06.Bonus.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# UNDERSTANDING CLASSIFICATION WITH TELCO CHURN PREDICTION DATASET\n",
    "\n",
    "**PROBLEM STATEMENT**\n",
    "\n",
    "*Predict behavior to retain customers. You can analyze all relevant customer data and develop focused customer retention programs.*\n",
    "\n",
    "*Each row represents a customer, each column contains customer’s attributes described on the column Metadata.*\n",
    "\n",
    "*The data set includes information about:*\n",
    "\n",
    "* *Customers who left within the last month – the column is called Churn*\n",
    "* *Services that each customer has signed up for – phone, multiple lines, internet, online security, online backup, device protection, tech support, and streaming TV and movies*\n",
    "* *Customer account information – how long they’ve been a customer, contract, payment method, paperless billing, monthly charges, and total charges*\n",
    "* *Demographic info about customers – gender, age range, and if they have partners and dependents*\n",
    "\n",
    "**Notebook Source** - https://github.com/jsaikmr/Telecom-Churn-Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from collections import Counter\n",
    "from imblearn.over_sampling import SMOTE, ADASYN\n",
    "\n",
    "from sklearn.model_selection import train_test_split, cross_val_score, StratifiedKFold, RandomizedSearchCV, GridSearchCV\n",
    "\n",
    "import statsmodels.api as sm\n",
    "from sklearn.linear_model import LogisticRegression, RidgeClassifier, SGDClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.svm import SVC, LinearSVC\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier, GradientBoostingClassifier, AdaBoostClassifier\n",
    "from xgboost import XGBClassifier\n",
    "\n",
    "from sklearn.metrics import confusion_matrix, classification_report, accuracy_score, precision_score, recall_score, roc_auc_score, roc_curve\n",
    "\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "%matplotlib inline\n",
    "RANDOM_STATE = 42"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mobile_number</th>\n",
       "      <td>7000842753</td>\n",
       "      <td>7001865778</td>\n",
       "      <td>7001625959</td>\n",
       "      <td>7001204172</td>\n",
       "      <td>7000142493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>circle_id</th>\n",
       "      <td>109</td>\n",
       "      <td>109</td>\n",
       "      <td>109</td>\n",
       "      <td>109</td>\n",
       "      <td>109</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loc_og_t2o_mou</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std_og_t2o_mou</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loc_ic_t2o_mou</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_date_of_month_6</th>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>6/30/2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_date_of_month_7</th>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_date_of_month_8</th>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "      <td>8/31/2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>last_date_of_month_9</th>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "      <td>9/30/2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arpu_6</th>\n",
       "      <td>197.385</td>\n",
       "      <td>34.047</td>\n",
       "      <td>167.69</td>\n",
       "      <td>221.338</td>\n",
       "      <td>261.636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arpu_7</th>\n",
       "      <td>214.816</td>\n",
       "      <td>355.074</td>\n",
       "      <td>189.058</td>\n",
       "      <td>251.102</td>\n",
       "      <td>309.876</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arpu_8</th>\n",
       "      <td>213.803</td>\n",
       "      <td>268.321</td>\n",
       "      <td>210.226</td>\n",
       "      <td>508.054</td>\n",
       "      <td>238.174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arpu_9</th>\n",
       "      <td>21.1</td>\n",
       "      <td>86.285</td>\n",
       "      <td>290.714</td>\n",
       "      <td>389.5</td>\n",
       "      <td>163.426</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onnet_mou_6</th>\n",
       "      <td>NaN</td>\n",
       "      <td>24.11</td>\n",
       "      <td>11.54</td>\n",
       "      <td>99.91</td>\n",
       "      <td>50.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onnet_mou_7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>78.68</td>\n",
       "      <td>55.24</td>\n",
       "      <td>54.39</td>\n",
       "      <td>149.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onnet_mou_8</th>\n",
