000   \n",
       "50%            0.000000         0.000000         0.000000         0.000000   \n",
       "75%            1.000000         1.000000         1.000000         1.000000   \n",
       "max           29.000000        35.000000        45.000000        49.000000   \n",
       "\n",
       "        av_rech_amt_data_6  av_rech_amt_data_7  av_rech_amt_data_8  \\\n",
       "count         25153.000000        25571.000000        26339.000000   \n",
       "unique                 NaN                 NaN                 NaN   \n",
       "top                    NaN                 NaN                 NaN   \n",
       "freq                   NaN                 NaN                 NaN   \n",
       "mean            192.600982          200.981292          197.526489   \n",
       "std             192.646318          196.791224          191.301305   \n",
       "min               1.000000            0.500000            0.500000   \n",
       "25%              82.000000           92.000000           87.000000   \n",
       "50%             154.000000          154.000000          154.000000   \n",
       "75%             252.000000          252.000000          252.000000   \n",
       "max            7546.000000         4365.000000         4076.000000   \n",
       "\n",
       "        av_rech_amt_data_9   vol_2g_mb_6   vol_2g_mb_7   vol_2g_mb_8  \\\n",
       "count         25922.000000  99999.000000  99999.000000  99999.000000   \n",
       "unique                 NaN           NaN           NaN           NaN   \n",
       "top                    NaN           NaN           NaN           NaN   \n",
       "freq                   NaN           NaN           NaN           NaN   \n",
       "mean            192.734315     51.904956     51.229937     50.170154   \n",
       "std             188.400286    213.356445    212.302217    212.347892   \n",
       "min               1.000000      0.000000      0.000000      0.000000   \n",
       "25%              69.000000      0.000000      0.000000      0.000000   \n",
       "50%             164.000000      0.000000      0.000000      0.000000   \n",
       "75%             252.000000      0.000000      0.000000      0.000000   \n",
       "max            4061.000000  10285.900000   7873.550000  11117.610000   \n",
       "\n",
       "         vol_2g_mb_9   vol_3g_mb_6   vol_3g_mb_7   vol_3g_mb_8   vol_3g_mb_9  \\\n",
       "count   99999.000000  99999.000000  99999.000000  99999.000000  99999.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean       44.719701    121.396219    128.995847    135.410689    136.056613   \n",
       "std       198.653570    544.247227    541.494013    558.775335    577.394194   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "75%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "max      8993.950000  45735.400000  28144.120000  30036.060000  39221.270000   \n",
       "\n",
       "           arpu_3g_6     arpu_3g_7     arpu_3g_8     arpu_3g_9     arpu_2g_6  \\\n",
       "count   25153.000000  25571.000000  26339.000000  25922.000000  25153.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean       89.555057     89.384120     91.173849    100.264116     86.398003   \n",
       "std       193.124653    195.893924    188.180936    216.291992    172.767523   \n",
       "min       -30.820000    -26.040000    -24.490000    -71.090000    -35.830000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         0.480000      0.420000      0.880000      2.605000     10.830000   \n",
       "75%       122.070000    119.560000    122.070000    140.010000    122.070000   \n",
       "max      6362.280000   4980.900000   3716.900000  13884.310000   6433.760000   \n",
       "\n",
       "           arpu_2g_7     arpu_2g_8     arpu_2g_9  night_pck_user_6  \\\n",
       "count   25571.000000  26339.000000  25922.000000      25153.000000   \n",
       "unique           NaN           NaN           NaN               NaN   \n",
       "top              NaN           NaN           NaN               NaN   \n",
       "freq             NaN           NaN           NaN               NaN   \n",
       "mean       85.914450     86.599478     93.712026          0.025086   \n",
       "std       176.379871    168.247852    171.384224          0.156391   \n",
       "min       -15.480000    -55.830000    -45.740000          0.000000   \n",
       "25%         0.000000      0.000000      0.000000          0.000000   \n",
       "50%         8.810000      9.270000     14.800000          0.000000   \n",
       "75%       122.070000    122.070000    140.010000          0.000000   \n",
       "max      4809.360000   3483.170000   3467.170000          1.000000   \n",
       "\n",
       "        night_pck_user_7  night_pck_user_8  night_pck_user_9  monthly_2g_6  \\\n",
       "count       25571.000000      26339.000000      25922.000000  99999.000000   \n",
       "unique               NaN               NaN               NaN           NaN   \n",
       "top                  NaN               NaN               NaN           NaN   \n",
       "freq                 NaN               NaN               NaN           NaN   \n",
       "mean            0.023034          0.020844          0.015971      0.079641   \n",
       "std             0.150014          0.142863          0.125366      0.295058   \n",
       "min             0.000000          0.000000          0.000000      0.000000   \n",
       "25%             0.000000          0.000000          0.000000      0.000000   \n",
       "50%             0.000000          0.000000          0.000000      0.000000   \n",
       "75%             0.000000          0.000000          0.000000      0.000000   \n",
       "max             1.000000          1.000000          1.000000      4.000000   \n",
       "\n",
       "        monthly_2g_7  monthly_2g_8  monthly_2g_9   sachet_2g_6   sachet_2g_7  \\\n",
       "count   99999.000000  99999.000000  99999.000000  99999.000000  99999.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean        0.083221      0.081001      0.068781      0.389384      0.439634   \n",
       "std         0.304395      0.299568      0.278120      1.497320      1.636230   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "75%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "max         5.000000      5.000000      4.000000     42.000000     48.000000   \n",
       "\n",
       "         sachet_2g_8   sachet_2g_9  monthly_3g_6  monthly_3g_7  monthly_3g_8  \\\n",
       "count   99999.000000  99999.000000  99999.000000  99999.000000  99999.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean        0.450075      0.393104      0.075921      0.078581      0.082941   \n",
       "std         1.630263      1.347140      0.363371      0.387231      0.384947   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "75%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "max        44.000000     40.000000     14.000000     16.000000     16.000000   \n",
       "\n",
       "        monthly_3g_9   sachet_3g_6   sachet_3g_7   sachet_3g_8   sachet_3g_9  \\\n",
       "count   99999.000000  99999.000000  99999.000000  99999.000000  99999.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean        0.086341      0.074781      0.080401      0.084501      0.084581   \n",
       "std         0.384978      0.568344      0.628334      0.660234      0.650457   \n",
       "min         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "25%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "50%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "75%         0.000000      0.000000      0.000000      0.000000      0.000000   \n",
       "max        11.000000     29.000000     35.000000     41.000000     49.000000   \n",
       "\n",
       "           fb_user_6     fb_user_7     fb_user_8     fb_user_9           aon  \\\n",
       "count   25153.000000  25571.000000  26339.000000  25922.000000  99999.000000   \n",
       "unique           NaN           NaN           NaN           NaN           NaN   \n",
       "top              NaN           NaN           NaN           NaN           NaN   \n",
       "freq             NaN           NaN           NaN           NaN           NaN   \n",
       "mean        0.914404      0.908764      0.890808      0.860968   1219.854749   \n",
       "std         0.279772      0.287950      0.311885      0.345987    954.733842   \n",
       "min         0.000000      0.000000      0.000000      0.000000    180.000000   \n",
       "25%         1.000000      1.000000      1.000000      1.000000    467.000000   \n",
       "50%         1.000000      1.000000      1.000000      1.000000    863.000000   \n",
       "75%         1.000000      1.000000      1.000000      1.000000   1807.500000   \n",
       "max         1.000000      1.000000      1.000000      1.000000   4337.000000   \n",
       "\n",
       "          aug_vbc_3g    jul_vbc_3g    jun_vbc_3g    sep_vbc_3g  \n",
       "count   99999.000000  99999.000000  99999.000000  99999.000000  \n",
       "unique           NaN           NaN           NaN           NaN  \n",
       "top              NaN           NaN           NaN           NaN  \n",
       "freq             NaN           NaN           NaN           NaN  \n",
       "mean       68.170248     66.839062     60.021204      3.299373  \n",
       "std       267.580450    271.201856    253.938223     32.408353  \n",
       "min         0.000000      0.000000      0.000000      0.000000  \n",
       "25%         0.000000      0.000000      0.000000      0.000000  \n",
       "50%         0.000000      0.000000      0.000000      0.000000  \n",
       "75%         0.000000      0.000000      0.000000      0.000000  \n",
       "max     12916.220000   9165.600000  11166.210000   2618.570000  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# look at data statistics\n",
    "churn.describe(include='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create backup of data\n",
    "original = churn.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "#ID cols: 2\n",
      "#Date cols:12\n",
      "#Numeric cols:204\n",
      "#Category cols:8\n",
      "True\n"
     ]
    }
   ],
   "source": [
    "# create column name list by types of columns\n",
    "id_cols = ['mobile_number', 'circle_id']\n",
    "\n",
    "date_cols = ['last_date_of_month_6',\n",
    "             'last_date_of_month_7',\n",
    "             'last_date_of_month_8',\n",
    "             'last_date_of_month_9',\n",
    "             'date_of_last_rech_6',\n",
    "             'date_of_last_rech_7',\n",
    "             'date_of_last_rech_8',\n",
    "             'date_of_last_rech_9',\n",
    "             'date_of_last_rech_data_6',\n",
    "             'date_of_last_rech_data_7',\n",
    "             'date_of_last_rech_data_8',\n",
    "             'date_of_last_rech_data_9'\n",
    "            ]\n",
    "\n",
    "cat_cols =  ['night_pck_user_6',\n",
    "             'night_pck_user_7',\n",
    "             'night_pck_user_8',\n",
    "             'night_pck_user_9',\n",
    "             'fb_user_6',\n",
    "             'fb_user_7',\n",
    "             'fb_user_8',\n",
    "             'fb_user_9'\n",
    "            ]\n",
    "\n",
    "num_cols = [column for column in churn.columns if column not in id_cols + date_cols + cat_cols]\n",
    "\n",
    "# print the number of columns in each list\n",
    "print(\"#ID cols: %d\\n#Date cols:%d\\n#Numeric cols:%d\\n#Category cols:%d\" % (len(id_cols), len(date_cols), len(num_cols), len(cat_cols)))\n",
    "\n",
    "# check if we have missed any column or not\n",
    "print(len(id_cols) + len(date_cols) + len(num_cols) + len(cat_cols) == churn.shape[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Handling missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "mobile_number                0.000000\n",
       "circle_id                    0.000000\n",
       "loc_og_t2o_mou               1.018010\n",
       "std_og_t2o_mou               1.018010\n",
       "loc_ic_t2o_mou               1.018010\n",
       "last_date_of_month_6         0.000000\n",
       "last_date_of_month_7         0.601006\n",
       "last_date_of_month_8         1.100011\n",
       "last_date_of_month_9         1.659017\n",
       "arpu_6                       0.000000\n",
       "arpu_7                       0.000000\n",
       "arpu_8                       0.000000\n",
       "arpu_9                       0.000000\n",
       "onnet_mou_6                  3.937039\n",
       "onnet_mou_7                  3.859039\n",
       "onnet_mou_8                  5.378054\n",
       "onnet_mou_9                  7.745077\n",
       "offnet_mou_6                 3.937039\n",
       "offnet_mou_7                 3.859039\n",
       "offnet_mou_8                 5.378054\n",
       "offnet_mou_9                 7.745077\n",
       "roam_ic_mou_6                3.937039\n",
       "roam_ic_mou_7                3.859039\n",
       "roam_ic_mou_8                5.378054\n",
       "roam_ic_mou_9                7.745077\n",
       "roam_og_mou_6                3.937039\n",
       "roam_og_mou_7                3.859039\n",
       "roam_og_mou_8                5.378054\n",
       "roam_og_mou_9                7.745077\n",
       "loc_og_t2t_mou_6             3.937039\n",
       "loc_og_t2t_mou_7             3.859039\n",
       "loc_og_t2t_mou_8             5.378054\n",
       "loc_og_t2t_mou_9             7.745077\n",
       "loc_og_t2m_mou_6             3.937039\n",
       "loc_og_t2m_mou_7             3.859039\n",
       "loc_og_t2m_mou_8             5.378054\n",
       "loc_og_t2m_mou_9             7.745077\n",
       "loc_og_t2f_mou_6             3.937039\n",
       "loc_og_t2f_mou_7             3.859039\n",
       "loc_og_t2f_mou_8             5.378054\n",
       "loc_og_t2f_mou_9             7.745077\n",
       "loc_og_t2c_mou_6             3.937039\n",
       "loc_og_t2c_mou_7             3.859039\n",
       "loc_og_t2c_mou_8             5.378054\n",
       "loc_og_t2c_mou_9             7.745077\n",
       "loc_og_mou_6                 3.937039\n",
       "loc_og_mou_7                 3.859039\n",
       "loc_og_mou_8                 5.378054\n",
       "loc_og_mou_9                 7.745077\n",
       "std_og_t2t_mou_6             3.937039\n",
       "std_og_t2t_mou_7             3.859039\n",
       "std_og_t2t_mou_8             5.378054\n",
       "std_og_t2t_mou_9             7.745077\n",
       "std_og_t2m_mou_6             3.937039\n",
       "std_og_t2m_mou_7             3.859039\n",
       "std_og_t2m_mou_8             5.378054\n",
       "std_og_t2m_mou_9             7.745077\n",
       "std_og_t2f_mou_6             3.937039\n",
       "std_og_t2f_mou_7             3.859039\n",
       "std_og_t2f_mou_8             5.378054\n",
       "std_og_t2f_mou_9             7.745077\n",
       "std_og_t2c_mou_6             3.937039\n",
       "std_og_t2c_mou_7             3.859039\n",
       "std_og_t2c_mou_8             5.378054\n",
       "std_og_t2c_mou_9             7.745077\n",
       "std_og_mou_6                 3.937039\n",
       "std_og_mou_7                 3.859039\n",
       "std_og_mou_8                 5.378054\n",
       "std_og_mou_9                 7.745077\n",
       "isd_og_mou_6                 3.937039\n",
       "isd_og_mou_7                 3.859039\n",
       "isd_og_mou_8                 5.378054\n",
       "isd_og_mou_9                 7.745077\n",
       "spl_og_mou_6                 3.937039\n",
       "spl_og_mou_7                 3.859039\n",
       "spl_og_mou_8                 5.378054\n",
       "spl_og_mou_9                 7.745077\n",
       "og_others_6                  3.937039\n",
       "og_others_7                  3.859039\n",
       "og_others_8                  5.378054\n",
       "og_others_9                  7.745077\n",
       "total_og_mou_6               0.000000\n",
       "total_og_mou_7               0.000000\n",
       "total_og_mou_8               0.000000\n",
       "total_og_mou_9               0.000000\n",
       "loc_ic_t2t_mou_6             3.937039\n",
       "loc_ic_t2t_mou_7             3.859039\n",
       "loc_ic_t2t_mou_8             5.378054\n",
       "loc_ic_t2t_mou_9             7.745077\n",
       "loc_ic_t2m_mou_6             3.937039\n",
       "loc_ic_t2m_mou_7             3.859039\n",
       "loc_ic_t2m_mou_8             5.378054\n",
       "loc_ic_t2m_mou_9             7.745077\n",
       "loc_ic_t2f_mou_6             3.937039\n",
       "loc_ic_t2f_mou_7             3.859039\n",
       "loc_ic_t2f_mou_8             5.378054\n",
       "loc_ic_t2f_mou_9             7.745077\n",
       "loc_ic_mou_6                 3.937039\n",
       "loc_ic_mou_7                 3.859039\n",
       "loc_ic_mou_8                 5.378054\n",
       "loc_ic_mou_9                 7.745077\n",
       "std_ic_t2t_mou_6             3.937039\n",
       "std_ic_t2t_mou_7             3.859039\n",
       "std_ic_t2t_mou_8             5.378054\n",
       "std_ic_t2t_mou_9             7.745077\n",
       "std_ic_t2m_mou_6             3.937039\n",
       "std_ic_t2m_mou_7             3.859039\n",
       "std_ic_t2m_mou_8             5.378054\n",
       "std_ic_t2m_mou_9             7.745077\n",
       "std_ic_t2f_mou_6             3.937039\n",
       "std_ic_t2f_mou_7             3.859039\n",
       "std_ic_t2f_mou_8             5.378054\n",
       "std_ic_t2f_mou_9             7.745077\n",
       "std_ic_t2o_mou_6             3.937039\n",
       "std_ic_t2o_mou_7             3.859039\n",
       "std_ic_t2o_mou_8             5.378054\n",
       "std_ic_t2o_mou_9             7.745077\n",
       "std_ic_mou_6                 3.937039\n",
       "std_ic_mou_7                 3.859039\n",
       "std_ic_mou_8                 5.378054\n",
       "std_ic_mou_9                 7.745077\n",
       "total_ic_mou_6               0.000000\n",
       "total_ic_mou_7               0.000000\n",
       "total_ic_mou_8               0.000000\n",
       "total_ic_mou_9               0.000000\n",
       "spl_ic_mou_6                 3.937039\n",
       "spl_ic_mou_7                 3.859039\n",
       "spl_ic_mou_8                 5.378054\n",
       "spl_ic_mou_9                 7.745077\n",
       "isd_ic_mou_6                 3.937039\n",
       "isd_ic_mou_7                 3.859039\n",
       "isd_ic_mou_8                 5.378054\n",
       "isd_ic_mou_9                 7.745077\n",
       "ic_others_6                  3.937039\n",
       "ic_others_7                  3.859039\n",
       "ic_others_8                  5.378054\n",
       "ic_others_9                  7.745077\n",
       "total_rech_num_6             0.000000\n",
       "total_rech_num_7             0.000000\n",
       "total_rech_num_8             0.000000\n",
       "total_rech_num_9             0.000000\n",
       "total_rech_amt_6             0.000000\n",
       "total_rech_amt_7             0.000000\n",
       "total_rech_amt_8             0.000000\n",
       "total_rech_amt_9             0.000000\n",
       "max_rech_amt_6               0.000000\n",
       "max_rech_amt_7               0.000000\n",
       "max_rech_amt_8               0.000000\n",
       "max_rech_amt_9               0.000000\n",
       "date_of_last_rech_6          1.607016\n",
       "date_of_last_rech_7          1.767018\n",
       "date_of_last_rech_8          3.622036\n",
       "date_of_last_rech_9          4.760048\n",
       "last_day_rch_amt_6           0.000000\n",
       "last_day_rch_amt_7           0.000000\n",
       "last_day_rch_amt_8           0.000000\n",
       "last_day_rch_amt_9           0.000000\n",
       "date_of_last_rech_data_6    74.846748\n",
       "date_of_last_rech_data_7    74.428744\n",
       "date_of_last_rech_data_8    73.660737\n",
       "date_of_last_rech_data_9    74.077741\n",
       "total_rech_data_6           74.846748\n",
       "total_rech_data_7           74.428744\n",
       "total_rech_data_8           73.660737\n",
       "total_rech_data_9           74.077741\n",
       "max_rech_data_6             74.846748\n",
       "max_rech_data_7             74.428744\n",
       "max_rech_data_8             73.660737\n",
       "max_rech_data_9             74.077741\n",
       "count_rech_2g_6             74.846748\n",
       "count_rech_2g_7             74.428744\n",
       "count_rech_2g_8             73.660737\n",
       "count_rech_2g_9             74.077741\n",
       "count_rech_3g_6             74.846748\n",
       "count_rech_3g_7             74.428744\n",
       "count_rech_3g_8             73.660737\n",
       "count_rech_3g_9             74.077741\n",
       "av_rech_amt_data_6          74.846748\n",
       "av_rech_amt_data_7          74.428744\n",
       "av_rech_amt_data_8          73.660737\n",
       "av_rech_amt_data_9          74.077741\n",
       "vol_2g_mb_6                  0.000000\n",
       "vol_2g_mb_7                  0.000000\n",
       "vol_2g_mb_8                  0.000000\n",
       "vol_2g_mb_9                  0.000000\n",
       "vol_3g_mb_6                  0.000000\n",
       "vol_3g_mb_7                  0.000000\n",
       "vol_3g_mb_8                  0.000000\n",
       "vol_3g_mb_9                  0.000000\n",
       "arpu_3g_6                   74.846748\n",
       "arpu_3g_7                   74.428744\n",
       "arpu_3g_8                   73.660737\n",
       "arpu_3g_9                   74.077741\n",
       "arpu_2g_6                   74.846748\n",
       "arpu_2g_7                   74.428744\n",
       "arpu_2g_8                   73.660737\n",
       "arpu_2g_9                   74.077741\n",
       "night_pck_user_6            74.846748\n",
       "night_pck_user_7            74.428744\n",
       "night_pck_user_8            73.660737\n",
       "night_pck_user_9            74.077741\n",
       "monthly_2g_6                 0.000000\n",
       "monthly_2g_7                 0.000000\n",
       "monthly_2g_8                 0.000000\n",
       "monthly_2g_9                 0.000000\n",
       "sachet_2g_6                  0.000000\n",
       "sachet_2g_7                  0.000000\n",
       "sachet_2g_8                  0.000000\n",
       "sachet_2g_9                  0.000000\n",
       "monthly_3g_6                 0.000000\n",
       "monthly_3g_7                 0.000000\n",
       "monthly_3g_8                 0.000000\n",
       "monthly_3g_9                 0.000000\n",
       "sachet_3g_6                  0.000000\n",
       "sachet_3g_7                  0.000000\n",
       "sachet_3g_8                  0.000000\n",
       "sachet_3g_9                  0.000000\n",
       "fb_user_6                   74.846748\n",
       "fb_user_7                   74.428744\n",
       "fb_user_8                   73.660737\n",
       "fb_user_9                   74.077741\n",
       "aon                          0.000000\n",
       "aug_vbc_3g                   0.000000\n",
       "jul_vbc_3g                   0.000000\n",
       "jun_vbc_3g                   0.000000\n",
       "sep_vbc_3g                   0.000000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# look at missing value ratio in each column\n",
    "churn.isnull().sum()*100/churn.shape[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# impute missing values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## i) Imputing with zeroes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_rech_data_6</th>\n",
       "      <th>total_rech_data_7</th>\n",
       "      <th>total_rech_data_8</th>\n",
       "      <th>total_rech_data_9</th>\n",
       "      <th>count_rech_2g_6</th>\n",
       "      <th>count_rech_2g_7</th>\n",
       "      <th>count_rech_2g_8</th>\n",
       "      <th>count_rech_2g_9</th>\n",
       "      <th>count_rech_3g_6</th>\n",
       "      <th>count_rech_3g_7</th>\n",
       "      <th>count_rech_3g_8</th>\n",
       "      <th>count_rech_3g_9</th>\n",
       "      <th>max_rech_data_6</th>\n",
       "      <th>max_rech_data_7</th>\n",
       "      <th>max_rech_data_8</th>\n",
       "      <th>max_rech_data_9</th>\n",
       "      <th>av_rech_amt_data_6</th>\n",
       "      <th>av_rech_amt_data_7</th>\n",