       "      <td>0</td>\n",
       "      <td>7.68</td>\n",
       "      <td>37.26</td>\n",
       "      <td>310.98</td>\n",
       "      <td>83.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onnet_mou_9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>18.34</td>\n",
       "      <td>74.81</td>\n",
       "      <td>241.71</td>\n",
       "      <td>58.78</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offnet_mou_6</th>\n",
       "      <td>NaN</td>\n",
       "      <td>15.74</td>\n",
       "      <td>143.33</td>\n",
       "      <td>123.31</td>\n",
       "      <td>76.96</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offnet_mou_7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>99.84</td>\n",
       "      <td>220.59</td>\n",
       "      <td>109.01</td>\n",
       "      <td>91.88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offnet_mou_8</th>\n",
       "      <td>0</td>\n",
       "      <td>304.76</td>\n",
       "      <td>208.36</td>\n",
       "      <td>71.68</td>\n",
       "      <td>124.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offnet_mou_9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>53.76</td>\n",
       "      <td>118.91</td>\n",
       "      <td>113.54</td>\n",
       "      <td>45.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roam_ic_mou_6</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roam_ic_mou_7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>54.86</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roam_ic_mou_8</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>44.38</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roam_ic_mou_9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>38.49</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roam_og_mou_6</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roam_og_mou_7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28.09</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roam_og_mou_8</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>39.04</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roam_og_mou_9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>70.94</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>loc_og_t2t_mou_6</th>\n",
       "      <td>NaN</td>\n",
       "      <td>23.88</td>\n",
       "      <td>7.19</td>\n",
       "      <td>73.68</td>\n",
       "      <td>50.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arpu_2g_9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>night_pck_user_6</th>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>night_pck_user_7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>night_pck_user_8</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>night_pck_user_9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>monthly_2g_6</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>monthly_2g_7</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>monthly_2g_8</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>monthly_2g_9</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sachet_2g_6</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sachet_2g_7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sachet_2g_8</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sachet_2g_9</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>monthly_3g_6</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>monthly_3g_7</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>monthly_3g_8</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>monthly_3g_9</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sachet_3g_6</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sachet_3g_7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sachet_3g_8</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sachet_3g_9</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fb_user_6</th>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fb_user_7</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fb_user_8</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fb_user_9</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aon</th>\n",
       "      <td>968</td>\n",
       "      <td>1006</td>\n",
       "      <td>1103</td>\n",
       "      <td>2491</td>\n",
       "      <td>1526</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aug_vbc_3g</th>\n",
       "      <td>30.4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>jul_vbc_3g</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>jun_vbc_3g</th>\n",
       "      <td>101.2</td>\n",
       "      <td>0</td>\n",