       "      <th>av_rech_amt_data_8</th>\n",
       "      <th>av_rech_amt_data_9</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.00000</td>\n",
       "      <td>25153.000000</td>\n",
       "      <td>25571.000000</td>\n",
       "      <td>26339.000000</td>\n",
       "      <td>25922.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>2.463802</td>\n",
       "      <td>2.666419</td>\n",
       "      <td>2.651999</td>\n",
       "      <td>2.441170</td>\n",
       "      <td>1.864668</td>\n",
       "      <td>2.044699</td>\n",
       "      <td>2.016288</td>\n",
       "      <td>1.781807</td>\n",
       "      <td>0.599133</td>\n",
       "      <td>0.621720</td>\n",
       "      <td>0.635711</td>\n",
       "      <td>0.659363</td>\n",
       "      <td>126.393392</td>\n",
       "      <td>126.729459</td>\n",
       "      <td>125.717301</td>\n",
       "      <td>124.94144</td>\n",
       "      <td>192.600982</td>\n",
       "      <td>200.981292</td>\n",
       "      <td>197.526489</td>\n",
       "      <td>192.734315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.789128</td>\n",
       "      <td>3.031593</td>\n",
       "      <td>3.074987</td>\n",
       "      <td>2.516339</td>\n",
       "      <td>2.570254</td>\n",
       "      <td>2.768332</td>\n",
       "      <td>2.720132</td>\n",
       "      <td>2.214701</td>\n",
       "      <td>1.274428</td>\n",
       "      <td>1.394524</td>\n",
       "      <td>1.422827</td>\n",
       "      <td>1.411513</td>\n",
       "      <td>108.477235</td>\n",
       "      <td>109.765267</td>\n",
       "      <td>109.437851</td>\n",
       "      <td>111.36376</td>\n",
       "      <td>192.646318</td>\n",
       "      <td>196.791224</td>\n",
       "      <td>191.301305</td>\n",
       "      <td>188.400286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.00000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>25.00000</td>\n",
       "      <td>82.000000</td>\n",
       "      <td>92.000000</td>\n",
       "      <td>87.000000</td>\n",
       "      <td>69.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>145.00000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>164.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>177.000000</td>\n",
       "      <td>177.000000</td>\n",
       "      <td>179.000000</td>\n",
       "      <td>179.00000</td>\n",
       "      <td>252.000000</td>\n",
       "      <td>252.000000</td>\n",
       "      <td>252.000000</td>\n",
       "      <td>252.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>61.000000</td>\n",
       "      <td>54.000000</td>\n",
       "      <td>60.000000</td>\n",
       "      <td>84.000000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>44.000000</td>\n",
       "      <td>40.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>49.000000</td>\n",
       "      <td>1555.000000</td>\n",
       "      <td>1555.000000</td>\n",
       "      <td>1555.000000</td>\n",
       "      <td>1555.00000</td>\n",
       "      <td>7546.000000</td>\n",
       "      <td>4365.000000</td>\n",
       "      <td>4076.000000</td>\n",
       "      <td>4061.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       total_rech_data_6  total_rech_data_7  total_rech_data_8  \\\n",
       "count       25153.000000       25571.000000       26339.000000   \n",
       "mean            2.463802           2.666419           2.651999   \n",
       "std             2.789128           3.031593           3.074987   \n",
       "min             1.000000           1.000000           1.000000   \n",
       "25%             1.000000           1.000000           1.000000   \n",
       "50%             1.000000           1.000000           1.000000   \n",
       "75%             3.000000           3.000000           3.000000   \n",
       "max            61.000000          54.000000          60.000000   \n",
       "\n",
       "       total_rech_data_9  count_rech_2g_6  count_rech_2g_7  count_rech_2g_8  \\\n",
       "count       25922.000000     25153.000000     25571.000000     26339.000000   \n",
       "mean            2.441170         1.864668         2.044699         2.016288   \n",
       "std             2.516339         2.570254         2.768332         2.720132   \n",
       "min             1.000000         0.000000         0.000000         0.000000   \n",
       "25%             1.000000         1.000000         1.000000         1.000000   \n",
       "50%             2.000000         1.000000         1.000000         1.000000   \n",
       "75%             3.000000         2.000000         2.000000         2.000000   \n",
       "max            84.000000        42.000000        48.000000        44.000000   \n",
       "\n",
       "       count_rech_2g_9  count_rech_3g_6  count_rech_3g_7  count_rech_3g_8  \\\n",
       "count     25922.000000     25153.000000     25571.000000     26339.000000   \n",
       "mean          1.781807         0.599133         0.621720         0.635711   \n",
       "std           2.214701         1.274428         1.394524         1.422827   \n",
       "min           0.000000         0.000000         0.000000         0.000000   \n",
       "25%           1.000000         0.000000         0.000000         0.000000   \n",
       "50%           1.000000         0.000000         0.000000         0.000000   \n",
       "75%           2.000000         1.000000         1.000000         1.000000   \n",
       "max          40.000000        29.000000        35.000000        45.000000   \n",
       "\n",
       "       count_rech_3g_9  max_rech_data_6  max_rech_data_7  max_rech_data_8  \\\n",
       "count     25922.000000     25153.000000     25571.000000     26339.000000   \n",
       "mean          0.659363       126.393392       126.729459       125.717301   \n",
       "std           1.411513       108.477235       109.765267       109.437851   \n",
       "min           0.000000         1.000000         1.000000         1.000000   \n",
       "25%           0.000000        25.000000        25.000000        25.000000   \n",
       "50%           0.000000       145.000000       145.000000       145.000000   \n",
       "75%           1.000000       177.000000       177.000000       179.000000   \n",
       "max          49.000000      1555.000000      1555.000000      1555.000000   \n",
       "\n",
       "       max_rech_data_9  av_rech_amt_data_6  av_rech_amt_data_7  \\\n",
       "count      25922.00000        25153.000000        25571.000000   \n",
       "mean         124.94144          192.600982          200.981292   \n",
       "std          111.36376          192.646318          196.791224   \n",
       "min            1.00000            1.000000            0.500000   \n",
       "25%           25.00000           82.000000           92.000000   \n",
       "50%          145.00000          154.000000          154.000000   \n",
       "75%          179.00000          252.000000          252.000000   \n",
       "max         1555.00000         7546.000000         4365.000000   \n",
       "\n",
       "       av_rech_amt_data_8  av_rech_amt_data_9  \n",
       "count        26339.000000        25922.000000  \n",
       "mean           197.526489          192.734315  \n",
       "std            191.301305          188.400286  \n",
       "min              0.500000            1.000000  \n",
       "25%             87.000000           69.000000  \n",
       "50%            154.000000          164.000000  \n",
       "75%            252.000000          252.000000  \n",
       "max           4076.000000         4061.000000  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# some recharge columns have minimum value of 1 while some don't\n",
    "recharge_cols = ['total_rech_data_6', 'total_rech_data_7', 'total_rech_data_8', 'total_rech_data_9',\n",
    "                 'count_rech_2g_6', 'count_rech_2g_7', 'count_rech_2g_8', 'count_rech_2g_9',\n",
    "                 'count_rech_3g_6', 'count_rech_3g_7', 'count_rech_3g_8', 'count_rech_3g_9',\n",
    "                 'max_rech_data_6', 'max_rech_data_7', 'max_rech_data_8', 'max_rech_data_9',\n",
    "                 'av_rech_amt_data_6', 'av_rech_amt_data_7', 'av_rech_amt_data_8', 'av_rech_amt_data_9',\n",
    "                 ]\n",
    "\n",
    "churn[recharge_cols].describe(include='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_rech_data_6</th>\n",
       "      <th>date_of_last_rech_data_6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    total_rech_data_6 date_of_last_rech_data_6\n",
       "1                 NaN                      NaN\n",
       "2                 NaN                      NaN\n",
       "3                 NaN                      NaN\n",
       "5                 NaN                      NaN\n",
       "6                 NaN                      NaN\n",
       "7                 NaN                      NaN\n",
       "8                 NaN                      NaN\n",
       "9                 NaN                      NaN\n",
       "10                NaN                      NaN\n",
       "11                NaN                      NaN\n",
       "12                NaN                      NaN\n",
       "13                NaN                      NaN\n",
       "14                NaN                      NaN\n",
       "15                NaN                      NaN\n",
       "16                NaN                      NaN\n",
       "17                NaN                      NaN\n",
       "18                NaN                      NaN\n",
       "20                NaN                      NaN\n",
       "21                NaN                      NaN\n",
       "22                NaN                      NaN"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# It is also observed that the recharge date and the recharge value are missing together which means the customer didn't recharge\n",
    "churn.loc[churn.total_rech_data_6.isnull() & churn.date_of_last_rech_data_6.isnull(), [\"total_rech_data_6\", \"date_of_last_rech_data_6\"]].head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the recharge variables where minumum value is 1, we can impute missing values with zeroes since it means customer didn't recharge their numbere that month."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create a list of recharge columns where we will impute missing values with zeroes\n",
    "zero_impute = ['total_rech_data_6', 'total_rech_data_7', 'total_rech_data_8', 'total_rech_data_9',\n",
    "        'av_rech_amt_data_6', 'av_rech_amt_data_7', 'av_rech_amt_data_8', 'av_rech_amt_data_9',\n",
    "        'max_rech_data_6', 'max_rech_data_7', 'max_rech_data_8', 'max_rech_data_9'\n",
    "       ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# impute missing values with 0\n",
    "churn[zero_impute] = churn[zero_impute].apply(lambda x: x.fillna(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing value ratio:\n",
      "\n",
      "total_rech_data_6     0.0\n",
      "total_rech_data_7     0.0\n",
      "total_rech_data_8     0.0\n",
      "total_rech_data_9     0.0\n",
      "av_rech_amt_data_6    0.0\n",
      "av_rech_amt_data_7    0.0\n",
      "av_rech_amt_data_8    0.0\n",
      "av_rech_amt_data_9    0.0\n",
      "max_rech_data_6       0.0\n",
      "max_rech_data_7       0.0\n",
      "max_rech_data_8       0.0\n",
      "max_rech_data_9       0.0\n",
      "dtype: float64\n",
      "\n",
      "\n",
      "Summary statistics\n",
      "\n",
      "       total_rech_data_6  total_rech_data_7  total_rech_data_8  \\\n",
      "count       99999.000000       99999.000000       99999.000000   \n",
      "mean            0.619726           0.681837           0.698517   \n",
      "std             1.760541           1.924382           1.963417   \n",
      "min             0.000000           0.000000           0.000000   \n",
      "25%             0.000000           0.000000           0.000000   \n",
      "50%             0.000000           0.000000           0.000000   \n",
      "75%             1.000000           1.000000           1.000000   \n",
      "max            61.000000          54.000000          60.000000   \n",
      "\n",
      "       total_rech_data_9  av_rech_amt_data_6  av_rech_amt_data_7  \\\n",
      "count       99999.000000        99999.000000        99999.000000   \n",
      "mean            0.632806           48.445409           51.393440   \n",
      "std             1.669040          127.743863          132.629365   \n",
      "min             0.000000            0.000000            0.000000   \n",
      "25%             0.000000            0.000000            0.000000   \n",
      "50%             0.000000            0.000000            0.000000   \n",
      "75%             1.000000            8.250000           17.000000   \n",
      "max            84.000000         7546.000000         4365.000000   \n",
      "\n",
      "       av_rech_amt_data_8  av_rech_amt_data_9  max_rech_data_6  \\\n",
      "count        99999.000000        99999.000000     99999.000000   \n",
      "mean            52.027022           49.961089        31.792048   \n",
      "std            131.182609          127.804280        77.248778   \n",
      "min              0.000000            0.000000         0.000000   \n",
      "25%              0.000000            0.000000         0.000000   \n",
      "50%              0.000000            0.000000         0.000000   \n",
      "75%             23.000000           17.000000         8.000000   \n",
      "max           4076.000000         4061.000000      1555.000000   \n",
      "\n",
      "       max_rech_data_7  max_rech_data_8  max_rech_data_9  \n",
      "count     99999.000000     99999.000000     99999.000000  \n",
      "mean         32.406314        33.113011        32.387644  \n",
      "std          78.342435        78.872739        78.818696  \n",
      "min           0.000000         0.000000         0.000000  \n",
      "25%           0.000000         0.000000         0.000000  \n",
      "50%           0.000000         0.000000         0.000000  \n",
      "75%          14.000000        17.000000        17.000000  \n",
      "max        1555.000000      1555.000000      1555.000000  \n"
     ]
    }
   ],
   "source": [
    "# now, let's make sure values are imputed correctly\n",
    "print(\"Missing value ratio:\\n\")\n",
    "print(churn[zero_impute].isnull().sum()*100/churn.shape[1])\n",
    "\n",
    "# summary\n",
    "print(\"\\n\\nSummary statistics\\n\")\n",
    "print(churn[zero_impute].describe(include='all'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape before dropping:  (99999, 226)\n",
      "Shape after dropping:  (99999, 212)\n"
     ]
    }
   ],
   "source": [
    "# drop id and date columns\n",
    "print(\"Shape before dropping: \", churn.shape)\n",
    "churn = churn.drop(id_cols + date_cols, axis=1)\n",
    "print(\"Shape after dropping: \", churn.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ii) Replace NaN values in categorical variables"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will replace missing values in the categorical values with '-1' where '-1' will be a new category."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# replace missing values with '-1' in categorical columns\n",
    "churn[cat_cols] = churn[cat_cols].apply(lambda x: x.fillna(-1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing value ratio:\n",
      "\n",
      "night_pck_user_6    0.0\n",
      "night_pck_user_7    0.0\n",
      "night_pck_user_8    0.0\n",
      "night_pck_user_9    0.0\n",
      "fb_user_6           0.0\n",
      "fb_user_7           0.0\n",
      "fb_user_8           0.0\n",
      "fb_user_9           0.0\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# missing value ratio\n",
    "print(\"Missing value ratio:\\n\")\n",
    "print(churn[cat_cols].isnull().sum()*100/churn.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## iii) Drop variables with more than a given threshold of missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>features</th>\n",
       "      <th>include</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>loc_og_t2o_mou</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>std_og_t2o_mou</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>loc_ic_t2o_mou</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>arpu_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>arpu_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>arpu_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>arpu_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>onnet_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>onnet_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>onnet_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>onnet_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>offnet_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>offnet_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>offnet_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>offnet_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>roam_ic_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>roam_ic_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>roam_ic_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>roam_ic_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>roam_og_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>roam_og_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>roam_og_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>roam_og_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>loc_og_t2t_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>loc_og_t2t_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>loc_og_t2t_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>loc_og_t2t_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>loc_og_t2m_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>loc_og_t2m_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>loc_og_t2m_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>loc_og_t2m_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>loc_og_t2f_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>loc_og_t2f_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>loc_og_t2f_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>loc_og_t2f_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>loc_og_t2c_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>loc_og_t2c_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>loc_og_t2c_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>loc_og_t2c_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>loc_og_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>loc_og_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>loc_og_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>loc_og_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>std_og_t2t_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>std_og_t2t_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>std_og_t2t_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>std_og_t2t_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>std_og_t2m_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>std_og_t2m_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>std_og_t2m_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>std_og_t2m_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>std_og_t2f_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>std_og_t2f_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>std_og_t2f_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>std_og_t2f_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>std_og_t2c_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>std_og_t2c_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>std_og_t2c_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>std_og_t2c_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>std_og_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>std_og_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>std_og_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>std_og_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>isd_og_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>isd_og_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>isd_og_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>isd_og_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>spl_og_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>spl_og_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>spl_og_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>spl_og_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>og_others_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>og_others_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>og_others_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>og_others_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>total_og_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>total_og_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>total_og_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>total_og_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>loc_ic_t2t_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>loc_ic_t2t_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>loc_ic_t2t_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>loc_ic_t2t_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>loc_ic_t2m_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>loc_ic_t2m_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>loc_ic_t2m_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>loc_ic_t2m_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>loc_ic_t2f_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>loc_ic_t2f_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>loc_ic_t2f_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>loc_ic_t2f_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>loc_ic_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>loc_ic_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>loc_ic_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>loc_ic_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>std_ic_t2t_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>std_ic_t2t_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>std_ic_t2t_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>std_ic_t2t_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>std_ic_t2m_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>std_ic_t2m_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>std_ic_t2m_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>std_ic_t2m_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>std_ic_t2f_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>std_ic_t2f_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>std_ic_t2f_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>std_ic_t2f_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>std_ic_t2o_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>std_ic_t2o_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>std_ic_t2o_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>std_ic_t2o_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>std_ic_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112</th>\n",
       "      <td>std_ic_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113</th>\n",
       "      <td>std_ic_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>std_ic_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>total_ic_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>total_ic_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>total_ic_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>total_ic_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>spl_ic_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>spl_ic_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>spl_ic_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>spl_ic_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123</th>\n",
       "      <td>isd_ic_mou_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>isd_ic_mou_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>isd_ic_mou_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>isd_ic_mou_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>ic_others_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>ic_others_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>ic_others_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>ic_others_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>total_rech_num_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>total_rech_num_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>total_rech_num_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>total_rech_num_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>total_rech_amt_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>total_rech_amt_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137</th>\n",
       "      <td>total_rech_amt_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>total_rech_amt_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>max_rech_amt_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>max_rech_amt_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>max_rech_amt_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>max_rech_amt_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>last_day_rch_amt_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>last_day_rch_amt_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>last_day_rch_amt_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>last_day_rch_amt_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>total_rech_data_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>total_rech_data_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>total_rech_data_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>total_rech_data_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>max_rech_data_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>max_rech_data_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>max_rech_data_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>max_rech_data_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>163</th>\n",
       "      <td>av_rech_amt_data_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>164</th>\n",
       "      <td>av_rech_amt_data_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>av_rech_amt_data_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>av_rech_amt_data_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>167</th>\n",
       "      <td>vol_2g_mb_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>vol_2g_mb_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>vol_2g_mb_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>vol_2g_mb_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>171</th>\n",
       "      <td>vol_3g_mb_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>172</th>\n",
       "      <td>vol_3g_mb_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>vol_3g_mb_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>174</th>\n",
       "      <td>vol_3g_mb_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>183</th>\n",
       "      <td>night_pck_user_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>night_pck_user_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>night_pck_user_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>186</th>\n",
       "      <td>night_pck_user_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>187</th>\n",
       "      <td>monthly_2g_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>188</th>\n",
       "      <td>monthly_2g_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>189</th>\n",
       "      <td>monthly_2g_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>190</th>\n",
       "      <td>monthly_2g_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>191</th>\n",
       "      <td>sachet_2g_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>sachet_2g_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>193</th>\n",
       "      <td>sachet_2g_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>194</th>\n",
       "      <td>sachet_2g_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195</th>\n",
       "      <td>monthly_3g_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>monthly_3g_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>monthly_3g_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>monthly_3g_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>sachet_3g_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>200</th>\n",
       "      <td>sachet_3g_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>201</th>\n",
       "      <td>sachet_3g_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>202</th>\n",
       "      <td>sachet_3g_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>203</th>\n",
       "      <td>fb_user_6</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204</th>\n",
       "      <td>fb_user_7</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>205</th>\n",
       "      <td>fb_user_8</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>206</th>\n",
       "      <td>fb_user_9</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>207</th>\n",
       "      <td>aon</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>208</th>\n",
       "      <td>aug_vbc_3g</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>209</th>\n",
       "      <td>jul_vbc_3g</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>210</th>\n",
       "      <td>jun_vbc_3g</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>211</th>\n",
       "      <td>sep_vbc_3g</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               features  include\n",
       "0        loc_og_t2o_mou     True\n",
       "1        std_og_t2o_mou     True\n",
       "2        loc_ic_t2o_mou     True\n",
       "3                arpu_6     True\n",
       "4                arpu_7     True\n",
       "5                arpu_8     True\n",
       "6                arpu_9     True\n",
       "7           onnet_mou_6     True\n",
       "8           onnet_mou_7     True\n",
       "9           onnet_mou_8     True\n",
       "10          onnet_mou_9     True\n",
       "11         offnet_mou_6     True\n",
       "12         offnet_mou_7     True\n",
       "13         offnet_mou_8     True\n",
       "14         offnet_mou_9     True\n",
       "15        roam_ic_mou_6     True\n",
       "16        roam_ic_mou_7     True\n",
       "17        roam_ic_mou_8     True\n",
       "18        roam_ic_mou_9     True\n",
       "19        roam_og_mou_6     True\n",
       "20        roam_og_mou_7     True\n",
       "21        roam_og_mou_8     True\n",
       "22        roam_og_mou_9     True\n",
       "23     loc_og_t2t_mou_6     True\n",
       "24     loc_og_t2t_mou_7     True\n",
       "25     loc_og_t2t_mou_8     True\n",
       "26     loc_og_t2t_mou_9     True\n",
       "27     loc_og_t2m_mou_6     True\n",
       "28     loc_og_t2m_mou_7     True\n",
       "29     loc_og_t2m_mou_8     True\n",
       "30     loc_og_t2m_mou_9     True\n",
       "31     loc_og_t2f_mou_6     True\n",
       "32     loc_og_t2f_mou_7     True\n",
       "33     loc_og_t2f_mou_8     True\n",
       "34     loc_og_t2f_mou_9     True\n",
       "35     loc_og_t2c_mou_6     True\n",
       "36     loc_og_t2c_mou_7     True\n",
       "37     loc_og_t2c_mou_8     True\n",
       "38     loc_og_t2c_mou_9     True\n",
       "39         loc_og_mou_6     True\n",
       "40         loc_og_mou_7     True\n",
       "41         loc_og_mou_8     True\n",
       "42         loc_og_mou_9     True\n",
       "43     std_og_t2t_mou_6     True\n",
       "44     std_og_t2t_mou_7     True\n",
       "45     std_og_t2t_mou_8     True\n",
       "46     std_og_t2t_mou_9     True\n",
       "47     std_og_t2m_mou_6     True\n",
       "48     std_og_t2m_mou_7     True\n",
       "49     std_og_t2m_mou_8     True\n",
       "50     std_og_t2m_mou_9     True\n",
       "51     std_og_t2f_mou_6     True\n",
       "52     std_og_t2f_mou_7     True\n",
       "53     std_og_t2f_mou_8     True\n",
       "54     std_og_t2f_mou_9     True\n",
       "55     std_og_t2c_mou_6     True\n",
       "56     std_og_t2c_mou_7     True\n",
       "57     std_og_t2c_mou_8     True\n",
       "58     std_og_t2c_mou_9     True\n",
       "59         std_og_mou_6     True\n",
       "60         std_og_mou_7     True\n",
       "61         std_og_mou_8     True\n",
       "62         std_og_mou_9     True\n",
       "63         isd_og_mou_6     True\n",
       "64         isd_og_mou_7     True\n",
       "65         isd_og_mou_8     True\n",
       "66         isd_og_mou_9     True\n",
       "67         spl_og_mou_6     True\n",
       "68         spl_og_mou_7     True\n",
       "69         spl_og_mou_8     True\n",
       "70         spl_og_mou_9     True\n",
       "71          og_others_6     True\n",
       "72          og_others_7     True\n",
       "73          og_others_8     True\n",
       "74          og_others_9     True\n",
       "75       total_og_mou_6     True\n",
       "76       total_og_mou_7     True\n",
       "77       total_og_mou_8     True\n",
       "78       total_og_mou_9     True\n",
       "79     loc_ic_t2t_mou_6     True\n",
       "80     loc_ic_t2t_mou_7     True\n",
       "81     loc_ic_t2t_mou_8     True\n",
       "82     loc_ic_t2t_mou_9     True\n",
       "83     loc_ic_t2m_mou_6     True\n",
       "84     loc_ic_t2m_mou_7     True\n",
       "85     loc_ic_t2m_mou_8     True\n",
       "86     loc_ic_t2m_mou_9     True\n",
       "87     loc_ic_t2f_mou_6     True\n",
       "88     loc_ic_t2f_mou_7     True\n",
       "89     loc_ic_t2f_mou_8     True\n",
       "90     loc_ic_t2f_mou_9     True\n",
       "91         loc_ic_mou_6     True\n",
       "92         loc_ic_mou_7     True\n",
       "93         loc_ic_mou_8     True\n",
       "94         loc_ic_mou_9     True\n",
       "95     std_ic_t2t_mou_6     True\n",
       "96     std_ic_t2t_mou_7     True\n",
       "97     std_ic_t2t_mou_8     True\n",
       "98     std_ic_t2t_mou_9     True\n",
       "99     std_ic_t2m_mou_6     True\n",
       "100    std_ic_t2m_mou_7     True\n",
       "101    std_ic_t2m_mou_8     True\n",
       "102    std_ic_t2m_mou_9     True\n",
       "103    std_ic_t2f_mou_6     True\n",
       "104    std_ic_t2f_mou_7     True\n",
       "105    std_ic_t2f_mou_8     True\n",
       "106    std_ic_t2f_mou_9     True\n",
       "107    std_ic_t2o_mou_6     True\n",
       "108    std_ic_t2o_mou_7     True\n",
       "109    std_ic_t2o_mou_8     True\n",
       "110    std_ic_t2o_mou_9     True\n",
       "111        std_ic_mou_6     True\n",
       "112        std_ic_mou_7     True\n",
       "113        std_ic_mou_8     True\n",
       "114        std_ic_mou_9     True\n",
       "115      total_ic_mou_6     True\n",
       "116      total_ic_mou_7     True\n",
       "117      total_ic_mou_8     True\n",
       "118      total_ic_mou_9     True\n",
       "119        spl_ic_mou_6     True\n",
       "120        spl_ic_mou_7     True\n",
       "121        spl_ic_mou_8     True\n",
       "122        spl_ic_mou_9     True\n",
       "123        isd_ic_mou_6     True\n",
       "124        isd_ic_mou_7     True\n",
       "125        isd_ic_mou_8     True\n",
       "126        isd_ic_mou_9     True\n",
       "127         ic_others_6     True\n",
       "128         ic_others_7     True\n",
       "129         ic_others_8     True\n",
       "130         ic_others_9     True\n",
       "131    total_rech_num_6     True\n",
       "132    total_rech_num_7     True\n",
       "133    total_rech_num_8     True\n",
       "134    total_rech_num_9     True\n",
       "135    total_rech_amt_6     True\n",
       "136    total_rech_amt_7     True\n",
       "137    total_rech_amt_8     True\n",
       "138    total_rech_amt_9     True\n",
       "139      max_rech_amt_6     True\n",
       "140      max_rech_amt_7     True\n",
       "141      max_rech_amt_8     True\n",
       "142      max_rech_amt_9     True\n",
       "143  last_day_rch_amt_6     True\n",
       "144  last_day_rch_amt_7     True\n",
       "145  last_day_rch_amt_8     True\n",
       "146  last_day_rch_amt_9     True\n",
       "147   total_rech_data_6     True\n",
       "148   total_rech_data_7     True\n",
       "149   total_rech_data_8     True\n",
       "150   total_rech_data_9     True\n",
       "151     max_rech_data_6     True\n",
       "152     max_rech_data_7     True\n",
       "153     max_rech_data_8     True\n",
       "154     max_rech_data_9     True\n",
       "163  av_rech_amt_data_6     True\n",
       "164  av_rech_amt_data_7     True\n",
       "165  av_rech_amt_data_8     True\n",
       "166  av_rech_amt_data_9     True\n",
       "167         vol_2g_mb_6     True\n",
       "168         vol_2g_mb_7     True\n",
       "169         vol_2g_mb_8     True\n",
       "170         vol_2g_mb_9     True\n",
       "171         vol_3g_mb_6     True\n",
       "172         vol_3g_mb_7     True\n",
       "173         vol_3g_mb_8     True\n",
       "174         vol_3g_mb_9     True\n",
       "183    night_pck_user_6     True\n",
       "184    night_pck_user_7     True\n",
       "185    night_pck_user_8     True\n",
       "186    night_pck_user_9     True\n",
       "187        monthly_2g_6     True\n",
       "188        monthly_2g_7     True\n",
       "189        monthly_2g_8     True\n",
       "190        monthly_2g_9     True\n",
       "191         sachet_2g_6     True\n",
       "192         sachet_2g_7     True\n",
       "193         sachet_2g_8     True\n",
       "194         sachet_2g_9     True\n",
       "195        monthly_3g_6     True\n",
       "196        monthly_3g_7     True\n",
       "197        monthly_3g_8     True\n",
       "198        monthly_3g_9     True\n",
       "199         sachet_3g_6     True\n",
       "200         sachet_3g_7     True\n",
       "201         sachet_3g_8     True\n",
       "202         sachet_3g_9     True\n",
       "203           fb_user_6     True\n",
       "204           fb_user_7     True\n",
       "205           fb_user_8     True\n",
       "206           fb_user_9     True\n",
       "207                 aon     True\n",
       "208          aug_vbc_3g     True\n",
       "209          jul_vbc_3g     True\n",
       "210          jun_vbc_3g     True\n",
       "211          sep_vbc_3g     True"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "initial_cols = churn.shape[1]\n",
    "\n",
    "MISSING_THRESHOLD = 0.7\n",
    "\n",
    "include_cols = list(churn.apply(lambda column: True if column.isnull().sum()/churn.shape[0] < MISSING_THRESHOLD else False))\n",
    "\n",
    "drop_missing = pd.DataFrame({'features':churn.columns , 'include': include_cols})\n",
    "drop_missing.loc[drop_missing.include == True,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-16 columns dropped.\n"
     ]
    }
   ],
   "source": [
    "# drop columns\n",
    "churn = churn.loc[:, include_cols]\n",
    "\n",
    "dropped_cols = churn.shape[1] - initial_cols\n",
    "print(\"{0} columns dropped.\".format(dropped_cols))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## iv) imputing using MICE"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "install fancyimpute package using [this](https://github.com/iskandr/fancyimpute) link and following the install instructions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Columns : 196\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MICE] Completing matrix with shape (99999, 196)\n",
      "[MICE] Starting imputation round 1/11, elapsed time 0.250\n",
      "[MICE] Starting imputation round 2/11, elapsed time 57.143\n",
      "[MICE] Starting imputation round 3/11, elapsed time 115.295\n",
      "[MICE] Starting imputation round 4/11, elapsed time 173.913\n",
      "[MICE] Starting imputation round 5/11, elapsed time 232.298\n",
      "[MICE] Starting imputation round 6/11, elapsed time 291.047\n",
      "[MICE] Starting imputation round 7/11, elapsed time 377.504\n",
      "[MICE] Starting imputation round 8/11, elapsed time 454.048\n",
      "[MICE] Starting imputation round 9/11, elapsed time 522.750\n",
      "[MICE] Starting imputation round 10/11, elapsed time 587.501\n",
      "[MICE] Starting imputation round 11/11, elapsed time 649.457\n"
     ]
    }
   ],
   "source": [
    "churn_cols = churn.columns\n",
    "print('Total Columns :',len(churn_cols))\n",
    "# using MICE technique to impute missing values in the rest of the columns\n",
    "from fancyimpute import MICE\n",
    "churn_imputed = MICE(n_imputations=1).complete(churn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loc_og_t2o_mou        0.0\n",
      "std_og_t2o_mou        0.0\n",
      "loc_ic_t2o_mou        0.0\n",
      "arpu_6                0.0\n",
      "arpu_7                0.0\n",
      "arpu_8                0.0\n",
      "arpu_9                0.0\n",
      "onnet_mou_6           0.0\n",
      "onnet_mou_7           0.0\n",
      "onnet_mou_8           0.0\n",
      "onnet_mou_9           0.0\n",
      "offnet_mou_6          0.0\n",
      "offnet_mou_7          0.0\n",
      "offnet_mou_8          0.0\n",
      "offnet_mou_9          0.0\n",
      "roam_ic_mou_6         0.0\n",
      "roam_ic_mou_7         0.0\n",
      "roam_ic_mou_8         0.0\n",
      "roam_ic_mou_9         0.0\n",
      "roam_og_mou_6         0.0\n",
      "roam_og_mou_7         0.0\n",
      "roam_og_mou_8         0.0\n",
      "roam_og_mou_9         0.0\n",
      "loc_og_t2t_mou_6      0.0\n",
      "loc_og_t2t_mou_7      0.0\n",
      "loc_og_t2t_mou_8      0.0\n",
      "loc_og_t2t_mou_9      0.0\n",
      "loc_og_t2m_mou_6      0.0\n",
      "loc_og_t2m_mou_7      0.0\n",
      "loc_og_t2m_mou_8      0.0\n",
      "loc_og_t2m_mou_9      0.0\n",
      "loc_og_t2f_mou_6      0.0\n",
      "loc_og_t2f_mou_7      0.0\n",
      "loc_og_t2f_mou_8      0.0\n",
      "loc_og_t2f_mou_9      0.0\n",
      "loc_og_t2c_mou_6      0.0\n",
      "loc_og_t2c_mou_7      0.0\n",
      "loc_og_t2c_mou_8      0.0\n",
      "loc_og_t2c_mou_9      0.0\n",
      "loc_og_mou_6          0.0\n",
      "loc_og_mou_7          0.0\n",
      "loc_og_mou_8          0.0\n",
      "loc_og_mou_9          0.0\n",
      "std_og_t2t_mou_6      0.0\n",
      "std_og_t2t_mou_7      0.0\n",
      "std_og_t2t_mou_8      0.0\n",
      "std_og_t2t_mou_9      0.0\n",
      "std_og_t2m_mou_6      0.0\n",
      "std_og_t2m_mou_7      0.0\n",
      "std_og_t2m_mou_8      0.0\n",
      "std_og_t2m_mou_9      0.0\n",
      "std_og_t2f_mou_6      0.0\n",
      "std_og_t2f_mou_7      0.0\n",
      "std_og_t2f_mou_8      0.0\n",
      "std_og_t2f_mou_9      0.0\n",
      "std_og_t2c_mou_6      0.0\n",
      "std_og_t2c_mou_7      0.0\n",
      "std_og_t2c_mou_8      0.0\n",
      "std_og_t2c_mou_9      0.0\n",
      "std_og_mou_6          0.0\n",
      "std_og_mou_7          0.0\n",
      "std_og_mou_8          0.0\n",
      "std_og_mou_9          0.0\n",
      "isd_og_mou_6          0.0\n",
      "isd_og_mou_7          0.0\n",
      "isd_og_mou_8          0.0\n",
      "isd_og_mou_9          0.0\n",
      "spl_og_mou_6          0.0\n",
      "spl_og_mou_7          0.0\n",
      "spl_og_mou_8          0.0\n",
      "spl_og_mou_9          0.0\n",
      "og_others_6           0.0\n",
      "og_others_7           0.0\n",
      "og_others_8           0.0\n",
      "og_others_9           0.0\n",
      "total_og_mou_6        0.0\n",
      "total_og_mou_7        0.0\n",
      "total_og_mou_8        0.0\n",
      "total_og_mou_9        0.0\n",
      "loc_ic_t2t_mou_6      0.0\n",
      "loc_ic_t2t_mou_7      0.0\n",
      "loc_ic_t2t_mou_8      0.0\n",
      "loc_ic_t2t_mou_9      0.0\n",
      "loc_ic_t2m_mou_6      0.0\n",
      "loc_ic_t2m_mou_7      0.0\n",
      "loc_ic_t2m_mou_8      0.0\n",
      "loc_ic_t2m_mou_9      0.0\n",
      "loc_ic_t2f_mou_6      0.0\n",
      "loc_ic_t2f_mou_7      0.0\n",
      "loc_ic_t2f_mou_8      0.0\n",
      "loc_ic_t2f_mou_9      0.0\n",
      "loc_ic_mou_6          0.0\n",
      "loc_ic_mou_7          0.0\n",
      "loc_ic_mou_8          0.0\n",
      "loc_ic_mou_9          0.0\n",
      "std_ic_t2t_mou_6      0.0\n",
      "std_ic_t2t_mou_7      0.0\n",
      "std_ic_t2t_mou_8      0.0\n",
      "std_ic_t2t_mou_9      0.0\n",
      "std_ic_t2m_mou_6      0.0\n",
      "std_ic_t2m_mou_7      0.0\n",
      "std_ic_t2m_mou_8      0.0\n",
      "std_ic_t2m_mou_9      0.0\n",
      "std_ic_t2f_mou_6      0.0\n",
      "std_ic_t2f_mou_7      0.0\n",
      "std_ic_t2f_mou_8      0.0\n",
      "std_ic_t2f_mou_9      0.0\n",
      "std_ic_t2o_mou_6      0.0\n",
      "std_ic_t2o_mou_7      0.0\n",
      "std_ic_t2o_mou_8      0.0\n",
      "std_ic_t2o_mou_9      0.0\n",
      "std_ic_mou_6          0.0\n",
      "std_ic_mou_7          0.0\n",
      "std_ic_mou_8          0.0\n",
      "std_ic_mou_9          0.0\n",
      "total_ic_mou_6        0.0\n",
      "total_ic_mou_7        0.0\n",
      "total_ic_mou_8        0.0\n",
      "total_ic_mou_9        0.0\n",
      "spl_ic_mou_6          0.0\n",
      "spl_ic_mou_7          0.0\n",
      "spl_ic_mou_8          0.0\n",
      "spl_ic_mou_9          0.0\n",
      "isd_ic_mou_6          0.0\n",
      "isd_ic_mou_7          0.0\n",
      "isd_ic_mou_8          0.0\n",
      "isd_ic_mou_9          0.0\n",
      "ic_others_6           0.0\n",
      "ic_others_7           0.0\n",
      "ic_others_8           0.0\n",
      "ic_others_9           0.0\n",
      "total_rech_num_6      0.0\n",
      "total_rech_num_7      0.0\n",
      "total_rech_num_8      0.0\n",
      "total_rech_num_9      0.0\n",
      "total_rech_amt_6      0.0\n",
      "total_rech_amt_7      0.0\n",
      "total_rech_amt_8      0.0\n",
      "total_rech_amt_9      0.0\n",
      "max_rech_amt_6        0.0\n",
      "max_rech_amt_7        0.0\n",
      "max_rech_amt_8        0.0\n",
      "max_rech_amt_9        0.0\n",
      "last_day_rch_amt_6    0.0\n",
      "last_day_rch_amt_7    0.0\n",
      "last_day_rch_amt_8    0.0\n",
      "last_day_rch_amt_9    0.0\n",
      "total_rech_data_6     0.0\n",
      "total_rech_data_7     0.0\n",
      "total_rech_data_8     0.0\n",
      "total_rech_data_9     0.0\n",
      "max_rech_data_6       0.0\n",
      "max_rech_data_7       0.0\n",
      "max_rech_data_8       0.0\n",
      "max_rech_data_9       0.0\n",
      "av_rech_amt_data_6    0.0\n",
      "av_rech_amt_data_7    0.0\n",
      "av_rech_amt_data_8    0.0\n",
      "av_rech_amt_data_9    0.0\n",
      "vol_2g_mb_6           0.0\n",
      "vol_2g_mb_7           0.0\n",
      "vol_2g_mb_8           0.0\n",
      "vol_2g_mb_9           0.0\n",
      "vol_3g_mb_6           0.0\n",
      "vol_3g_mb_7           0.0\n",
      "vol_3g_mb_8           0.0\n",
      "vol_3g_mb_9           0.0\n",
      "night_pck_user_6      0.0\n",
      "night_pck_user_7      0.0\n",
      "night_pck_user_8      0.0\n",
      "night_pck_user_9      0.0\n",
      "monthly_2g_6          0.0\n",
      "monthly_2g_7          0.0\n",
      "monthly_2g_8          0.0\n",
      "monthly_2g_9          0.0\n",
      "sachet_2g_6           0.0\n",
      "sachet_2g_7           0.0\n",
      "sachet_2g_8           0.0\n",
      "sachet_2g_9           0.0\n",
      "monthly_3g_6          0.0\n",
      "monthly_3g_7          0.0\n",
      "monthly_3g_8          0.0\n",
      "monthly_3g_9          0.0\n",
      "sachet_3g_6           0.0\n",
      "sachet_3g_7           0.0\n",
      "sachet_3g_8           0.0\n",
      "sachet_3g_9           0.0\n",
      "fb_user_6             0.0\n",
      "fb_user_7             0.0\n",
      "fb_user_8             0.0\n",
      "fb_user_9             0.0\n",
      "aon                   0.0\n",
      "aug_vbc_3g            0.0\n",
      "jul_vbc_3g            0.0\n",
      "jun_vbc_3g            0.0\n",
      "sep_vbc_3g            0.0\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# convert imputed numpy array to pandas dataframe\n",
    "churn = pd.DataFrame(churn_imputed, columns=churn_cols)\n",