       "      <td>4.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sep_vbc_3g</th>\n",
       "      <td>3.58</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>226 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                               0           1           2           3  \\\n",
       "mobile_number         7000842753  7001865778  7001625959  7001204172   \n",
       "circle_id                    109         109         109         109   \n",
       "loc_og_t2o_mou                 0           0           0           0   \n",
       "std_og_t2o_mou                 0           0           0           0   \n",
       "loc_ic_t2o_mou                 0           0           0           0   \n",
       "last_date_of_month_6   6/30/2014   6/30/2014   6/30/2014   6/30/2014   \n",
       "last_date_of_month_7   7/31/2014   7/31/2014   7/31/2014   7/31/2014   \n",
       "last_date_of_month_8   8/31/2014   8/31/2014   8/31/2014   8/31/2014   \n",
       "last_date_of_month_9   9/30/2014   9/30/2014   9/30/2014   9/30/2014   \n",
       "arpu_6                   197.385      34.047      167.69     221.338   \n",
       "arpu_7                   214.816     355.074     189.058     251.102   \n",
       "arpu_8                   213.803     268.321     210.226     508.054   \n",
       "arpu_9                      21.1      86.285     290.714       389.5   \n",
       "onnet_mou_6                  NaN       24.11       11.54       99.91   \n",
       "onnet_mou_7                  NaN       78.68       55.24       54.39   \n",
       "onnet_mou_8                    0        7.68       37.26      310.98   \n",
       "onnet_mou_9                  NaN       18.34       74.81      241.71   \n",
       "offnet_mou_6                 NaN       15.74      143.33      123.31   \n",
       "offnet_mou_7                 NaN       99.84      220.59      109.01   \n",
       "offnet_mou_8                   0      304.76      208.36       71.68   \n",
       "offnet_mou_9                 NaN       53.76      118.91      113.54   \n",
       "roam_ic_mou_6                NaN           0           0           0   \n",
       "roam_ic_mou_7                NaN           0           0       54.86   \n",
       "roam_ic_mou_8                  0           0           0       44.38   \n",
       "roam_ic_mou_9                NaN           0       38.49           0   \n",
       "roam_og_mou_6                NaN           0           0           0   \n",
       "roam_og_mou_7                NaN           0           0       28.09   \n",
       "roam_og_mou_8                  0           0           0       39.04   \n",
       "roam_og_mou_9                NaN           0       70.94           0   \n",
       "loc_og_t2t_mou_6             NaN       23.88        7.19       73.68   \n",
       "...                          ...         ...         ...         ...   \n",
       "arpu_2g_9                    NaN         NaN           0         NaN   \n",
       "night_pck_user_6               0         NaN         NaN         NaN   \n",
       "night_pck_user_7               0           0         NaN         NaN   \n",
       "night_pck_user_8               0           0         NaN         NaN   \n",
       "night_pck_user_9             NaN         NaN           0         NaN   \n",
       "monthly_2g_6                   0           0           0           0   \n",
       "monthly_2g_7                   0           1           0           0   \n",
       "monthly_2g_8                   0           0           0           0   \n",
       "monthly_2g_9                   0           0           0           0   \n",
       "sachet_2g_6                    0           0           0           0   \n",
       "sachet_2g_7                    0           0           0           0   \n",
       "sachet_2g_8                    0           2           0           0   \n",
       "sachet_2g_9                    0           0           1           0   \n",
       "monthly_3g_6                   1           0           0           0   \n",
       "monthly_3g_7                   1           0           0           0   \n",
       "monthly_3g_8                   1           0           0           0   \n",
       "monthly_3g_9                   0           0           0           0   \n",
       "sachet_3g_6                    0           0           0           0   \n",
       "sachet_3g_7                    0           0           0           0   \n",
       "sachet_3g_8                    0           0           0           0   \n",
       "sachet_3g_9                    0           0           0           0   \n",