    "print(churn.isnull().sum()*100/churn.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# filter high-value customers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### calculate total data recharge amount"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# calculate the total data recharge amount for June and July --> number of recharges * average recharge amount\n",
    "churn['total_data_rech_6'] = churn.total_rech_data_6 * churn.av_rech_amt_data_6\n",
    "churn['total_data_rech_7'] = churn.total_rech_data_7 * churn.av_rech_amt_data_7"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### add total data recharge and total recharge to get total combined recharge amount for a month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# calculate total recharge amount for June and July --> call recharge amount + data recharge amount\n",
    "churn['amt_data_6'] = churn.total_rech_amt_6 + churn.total_data_rech_6\n",
    "churn['amt_data_7'] = churn.total_rech_amt_7 + churn.total_data_rech_7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# calculate average recharge done by customer in June and July\n",
    "churn['av_amt_data_6_7'] = (churn.amt_data_6 + churn.amt_data_7)/2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recharge amount at 70th percentile: 478.0\n"
     ]
    }
   ],
   "source": [
    "# look at the 70th percentile recharge amount\n",
    "print(\"Recharge amount at 70th percentile: {0}\".format(churn.av_amt_data_6_7.quantile(0.7)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(30001, 201)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# retain only those customers who have recharged their mobiles with more than or equal to 70th percentile amount\n",
    "churn_filtered = churn.loc[churn.av_amt_data_6_7 >= churn.av_amt_data_6_7.quantile(0.7), :]\n",
    "churn_filtered = churn_filtered.reset_index(drop=True)\n",
    "churn_filtered.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(30001, 196)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# delete variables created to filter high-value customers\n",
    "churn_filtered = churn_filtered.drop(['total_data_rech_6', 'total_data_rech_7',\n",
    "                                      'amt_data_6', 'amt_data_7', 'av_amt_data_6_7'], axis=1)\n",
    "churn_filtered.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We're left with 30,001 rows after selecting the customers who have provided recharge value of more than or equal to the recharge value of the 70th percentile customer."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# derive churn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# calculate total incoming and outgoing minutes of usage\n",
    "churn_filtered['total_calls_mou_9'] = churn_filtered.total_ic_mou_9 + churn_filtered.total_og_mou_9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# calculate 2g and 3g data consumption\n",
    "churn_filtered['total_internet_mb_9'] =  churn_filtered.vol_2g_mb_9 + churn_filtered.vol_3g_mb_9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create churn variable: those who have not used either calls or internet in the month of September are customers who have churned\n",
    "\n",
    "# 0 - not churn, 1 - churn\n",
    "churn_filtered['churn'] = churn_filtered.apply(lambda row: 1 if (row.total_calls_mou_9 == 0 and row.total_internet_mb_9 == 0) else 0, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# delete derived variables\n",
    "churn_filtered = churn_filtered.drop(['total_calls_mou_9', 'total_internet_mb_9'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Churn Ratio:\n",
      "0    91.863605\n",
      "1     8.136395\n",
      "Name: churn, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# change data type to category\n",
    "churn_filtered.churn = churn_filtered.churn.astype(\"category\")\n",
    "\n",
    "# print churn ratio\n",
    "print(\"Churn Ratio:\")\n",
    "print(churn_filtered.churn.value_counts()*100/churn_filtered.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Calculate difference between 8th and previous months"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's derive some variables. The most important feature, in this situation, can be the difference between the 8th month and the previous months. The difference can be in patterns such as usage difference or recharge value difference. Let's calculate difference variable as the difference between 8th month and the average of 6th and 7th month."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "churn_filtered['arpu_diff'] = churn_filtered.arpu_8 - ((churn_filtered.arpu_6 + churn_filtered.arpu_7)/2)\n",
    "\n",
    "churn_filtered['onnet_mou_diff'] = churn_filtered.onnet_mou_8 - ((churn_filtered.onnet_mou_6 + churn_filtered.onnet_mou_7)/2)\n",
    "\n",
    "churn_filtered['offnet_mou_diff'] = churn_filtered.offnet_mou_8 - ((churn_filtered.offnet_mou_6 + churn_filtered.offnet_mou_7)/2)\n",
    "\n",
    "churn_filtered['roam_ic_mou_diff'] = churn_filtered.roam_ic_mou_8 - ((churn_filtered.roam_ic_mou_6 + churn_filtered.roam_ic_mou_7)/2)\n",
    "\n",
    "churn_filtered['roam_og_mou_diff'] = churn_filtered.roam_og_mou_8 - ((churn_filtered.roam_og_mou_6 + churn_filtered.roam_og_mou_7)/2)\n",
    "\n",
    "churn_filtered['loc_og_mou_diff'] = churn_filtered.loc_og_mou_8 - ((churn_filtered.loc_og_mou_6 + churn_filtered.loc_og_mou_7)/2)\n",
    "\n",
    "churn_filtered['std_og_mou_diff'] = churn_filtered.std_og_mou_8 - ((churn_filtered.std_og_mou_6 + churn_filtered.std_og_mou_7)/2)\n",
    "\n",
    "churn_filtered['isd_og_mou_diff'] = churn_filtered.isd_og_mou_8 - ((churn_filtered.isd_og_mou_6 + churn_filtered.isd_og_mou_7)/2)\n",
    "\n",
    "churn_filtered['spl_og_mou_diff'] = churn_filtered.spl_og_mou_8 - ((churn_filtered.spl_og_mou_6 + churn_filtered.spl_og_mou_7)/2)\n",
    "\n",
    "churn_filtered['total_og_mou_diff'] = churn_filtered.total_og_mou_8 - ((churn_filtered.total_og_mou_6 + churn_filtered.total_og_mou_7)/2)\n",
    "\n",
    "churn_filtered['loc_ic_mou_diff'] = churn_filtered.loc_ic_mou_8 - ((churn_filtered.loc_ic_mou_6 + churn_filtered.loc_ic_mou_7)/2)\n",
    "\n",
    "churn_filtered['std_ic_mou_diff'] = churn_filtered.std_ic_mou_8 - ((churn_filtered.std_ic_mou_6 + churn_filtered.std_ic_mou_7)/2)\n",
    "\n",
    "churn_filtered['isd_ic_mou_diff'] = churn_filtered.isd_ic_mou_8 - ((churn_filtered.isd_ic_mou_6 + churn_filtered.isd_ic_mou_7)/2)\n",
    "\n",
    "churn_filtered['spl_ic_mou_diff'] = churn_filtered.spl_ic_mou_8 - ((churn_filtered.spl_ic_mou_6 + churn_filtered.spl_ic_mou_7)/2)\n",
    "\n",
    "churn_filtered['total_ic_mou_diff'] = churn_filtered.total_ic_mou_8 - ((churn_filtered.total_ic_mou_6 + churn_filtered.total_ic_mou_7)/2)\n",
    "\n",
    "churn_filtered['total_rech_num_diff'] = churn_filtered.total_rech_num_8 - ((churn_filtered.total_rech_num_6 + churn_filtered.total_rech_num_7)/2)\n",
    "\n",
    "churn_filtered['total_rech_amt_diff'] = churn_filtered.total_rech_amt_8 - ((churn_filtered.total_rech_amt_6 + churn_filtered.total_rech_amt_7)/2)\n",
    "\n",
    "churn_filtered['max_rech_amt_diff'] = churn_filtered.max_rech_amt_8 - ((churn_filtered.max_rech_amt_6 + churn_filtered.max_rech_amt_7)/2)\n",
    "\n",
    "churn_filtered['total_rech_data_diff'] = churn_filtered.total_rech_data_8 - ((churn_filtered.total_rech_data_6 + churn_filtered.total_rech_data_7)/2)\n",
    "\n",
    "churn_filtered['max_rech_data_diff'] = churn_filtered.max_rech_data_8 - ((churn_filtered.max_rech_data_6 + churn_filtered.max_rech_data_7)/2)\n",
    "\n",
    "churn_filtered['av_rech_amt_data_diff'] = churn_filtered.av_rech_amt_data_8 - ((churn_filtered.av_rech_amt_data_6 + churn_filtered.av_rech_amt_data_7)/2)\n",
    "\n",
    "churn_filtered['vol_2g_mb_diff'] = churn_filtered.vol_2g_mb_8 - ((churn_filtered.vol_2g_mb_6 + churn_filtered.vol_2g_mb_7)/2)\n",
    "\n",
    "churn_filtered['vol_3g_mb_diff'] = churn_filtered.vol_3g_mb_8 - ((churn_filtered.vol_3g_mb_6 + churn_filtered.vol_3g_mb_7)/2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    30001.000000\n",
       "mean       -67.437337\n",
       "std        502.630069\n",
       "min      -7213.410000\n",
       "25%       -168.025000\n",
       "50%        -14.625000\n",
       "75%         67.915000\n",
       "max      12768.705000\n",
       "Name: total_og_mou_diff, dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# let's look at summary of one of the difference variables\n",
    "churn_filtered['total_og_mou_diff'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## delete columns that belong to the churn month (9th month)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(30001, 173)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# delete all variables relating to 9th month\n",
    "churn_filtered = churn_filtered.filter(regex='[^9]$', axis=1)\n",
    "churn_filtered.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# extract all names that end with 9\n",
    "col_9_names = churn.filter(regex='9$', axis=1).columns\n",
    "\n",
    "# update num_cols and cat_cols column name list\n",
    "cat_cols = [col for col in cat_cols if col not in col_9_names]\n",
    "cat_cols.append('churn')\n",
    "num_cols = [col for col in churn_filtered.columns if col not in cat_cols]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## visualise data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# change columns types\n",
    "churn_filtered[num_cols] = churn_filtered[num_cols].apply(pd.to_numeric)\n",
    "churn_filtered[cat_cols] = churn_filtered[cat_cols].apply(lambda column: column.astype(\"category\"), axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create plotting functions\n",
    "def data_type(variable):\n",
    "    if variable.dtype == np.int64 or variable.dtype == np.float64:\n",
    "        return 'numerical'\n",
    "    elif variable.dtype == 'category':\n",
    "        return 'categorical'\n",
    "    \n",
    "def univariate(variable, stats=True):\n",
    "    \n",
    "    if data_type(variable) == 'numerical':\n",
    "        sns.distplot(variable)\n",
    "        if stats == True:\n",
    "            print(variable.describe())\n",
    "    \n",
    "    elif data_type(variable) == 'categorical':\n",
    "        sns.countplot(variable)\n",
    "        if stats == True:\n",
    "            print(variable.value_counts())\n",
    "            \n",
    "    else:\n",
    "        print(\"Invalid variable passed: either pass a numeric variable or a categorical vairable.\")\n",
    "        \n",
    "def bivariate(var1, var2):\n",
    "    if data_type(var1) == 'numerical' and data_type(var2) == 'numerical':\n",
    "        sns.regplot(var1, var2)\n",
    "    elif (data_type(var1) == 'categorical' and data_type(var2) == 'numerical') or (data_type(var1) == 'numerical' and data_type(var2) == 'categorical'):        \n",
    "        sns.boxplot(var1, var2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Univariate EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "count    99999.000000\n",
      "mean       282.987358\n",
      "std        328.439770\n",
      "min      -2258.709000\n",
      "25%         93.411500\n",
      "50%        197.704000\n",
      "75%        371.060000\n",
      "max      27731.088000\n",
      "Name: arpu_6, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "univariate(churn.arpu_6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "count    99999.000000\n",
      "mean        -0.000001\n",
      "std          0.000316\n",
      "min         -0.009597\n",
      "25%          0.000000\n",
      "50%          0.000000\n",
      "75%          0.000000\n",
      "max          0.010082\n",
      "Name: loc_og_t2o_mou, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "univariate(churn.loc_og_t2o_mou)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "count    9.999900e+04\n",
      "mean     8.859026e-07\n",
      "std      3.331207e-04\n",
      "min     -1.246070e-02\n",
      "25%      0.000000e+00\n",
      "50%      0.000000e+00\n",
      "75%      0.000000e+00\n",
      "max      1.080637e-02\n",
      "Name: std_og_t2o_mou, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAELCAYAAAAybErdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAHOlJREFUeJzt3X+QZWV95/H35952BmWNDDDL4gzZmayTbA2aGJwAKYyrEGFgsw5bq9YQK4xI7exuMBvN7iqsVcHlR5XE3SWyRFKsjA6o/JCNxZSiZBwgmmwYGOQ3iLT8kBn50TKAQVb03vvdP85ze07fvv2De7vPebrn86rq6nOf85xzzz3dPZ95znOe5ygiMDMzK2vUfQBmZpYfh4OZmU3icDAzs0kcDmZmNonDwczMJnE4mJnZJA4HMzObxOFgZmaTOBzMzGySkboPYFCHHnporFq1qu7DMDNbUO68884fR8Tymeot2HBYtWoVu3btqvswzMwWFElPzKaeLyuZmdkkDgczM5tkxnCQtEXSs5LuL5V9WtL3JN0r6auSDiqtO0fSqKSHJZ1UKl+fykYlnV0qXy1pZyq/VtKSufyAZmb26s2m5fAFYH1P2XbgzRHx68D3gXMAJK0FNgJHpm0+K6kpqQn8BXAysBY4LdUFuAi4OCLeBDwPnDnUJzIzs6HNGA4R8W1gb0/ZX0dEK728DViZljcA10TEKxHxGDAKHJ2+RiPi0Yj4OXANsEGSgOOB69P2W4FTh/xMZmY2pLnoc/gQ8I20vAJ4srRudyqbqvwQ4IVS0HTLzcysRkOFg6RPAC3gS3NzODO+32ZJuyTtGhsbq+Itzcz2SwOHg6QPAr8HfCD2PWt0D3BEqdrKVDZV+XPAQZJGesr7iojLI2JdRKxbvnzGMRxmZjaggcJB0nrgY8B7IuLl0qptwEZJSyWtBtYAtwN3AGvSnUlLKDqtt6VQuQV4b9p+E3DDYB/FbGH5k2vv5urbf1j3YZj1NZtbWa8G/h74NUm7JZ0JXAq8Htgu6W5JfwkQEQ8A1wEPAt8EzoqIdupT+DBwE/AQcF2qC/Bx4E8kjVL0QVwxp5/QLFPfGf0x333i+boPw6yvGafPiIjT+hRP+Q94RFwIXNin/Ebgxj7lj1LczWS2X2l3gnYnZq5oVgOPkDarSavdoeVwsEw5HMxq4paD5czhYFaTdjgcLF8OB7OatDvhy0qWLYeDWU1anaDd6dR9GGZ9ORzMatDpBBG45WDZcjiY1aAbCu5zsFw5HMxq0AmHg+XN4WBWA7ccLHcOB7MatNtFKLjPwXLlcDCrQSvdpeSWg+XK4WBWg24ouOVguXI4mNWgnTqkOw4Hy5TDwawGrfE+Bw+Cszw5HMxq0PbdSpY5h4NZDVruc7DMORzMauCWg+XO4WBWA4eD5c7hYFYDh4PlzuFgVoPuXUruc7BcORzMauCWg+XO4WBWg313K3mcg+XJ4WBWg+7IaGeD5crhYFYDtxwsdw4Hsxp0+xo64fmVLE8OB7MalO9S6k7CZ5aTGcNB0hZJz0q6v1R2sKTtkh5J35elckm6RNKopHslHVXaZlOq/4ikTaXyt0m6L21ziSTN9Yc0y027dDnJdyxZjmbTcvgCsL6n7GxgR0SsAXak1wAnA2vS12bgMijCBDgXOAY4Gji3Gyipzr8tbdf7XmaLTrtTXnY4WH5mDIeI+Dawt6d4A7A1LW8FTi2VXxmF24CDJB0OnARsj4i9EfE8sB1Yn9b9UkTcFhEBXFnal9miVe6I9kA4y9GgfQ6HRcRTaflp4LC0vAJ4slRvdyqbrnx3n3KzRa3cWnDLwXI0dId0+h9/Jb/dkjZL2iVp19jYWBVvaTYvyq0F385qORo0HJ5Jl4RI359N5XuAI0r1Vqay6cpX9invKyIuj4h1EbFu+fLlAx66Wf3ccrDcDRoO24DuHUebgBtK5aenu5aOBV5Ml59uAk6UtCx1RJ8I3JTW/UTSsekupdNL+zJbtBwOlruRmSpIuhp4J3CopN0Udx19CrhO0pnAE8D7U/UbgVOAUeBl4AyAiNgr6XzgjlTvvIjodnL/IcUdUa8FvpG+zBY1h4PlbsZwiIjTplh1Qp+6AZw1xX62AFv6lO8C3jzTcZgtJhP7HBwOlh+PkDargQfBWe4cDmY1mNByaDscLD8OB7MalCfb63huJcuQw8GsBu5zsNw5HMxqMPFuJQ+Cs/w4HMxq4D4Hy53DwawGHudguXM4mNWg7Yf9WOYcDmY1aLtD2jLncDCrQXkm1rb7HCxDDgezGrjlYLlzOJjVoHyHkjukLUcOB7MalDuh3SFtOXI4mNXAg+Asdw4Hsxq0OsHSkeLPz4PgLEcOB7MatNv7wsF9DpYjh4NZDVqdYMlIc3zZLDcOB7MadGJfy8FTdluOHA5mNXCfg+XO4WBWg3anwxL3OVjGHA5mNWi1g6WvcZ+D5cvhYFaDdidY2uy2HDzOwfLjcDCrQTuidFmp5oMx68PhYFaDdid4TVNIbjlYnhwOZjVotYNmo8FIQ+5zsCw5HMxq0O4EIw3RbMh3K1mWhgoHSR+V9ICk+yVdLekASasl7ZQ0KulaSUtS3aXp9Whav6q0n3NS+cOSThruI5nlr9Xp0GyIptxysDwNHA6SVgD/EVgXEW8GmsBG4CLg4oh4E/A8cGba5Ezg+VR+caqHpLVpuyOB9cBnJTUHPS6zhaATFOHgloNlatjLSiPAayWNAK8DngKOB65P67cCp6blDek1af0JkpTKr4mIVyLiMWAUOHrI4zLLWqvTYaQhRpoNh4NlaeBwiIg9wH8HfkgRCi8CdwIvREQrVdsNrEjLK4An07atVP+QcnmfbcwWpXY7xlsOvqxkORrmstIyiv/1rwbeCBxIcVlo3kjaLGmXpF1jY2Pz+VZm86rVCUaaYqQh38pqWRrmstLvAo9FxFhE/AL4K+A44KB0mQlgJbAnLe8BjgBI698APFcu77PNBBFxeUSsi4h1y5cvH+LQzerV7gQNiYY7pC1Tw4TDD4FjJb0u9R2cADwI3AK8N9XZBNyQlrel16T1N0dEpPKN6W6m1cAa4PYhjssse+2I1OcgOg4Hy9DIzFX6i4idkq4Hvgu0gLuAy4GvA9dIuiCVXZE2uQK4StIosJfiDiUi4gFJ11EESws4KyLagx6X2ULQToPg3OdguRo4HAAi4lzg3J7iR+lzt1FE/Ax43xT7uRC4cJhjMVtIJvY5OBwsPx4hbVaDdqd7t1LDLQfLksPBrAatToemRLPhh/1YnhwOZhWLiNIIaQ+Cszw5HMwq1g2DkYb7HCxfDgezinX7GJrN7ghpD4Kz/DgczCrmloMtBA4Hs4p1Ww4NeW4ly5fDwaxinVLLodnwCGnLk8PBrGL7+hz8mFDLl8PBrGLtnpaD+xwsRw4Hs4p1705qNsSIR0hbphwOZhXrthSaEg23HCxTDgezio1fVvLEe5Yxh4NZxcZbDu5zsIw5HMwq1uoZBOcR0pYjh4NZxfa1HBpuOVi2HA5mFRsf59DAI6QtWw4Hs4q55WALgcPBrGKeeM8WAoeDWcXKg+D8mFDLlcPBrGJuOdhC4HAwq9j4lN2NfSOkIxwQlheHg1nFOj0tBwA3Hiw3DgezirV6RkgXZR4IZ3lxOJhVbF+fQ2O85eB+B8uNw8GsYv1bDg4Hy8tQ4SDpIEnXS/qepIck/bakgyVtl/RI+r4s1ZWkSySNSrpX0lGl/WxK9R+RtGnYD2WWs/aEW1lTy6HtcLC8DNty+AzwzYj458BvAA8BZwM7ImINsCO9BjgZWJO+NgOXAUg6GDgXOAY4Gji3Gyhmi1E7dS+UO6TbvlvJMjNwOEh6A/AO4AqAiPh5RLwAbAC2pmpbgVPT8gbgyijcBhwk6XDgJGB7ROyNiOeB7cD6QY/LLHftnkFwRZnDwfIyTMthNTAGfF7SXZI+J+lA4LCIeCrVeRo4LC2vAJ4sbb87lU1VbrYo9U7ZXS4zy8Uw4TACHAVcFhG/CfyUfZeQAIhiZM+c/dZL2ixpl6RdY2Njc7Vbs0r1PuwH3Odg+RkmHHYDuyNiZ3p9PUVYPJMuF5G+P5vW7wGOKG2/MpVNVT5JRFweEesiYt3y5cuHOHSz+rTaHudg+Rs4HCLiaeBJSb+Wik4AHgS2Ad07jjYBN6TlbcDp6a6lY4EX0+Wnm4ATJS1LHdEnpjKzRakTk8Oh4w5py8zIkNv/EfAlSUuAR4EzKALnOklnAk8A7091bwROAUaBl1NdImKvpPOBO1K98yJi75DHZZatVp9BcO5zsNwMFQ4RcTewrs+qE/rUDeCsKfazBdgyzLGYLRT9+hxa7nOwzHiEtFnFukEw0hAjTU+fYXlyOJhVrDvOodEQDfmykuXJ4WBWsXbEeF/DSBoE5w5py43DwaxirU6M9zW4z8Fy5XAwq1i7XWo5uM/BMuVwMKtY35aDB8FZZhwOZhVrl8NBbjlYnhwOZhVrR4zPxtr0k+AsUw4Hs4q5z8EWAoeDWcXKfQ6ePsNy5XAwq1i70xlvMfhhP5Yrh4NZxVqdGO+IbnqEtGXK4WBWsU6U7lZKLYiOw8Ey43Awq1ir7T4Hy5/Dwaxi7U6U+hy6dyt5EJzlxeFgVrHibqXiT88tB8uVw8GsYu1OkBoONDwIzjLlcDCrWLsT41N1jzgcLFMOB7OKtftOvOdwsLw4HMwq1ioNghvxIDjLlMPBrGLllkP65paDZcfhYFax8ghpSTQb8q2slh2Hg1nFyi0HIIVDjQdk1ofDwaxi5UFwUNyx5JaD5cbhYFaxdmkQHBQtB/c5WG4cDmYVa3X2PewHui0Hh4PlZehwkNSUdJekr6XXqyXtlDQq6VpJS1L50vR6NK1fVdrHOan8YUknDXtMZjlrd4KGyn0ODbccLDtz0XL4Y+Ch0uuLgIsj4k3A88CZqfxM4PlUfnGqh6S1wEbgSGA98FlJzTk4LrMstXtaDs2Gp+y2/AwVDpJWAv8S+Fx6LeB44PpUZStwalrekF6T1p+Q6m8AromIVyLiMWAUOHqY4zLLWasT489xgGIgnFsOlpthWw5/DnwM6N5qcQjwQkS00uvdwIq0vAJ4EiCtfzHVHy/vs43ZotPudHpaDu5zsPwMHA6Sfg94NiLunMPjmek9N0vaJWnX2NhYVW9rNqdaPeMcRny3kmVomJbDccB7JD0OXENxOekzwEGSRlKdlcCetLwHOAIgrX8D8Fy5vM82E0TE5RGxLiLWLV++fIhDN6tPuzRCGvAIacvSwOEQEedExMqIWEXRoXxzRHwAuAV4b6q2CbghLW9Lr0nrb46ISOUb091Mq4E1wO2DHpdZ7to9fQ6+rGQ5Gpm5yqv2ceAaSRcAdwFXpPIrgKskjQJ7KQKFiHhA0nXAg0ALOCsi2vNwXGZZmHy3ksPB8jMn4RARtwK3puVH6XO3UUT8DHjfFNtfCFw4F8dilrOImPCYUHCfg+XJI6TNKtTNALccLHcOB7MKtVLH88S7lRq02g4Hy4vDwaxC3ZuSyuHQaEA7HA6WF4eDWYW6LYeRnpaDLytZbhwOZhXqhkDvw37cIW25cTiYVagbApOn7PYgOMuLw8GsQt2WQ6O35eAOacuMw8GsQu0+LYdmQ3TcIW2ZcTiYVWhfn4MfE2p5cziYVWjqPgeHg+XF4WBWoXafQXBND4KzDDkczCrU6nMrq1sOliOHg1mF+o1zaDTkEdKWHYeDWYX63a3kloPlyOFgVqF+l5WKcQ4eBGd5cTiYVWhfy2Hi8xzccrDcOBzMKtS9K6mUDTSbHudg+XE4mFWoOxK63HJoyiOkLT8OB7MKTXUrq1sOlhuHg1mF2n2e59BsNIiAjgPCMuJwMKtQt89hQsuhWSy79WA5cTiYVWiqh/2U15nlwOFgVqF29JmyW5qwziwHDgezCk3bcvDke5YRh4NZhbp9DhMGwY33OXiUtOXD4WBWofGWQ9N9Dpa3gcNB0hGSbpH0oKQHJP1xKj9Y0nZJj6Tvy1K5JF0iaVTSvZKOKu1rU6r/iKRNw38sszxN9bCf8jqzHAzTcmgB/yki1gLHAmdJWgucDeyIiDXAjvQa4GRgTfraDFwGRZgA5wLHAEcD53YDxWyx6XY6N1SasltuOVh+RgbdMCKeAp5Ky/8g6SFgBbABeGeqthW4Ffh4Kr8yIgK4TdJBkg5PdbdHxF4ASduB9cDVgx6bWY6+vPOH7Hz0OQBuuGsPr1ta/Pm9dknxfzSHg+VkTvocJK0CfhPYCRyWggPgaeCwtLwCeLK02e5UNlW52aLT/fe/0TNCGnxZyfIydDhI+kfA/wE+EhE/Ka9LrYQ5+42XtFnSLkm7xsbG5mq3ZpXpTpFRvqw04g5py9BQ4SDpNRTB8KWI+KtU/Ey6XET6/mwq3wMcUdp8ZSqbqnySiLg8ItZFxLrly5cPc+hmteiM9znsK2s2fCur5WeYu5UEXAE8FBH/s7RqG9C942gTcEOp/PR019KxwIvp8tNNwImSlqWO6BNTmdmi020cSJNHSDsbLCcDd0gDxwF/ANwn6e5U9l+BTwHXSToTeAJ4f1p3I3AKMAq8DJwBEBF7JZ0P3JHqndftnDZbbPq2HDwIzjI0zN1KfwtoitUn9KkfwFlT7GsLsGXQYzFbKDoRNDSx5eA+B8uRR0ibVajTmdgZDeU+B4eD5cPhYFahouUwMRy68yy55WA5cTiYVSgi6MkGmumv0OFgOXE4mFWoHf0uK7nlYPlxOJhVqBMxYXQ0eOI9y5PDwaxCnU7QnHRZqXu3km9ltXw4HMwq1OlzWcktB8uRw8GsQv06pBse52AZcjiYVajd91ZWh4Plx+FgVqFOMKlD2oPgLEcOB7MKFR3SHgRn+XM4mFWoO7dSmVsOliOHg1mFIiZOugf7wqHjcLCMOBzMKuSWgy0UDgezCrWnGSHtQXCWE4eDWYU8ZbctFA4Hswp1ot/dSqnl0HY4WD4cDmYV6j9ldwqHcDhYPhwOZhXqN7eSJBryOAfLi8PBrEL9puyGYiCc+xwsJw4Hswq1O5NvZYXi0pJbDpYTh4NZhfpdVoKiU7rlDmnLiMPBrELRZxAcFJPxddwhbRlxOJhVqNNnym5ILQcPgrOMOBzMKtRvym5wn4Plx+FgVqFOZ5qWg/scLCPZhIOk9ZIeljQq6ey6j8dsPrSn6HNoNt1ysLxkEQ6SmsBfACcDa4HTJK2t96jM5l5McbdSU/IIactKFuEAHA2MRsSjEfFz4BpgQ83HZDbn+k3ZDUWfgwfBWU5G6j6AZAXwZOn1buCY+XijTVtu54nnfjofu87adP/saBZ1Ju3vVVSO0p77bdfnP9Kv6n3K28/2uGaqF6UKvQ/nmWmb8q67rQQJfvpKi1danSlHSO946BmO+9TNSEX97iEM0qDoHvJU287yI836fWYy3c9dDHYw0mC/u6/6ffqUzebvaT7d9NF3sHSkOa/vkUs4zIqkzcDm9PIlSQ/XeTxDOhT4cd0HUbP98hw8AXxx38tDP7AfnoMe++XvQY9XdQ4O+NhQ7/VPZ1Mpl3DYAxxRer0ylU0QEZcDl1d1UPNJ0q6IWFf3cdTJ58DnAHwOIM9zkEufwx3AGkmrJS0BNgLbaj4mM7P9VhYth4hoSfowcBPQBLZExAM1H5aZ2X4ri3AAiIgbgRvrPo4KLYrLY0PyOfA5AJ8DyPAcKHxvtZmZ9cilz8HMzDLicJhjkg6WtF3SI+n7sinqbUp1HpG0qVR+oaQnJb3UU3+ppGvT9CI7Ja2a308yuDk4B2+TdF/6rJcoDTSQ9ElJeyTdnb5OqeozzcZMU8BM9zOUdE4qf1jSSbPdZ27m6Rw8nn4f7pa0q5pPMrhBz4GkQyTdIuklSZf2bNP3b2JeRYS/5vAL+DPg7LR8NnBRnzoHA4+m78vS8rK07ljgcOClnm3+EPjLtLwRuLbuzzqP5+D2dB4EfAM4OZV/EvjPdX++KT5zE/gB8CvAEuAeYO1sfoYUU8bcAywFVqf9NGezz5y+5uMcpHWPA4fW/fkqOAcHAm8H/j1wac82ff8m5vPLLYe5twHYmpa3Aqf2qXMSsD0i9kbE88B2YD1ARNwWEU/NsN/rgRMq+d/DYAY+B5IOB34pnYcArpxi+9zMZgqYqX6GG4BrIuKViHgMGE37W2jTyszHOVhoBj4HEfHTiPhb4GflynX9TTgc5t5hpX/cnwYO61On33QhK2bY7/g2EdECXgQOGe5Q580w52BFWu4t7/qwpHslbZnqclVNZvMznepnON25eLW/J3Waj3MAxWwVfy3pzjRLQs6GOQfT7XO6v4l5kc2trAuJpG8B/6TPqk+UX0RESFqUt4PVdA4uA86n+MfifOB/AB+ao31bvt4eEXsk/WNgu6TvRcS36z6oxc7hMICI+N2p1kl6RtLhEfFUag4+26faHuCdpdcrgVtneNvuFCO7JY0AbwCeezXHPZfm8RzsScvl8j3pPZ8pvcf/Br426PHPg9lMATPVz3C6bWecViYj83IOIqL7/VlJX6W4dJNrOAxzDqbbZ9+/ifnky0pzbxvQvfNmE3BDnzo3ASdKWpYujZyYyma73/cCN6frjzka+Byky1E/kXRsuhZ9enf7FDRd/xq4f74+wABmMwXMVD/DbcDGdBfLamANRQfkQptWZs7PgaQDJb0eQNKBFL8nOf3cew1zDvqa7m9iXtXdu7/YviiuHe4AHgG+BRycytcBnyvV+xBFp9socEap/M8oril20vdPpvIDgK+k+rcDv1L3Z53Hc7CO4h+AHwCXsm+w5lXAfcC9FH9gh9f9WXs+9ynA99NxfyKVnQe8Z6afIcXluB8AD1O6E6XfPnP+mutzQHHXzz3p64H94Bw8DuwFXkp//2tTed+/ifn88ghpMzObxJeVzMxsEoeDmZlN4nAwM7NJHA5mZjaJw8HMzCZxOJiZ2SQOB1sUJH1E0uumWPfB3imQ5+H9V0n6/dLrd6e5gO5L34+fz/c3m2sOB1ssPgL0DYeKrAJ+v/T6x8C/ioi3UIyGvaqOgzIblMPBFpw0pcLXJd0j6X5J5wJvBG6RdEuqc4ak70u6HThuhv2tknRzmu11h6RfTuX/TNJt6X//F6jnAUw9PgX8TnogzUcj4q6I+FFa9wDwWklL035PS/u8X9JFMxzbS5I+LekBSd+SdLSkWyU9Kuk9qc4Bkj6f9nmXpHel8gktJklfk/TO6d7PrMvhYAvReuBHEfEbEfFm4M+BHwHvioh3pTmY/htFKLyd4kEy0/lfwNaI+HXgS8AlqfwzwGfS//53T7VxcjbwnYh4a0Rc3LPu3wDfjYhXJL0RuAg4Hngr8FuSppub/0CKuXeOBP4BuAB4N8XcUuelOmdRTID7FuA0YKukA2Y4XrNpORxsIboPeLekiyT9TkS82LP+GODWiBiL4oEr186wv98GvpyWr6IIlG75V9Lyl3s3mg1JR1KEwb9LRb9VOrYWRRi9Y5pd/Bz4Zlq+D/ibiPhFWl6Vyt8OfBEgIr4HPAH86iDHa9blKbttwYmI70s6imKCswsk7aj7mPqRtBL4KnB6RPxgwN38IvZNgNYBXgGIiE6a7nk6LSb+B9CtCZs1txxswUmXZl6OiC8CnwaOorjk8vpUZSfwL1Q8sP01wPtm2OX/pZhaGeADwHfS8m0Ul4QorZ9K+f2RdBDwdYpnaf9dqd7t6dgOldSkuAz0NzPseybfSceNpF8FfpliZtPHgbdKakg6goX52E2riVsOthC9Bfi0pA7wC+A/UFwC+qakH6V+h08Cfw+8ANw9w/7+CPi8pP8CjAFnpPKPAF+U9AmKSzu9l6/K7gXaku4BvkDRV/Am4E8l/Wmqc2IUD0A6G7iF4mHxX4+IYefm/yxwmaT7KFoLH0z9G38HPAY8CDwEfHfI97H9iKfsNptCGjfx/yIiJG0ETouI3ofFmy1KbjmYTe1twKXp6Vsv4OdV237ELQfbb6TLQ739D1+JiAtfxT7ewuQBba9ExDFDHttOYGlP8R9ExH3D7NdsUA4HMzObxHcrmZnZJA4HMzObxOFgZmaTOBzMzGwSh4OZmU3y/wGLY55At4ypmwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "univariate(churn.std_og_t2o_mou)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "count    99999.000000\n",
      "mean       125.995363\n",
      "std        302.796523\n",
      "min       -707.336289\n",
      "25%          5.490000\n",
      "50%         30.780000\n",
      "75%        109.581675\n",
      "max      10752.560000\n",
      "Name: onnet_mou_8, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "univariate(churn.onnet_mou_8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "count    99999.000000\n",
      "mean       175.999079\n",
      "std        312.011491\n",
      "min      -1389.344441\n",
      "25%         21.880000\n",
      "50%         78.410000\n",
      "75%        204.295000\n",
      "max      10310.760000\n",
      "Name: offnet_mou_9, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "univariate(churn.offnet_mou_9)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Variables are very **skewed** towards the left."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bivariate EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "bivariate(churn_filtered.churn, churn_filtered.aon)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "bivariate(churn_filtered.sep_vbc_3g, churn_filtered.churn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "bivariate(churn_filtered.spl_og_mou_8, churn_filtered.churn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>night_pck_user_8</th>\n",
       "      <th>-1.0</th>\n",
       "      <th>0.0</th>\n",
       "      <th>1.0</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>churn</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>85.89123</td>\n",
       "      <td>97.117602</td>\n",
       "      <td>97.360704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>14.10877</td>\n",
       "      <td>2.882398</td>\n",
       "      <td>2.639296</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "night_pck_user_8      -1.0        0.0        1.0\n",
       "churn                                           \n",
       "0                 85.89123  97.117602  97.360704\n",