       "fb_user_6                      1         NaN         NaN         NaN   \n",
       "fb_user_7                      1           1         NaN         NaN   \n",
       "fb_user_8                      1           1         NaN         NaN   \n",
       "fb_user_9                    NaN         NaN           1         NaN   \n",
       "aon                          968        1006        1103        2491   \n",
       "aug_vbc_3g                  30.4           0           0           0   \n",
       "jul_vbc_3g                     0           0           0           0   \n",
       "jun_vbc_3g                 101.2           0        4.17           0   \n",
       "sep_vbc_3g                  3.58           0           0           0   \n",
       "\n",
       "                               4  \n",
       "mobile_number         7000142493  \n",
       "circle_id                    109  \n",
       "loc_og_t2o_mou                 0  \n",
       "std_og_t2o_mou                 0  \n",
       "loc_ic_t2o_mou                 0  \n",
       "last_date_of_month_6   6/30/2014  \n",
       "last_date_of_month_7   7/31/2014  \n",
       "last_date_of_month_8   8/31/2014  \n",
       "last_date_of_month_9   9/30/2014  \n",
       "arpu_6                   261.636  \n",
       "arpu_7                   309.876  \n",
       "arpu_8                   238.174  \n",
       "arpu_9                   163.426  \n",
       "onnet_mou_6                50.31  \n",
       "onnet_mou_7               149.44  \n",
       "onnet_mou_8                83.89  \n",
       "onnet_mou_9                58.78  \n",
       "offnet_mou_6               76.96  \n",
       "offnet_mou_7               91.88  \n",
       "offnet_mou_8              124.26  \n",
       "offnet_mou_9               45.81  \n",
       "roam_ic_mou_6                  0  \n",
       "roam_ic_mou_7                  0  \n",
       "roam_ic_mou_8                  0  \n",
       "roam_ic_mou_9                  0  \n",
       "roam_og_mou_6                  0  \n",
       "roam_og_mou_7                  0  \n",
       "roam_og_mou_8                  0  \n",
       "roam_og_mou_9                  0  \n",
       "loc_og_t2t_mou_6           50.31  \n",
       "...                          ...  \n",
       "arpu_2g_9                    NaN  \n",
       "night_pck_user_6               0  \n",
       "night_pck_user_7             NaN  \n",
       "night_pck_user_8             NaN  \n",
       "night_pck_user_9             NaN  \n",
       "monthly_2g_6                   0  \n",
       "monthly_2g_7                   0  \n",
       "monthly_2g_8                   0  \n",
       "monthly_2g_9                   0  \n",
       "sachet_2g_6                    1  \n",
       "sachet_2g_7                    0  \n",
       "sachet_2g_8                    0  \n",
       "sachet_2g_9                    0  \n",
       "monthly_3g_6                   0  \n",
       "monthly_3g_7                   0  \n",
       "monthly_3g_8                   0  \n",
       "monthly_3g_9                   0  \n",
       "sachet_3g_6                    0  \n",
       "sachet_3g_7                    0  \n",
       "sachet_3g_8                    0  \n",
       "sachet_3g_9                    0  \n",
       "fb_user_6                      0  \n",
       "fb_user_7                    NaN  \n",
       "fb_user_8                    NaN  \n",
       "fb_user_9                    NaN  \n",
       "aon                         1526  \n",
       "aug_vbc_3g                     0  \n",
       "jul_vbc_3g                     0  \n",
       "jun_vbc_3g                     0  \n",
       "sep_vbc_3g                     0  \n",
       "\n",
       "[226 rows x 5 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv('telecom_churn_data.csv')\n",
    "data.head().T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(99999, 226)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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": [
    "data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_cols_split(df):\n",
    "\n",
    "    col_len = len(df.columns)\n",
    "\n",
    "    jun_cols = []\n",
    "    jul_cols = []\n",
    "    aug_cols = []\n",
    "    sep_cols = []\n",
    "    common_cols = []\n",
    "    date_cols = []\n",
    "    \n",
    "    for i in range(0, col_len):\n",
    "        if any(pd.Series(df.columns[i]).str.contains('_6|jun')):\n",
    "            jun_cols.append(df.columns[i])\n",
    "        elif any(pd.Series(df.columns[i]).str.contains('_7|jul')):\n",
    "            jul_cols.append(df.columns[i])\n",
    "        elif any(pd.Series(df.columns[i]).str.contains('_8|aug')):\n",