       "1                 14.10877   2.882398   2.639296"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(churn_filtered.churn, churn_filtered.night_pck_user_8, normalize='columns')*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>sachet_3g_8</th>\n",
       "      <th>0.0</th>\n",
       "      <th>1.0</th>\n",
       "      <th>2.0</th>\n",
       "      <th>3.0</th>\n",
       "      <th>4.0</th>\n",
       "      <th>5.0</th>\n",
       "      <th>6.0</th>\n",
       "      <th>7.0</th>\n",
       "      <th>8.0</th>\n",
       "      <th>9.0</th>\n",
       "      <th>10.0</th>\n",
       "      <th>11.0</th>\n",
       "      <th>12.0</th>\n",
       "      <th>13.0</th>\n",
       "      <th>14.0</th>\n",
       "      <th>15.0</th>\n",
       "      <th>16.0</th>\n",
       "      <th>17.0</th>\n",
       "      <th>18.0</th>\n",
       "      <th>19.0</th>\n",
       "      <th>20.0</th>\n",
       "      <th>21.0</th>\n",
       "      <th>23.0</th>\n",
       "      <th>25.0</th>\n",
       "      <th>27.0</th>\n",
       "      <th>29.0</th>\n",
       "      <th>30.0</th>\n",
       "      <th>38.0</th>\n",
       "      <th>41.0</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>churn</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>24972</td>\n",
       "      <td>1609</td>\n",
       "      <td>399</td>\n",
       "      <td>184</td>\n",
       "      <td>106</td>\n",
       "      <td>86</td>\n",
       "      <td>43</td>\n",
       "      <td>35</td>\n",
       "      <td>28</td>\n",
       "      <td>19</td>\n",
       "      <td>15</td>\n",
       "      <td>8</td>\n",
       "      <td>11</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2369</td>\n",
       "      <td>48</td>\n",
       "      <td>5</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "sachet_3g_8   0.0   1.0   2.0   3.0   4.0   5.0   6.0   7.0   8.0   9.0   \\\n",
       "churn                                                                      \n",
       "0            24972  1609   399   184   106    86    43    35    28    19   \n",
       "1             2369    48     5     8     4     2     1     0     2     0   \n",
       "\n",
       "sachet_3g_8  10.0  11.0  12.0  13.0  14.0  15.0  16.0  17.0  18.0  19.0  20.0  \\\n",
       "churn                                                                           \n",
       "0              15     8    11    10     6     6     2     2     3     1     3   \n",
       "1               0     0     0     1     0     0     1     0     0     0     0   \n",
       "\n",
       "sachet_3g_8  21.0  23.0  25.0  27.0  29.0  30.0  38.0  41.0  \n",
       "churn                                                        \n",
       "0               3     2     1     1     1     2     1     1  \n",
       "1               0     0     0     0     0     0     0     0  "
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(churn_filtered.churn, churn_filtered.sachet_3g_8)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cap outliers in all numeric variables with k-sigma technique"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def cap_outliers(array, k=3):\n",
    "    upper_limit = array.mean() + k*array.std()\n",
    "    lower_limit = array.mean() - k*array.std()\n",
    "    array[array<lower_limit] = lower_limit\n",
    "    array[array>upper_limit] = upper_limit\n",
    "    return array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Array after capping outliers: \n",
      " [-2780     1     2     3     4     5     6     7     8     9    10    11\n",
      "    12    13    14    15    16    17    18    19    20    21    22    23\n",
      "    24    25    26    27    28    29    30    31    32    33    34    35\n",
      "    36    37    38    39    40    41    42    43    44    45    46    47\n",
      "    48    49    50    51    52    53    54    55    56    57    58    59\n",
      "    60    61    62    63    64    65    66    67    68    69    70    71\n",
      "    72    73    74    75    76    77    78    79    80    81    82    83\n",
      "    84    85    86    87    88    89    90    91    92    93    94    95\n",
      "    96    97    98  2877]\n"
     ]
    }
   ],
   "source": [
    "# example of capping\n",
    "sample_array = list(range(100))\n",
    "\n",
    "# add outliers to the data\n",
    "sample_array[0] = -9999\n",
    "sample_array[99] = 9999\n",
    "\n",
    "# cap outliers\n",
    "sample_array = np.array(sample_array)\n",
    "print(\"Array after capping outliers: \\n\", cap_outliers(sample_array, k=2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# cap outliers in the numeric columns\n",
    "churn_filtered[num_cols] = churn_filtered[num_cols].apply(cap_outliers, axis=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Modelling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## i) Making predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# import required libraries\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import FeatureUnion\n",
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from imblearn.metrics import sensitivity_specificity_support\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from sklearn.svm import SVC"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# change churn to numeric\n",
    "churn_filtered['churn'] = pd.to_numeric(churn_filtered['churn'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Test split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# divide data into train and test\n",
    "X = churn_filtered.drop(\"churn\", axis = 1)\n",
    "y = churn_filtered.churn\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.25, random_state = 4, stratify = y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(22500, 172)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "(22500,)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "(7501, 172)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "(7501,)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# print shapes of train and test sets\n",
    "X_train.shape\n",
    "y_train.shape\n",
    "X_test.shape\n",
    "y_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Aggregating the categorical columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "night_pck_user_6\n",
       "-1.0    0.099165\n",
       " 0.0    0.066797\n",
       " 1.0    0.087838\n",
       "Name: churn, dtype: float64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "night_pck_user_7\n",
       "-1.0    0.115746\n",
       " 0.0    0.055494\n",
       " 1.0    0.051282\n",
       "Name: churn, dtype: float64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "night_pck_user_8\n",
       "-1.0    0.141108\n",
       " 0.0    0.029023\n",
       " 1.0    0.016194\n",
       "Name: churn, dtype: float64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "fb_user_6\n",
       "-1.0    0.099165\n",
       " 0.0    0.069460\n",
       " 1.0    0.067124\n",
       "Name: churn, dtype: float64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "fb_user_7\n",
       "-1.0    0.115746\n",
       " 0.0    0.059305\n",
       " 1.0    0.055082\n",
       "Name: churn, dtype: float64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "fb_user_8\n",
       "-1.0    0.141108\n",
       " 0.0    0.066887\n",
       " 1.0    0.024463\n",
       "Name: churn, dtype: float64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train = pd.concat([X_train, y_train], axis=1)\n",
    "\n",
    "# aggregate the categorical variables\n",
    "train.groupby('night_pck_user_6').churn.mean()\n",
    "train.groupby('night_pck_user_7').churn.mean()\n",
    "train.groupby('night_pck_user_8').churn.mean()\n",
    "train.groupby('fb_user_6').churn.mean()\n",
    "train.groupby('fb_user_7').churn.mean()\n",
    "train.groupby('fb_user_8').churn.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# replace categories with aggregated values in each categorical column\n",
    "mapping = {'night_pck_user_6' : {-1: 0.099165, 0: 0.066797, 1: 0.087838},\n",
    "           'night_pck_user_7' : {-1: 0.115746, 0: 0.055494, 1: 0.051282},\n",
    "           'night_pck_user_8' : {-1: 0.141108, 0: 0.029023, 1: 0.016194},\n",
    "           'fb_user_6'        : {-1: 0.099165, 0: 0.069460, 1: 0.067124},\n",
    "           'fb_user_7'        : {-1: 0.115746, 0: 0.059305, 1: 0.055082},\n",
    "           'fb_user_8'        : {-1: 0.141108, 0: 0.066887, 1: 0.024463}\n",
    "          }\n",
    "X_train.replace(mapping, inplace = True)\n",
    "X_test.replace(mapping, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 22500 entries, 4525 to 22754\n",
      "Data columns (total 6 columns):\n",
      "night_pck_user_6    22500 non-null float64\n",
      "night_pck_user_7    22500 non-null float64\n",
      "night_pck_user_8    22500 non-null float64\n",
      "fb_user_6           22500 non-null float64\n",
      "fb_user_7           22500 non-null float64\n",
      "fb_user_8           22500 non-null float64\n",
      "dtypes: float64(6)\n",
      "memory usage: 1.2 MB\n"
     ]
    }
   ],
   "source": [
    "# check data type of categorical columns - make sure they are numeric\n",
    "X_train[[col for col in cat_cols if col not in ['churn']]].info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# apply pca to train data\n",
    "pca = Pipeline([('scaler', StandardScaler()), ('pca', PCA())])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('scaler',\n",
       "                 StandardScaler(copy=True, with_mean=True, with_std=True)),\n",
       "                ('pca',\n",
       "                 PCA(copy=True, iterated_power='auto', n_components=None,\n",
       "                     random_state=None, svd_solver='auto', tol=0.0,\n",
       "                     whiten=False))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca.fit(X_train)\n",
    "churn_pca = pca.fit_transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0       10.91\n",
      "1       19.80\n",
      "2       25.31\n",
      "3       29.96\n",
      "4       33.81\n",
      "5       37.30\n",
      "6       39.90\n",
      "7       42.33\n",
      "8       44.51\n",
      "9       46.52\n",
      "10      48.41\n",
      "11      50.23\n",
      "12      51.91\n",
      "13      53.53\n",
      "14      55.02\n",
      "15      56.39\n",
      "16      57.68\n",
      "17      58.94\n",
      "18      60.18\n",
      "19      61.40\n",
      "20      62.52\n",
      "21      63.63\n",
      "22      64.69\n",
      "23      65.71\n",
      "24      66.69\n",
      "25      67.59\n",
      "26      68.48\n",
      "27      69.33\n",
      "28      70.16\n",
      "29      70.98\n",
      "30      71.75\n",
      "31      72.52\n",
      "32      73.27\n",
      "33      74.00\n",
      "34      74.71\n",
      "35      75.40\n",
      "36      76.08\n",
      "37      76.76\n",
      "38      77.42\n",
      "39      78.07\n",
      "40      78.70\n",
      "41      79.33\n",
      "42      79.95\n",
      "43      80.55\n",
      "44      81.13\n",
      "45      81.70\n",
      "46      82.26\n",
      "47      82.80\n",
      "48      83.34\n",
      "49      83.86\n",
      "50      84.37\n",
      "51      84.87\n",
      "52      85.37\n",
      "53      85.87\n",
      "54      86.35\n",
      "55      86.81\n",
      "56      87.28\n",
      "57      87.72\n",
      "58      88.17\n",
      "59      88.59\n",
      "60      89.00\n",
      "61      89.39\n",
      "62      89.76\n",
      "63      90.12\n",
      "64      90.46\n",
      "65      90.80\n",
      "66      91.13\n",
      "67      91.44\n",
      "68      91.74\n",
      "69      92.04\n",
      "70      92.33\n",
      "71      92.61\n",
      "72      92.88\n",
      "73      93.15\n",
      "74      93.42\n",
      "75      93.68\n",
      "76      93.92\n",
      "77      94.16\n",
      "78      94.40\n",
      "79      94.63\n",
      "80      94.85\n",
      "81      95.05\n",
      "82      95.24\n",
      "83      95.43\n",
      "84      95.61\n",
      "85      95.79\n",
      "86      95.97\n",
      "87      96.14\n",
      "88      96.31\n",
      "89      96.48\n",
      "90      96.64\n",
      "91      96.80\n",
      "92      96.96\n",
      "93      97.11\n",
      "94      97.25\n",
      "95      97.39\n",
      "96      97.53\n",
      "97      97.66\n",
      "98      97.78\n",
      "99      97.90\n",
      "100     98.03\n",
      "101     98.14\n",
      "102     98.25\n",
      "103     98.34\n",
      "104     98.43\n",
      "105     98.52\n",
      "106     98.60\n",
      "107     98.68\n",
      "108     98.76\n",
      "109     98.83\n",
      "110     98.90\n",
      "111     98.96\n",
      "112     99.01\n",
      "113     99.06\n",
      "114     99.12\n",
      "115     99.17\n",
      "116     99.22\n",
      "117     99.26\n",
      "118     99.31\n",
      "119     99.34\n",
      "120     99.38\n",
      "121     99.42\n",
      "122     99.45\n",
      "123     99.48\n",
      "124     99.51\n",
      "125     99.53\n",
      "126     99.56\n",
      "127     99.59\n",
      "128     99.61\n",
      "129     99.63\n",
      "130     99.65\n",
      "131     99.67\n",
      "132     99.69\n",
      "133     99.71\n",
      "134     99.73\n",
      "135     99.74\n",
      "136     99.76\n",
      "137     99.77\n",
      "138     99.79\n",
      "139     99.80\n",
      "140     99.82\n",
      "141     99.83\n",
      "142     99.84\n",
      "143     99.85\n",
      "144     99.86\n",
      "145     99.87\n",
      "146     99.88\n",
      "147     99.89\n",
      "148     99.90\n",
      "149     99.91\n",
      "150     99.92\n",
      "151     99.92\n",
      "152     99.93\n",
      "153     99.94\n",
      "154     99.94\n",
      "155     99.95\n",
      "156     99.95\n",
      "157     99.96\n",
      "158     99.97\n",
      "159     99.97\n",
      "160     99.97\n",
      "161     99.98\n",
      "162     99.98\n",
      "163     99.99\n",
      "164     99.99\n",
      "165     99.99\n",
      "166     99.99\n",
      "167    100.00\n",
      "168    100.00\n",
      "169    100.00\n",
      "170    100.00\n",
      "171    100.00\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# extract pca model from pipeline\n",
    "pca = pca.named_steps['pca']\n",
    "\n",
    "# look at explainded variance of PCA components\n",
    "print(pd.Series(np.round(pca.explained_variance_ratio_.cumsum(), 4)*100))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "~ 60 components explain 90% variance\n",
    "\n",
    "~ 80 components explain 95% variance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 630x360 with 0 Axes>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f50acd63f98>]"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 630x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot feature variance\n",
    "features = range(pca.n_components_)\n",
    "cumulative_variance = np.round(np.cumsum(pca.explained_variance_ratio_)*100, decimals=4)\n",
    "plt.figure(figsize=(175/20,100/20)) # 100 elements on y-axis; 175 elements on x-axis; 20 is normalising factor\n",
    "plt.plot(cumulative_variance)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PCA and Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create pipeline\n",
    "PCA_VARS = 60\n",
    "steps = [('scaler', StandardScaler()),\n",
    "         (\"pca\", PCA(n_components=PCA_VARS)),\n",
    "         (\"logistic\", LogisticRegression(class_weight='balanced'))\n",
    "        ]\n",
    "pipeline = Pipeline(steps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('scaler',\n",
       "                 StandardScaler(copy=True, with_mean=True, with_std=True)),\n",
       "                ('pca',\n",
       "                 PCA(copy=True, iterated_power='auto', n_components=60,\n",
       "                     random_state=None, svd_solver='auto', tol=0.0,\n",
       "                     whiten=False)),\n",
       "                ('logistic',\n",
       "                 LogisticRegression(C=1.0, class_weight='balanced', dual=False,\n",
       "                                    fit_intercept=True, intercept_scaling=1,\n",
       "                                    l1_ratio=None, max_iter=100,\n",
       "                                    multi_class='auto', n_jobs=None,\n",
       "                                    penalty='l2', random_state=None,\n",
       "                                    solver='lbfgs', tol=0.0001, verbose=0,\n",
       "                                    warm_start=False))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "0.8205333333333333"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit model\n",
    "pipeline.fit(X_train, y_train)\n",
    "\n",
    "# check score on train data\n",
    "pipeline.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate on test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[5614 1277]\n",
      " [  95  515]]\n",
      "Sensitivity: \t0.84\n",
      "Specificity: \t0.81\n",
      "AUC:    \t 0.9\n"
     ]
    }
   ],
   "source": [
    "# predict churn on test data\n",
    "y_pred = pipeline.predict(X_test)\n",
    "\n",
    "# create onfusion matrix\n",
    "cm = confusion_matrix(y_test, y_pred)\n",
    "print(cm)\n",
    "\n",
    "# check sensitivity and specificity\n",
    "sensitivity, specificity, _ = sensitivity_specificity_support(y_test, y_pred, average='binary')\n",
    "print(\"Sensitivity: \\t\", round(sensitivity, 2), \"\\n\", \"Specificity: \\t\", round(specificity, 2), sep='')\n",
    "\n",
    "# check area under curve\n",
    "y_pred_prob = pipeline.predict_proba(X_test)[:, 1]\n",
    "print(\"AUC:    \\t\", round(roc_auc_score(y_test, y_pred_prob),2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyperparameter tuning - PCA and Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0.918622\n",
       "1    0.081378\n",
       "Name: churn, dtype: float64"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# class imbalance\n",
    "y_train.value_counts()/y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# PCA\n",
    "pca = PCA()\n",
    "\n",
    "# logistic regression - the class weight is used to handle class imbalance - it adjusts the cost function\n",
    "logistic = LogisticRegression(class_weight={0:0.1, 1: 0.9})\n",
    "\n",
    "# create pipeline\n",
    "steps = [(\"scaler\", StandardScaler()), \n",
    "         (\"pca\", pca),\n",
    "         (\"logistic\", logistic)\n",
    "        ]\n",
    "\n",
    "# compile pipeline\n",
    "pca_logistic = Pipeline(steps)\n",
    "\n",
    "# hyperparameter space\n",
    "params = {'pca__n_components': [60, 80], 'logistic__C': [0.1, 0.5, 1, 2, 3, 4, 5, 10], 'logistic__penalty': ['l1', 'l2']}\n",
    "\n",
    "# create 5 folds\n",
    "folds = StratifiedKFold(n_splits = 5, shuffle = True, random_state = 4)\n",
    "\n",
    "# create gridsearch object\n",
    "model = GridSearchCV(estimator=pca_logistic, cv=folds, param_grid=params, scoring='roc_auc', n_jobs=-1, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 32 candidates, totalling 160 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  34 tasks      | elapsed:   13.6s\n",
      "[Parallel(n_jobs=-1)]: Done 160 out of 160 | elapsed:   51.4s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=StratifiedKFold(n_splits=5, random_state=4, shuffle=True),\n",
       "             error_score=nan,\n",
       "             estimator=Pipeline(memory=None,\n",
       "                                steps=[('scaler',\n",
       "                                        StandardScaler(copy=True,\n",
       "                                                       with_mean=True,\n",
       "                                                       with_std=True)),\n",
       "                                       ('pca',\n",
       "                                        PCA(copy=True, iterated_power='auto',\n",
       "                                            n_components=None,\n",
       "                                            random_state=None,\n",
       "                                            svd_solver='auto', tol=0.0,\n",
       "                                            whiten=False)),\n",
       "                                       ('logistic',\n",
       "                                        LogisticRegression(C=1.0,\n",
       "                                                           c...\n",
       "                                                           multi_class='auto',\n",
       "                                                           n_jobs=None,\n",
       "                                                           penalty='l2',\n",
       "                                                           random_state=None,\n",
       "                                                           solver='lbfgs',\n",
       "                                                           tol=0.0001,\n",
       "                                                           verbose=0,\n",
       "                                                           warm_start=False))],\n",
       "                                verbose=False),\n",
       "             iid='deprecated', n_jobs=-1,\n",
       "             param_grid={'logistic__C': [0.1, 0.5, 1, 2, 3, 4, 5, 10],\n",
       "                         'logistic__penalty': ['l1', 'l2'],\n",
       "                         'pca__n_components': [60, 80]},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring='roc_auc', verbose=1)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit model\n",
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_fit_time</th>\n",
       "      <th>mean_score_time</th>\n",
       "      <th>mean_test_score</th>\n",
       "      <th>param_logistic__C</th>\n",
       "      <th>param_logistic__penalty</th>\n",
       "      <th>param_pca__n_components</th>\n",
       "      <th>params</th>\n",
       "      <th>rank_test_score</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "      <th>split3_test_score</th>\n",
       "      <th>split4_test_score</th>\n",
       "      <th>std_fit_time</th>\n",
       "      <th>std_score_time</th>\n",
       "      <th>std_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2.072347</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.1</td>\n",
       "      <td>l1</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>32</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.179106</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.450242</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.1</td>\n",
       "      <td>l1</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.068343</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.323431</td>\n",
       "      <td>0.045277</td>\n",
       "      <td>0.888393</td>\n",
       "      <td>0.1</td>\n",
       "      <td>l2</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>15</td>\n",
       "      <td>0.887588</td>\n",
       "      <td>0.875877</td>\n",
       "      <td>0.889909</td>\n",
       "      <td>0.906372</td>\n",
       "      <td>0.882219</td>\n",
       "      <td>0.028023</td>\n",
       "      <td>0.006484</td>\n",
       "      <td>0.010207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2.897196</td>\n",
       "      <td>0.048496</td>\n",
       "      <td>0.892046</td>\n",
       "      <td>0.1</td>\n",
       "      <td>l2</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>1</td>\n",
       "      <td>0.893150</td>\n",
       "      <td>0.880796</td>\n",
       "      <td>0.890025</td>\n",
       "      <td>0.906630</td>\n",
       "      <td>0.889626</td>\n",
       "      <td>0.042294</td>\n",
       "      <td>0.007149</td>\n",
       "      <td>0.008372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.905288</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.5</td>\n",
       "      <td>l1</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.170325</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2.283087</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.5</td>\n",
       "      <td>l1</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>27</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.134897</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2.223308</td>\n",
       "      <td>0.045896</td>\n",
       "      <td>0.888674</td>\n",
       "      <td>0.5</td>\n",
       "      <td>l2</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>11</td>\n",
       "      <td>0.887447</td>\n",
       "      <td>0.875879</td>\n",
       "      <td>0.891028</td>\n",
       "      <td>0.906559</td>\n",
       "      <td>0.882456</td>\n",
       "      <td>0.115637</td>\n",
       "      <td>0.006426</td>\n",
       "      <td>0.010289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2.918834</td>\n",
       "      <td>0.052805</td>\n",
       "      <td>0.891477</td>\n",
       "      <td>0.5</td>\n",
       "      <td>l2</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>8</td>\n",
       "      <td>0.893332</td>\n",
       "      <td>0.880602</td>\n",
       "      <td>0.888948</td>\n",
       "      <td>0.905079</td>\n",
       "      <td>0.889425</td>\n",
       "      <td>0.055376</td>\n",
       "      <td>0.008769</td>\n",
       "      <td>0.007967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1.742697</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>l1</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>19</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.165929</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2.463538</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>l1</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.035041</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2.286158</td>\n",
       "      <td>0.051844</td>\n",
       "      <td>0.888651</td>\n",
       "      <td>1</td>\n",
       "      <td>l2</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>12</td>\n",
       "      <td>0.888083</td>\n",
       "      <td>0.876299</td>\n",
       "      <td>0.890252</td>\n",
       "      <td>0.905843</td>\n",
       "      <td>0.882776</td>\n",
       "      <td>0.051910</td>\n",
       "      <td>0.011756</td>\n",
       "      <td>0.009854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2.867383</td>\n",
       "      <td>0.038958</td>\n",
       "      <td>0.891716</td>\n",
       "      <td>1</td>\n",
       "      <td>l2</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>5</td>\n",
       "      <td>0.893352</td>\n",
       "      <td>0.881054</td>\n",
       "      <td>0.888762</td>\n",
       "      <td>0.905991</td>\n",
       "      <td>0.889419</td>\n",
       "      <td>0.042032</td>\n",
       "      <td>0.004801</td>\n",
       "      <td>0.008175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1.683732</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>l1</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.100971</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>2.326292</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>l1</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>31</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.068712</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2.196153</td>\n",
       "      <td>0.047117</td>\n",
       "      <td>0.888543</td>\n",
       "      <td>2</td>\n",
       "      <td>l2</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>13</td>\n",
       "      <td>0.887622</td>\n",
       "      <td>0.875647</td>\n",
       "      <td>0.890623</td>\n",
       "      <td>0.906187</td>\n",
       "      <td>0.882637</td>\n",
       "      <td>0.083508</td>\n",
       "      <td>0.008024</td>\n",
       "      <td>0.010175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>2.886672</td>\n",
       "      <td>0.052928</td>\n",
       "      <td>0.891727</td>\n",
       "      <td>2</td>\n",
       "      <td>l2</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>4</td>\n",
       "      <td>0.893335</td>\n",
       "      <td>0.880494</td>\n",
       "      <td>0.889003</td>\n",
       "      <td>0.906268</td>\n",
       "      <td>0.889535</td>\n",
       "      <td>0.074145</td>\n",
       "      <td>0.008369</td>\n",
       "      <td>0.008396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.741456</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>l1</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.105221</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>2.376044</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>l1</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.051659</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>2.257523</td>\n",
       "      <td>0.047837</td>\n",
       "      <td>0.888475</td>\n",
       "      <td>3</td>\n",
       "      <td>l2</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>14</td>\n",
       "      <td>0.887593</td>\n",
       "      <td>0.876824</td>\n",
       "      <td>0.890079</td>\n",
       "      <td>0.905703</td>\n",
       "      <td>0.882176</td>\n",
       "      <td>0.127460</td>\n",
       "      <td>0.010735</td>\n",
       "      <td>0.009753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>2.872739</td>\n",
       "      <td>0.054852</td>\n",
       "      <td>0.891881</td>\n",
       "      <td>3</td>\n",
       "      <td>l2</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>2</td>\n",
       "      <td>0.893473</td>\n",
       "      <td>0.881179</td>\n",
       "      <td>0.889009</td>\n",
       "      <td>0.906119</td>\n",
       "      <td>0.889626</td>\n",
       "      <td>0.031622</td>\n",
       "      <td>0.018015</td>\n",
       "      <td>0.008162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1.677792</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>l1</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.080890</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>2.353303</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4</td>\n",
       "      <td>l1</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>25</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.095411</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>2.230132</td>\n",
       "      <td>0.057473</td>\n",
       "      <td>0.888268</td>\n",
       "      <td>4</td>\n",
       "      <td>l2</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>16</td>\n",
       "      <td>0.887541</td>\n",
       "      <td>0.875820</td>\n",
       "      <td>0.890402</td>\n",
       "      <td>0.905956</td>\n",
       "      <td>0.881619</td>\n",
       "      <td>0.031008</td>\n",
       "      <td>0.009995</td>\n",
       "      <td>0.010170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>2.879795</td>\n",
       "      <td>0.049074</td>\n",
       "      <td>0.891777</td>\n",
       "      <td>4</td>\n",
       "      <td>l2</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>3</td>\n",
       "      <td>0.893217</td>\n",
       "      <td>0.881290</td>\n",
       "      <td>0.888364</td>\n",
       "      <td>0.906204</td>\n",
       "      <td>0.889809</td>\n",
       "      <td>0.045812</td>\n",
       "      <td>0.006387</td>\n",
       "      <td>0.008194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1.657990</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5</td>\n",
       "      <td>l1</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.084555</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>2.460682</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5</td>\n",
       "      <td>l1</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.055552</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>2.274911</td>\n",
       "      <td>0.061866</td>\n",
       "      <td>0.888798</td>\n",
       "      <td>5</td>\n",
       "      <td>l2</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>9</td>\n",
       "      <td>0.887991</td>\n",
       "      <td>0.876833</td>\n",
       "      <td>0.890735</td>\n",
       "      <td>0.906306</td>\n",
       "      <td>0.882126</td>\n",
       "      <td>0.068856</td>\n",
       "      <td>0.007284</td>\n",
       "      <td>0.009986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>2.847522</td>\n",
       "      <td>0.047065</td>\n",
       "      <td>0.891590</td>\n",
       "      <td>5</td>\n",
       "      <td>l2</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>7</td>\n",
       "      <td>0.893240</td>\n",
       "      <td>0.881528</td>\n",
       "      <td>0.888193</td>\n",
       "      <td>0.905491</td>\n",
       "      <td>0.889501</td>\n",
       "      <td>0.066800</td>\n",
       "      <td>0.005178</td>\n",
       "      <td>0.007913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1.702650</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10</td>\n",
       "      <td>l1</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>18</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.078606</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>2.422845</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10</td>\n",
       "      <td>l1</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.041243</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>2.194749</td>\n",
       "      <td>0.043969</td>\n",
       "      <td>0.888769</td>\n",
       "      <td>10</td>\n",
       "      <td>l2</td>\n",
       "      <td>60</td>\n",
       "      <td>{'pca__n_components': 60, 'logistic__penalty':...</td>\n",
       "      <td>10</td>\n",
       "      <td>0.887981</td>\n",
       "      <td>0.876398</td>\n",
       "      <td>0.890943</td>\n",
       "      <td>0.906101</td>\n",
       "      <td>0.882423</td>\n",
       "      <td>0.171670</td>\n",
       "      <td>0.013261</td>\n",
       "      <td>0.009990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>2.250251</td>\n",
       "      <td>0.022688</td>\n",
       "      <td>0.891670</td>\n",
       "      <td>10</td>\n",
       "      <td>l2</td>\n",
       "      <td>80</td>\n",
       "      <td>{'pca__n_components': 80, 'logistic__penalty':...</td>\n",
       "      <td>6</td>\n",
       "      <td>0.893382</td>\n",
       "      <td>0.880989</td>\n",
       "      <td>0.889213</td>\n",
       "      <td>0.905230</td>\n",
       "      <td>0.889533</td>\n",
       "      <td>0.246210</td>\n",
       "      <td>0.004502</td>\n",
       "      <td>0.007893</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_fit_time  mean_score_time  mean_test_score param_logistic__C  \\\n",
       "0        2.072347         0.000000              NaN               0.1   \n",
       "1        2.450242         0.000000              NaN               0.1   \n",
       "2        2.323431         0.045277         0.888393               0.1   \n",
       "3        2.897196         0.048496         0.892046               0.1   \n",
       "4        1.905288         0.000000              NaN               0.5   \n",
       "5        2.283087         0.000000              NaN               0.5   \n",
       "6        2.223308         0.045896         0.888674               0.5   \n",
       "7        2.918834         0.052805         0.891477               0.5   \n",
       "8        1.742697         0.000000              NaN                 1   \n",
       "9        2.463538         0.000000              NaN                 1   \n",
       "10       2.286158         0.051844         0.888651                 1   \n",
       "11       2.867383         0.038958         0.891716                 1   \n",
       "12       1.683732         0.000000              NaN                 2   \n",
       "13       2.326292         0.000000              NaN                 2   \n",
       "14       2.196153         0.047117         0.888543                 2   \n",
       "15       2.886672         0.052928         0.891727                 2   \n",
       "16       1.741456         0.000000              NaN                 3   \n",
       "17       2.376044         0.000000              NaN                 3   \n",
       "18       2.257523         0.047837         0.888475                 3   \n",
       "19       2.872739         0.054852         0.891881                 3   \n",
       "20       1.677792         0.000000              NaN                 4   \n",
       "21       2.353303         0.000000              NaN                 4   \n",
       "22       2.230132         0.057473         0.888268                 4   \n",
       "23       2.879795         0.049074         0.891777                 4   \n",
       "24       1.657990         0.000000              NaN                 5   \n",
       "25       2.460682         0.000000              NaN                 5   \n",
       "26       2.274911         0.061866         0.888798                 5   \n",
       "27       2.847522         0.047065         0.891590                 5   \n",
       "28       1.702650         0.000000              NaN                10   \n",
       "29       2.422845         0.000000              NaN                10   \n",
       "30       2.194749         0.043969         0.888769                10   \n",
       "31       2.250251         0.022688         0.891670                10   \n",
       "\n",
       "   param_logistic__penalty param_pca__n_components  \\\n",
       "0                       l1                      60   \n",
       "1                       l1                      80   \n",
       "2                       l2                      60   \n",
       "3                       l2                      80   \n",
       "4                       l1                      60   \n",
       "5                       l1                      80   \n",
       "6                       l2                      60   \n",
       "7                       l2                      80   \n",
       "8                       l1                      60   \n",
       "9                       l1                      80   \n",
       "10                      l2                      60   \n",
       "11                      l2                      80   \n",
       "12                      l1                      60   \n",
       "13                      l1                      80   \n",
       "14                      l2                      60   \n",
       "15                      l2                      80   \n",
       "16                      l1                      60   \n",
       "17                      l1                      80   \n",
       "18                      l2                      60   \n",
       "19                      l2                      80   \n",
       "20                      l1                      60   \n",
       "21                      l1                      80   \n",
       "22                      l2                      60   \n",
       "23                      l2                      80   \n",
       "24                      l1                      60   \n",
       "25                      l1                      80   \n",
       "26                      l2                      60   \n",
       "27                      l2                      80   \n",
       "28                      l1                      60   \n",
       "29                      l1                      80   \n",
       "30                      l2                      60   \n",
       "31                      l2                      80   \n",
       "\n",
       "                                               params  rank_test_score  \\\n",
       "0   {'pca__n_components': 60, 'logistic__penalty':...               32   \n",
       "1   {'pca__n_components': 80, 'logistic__penalty':...               23   \n",
       "2   {'pca__n_components': 60, 'logistic__penalty':...               15   \n",
       "3   {'pca__n_components': 80, 'logistic__penalty':...                1   \n",
       "4   {'pca__n_components': 60, 'logistic__penalty':...               30   \n",
       "5   {'pca__n_components': 80, 'logistic__penalty':...               27   \n",
       "6   {'pca__n_components': 60, 'logistic__penalty':...               11   \n",
       "7   {'pca__n_components': 80, 'logistic__penalty':...                8   \n",
       "8   {'pca__n_components': 60, 'logistic__penalty':...               19   \n",
       "9   {'pca__n_components': 80, 'logistic__penalty':...               20   \n",
       "10  {'pca__n_components': 60, 'logistic__penalty':...               12   \n",
       "11  {'pca__n_components': 80, 'logistic__penalty':...                5   \n",
       "12  {'pca__n_components': 60, 'logistic__penalty':...               24   \n",
       "13  {'pca__n_components': 80, 'logistic__penalty':...               31   \n",
       "14  {'pca__n_components': 60, 'logistic__penalty':...               13   \n",
       "15  {'pca__n_components': 80, 'logistic__penalty':...                4   \n",
       "16  {'pca__n_components': 60, 'logistic__penalty':...               29   \n",
       "17  {'pca__n_components': 80, 'logistic__penalty':...               28   \n",
       "18  {'pca__n_components': 60, 'logistic__penalty':...               14   \n",
       "19  {'pca__n_components': 80, 'logistic__penalty':...                2   \n",
       "20  {'pca__n_components': 60, 'logistic__penalty':...               26   \n",
       "21  {'pca__n_components': 80, 'logistic__penalty':...               25   \n",
       "22  {'pca__n_components': 60, 'logistic__penalty':...               16   \n",
       "23  {'pca__n_components': 80, 'logistic__penalty':...                3   \n",
       "24  {'pca__n_components': 60, 'logistic__penalty':...               22   \n",
       "25  {'pca__n_components': 80, 'logistic__penalty':...               21   \n",
       "26  {'pca__n_components': 60, 'logistic__penalty':...                9   \n",
       "27  {'pca__n_components': 80, 'logistic__penalty':...                7   \n",
       "28  {'pca__n_components': 60, 'logistic__penalty':...               18   \n",
       "29  {'pca__n_components': 80, 'logistic__penalty':...               17   \n",
       "30  {'pca__n_components': 60, 'logistic__penalty':...               10   \n",
       "31  {'pca__n_components': 80, 'logistic__penalty':...                6   \n",
       "\n",
       "    split0_test_score  split1_test_score  split2_test_score  \\\n",
       "0                 NaN                NaN                NaN   \n",
       "1                 NaN                NaN                NaN   \n",
       "2            0.887588           0.875877           0.889909   \n",
       "3            0.893150           0.880796           0.890025   \n",
       "4                 NaN                NaN                NaN   \n",
       "5                 NaN                NaN                NaN   \n",
       "6            0.887447           0.875879           0.891028   \n",
       "7            0.893332           0.880602           0.888948   \n",
       "8                 NaN                NaN                NaN   \n",
       "9                 NaN                NaN                NaN   \n",
       "10           0.888083           0.876299           0.890252   \n",
       "11           0.893352           0.881054           0.888762   \n",
       "12                NaN                NaN                NaN   \n",
       "13                NaN                NaN                NaN   \n",
       "14           0.887622           0.875647           0.890623   \n",
       "15           0.893335           0.880494           0.889003   \n",
       "16                NaN                NaN                NaN   \n",
       "17                NaN                NaN                NaN   \n",
       "18           0.887593           0.876824           0.890079   \n",
       "19           0.893473           0.881179           0.889009   \n",
       "20                NaN                NaN                NaN   \n",
       "21                NaN                NaN                NaN   \n",
       "22           0.887541           0.875820           0.890402   \n",
       "23           0.893217           0.881290           0.888364   \n",
       "24                NaN                NaN                NaN   \n",
       "25                NaN                NaN                NaN   \n",
       "26           0.887991           0.876833           0.890735   \n",
       "27           0.893240           0.881528           0.888193   \n",
       "28                NaN                NaN                NaN   \n",
       "29                NaN                NaN                NaN   \n",
       "30           0.887981           0.876398           0.890943   \n",
       "31           0.893382           0.880989           0.889213   \n",
       "\n",
       "    split3_test_score  split4_test_score  std_fit_time  std_score_time  \\\n",
       "0                 NaN                NaN      0.179106        0.000000   \n",
       "1                 NaN                NaN      0.068343        0.000000   \n",
       "2            0.906372           0.882219      0.028023        0.006484   \n",
       "3            0.906630           0.889626      0.042294        0.007149   \n",
       "4                 NaN                NaN      0.170325        0.000000   \n",
       "5                 NaN                NaN      0.134897        0.000000   \n",
       "6            0.906559           0.882456      0.115637        0.006426   \n",
       "7            0.905079           0.889425      0.055376        0.008769   \n",
       "8                 NaN                NaN      0.165929        0.000000   \n",
       "9                 NaN                NaN      0.035041        0.000000   \n",
       "10           0.905843           0.882776      0.051910        0.011756   \n",
       "11           0.905991           0.889419      0.042032        0.004801   \n",
       "12                NaN                NaN      0.100971        0.000000   \n",
       "13                NaN                NaN      0.068712        0.000000   \n",
       "14           0.906187           0.882637      0.083508        0.008024   \n",
       "15           0.906268           0.889535      0.074145        0.008369   \n",
       "16                NaN                NaN      0.105221        0.000000   \n",
       "17                NaN                NaN      0.051659        0.000000   \n",
       "18           0.905703           0.882176      0.127460        0.010735   \n",
       "19           0.906119           0.889626      0.031622        0.018015   \n",
       "20                NaN                NaN      0.080890        0.000000   \n",
       "21                NaN                NaN      0.095411        0.000000   \n",
       "22           0.905956           0.881619      0.031008        0.009995   \n",
       "23           0.906204           0.889809      0.045812        0.006387   \n",
       "24                NaN                NaN      0.084555        0.000000   \n",
       "25                NaN                NaN      0.055552        0.000000   \n",
       "26           0.906306           0.882126      0.068856        0.007284   \n",
       "27           0.905491           0.889501      0.066800        0.005178   \n",
       "28                NaN                NaN      0.078606        0.000000   \n",
       "29                NaN                NaN      0.041243        0.000000   \n",
       "30           0.906101           0.882423      0.171670        0.013261   \n",
       "31           0.905230           0.889533      0.246210        0.004502   \n",
       "\n",
       "    std_test_score  \n",
       "0              NaN  \n",
       "1              NaN  \n",
       "2         0.010207  \n",
       "3         0.008372  \n",
       "4              NaN  \n",
       "5              NaN  \n",
       "6         0.010289  \n",
       "7         0.007967  \n",
       "8              NaN  \n",
       "9              NaN  \n",
       "10        0.009854  \n",
       "11        0.008175  \n",
       "12             NaN  \n",
       "13             NaN  \n",
       "14        0.010175  \n",
       "15        0.008396  \n",
       "16             NaN  \n",
       "17             NaN  \n",
       "18        0.009753  \n",
       "19        0.008162  \n",
       "20             NaN  \n",