    "            aug_cols.append(df.columns[i])\n",
    "        elif any(pd.Series(df.columns[i]).str.contains('_9|sep')):\n",
    "            sep_cols.append(df.columns[i])\n",
    "        else:\n",
    "            common_cols.append(df.columns[i])\n",
    "        \n",
    "        if any(pd.Series(df.columns[i]).str.contains('date')):\n",
    "            date_cols.append(df.columns[i])\n",
    "            \n",
    "    return jun_cols,jul_cols,aug_cols,sep_cols,common_cols,date_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_cols_sub_split(col_list):\n",
    "    call_usage_cols = []\n",
    "    recharge_cols = []\n",
    "    ic_usage_cols = []\n",
    "    og_usage_cols = []\n",
    "\n",
    "    call_usage_search_for = ['og','ic','mou']\n",
    "\n",
    "    for i in range(0, len(col_list)):\n",
    "        if any(pd.Series(col_list[i]).str.contains('|'.join(['rech','rch']))):\n",
    "            recharge_cols.append(col_list[i])\n",
    "        elif any(pd.Series(col_list[i]).str.contains('|'.join(call_usage_search_for))):\n",
    "            call_usage_cols.append(col_list[i])\n",
    "\n",
    "        if any(pd.Series(col_list[i]).str.contains('ic')):\n",
    "            ic_usage_cols.append(col_list[i])\n",
    "        elif any(pd.Series(col_list[i]).str.contains('og')):\n",
    "            og_usage_cols.append(col_list[i])\n",
    "            \n",
    "    return call_usage_cols,recharge_cols,ic_usage_cols,og_usage_cols            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(29979, 226)\n"
     ]
    }
   ],
   "source": [
    "data['avg_rech_amt_6_7'] = ( data['total_rech_amt_6'] + data['total_rech_amt_7'] ) / 2\n",
    "\n",
    "\n",
    "data = data.loc[(data['avg_rech_amt_6_7'] > np.percentile(data['avg_rech_amt_6_7'], 70))]\n",
    "\n",
    "\n",
    "data.drop(['avg_rech_amt_6_7'], axis=1, inplace=True)\n",
    "\n",
    "print(data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "tag_churn_cols = ['total_ic_mou_9', 'total_og_mou_9', 'vol_2g_mb_9', 'vol_3g_mb_9']\n",
    "data['churn'] = np.where(data[tag_churn_cols].sum(axis=1) == 0, 1, 0 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    27390\n",
       "1     2589\n",
       "Name: churn, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['churn'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Churn Rate : 8.64%\n"
     ]
    }
   ],
   "source": [
    "print('Churn Rate : {0}%'.format(round(((sum(data['churn'])/len(data['churn']))*100),2)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "jun_cols, jul_cols, aug_cols, sep_cols, common_cols, date_cols = get_cols_split(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop(sep_cols, axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mobile_number 29979\n",
      "circle_id 1\n",
      "loc_og_t2o_mou 2\n",
      "std_og_t2o_mou 2\n",
      "loc_ic_t2o_mou 2\n",
      "last_date_of_month_6 1\n",
      "last_date_of_month_7 2\n",
      "last_date_of_month_8 2\n",
      "arpu_6 29230\n",
      "arpu_7 29228\n",
      "arpu_8 28376\n",
      "onnet_mou_6 18806\n",
      "onnet_mou_7 18934\n",
      "onnet_mou_8 17598\n",
      "offnet_mou_6 22441\n",
      "offnet_mou_7 22639\n",
      "offnet_mou_8 21500\n",
      "roam_ic_mou_6 4338\n",
      "roam_ic_mou_7 3649\n",
      "roam_ic_mou_8 3654\n",
      "roam_og_mou_6 5174\n",
      "roam_og_mou_7 4431\n",
      "roam_og_mou_8 4382\n",
      "loc_og_t2t_mou_6 11150\n",
      "loc_og_t2t_mou_7 11152\n",
      "loc_og_t2t_mou_8 10770\n",
      "loc_og_t2m_mou_6 16739\n",
      "loc_og_t2m_mou_7 16865\n",
      "loc_og_t2m_mou_8 16155\n",
      "loc_og_t2f_mou_6 3252\n",
      "loc_og_t2f_mou_7 3267\n",
      "loc_og_t2f_mou_8 3124\n",
      "loc_og_t2c_mou_6 1659\n",
      "loc_og_t2c_mou_7 1750\n",
      "loc_og_t2c_mou_8 1731\n",
      "loc_og_mou_6 19677\n",
      "loc_og_mou_7 19867\n",
      "loc_og_mou_8 18872\n",
      "std_og_t2t_mou_6 12773\n",
      "std_og_t2t_mou_7 12982\n",
      "std_og_t2t_mou_8 11781\n",
      "std_og_t2m_mou_6 14512\n",
      "std_og_t2m_mou_7 14583\n",
      "std_og_t2m_mou_8 13320\n",
      "std_og_t2f_mou_6 1774\n",
      "std_og_t2f_mou_7 1715\n",
      "std_og_t2f_mou_8 1627\n",
      "std_og_t2c_mou_6 2\n",
      "std_og_t2c_mou_7 2\n",
      "std_og_t2c_mou_8 2\n",
      "std_og_mou_6 18316\n",
      "std_og_mou_7 18434\n",
      "std_og_mou_8 16859\n",
      "isd_og_mou_6 1114\n",
      "isd_og_mou_7 1125\n",
      "isd_og_mou_8 940\n",
      "spl_og_mou_6 3054\n",
      "spl_og_mou_7 3400\n",
      "spl_og_mou_8 3239\n",
      "og_others_6 863\n",
      "og_others_7 124\n",
      "og_others_8 134\n",
      "total_og_mou_6 24587\n",