       "21             NaN  \n",
       "22        0.010170  \n",
       "23        0.008194  \n",
       "24             NaN  \n",
       "25             NaN  \n",
       "26        0.009986  \n",
       "27        0.007913  \n",
       "28             NaN  \n",
       "29             NaN  \n",
       "30        0.009990  \n",
       "31        0.007893  "
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# cross validation results\n",
    "pd.DataFrame(model.cv_results_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best AUC:  0.8920455423137623\n",
      "Best hyperparameters:  {'pca__n_components': 80, 'logistic__penalty': 'l2', 'logistic__C': 0.1}\n"
     ]
    }
   ],
   "source": [
    "# print best hyperparameters\n",
    "print(\"Best AUC: \", model.best_score_)\n",
    "print(\"Best hyperparameters: \", model.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[5874 1017]\n",
      " [ 111  499]]\n",
      "Sensitivity: \t0.82\n",
      "Specificity: \t0.85\n",
      "AUC:    \t 0.9\n"
     ]
    }
   ],
   "source": [
    "# predict churn on test data\n",
    "y_pred = model.predict(X_test)\n",
    "\n",
    "# create onfusion matrix\n",
    "cm = confusion_matrix(y_test, y_pred)\n",
    "print(cm)\n",
    "\n",
    "# check sensitivity and specificity\n",
    "sensitivity, specificity, _ = sensitivity_specificity_support(y_test, y_pred, average='binary')\n",
    "print(\"Sensitivity: \\t\", round(sensitivity, 2), \"\\n\", \"Specificity: \\t\", round(specificity, 2), sep='')\n",
    "\n",
    "# check area under curve\n",
    "y_pred_prob = model.predict_proba(X_test)[:, 1]\n",
    "print(\"AUC:    \\t\", round(roc_auc_score(y_test, y_pred_prob),2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# random forest - the class weight is used to handle class imbalance - it adjusts the cost function\n",
    "forest = RandomForestClassifier(class_weight={0:0.1, 1: 0.9}, n_jobs = -1)\n",
    "\n",
    "# hyperparameter space\n",
    "params = {\"criterion\": ['gini', 'entropy'], \"max_features\": ['auto', 0.4]}\n",
    "\n",
    "# create 5 folds\n",
    "folds = StratifiedKFold(n_splits = 5, shuffle = True, random_state = 4)\n",
    "\n",
    "# create gridsearch object\n",
    "model = GridSearchCV(estimator=forest, cv=folds, param_grid=params, scoring='roc_auc', n_jobs=-1, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 4 candidates, totalling 20 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  20 out of  20 | elapsed:  3.2min finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=StratifiedKFold(n_splits=5, random_state=4, shuffle=True),\n",
       "             error_score=nan,\n",
       "             estimator=RandomForestClassifier(bootstrap=True, ccp_alpha=0.0,\n",
       "                                              class_weight={0: 0.1, 1: 0.9},\n",
       "                                              criterion='gini', max_depth=None,\n",
       "                                              max_features='auto',\n",
       "                                              max_leaf_nodes=None,\n",
       "                                              max_samples=None,\n",
       "                                              min_impurity_decrease=0.0,\n",
       "                                              min_impurity_split=None,\n",
       "                                              min_samples_leaf=1,\n",
       "                                              min_samples_split=2,\n",
       "                                              min_weight_fraction_leaf=0.0,\n",
       "                                              n_estimators=100, n_jobs=-1,\n",
       "                                              oob_score=False,\n",
       "                                              random_state=None, verbose=0,\n",
       "                                              warm_start=False),\n",
       "             iid='deprecated', n_jobs=-1,\n",
       "             param_grid={'criterion': ['gini', 'entropy'],\n",
       "                         'max_features': ['auto', 0.4]},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring='roc_auc', verbose=1)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit model\n",
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best AUC:  0.9270687272272824\n",
      "Best hyperparameters:  {'criterion': 'entropy', 'max_features': 0.4}\n"
     ]
    }
   ],
   "source": [
    "# print best hyperparameters\n",
    "print(\"Best AUC: \", model.best_score_)\n",
    "print(\"Best hyperparameters: \", model.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[6773  118]\n",
      " [ 305  305]]\n",
      "Sensitivity: \t0.5\n",
      "Specificity: \t0.98\n",
      "AUC:    \t 0.93\n"
     ]
    }
   ],
   "source": [
    "# predict churn on test data\n",
    "y_pred = model.predict(X_test)\n",
    "\n",
    "# create onfusion matrix\n",
    "cm = confusion_matrix(y_test, y_pred)\n",
    "print(cm)\n",
    "\n",
    "# check sensitivity and specificity\n",
    "sensitivity, specificity, _ = sensitivity_specificity_support(y_test, y_pred, average='binary')\n",
    "print(\"Sensitivity: \\t\", round(sensitivity, 2), \"\\n\", \"Specificity: \\t\", round(specificity, 2), sep='')\n",
    "\n",
    "# check area under curve\n",
    "y_pred_prob = model.predict_proba(X_test)[:, 1]\n",
    "print(\"AUC:    \\t\", round(roc_auc_score(y_test, y_pred_prob),2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Poor sensitivity. The best model is PCA along with Logistic regression."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ii) Choosing best features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13\n"
     ]
    }
   ],
   "source": [
    "# run a random forest model on train data\n",
    "max_features = int(round(np.sqrt(X_train.shape[1])))    # number of variables to consider to split each node\n",
    "print(max_features)\n",
    "\n",
    "rf_model = RandomForestClassifier(n_estimators=100, max_features=max_features, class_weight={0:0.1, 1: 0.9}, oob_score=True, random_state=4, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:   12.7s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, ccp_alpha=0.0,\n",
       "                       class_weight={0: 0.1, 1: 0.9}, criterion='gini',\n",
       "                       max_depth=None, max_features=13, max_leaf_nodes=None,\n",
       "                       max_samples=None, min_impurity_decrease=0.0,\n",
       "                       min_impurity_split=None, min_samples_leaf=1,\n",
       "                       min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "                       n_estimators=100, n_jobs=None, oob_score=True,\n",
       "                       random_state=4, verbose=1, warm_start=False)"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit model\n",
    "rf_model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9421333333333334"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# OOB score\n",
    "rf_model.oob_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "172"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(X_train.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[6781  110]\n",
      " [ 329  281]]\n",
      "Sensitivity: \t0.46\n",
      "Specificity: \t0.98\n",
      "ROC:    \t 0.93\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 100 out of 100 | elapsed:    0.1s finished\n"
     ]
    }
   ],
   "source": [
    "# predict churn on test data\n",
    "y_pred = rf_model.predict(X_test)\n",
    "\n",
    "# create onfusion matrix\n",
    "cm = confusion_matrix(y_test, y_pred)\n",
    "print(cm)\n",
    "\n",
    "# check sensitivity and specificity\n",
    "sensitivity, specificity, _ = sensitivity_specificity_support(y_test, y_pred, average='binary')\n",
    "print(\"Sensitivity: \\t\", round(sensitivity, 2), \"\\n\", \"Specificity: \\t\", round(specificity, 2), sep='')\n",
    "\n",
    "# check area under curve\n",
    "y_pred_prob = rf_model.predict_proba(X_test)[:, 1]\n",
    "print(\"ROC:    \\t\", round(roc_auc_score(y_test, y_pred_prob),2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sum of importance= 100.00000000000001\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>variables</th>\n",
       "      <th>importance_percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>total_ic_mou_8</td>\n",
       "      <td>6.823264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>total_og_mou_8</td>\n",
       "      <td>4.690836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>last_day_rch_amt_8</td>\n",
       "      <td>3.456561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>total_rech_amt_diff</td>\n",
       "      <td>3.195499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>arpu_8</td>\n",
       "      <td>2.789690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>max_rech_amt_8</td>\n",
       "      <td>2.746171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>total_rech_amt_8</td>\n",
       "      <td>2.665804</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>roam_og_mou_8</td>\n",
       "      <td>2.659917</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>arpu_diff</td>\n",
       "      <td>2.608655</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>loc_ic_mou_8</td>\n",
       "      <td>2.319111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>loc_ic_t2m_mou_8</td>\n",
       "      <td>2.246671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>loc_ic_t2t_mou_8</td>\n",
       "      <td>2.207646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>roam_ic_mou_8</td>\n",
       "      <td>2.047148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>roam_og_mou_diff</td>\n",
       "      <td>1.692025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>total_ic_mou_diff</td>\n",
       "      <td>1.607144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>total_rech_num_diff</td>\n",
       "      <td>1.451340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>av_rech_amt_data_8</td>\n",
       "      <td>1.418172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>total_og_mou_diff</td>\n",
       "      <td>1.394561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>std_og_mou_diff</td>\n",
       "      <td>1.290633</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>loc_og_t2m_mou_8</td>\n",
       "      <td>1.190306</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>loc_ic_mou_diff</td>\n",
       "      <td>1.054936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>loc_og_mou_8</td>\n",
       "      <td>1.053013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>roam_ic_mou_diff</td>\n",
       "      <td>0.925682</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>offnet_mou_diff</td>\n",
       "      <td>0.900222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>fb_user_8</td>\n",
       "      <td>0.869495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>vol_2g_mb_8</td>\n",
       "      <td>0.852017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>offnet_mou_8</td>\n",
       "      <td>0.801113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>night_pck_user_8</td>\n",
       "      <td>0.783681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>max_rech_amt_diff</td>\n",
       "      <td>0.781095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>std_og_t2m_mou_8</td>\n",
       "      <td>0.723541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>std_ic_mou_8</td>\n",
       "      <td>0.715820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>max_rech_data_8</td>\n",
       "      <td>0.691347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>aon</td>\n",
       "      <td>0.653083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>loc_og_t2t_mou_8</td>\n",
       "      <td>0.632173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>loc_og_mou_diff</td>\n",
       "      <td>0.622913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>total_ic_mou_7</td>\n",
       "      <td>0.596221</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>total_rech_data_8</td>\n",
       "      <td>0.594567</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>loc_ic_mou_7</td>\n",
       "      <td>0.587313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>total_rech_num_8</td>\n",
       "      <td>0.560454</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>onnet_mou_diff</td>\n",
       "      <td>0.560398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>std_ic_mou_diff</td>\n",
       "      <td>0.558221</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>arpu_7</td>\n",
       "      <td>0.538724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>loc_ic_t2m_mou_7</td>\n",
       "      <td>0.537881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>onnet_mou_8</td>\n",
       "      <td>0.533674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>offnet_mou_7</td>\n",
       "      <td>0.528897</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>total_ic_mou_6</td>\n",
       "      <td>0.501062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>total_og_mou_7</td>\n",
       "      <td>0.500877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>loc_ic_t2f_mou_8</td>\n",
       "      <td>0.499444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>std_og_mou_8</td>\n",
       "      <td>0.498551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>arpu_6</td>\n",
       "      <td>0.497737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>loc_ic_mou_6</td>\n",
       "      <td>0.496871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>std_ic_t2m_mou_8</td>\n",
       "      <td>0.490351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>std_ic_t2t_mou_8</td>\n",
       "      <td>0.487406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>loc_og_mou_6</td>\n",
       "      <td>0.487014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>loc_ic_t2t_mou_7</td>\n",
       "      <td>0.480710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>total_rech_amt_7</td>\n",
       "      <td>0.464849</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>loc_og_mou_7</td>\n",
       "      <td>0.462730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>offnet_mou_6</td>\n",
       "      <td>0.456438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>loc_og_t2m_mou_6</td>\n",
       "      <td>0.452379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>max_rech_data_diff</td>\n",
       "      <td>0.450462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>spl_og_mou_diff</td>\n",
       "      <td>0.443710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>loc_ic_t2m_mou_6</td>\n",
       "      <td>0.439872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>loc_ic_t2t_mou_6</td>\n",
       "      <td>0.431444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>loc_og_t2m_mou_7</td>\n",
       "      <td>0.429476</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>vol_2g_mb_diff</td>\n",
       "      <td>0.423897</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>total_og_mou_6</td>\n",
       "      <td>0.419840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>total_rech_amt_6</td>\n",
       "      <td>0.417555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>av_rech_amt_data_diff</td>\n",
       "      <td>0.413075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>loc_og_t2t_mou_7</td>\n",
       "      <td>0.408731</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>vol_3g_mb_8</td>\n",
       "      <td>0.403165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>onnet_mou_6</td>\n",
       "      <td>0.390598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>onnet_mou_7</td>\n",
       "      <td>0.383758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>total_rech_num_7</td>\n",
       "      <td>0.382995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>aug_vbc_3g</td>\n",
       "      <td>0.381273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>loc_ic_t2f_mou_7</td>\n",
       "      <td>0.381029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>std_og_mou_6</td>\n",
       "      <td>0.375943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>isd_ic_mou_diff</td>\n",
       "      <td>0.372141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>std_ic_mou_7</td>\n",
       "      <td>0.370992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>loc_og_t2t_mou_6</td>\n",
       "      <td>0.370724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>total_rech_num_6</td>\n",
       "      <td>0.364863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>std_og_mou_7</td>\n",
       "      <td>0.360893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>std_og_t2t_mou_7</td>\n",
       "      <td>0.355739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>std_og_t2m_mou_6</td>\n",
       "      <td>0.352247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>std_ic_mou_6</td>\n",
       "      <td>0.348802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>max_rech_amt_6</td>\n",
       "      <td>0.346837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>spl_ic_mou_diff</td>\n",
       "      <td>0.345981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>roam_og_mou_7</td>\n",
       "      <td>0.344065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>std_ic_t2m_mou_7</td>\n",
       "      <td>0.333720</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>std_og_t2m_mou_7</td>\n",
       "      <td>0.324795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>std_ic_t2m_mou_6</td>\n",
       "      <td>0.318996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>max_rech_amt_7</td>\n",
       "      <td>0.311066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>spl_og_mou_6</td>\n",
       "      <td>0.310650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>last_day_rch_amt_7</td>\n",
       "      <td>0.309911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>spl_og_mou_7</td>\n",
       "      <td>0.306326</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>std_og_t2t_mou_6</td>\n",
       "      <td>0.304446</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>std_og_t2t_mou_8</td>\n",
       "      <td>0.303592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>vol_3g_mb_diff</td>\n",
       "      <td>0.303344</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>roam_ic_mou_7</td>\n",
       "      <td>0.302379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>total_rech_data_diff</td>\n",
       "      <td>0.300898</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>loc_ic_t2f_mou_6</td>\n",
       "      <td>0.290628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>loc_og_t2f_mou_6</td>\n",
       "      <td>0.283906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>last_day_rch_amt_6</td>\n",
       "      <td>0.276142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>std_ic_t2t_mou_7</td>\n",
       "      <td>0.272297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>isd_ic_mou_8</td>\n",
       "      <td>0.267061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>av_rech_amt_data_7</td>\n",
       "      <td>0.263161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>loc_og_t2f_mou_8</td>\n",
       "      <td>0.253972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>vol_2g_mb_6</td>\n",
       "      <td>0.253849</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>std_ic_t2t_mou_6</td>\n",
       "      <td>0.250764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>loc_og_t2f_mou_7</td>\n",
       "      <td>0.248779</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>spl_og_mou_8</td>\n",
       "      <td>0.244960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>roam_og_mou_6</td>\n",
       "      <td>0.242051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>og_others_6</td>\n",
       "      <td>0.241792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112</th>\n",
       "      <td>std_ic_t2f_mou_8</td>\n",
       "      <td>0.238772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113</th>\n",
       "      <td>vol_2g_mb_7</td>\n",
       "      <td>0.235147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>vol_3g_mb_7</td>\n",
       "      <td>0.220906</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>isd_og_mou_diff</td>\n",
       "      <td>0.218017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>ic_others_8</td>\n",
       "      <td>0.217148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>roam_ic_mou_6</td>\n",
       "      <td>0.212074</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>jul_vbc_3g</td>\n",
       "      <td>0.206196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>spl_ic_mou_6</td>\n",
       "      <td>0.200643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>loc_og_t2c_mou_8</td>\n",
       "      <td>0.198151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>av_rech_amt_data_6</td>\n",
       "      <td>0.196566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>ic_others_6</td>\n",
       "      <td>0.195205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123</th>\n",
       "      <td>loc_og_t2c_mou_7</td>\n",
       "      <td>0.193448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>loc_og_t2c_mou_6</td>\n",
       "      <td>0.187509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>ic_others_7</td>\n",
       "      <td>0.186782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>jun_vbc_3g</td>\n",
       "      <td>0.185713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>max_rech_data_7</td>\n",
       "      <td>0.182765</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>isd_ic_mou_7</td>\n",
       "      <td>0.181296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>spl_ic_mou_8</td>\n",
       "      <td>0.167775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>isd_ic_mou_6</td>\n",
       "      <td>0.167557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>max_rech_data_6</td>\n",
       "      <td>0.165927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>vol_3g_mb_6</td>\n",
       "      <td>0.163036</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>sep_vbc_3g</td>\n",
       "      <td>0.158233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>std_ic_t2f_mou_7</td>\n",
       "      <td>0.146391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>total_rech_data_7</td>\n",
       "      <td>0.136823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>std_og_t2f_mou_7</td>\n",
       "      <td>0.133265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137</th>\n",
       "      <td>std_ic_t2f_mou_6</td>\n",
       "      <td>0.127915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>isd_og_mou_6</td>\n",
       "      <td>0.126838</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>isd_og_mou_8</td>\n",
       "      <td>0.121602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>std_ic_t2o_mou_8</td>\n",
       "      <td>0.118197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>total_rech_data_6</td>\n",
       "      <td>0.117570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>sachet_2g_7</td>\n",
       "      <td>0.114033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>sachet_2g_8</td>\n",
       "      <td>0.110460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>std_og_t2f_mou_6</td>\n",
       "      <td>0.109702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>std_og_t2f_mou_8</td>\n",
       "      <td>0.103984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>spl_ic_mou_7</td>\n",
       "      <td>0.102997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>isd_og_mou_7</td>\n",
       "      <td>0.094254</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>148</th>\n",
       "      <td>sachet_2g_6</td>\n",
       "      <td>0.087543</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>og_others_8</td>\n",
       "      <td>0.086379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>sachet_3g_7</td>\n",
       "      <td>0.080858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>std_og_t2c_mou_8</td>\n",