      "total_og_mou_7 24892\n",
      "total_og_mou_8 23624\n",
      "loc_ic_t2t_mou_6 9870\n",
      "loc_ic_t2t_mou_7 9959\n",
      "loc_ic_t2t_mou_8 9670\n",
      "loc_ic_t2m_mou_6 16011\n",
      "loc_ic_t2m_mou_7 16063\n",
      "loc_ic_t2m_mou_8 15589\n",
      "loc_ic_t2f_mou_6 4817\n",
      "loc_ic_t2f_mou_7 4897\n",
      "loc_ic_t2f_mou_8 4705\n",
      "loc_ic_mou_6 19123\n",
      "loc_ic_mou_7 19021\n",
      "loc_ic_mou_8 18564\n",
      "std_ic_t2t_mou_6 4607\n",
      "std_ic_t2t_mou_7 4707\n",
      "std_ic_t2t_mou_8 4485\n",
      "std_ic_t2m_mou_6 6679\n",
      "std_ic_t2m_mou_7 6747\n",
      "std_ic_t2m_mou_8 6416\n",
      "std_ic_t2f_mou_6 2034\n",
      "std_ic_t2f_mou_7 2076\n",
      "std_ic_t2f_mou_8 1942\n",
      "std_ic_t2o_mou_6 2\n",
      "std_ic_t2o_mou_7 2\n",
      "std_ic_t2o_mou_8 2\n",
      "std_ic_mou_6 8389\n",
      "std_ic_mou_7 8542\n",
      "std_ic_mou_8 8031\n",
      "total_ic_mou_6 20592\n",
      "total_ic_mou_7 20696\n",
      "total_ic_mou_8 20083\n",
      "spl_ic_mou_6 79\n",
      "spl_ic_mou_7 94\n",
      "spl_ic_mou_8 86\n",
      "isd_ic_mou_6 3427\n",
      "isd_ic_mou_7 3636\n",
      "isd_ic_mou_8 3490\n",
      "ic_others_6 1228\n",
      "ic_others_7 1372\n",
      "ic_others_8 1259\n",
      "total_rech_num_6 102\n",
      "total_rech_num_7 101\n",
      "total_rech_num_8 96\n",
      "total_rech_amt_6 2241\n",
      "total_rech_amt_7 2265\n",
      "total_rech_amt_8 2299\n",
      "max_rech_amt_6 170\n",
      "max_rech_amt_7 151\n",
      "max_rech_amt_8 182\n",
      "date_of_last_rech_6 31\n",
      "date_of_last_rech_7 32\n",
      "date_of_last_rech_8 32\n",
      "last_day_rch_amt_6 158\n",
      "last_day_rch_amt_7 149\n",
      "last_day_rch_amt_8 179\n",
      "date_of_last_rech_data_6 31\n",
      "date_of_last_rech_data_7 32\n",
      "date_of_last_rech_data_8 32\n",
      "total_rech_data_6 37\n",
      "total_rech_data_7 41\n",
      "total_rech_data_8 46\n",
      "max_rech_data_6 44\n",
      "max_rech_data_7 47\n",
      "max_rech_data_8 48\n",
      "count_rech_2g_6 31\n",
      "count_rech_2g_7 36\n",
      "count_rech_2g_8 34\n",
      "count_rech_3g_6 26\n",
      "count_rech_3g_7 29\n",
      "count_rech_3g_8 30\n",
      "av_rech_amt_data_6 795\n",
      "av_rech_amt_data_7 864\n",
      "av_rech_amt_data_8 838\n",
      "vol_2g_mb_6 7803\n",
      "vol_2g_mb_7 7808\n",
      "vol_2g_mb_8 7304\n",
      "vol_3g_mb_6 7039\n",
      "vol_3g_mb_7 7435\n",
      "vol_3g_mb_8 7143\n",
      "arpu_3g_6 4872\n",
      "arpu_3g_7 4859\n",
      "arpu_3g_8 4767\n",
      "arpu_2g_6 4500\n",
      "arpu_2g_7 4217\n",
      "arpu_2g_8 3851\n",
      "night_pck_user_6 3\n",
      "night_pck_user_7 3\n",
      "night_pck_user_8 3\n",
      "monthly_2g_6 5\n",
      "monthly_2g_7 6\n",
      "monthly_2g_8 6\n",
      "sachet_2g_6 30\n",
      "sachet_2g_7 34\n",
      "sachet_2g_8 34\n",
      "monthly_3g_6 12\n",
      "monthly_3g_7 15\n",
      "monthly_3g_8 12\n",
      "sachet_3g_6 25\n",
      "sachet_3g_7 27\n",
      "sachet_3g_8 29\n",
      "fb_user_6 3\n",
      "fb_user_7 3\n",
      "fb_user_8 3\n",
      "aon 3321\n",
      "aug_vbc_3g 7284\n",
      "jul_vbc_3g 7312\n",
      "jun_vbc_3g 6858\n",
      "churn 2\n"
     ]
    }
   ],
   "source": [
    "for col in data.columns:\n",
    "    print(col, len(data[col].unique()))"
   ]
  },
  {
   "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>mobile_number</th>\n",
       "      <th>circle_id</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>loc_og_t2o_mou</th>\n",
       "      <th>std_og_t2o_mou</th>\n",
       "      <th>loc_ic_t2o_mou</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_ic_t2o_mou_6</th>\n",
       "      <th>std_ic_t2o_mou_7</th>\n",
       "      <th>std_ic_t2o_mou_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7000701601</td>\n",
       "      <td>109</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>7001524846</td>\n",
       "      <td>109</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>7002191713</td>\n",
       "      <td>109</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>7000875565</td>\n",
       "      <td>109</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>7000187447</td>\n",
       "      <td>109</td>\n",
       "      <td>6/30/2014</td>\n",
       "      <td>7/31/2014</td>\n",
       "      <td>8/31/2014</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mobile_number  circle_id last_date_of_month_6 last_date_of_month_7  \\\n",
       "7      7000701601        109            6/30/2014            7/31/2014   \n",
       "8      7001524846        109            6/30/2014            7/31/2014   \n",
       "13     7002191713        109            6/30/2014            7/31/2014   \n",
       "16     7000875565        109            6/30/2014            7/31/2014   \n",
       "17     7000187447        109            6/30/2014            7/31/2014   \n",
       "\n",
       "   last_date_of_month_8  loc_og_t2o_mou  std_og_t2o_mou  loc_ic_t2o_mou  \\\n",
       "7             8/31/2014             0.0             0.0             0.0   \n",
       "8             8/31/2014             0.0             0.0             0.0   \n",
       "13            8/31/2014             0.0             0.0             0.0   \n",
       "16            8/31/2014             0.0             0.0             0.0   \n",
       "17            8/31/2014             0.0             0.0             0.0   \n",
       "\n",
       "    std_og_t2c_mou_6  std_og_t2c_mou_7  std_og_t2c_mou_8  std_ic_t2o_mou_6  \\\n",
       "7                0.0               0.0               0.0               0.0   \n",
       "8                0.0               0.0               0.0               0.0   \n",
       "13               0.0               0.0               0.0               0.0   \n",
       "16               0.0               0.0               0.0               0.0   \n",
       "17               0.0               0.0               0.0               0.0   \n",
       "\n",
       "    std_ic_t2o_mou_7  std_ic_t2o_mou_8  \n",
       "7                0.0               0.0  \n",
       "8                0.0               0.0  \n",
       "13               0.0               0.0  \n",
       "16               0.0               0.0  \n",
       "17               0.0               0.0  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[['mobile_number','circle_id','last_date_of_month_6','last_date_of_month_7','last_date_of_month_8',\\\n",
    "           'loc_og_t2o_mou', 'std_og_t2o_mou', 'loc_ic_t2o_mou','std_og_t2c_mou_6','std_og_t2c_mou_7','std_og_t2c_mou_8',\\\n",
    "           'std_ic_t2o_mou_6','std_ic_t2o_mou_7','std_ic_t2o_mou_8']].head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.drop(['mobile_number','circle_id','last_date_of_month_6','last_date_of_month_7','last_date_of_month_8',\\\n",
    "           'loc_og_t2o_mou', 'std_og_t2o_mou', 'loc_ic_t2o_mou','std_og_t2c_mou_6','std_og_t2c_mou_7','std_og_t2c_mou_8',\\\n",
    "           'std_ic_t2o_mou_6','std_ic_t2o_mou_7','std_ic_t2o_mou_8'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>av_rech_amt_data_6</th>\n",
       "      <th>max_rech_data_6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>NaN</td>\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",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>NaN</td>\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",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    total_rech_data_6  av_rech_amt_data_6  max_rech_data_6\n",
       "7                 NaN                 NaN              NaN\n",
       "8                 NaN                 NaN              NaN\n",
       "13                NaN                 NaN              NaN\n",
       "16                NaN                 NaN              NaN\n",
       "17                NaN                 NaN              NaN"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[['total_rech_data_6','av_rech_amt_data_6','max_rech_data_6']].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.rename(columns={'av_rech_amt_data_6':'total_rech_amt_data_6','av_rech_amt_data_7':'total_rech_amt_data_7','av_rech_amt_data_8':'total_rech_amt_data_8'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": false
   },
   "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>index</th>\n",
       "      <th>missing_cnt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>count_rech_2g_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>total_rech_data_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>total_rech_amt_data_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>arpu_3g_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>max_rech_data_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>arpu_2g_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>night_pck_user_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>count_rech_3g_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>fb_user_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>date_of_last_rech_data_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>136</th>\n",