       "      <td>0.076194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>sachet_3g_6</td>\n",
       "      <td>0.069551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>monthly_3g_7</td>\n",
       "      <td>0.058653</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>std_og_t2c_mou_6</td>\n",
       "      <td>0.058549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>night_pck_user_7</td>\n",
       "      <td>0.058299</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>156</th>\n",
       "      <td>fb_user_7</td>\n",
       "      <td>0.052356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>157</th>\n",
       "      <td>monthly_2g_7</td>\n",
       "      <td>0.049973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>og_others_7</td>\n",
       "      <td>0.045963</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>sachet_3g_8</td>\n",
       "      <td>0.043989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>160</th>\n",
       "      <td>monthly_2g_6</td>\n",
       "      <td>0.041783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>161</th>\n",
       "      <td>fb_user_6</td>\n",
       "      <td>0.040493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>162</th>\n",
       "      <td>std_ic_t2o_mou_6</td>\n",
       "      <td>0.039386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>163</th>\n",
       "      <td>monthly_3g_6</td>\n",
       "      <td>0.038646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>164</th>\n",
       "      <td>monthly_3g_8</td>\n",
       "      <td>0.035726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>night_pck_user_6</td>\n",
       "      <td>0.032310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>166</th>\n",
       "      <td>monthly_2g_8</td>\n",
       "      <td>0.030092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>167</th>\n",
       "      <td>std_og_t2c_mou_7</td>\n",
       "      <td>0.023321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>std_ic_t2o_mou_7</td>\n",
       "      <td>0.014360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>std_og_t2o_mou</td>\n",
       "      <td>0.012859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>loc_ic_t2o_mou</td>\n",
       "      <td>0.009765</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>171</th>\n",
       "      <td>loc_og_t2o_mou</td>\n",
       "      <td>0.006404</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 variables  importance_percentage\n",
       "0           total_ic_mou_8               6.823264\n",
       "1           total_og_mou_8               4.690836\n",
       "2       last_day_rch_amt_8               3.456561\n",
       "3      total_rech_amt_diff               3.195499\n",
       "4                   arpu_8               2.789690\n",
       "5           max_rech_amt_8               2.746171\n",
       "6         total_rech_amt_8               2.665804\n",
       "7            roam_og_mou_8               2.659917\n",
       "8                arpu_diff               2.608655\n",
       "9             loc_ic_mou_8               2.319111\n",
       "10        loc_ic_t2m_mou_8               2.246671\n",
       "11        loc_ic_t2t_mou_8               2.207646\n",
       "12           roam_ic_mou_8               2.047148\n",
       "13        roam_og_mou_diff               1.692025\n",
       "14       total_ic_mou_diff               1.607144\n",
       "15     total_rech_num_diff               1.451340\n",
       "16      av_rech_amt_data_8               1.418172\n",
       "17       total_og_mou_diff               1.394561\n",
       "18         std_og_mou_diff               1.290633\n",
       "19        loc_og_t2m_mou_8               1.190306\n",
       "20         loc_ic_mou_diff               1.054936\n",
       "21            loc_og_mou_8               1.053013\n",
       "22        roam_ic_mou_diff               0.925682\n",
       "23         offnet_mou_diff               0.900222\n",
       "24               fb_user_8               0.869495\n",
       "25             vol_2g_mb_8               0.852017\n",
       "26            offnet_mou_8               0.801113\n",
       "27        night_pck_user_8               0.783681\n",
       "28       max_rech_amt_diff               0.781095\n",
       "29        std_og_t2m_mou_8               0.723541\n",
       "30            std_ic_mou_8               0.715820\n",
       "31         max_rech_data_8               0.691347\n",
       "32                     aon               0.653083\n",
       "33        loc_og_t2t_mou_8               0.632173\n",
       "34         loc_og_mou_diff               0.622913\n",
       "35          total_ic_mou_7               0.596221\n",
       "36       total_rech_data_8               0.594567\n",
       "37            loc_ic_mou_7               0.587313\n",
       "38        total_rech_num_8               0.560454\n",
       "39          onnet_mou_diff               0.560398\n",
       "40         std_ic_mou_diff               0.558221\n",
       "41                  arpu_7               0.538724\n",
       "42        loc_ic_t2m_mou_7               0.537881\n",
       "43             onnet_mou_8               0.533674\n",
       "44            offnet_mou_7               0.528897\n",
       "45          total_ic_mou_6               0.501062\n",
       "46          total_og_mou_7               0.500877\n",
       "47        loc_ic_t2f_mou_8               0.499444\n",
       "48            std_og_mou_8               0.498551\n",
       "49                  arpu_6               0.497737\n",
       "50            loc_ic_mou_6               0.496871\n",
       "51        std_ic_t2m_mou_8               0.490351\n",
       "52        std_ic_t2t_mou_8               0.487406\n",
       "53            loc_og_mou_6               0.487014\n",
       "54        loc_ic_t2t_mou_7               0.480710\n",
       "55        total_rech_amt_7               0.464849\n",
       "56            loc_og_mou_7               0.462730\n",
       "57            offnet_mou_6               0.456438\n",
       "58        loc_og_t2m_mou_6               0.452379\n",
       "59      max_rech_data_diff               0.450462\n",
       "60         spl_og_mou_diff               0.443710\n",
       "61        loc_ic_t2m_mou_6               0.439872\n",
       "62        loc_ic_t2t_mou_6               0.431444\n",
       "63        loc_og_t2m_mou_7               0.429476\n",
       "64          vol_2g_mb_diff               0.423897\n",
       "65          total_og_mou_6               0.419840\n",
       "66        total_rech_amt_6               0.417555\n",
       "67   av_rech_amt_data_diff               0.413075\n",
       "68        loc_og_t2t_mou_7               0.408731\n",
       "69             vol_3g_mb_8               0.403165\n",
       "70             onnet_mou_6               0.390598\n",
       "71             onnet_mou_7               0.383758\n",
       "72        total_rech_num_7               0.382995\n",
       "73              aug_vbc_3g               0.381273\n",
       "74        loc_ic_t2f_mou_7               0.381029\n",
       "75            std_og_mou_6               0.375943\n",
       "76         isd_ic_mou_diff               0.372141\n",
       "77            std_ic_mou_7               0.370992\n",
       "78        loc_og_t2t_mou_6               0.370724\n",
       "79        total_rech_num_6               0.364863\n",
       "80            std_og_mou_7               0.360893\n",
       "81        std_og_t2t_mou_7               0.355739\n",
       "82        std_og_t2m_mou_6               0.352247\n",
       "83            std_ic_mou_6               0.348802\n",
       "84          max_rech_amt_6               0.346837\n",
       "85         spl_ic_mou_diff               0.345981\n",
       "86           roam_og_mou_7               0.344065\n",
       "87        std_ic_t2m_mou_7               0.333720\n",
       "88        std_og_t2m_mou_7               0.324795\n",
       "89        std_ic_t2m_mou_6               0.318996\n",
       "90          max_rech_amt_7               0.311066\n",
       "91            spl_og_mou_6               0.310650\n",
       "92      last_day_rch_amt_7               0.309911\n",
       "93            spl_og_mou_7               0.306326\n",
       "94        std_og_t2t_mou_6               0.304446\n",
       "95        std_og_t2t_mou_8               0.303592\n",
       "96          vol_3g_mb_diff               0.303344\n",
       "97           roam_ic_mou_7               0.302379\n",
       "98    total_rech_data_diff               0.300898\n",
       "99        loc_ic_t2f_mou_6               0.290628\n",
       "100       loc_og_t2f_mou_6               0.283906\n",
       "101     last_day_rch_amt_6               0.276142\n",
       "102       std_ic_t2t_mou_7               0.272297\n",
       "103           isd_ic_mou_8               0.267061\n",
       "104     av_rech_amt_data_7               0.263161\n",
       "105       loc_og_t2f_mou_8               0.253972\n",
       "106            vol_2g_mb_6               0.253849\n",
       "107       std_ic_t2t_mou_6               0.250764\n",
       "108       loc_og_t2f_mou_7               0.248779\n",
       "109           spl_og_mou_8               0.244960\n",
       "110          roam_og_mou_6               0.242051\n",
       "111            og_others_6               0.241792\n",
       "112       std_ic_t2f_mou_8               0.238772\n",
       "113            vol_2g_mb_7               0.235147\n",
       "114            vol_3g_mb_7               0.220906\n",
       "115        isd_og_mou_diff               0.218017\n",
       "116            ic_others_8               0.217148\n",
       "117          roam_ic_mou_6               0.212074\n",
       "118             jul_vbc_3g               0.206196\n",
       "119           spl_ic_mou_6               0.200643\n",
       "120       loc_og_t2c_mou_8               0.198151\n",
       "121     av_rech_amt_data_6               0.196566\n",
       "122            ic_others_6               0.195205\n",
       "123       loc_og_t2c_mou_7               0.193448\n",
       "124       loc_og_t2c_mou_6               0.187509\n",
       "125            ic_others_7               0.186782\n",
       "126             jun_vbc_3g               0.185713\n",
       "127        max_rech_data_7               0.182765\n",
       "128           isd_ic_mou_7               0.181296\n",
       "129           spl_ic_mou_8               0.167775\n",
       "130           isd_ic_mou_6               0.167557\n",
       "131        max_rech_data_6               0.165927\n",
       "132            vol_3g_mb_6               0.163036\n",
       "133             sep_vbc_3g               0.158233\n",
       "134       std_ic_t2f_mou_7               0.146391\n",
       "135      total_rech_data_7               0.136823\n",
       "136       std_og_t2f_mou_7               0.133265\n",
       "137       std_ic_t2f_mou_6               0.127915\n",
       "138           isd_og_mou_6               0.126838\n",
       "139           isd_og_mou_8               0.121602\n",
       "140       std_ic_t2o_mou_8               0.118197\n",
       "141      total_rech_data_6               0.117570\n",
       "142            sachet_2g_7               0.114033\n",
       "143            sachet_2g_8               0.110460\n",
       "144       std_og_t2f_mou_6               0.109702\n",
       "145       std_og_t2f_mou_8               0.103984\n",
       "146           spl_ic_mou_7               0.102997\n",
       "147           isd_og_mou_7               0.094254\n",
       "148            sachet_2g_6               0.087543\n",
       "149            og_others_8               0.086379\n",
       "150            sachet_3g_7               0.080858\n",
       "151       std_og_t2c_mou_8               0.076194\n",
       "152            sachet_3g_6               0.069551\n",
       "153           monthly_3g_7               0.058653\n",
       "154       std_og_t2c_mou_6               0.058549\n",
       "155       night_pck_user_7               0.058299\n",
       "156              fb_user_7               0.052356\n",
       "157           monthly_2g_7               0.049973\n",
       "158            og_others_7               0.045963\n",
       "159            sachet_3g_8               0.043989\n",
       "160           monthly_2g_6               0.041783\n",
       "161              fb_user_6               0.040493\n",
       "162       std_ic_t2o_mou_6               0.039386\n",
       "163           monthly_3g_6               0.038646\n",
       "164           monthly_3g_8               0.035726\n",
       "165       night_pck_user_6               0.032310\n",
       "166           monthly_2g_8               0.030092\n",
       "167       std_og_t2c_mou_7               0.023321\n",
       "168       std_ic_t2o_mou_7               0.014360\n",
       "169         std_og_t2o_mou               0.012859\n",
       "170         loc_ic_t2o_mou               0.009765\n",
       "171         loc_og_t2o_mou               0.006404"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# predictors\n",
    "features = churn_filtered.drop('churn', axis=1).columns\n",
    "\n",
    "# feature_importance\n",
    "importance = rf_model.feature_importances_\n",
    "\n",
    "# create dataframe\n",
    "feature_importance = pd.DataFrame({'variables': features, 'importance_percentage': importance*100})\n",
    "feature_importance = feature_importance[['variables', 'importance_percentage']]\n",
    "\n",
    "# sort features\n",
    "feature_importance = feature_importance.sort_values('importance_percentage', ascending=False).reset_index(drop=True)\n",
    "print(\"Sum of importance=\", feature_importance.importance_percentage.sum())\n",
    "feature_importance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Extracting top 30 features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# extract top 'n' features\n",
    "top_n = 30\n",
    "top_features = feature_importance.variables[0:top_n]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f50acdf1518>"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot feature correlation\n",
    "import seaborn as sns\n",
    "plt.rcParams[\"figure.figsize\"] =(10,10)\n",
    "mycmap = sns.diverging_palette(199, 359, s=99, center=\"light\", as_cmap=True)\n",
    "sns.heatmap(data=X_train[top_features].corr(), center=0.0, cmap=mycmap)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "top_features = ['total_ic_mou_8', 'total_rech_amt_diff', 'total_og_mou_8', 'arpu_8', 'roam_ic_mou_8', 'roam_og_mou_8', \n",
    "                'std_ic_mou_8', 'av_rech_amt_data_8', 'std_og_mou_8']\n",
    "X_train = X_train[top_features]\n",
    "X_test = X_test[top_features]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# logistic regression\n",
    "steps = [('scaler', StandardScaler()), \n",
    "         (\"logistic\", LogisticRegression(class_weight={0:0.1, 1:0.9}))\n",
    "        ]\n",
    "\n",
    "# compile pipeline\n",
    "logistic = Pipeline(steps)\n",
    "\n",
    "# hyperparameter space\n",
    "params = {'logistic__C': [0.1, 0.5, 1, 2, 3, 4, 5, 10], 'logistic__penalty': ['l1', 'l2']}\n",
    "\n",
    "# create 5 folds\n",
    "folds = StratifiedKFold(n_splits = 5, shuffle = True, random_state = 4)\n",
    "\n",
    "# create gridsearch object\n",
    "model = GridSearchCV(estimator=logistic, cv=folds, param_grid=params, scoring='roc_auc', n_jobs=-1, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 16 candidates, totalling 80 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done  80 out of  80 | elapsed:    1.0s finished\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=StratifiedKFold(n_splits=5, random_state=4, shuffle=True),\n",
       "             error_score=nan,\n",
       "             estimator=Pipeline(memory=None,\n",
       "                                steps=[('scaler',\n",
       "                                        StandardScaler(copy=True,\n",
       "                                                       with_mean=True,\n",
       "                                                       with_std=True)),\n",
       "                                       ('logistic',\n",
       "                                        LogisticRegression(C=1.0,\n",
       "                                                           class_weight={0: 0.1,\n",
       "                                                                         1: 0.9},\n",
       "                                                           dual=False,\n",
       "                                                           fit_intercept=True,\n",
       "                                                           intercept_scaling=1,\n",
       "                                                           l1_ratio=None,\n",
       "                                                           max_iter=100,\n",
       "                                                           multi_class='auto',\n",
       "                                                           n_jobs=None,\n",
       "                                                           penalty='l2',\n",
       "                                                           random_state=None,\n",
       "                                                           solver='lbfgs',\n",
       "                                                           tol=0.0001,\n",
       "                                                           verbose=0,\n",
       "                                                           warm_start=False))],\n",
       "                                verbose=False),\n",
       "             iid='deprecated', n_jobs=-1,\n",
       "             param_grid={'logistic__C': [0.1, 0.5, 1, 2, 3, 4, 5, 10],\n",
       "                         'logistic__penalty': ['l1', 'l2']},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring='roc_auc', verbose=1)"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit model\n",
    "model.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best AUC:  0.8736904286992818\n",
      "Best hyperparameters:  {'logistic__penalty': 'l2', 'logistic__C': 10}\n"
     ]
    }
   ],
   "source": [
    "# print best hyperparameters\n",
    "print(\"Best AUC: \", model.best_score_)\n",
    "print(\"Best hyperparameters: \", model.best_params_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save the ML Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['models/rf_model.pkl']"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Save the model\n",
    "from sklearn.externals import joblib \n",
    "\n",
    "# Save the model as a pickle in a file \n",
    "joblib.dump(model, 'models/rf_model.pkl') "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load the ML Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save the model\n",
    "from sklearn.externals import joblib\n",
    "\n",
    "# Load the model from the file \n",
    "rf_pickle_model = joblib.load('models/rf_model.pkl') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use the loaded model to make predictions\n",
    "response_rf_model = rf_pickle_model.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate the Loaded Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[5760 1131]\n",
      " [ 122  488]]\n",
      "Sensitivity: \t0.8\n",
      "Specificity: \t0.84\n",
      "ROC:    \t 0.88\n"
     ]
    }
   ],
   "source": [
    "# predict churn on test data\n",
    "y_pred_load = rf_pickle_model.predict(X_test)\n",
    "\n",
    "# create onfusion matrix\n",
    "cm_load = confusion_matrix(y_test, y_pred_load)\n",
    "print(cm_load)\n",
    "\n",
    "# check sensitivity and specificity\n",
    "sensitivity_load, specificity_load, _ = sensitivity_specificity_support(y_test, y_pred, average='binary')\n",
    "print(\"Sensitivity: \\t\", round(sensitivity_load, 2), \"\\n\", \"Specificity: \\t\", round(specificity_load, 2), sep='')\n",
    "\n",
    "# check area under curve\n",
    "y_pred_prob_load = rf_pickle_model.predict_proba(X_test)[:, 1]\n",
    "print(\"ROC:    \\t\", round(roc_auc_score(y_test, y_pred_prob_load),2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate the trained model object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[5760 1131]\n",
      " [ 122  488]]\n",
      "Sensitivity: \t0.8\n",
      "Specificity: \t0.84\n",
      "ROC:    \t 0.88\n"
     ]
    }
   ],
   "source": [
    "# predict churn on test data\n",
    "y_pred = model.predict(X_test)\n",
    "\n",
    "# create onfusion matrix\n",
    "cm = confusion_matrix(y_test, y_pred)\n",
    "print(cm)\n",
    "\n",
    "# check sensitivity and specificity\n",
    "sensitivity, specificity, _ = sensitivity_specificity_support(y_test, y_pred, average='binary')\n",
    "print(\"Sensitivity: \\t\", round(sensitivity, 2), \"\\n\", \"Specificity: \\t\", round(specificity, 2), sep='')\n",
    "\n",
    "# check area under curve\n",
    "y_pred_prob = model.predict_proba(X_test)[:, 1]\n",
    "print(\"ROC:    \\t\", round(roc_auc_score(y_test, y_pred_prob),2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Extract the intercept and the coefficients from the logistic model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "logistic_model = model.best_estimator_.named_steps['logistic']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# intercept\n",
    "intercept_df = pd.DataFrame(logistic_model.intercept_.reshape((1,1)), columns = ['intercept'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# coefficients\n",
    "coefficients = logistic_model.coef_.reshape((9, 1)).tolist()\n",
    "coefficients = [val for sublist in coefficients for val in sublist]\n",
    "coefficients = [round(coefficient, 3) for coefficient in coefficients]\n",
    "\n",
    "logistic_features = list(X_train.columns)\n",
    "coefficients_df = pd.DataFrame(logistic_model.coef_, columns=logistic_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>intercept</th>\n",
       "      <th>total_ic_mou_8</th>\n",
       "      <th>total_rech_amt_diff</th>\n",
       "      <th>total_og_mou_8</th>\n",
       "      <th>arpu_8</th>\n",
       "      <th>roam_ic_mou_8</th>\n",
       "      <th>roam_og_mou_8</th>\n",
       "      <th>std_ic_mou_8</th>\n",
       "      <th>av_rech_amt_data_8</th>\n",
       "      <th>std_og_mou_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-1.46829</td>\n",
       "      <td>-1.227629</td>\n",
       "      <td>-0.68268</td>\n",
       "      <td>-1.067155</td>\n",
       "      <td>0.169269</td>\n",
       "      <td>0.042355</td>\n",
       "      <td>0.194784</td>\n",
       "      <td>0.075685</td>\n",
       "      <td>-0.794502</td>\n",
       "      <td>0.56011</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   intercept  total_ic_mou_8  total_rech_amt_diff  total_og_mou_8    arpu_8  \\\n",
       "0   -1.46829       -1.227629             -0.68268       -1.067155  0.169269   \n",
       "\n",
       "   roam_ic_mou_8  roam_og_mou_8  std_ic_mou_8  av_rech_amt_data_8  \\\n",
       "0       0.042355       0.194784      0.075685           -0.794502   \n",
       "\n",
       "   std_og_mou_8  \n",
       "0       0.56011  "
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# concatenate dataframes\n",
    "coefficients = pd.concat([intercept_df, coefficients_df], axis=1)\n",
    "coefficients"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Business Insights\n",
    "\n",
    "* Telecom company needs to pay attention to the roaming rates. They need to provide good offers to the customers who are using services from a roaming zone.\n",
    "* The company needs to focus on the STD and ISD rates. Perhaps, the rates are too high. Provide them with some kind of STD and ISD packages.\n",
    "* To look into both of the issues stated above, it is desired that the telecom company collects customer query and complaint data and work on their services according to the needs of customers. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "py3",
   "language": "python",
   "name": "py3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}