       "      <td>night_pck_user_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>total_rech_data_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>count_rech_3g_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>date_of_last_rech_data_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112</th>\n",
       "      <td>max_rech_data_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>arpu_3g_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>fb_user_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>count_rech_2g_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>total_rech_amt_data_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>arpu_2g_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>date_of_last_rech_data_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>total_rech_data_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113</th>\n",
       "      <td>max_rech_data_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>count_rech_2g_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>152</th>\n",
       "      <td>fb_user_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>count_rech_3g_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>total_rech_amt_data_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>arpu_2g_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>arpu_3g_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137</th>\n",
       "      <td>night_pck_user_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>offnet_mou_6</td>\n",
       "      <td>316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>loc_og_t2m_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>loc_og_t2t_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>loc_og_t2f_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>loc_og_t2c_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>roam_og_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>loc_og_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>roam_ic_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>offnet_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>std_og_t2t_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>std_ic_t2m_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>std_og_t2m_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>loc_ic_t2m_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>std_ic_t2t_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>std_ic_t2f_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>loc_ic_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>std_ic_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>onnet_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>loc_ic_t2f_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>spl_ic_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>isd_ic_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>std_og_t2f_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>loc_ic_t2t_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>ic_others_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>og_others_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>spl_og_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>isd_og_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>std_og_mou_7</td>\n",
       "      <td>303</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>date_of_last_rech_7</td>\n",
       "      <td>114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>date_of_last_rech_6</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>114 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        index  missing_cnt\n",
       "114           count_rech_2g_6        18592\n",
       "108         total_rech_data_6        18592\n",
       "120     total_rech_amt_data_6        18592\n",
       "129                 arpu_3g_6        18592\n",
       "111           max_rech_data_6        18592\n",
       "132                 arpu_2g_6        18592\n",
       "135          night_pck_user_6        18592\n",
       "117           count_rech_3g_6        18592\n",
       "150                 fb_user_6        18592\n",
       "105  date_of_last_rech_data_6        18592\n",
       "136          night_pck_user_7        18327\n",
       "109         total_rech_data_7        18327\n",
       "118           count_rech_3g_7        18327\n",
       "106  date_of_last_rech_data_7        18327\n",
       "112           max_rech_data_7        18327\n",
       "130                 arpu_3g_7        18327\n",
       "151                 fb_user_7        18327\n",
       "115           count_rech_2g_7        18327\n",
       "121     total_rech_amt_data_7        18327\n",
       "133                 arpu_2g_7        18327\n",
       "107  date_of_last_rech_data_8        18238\n",
       "110         total_rech_data_8        18238\n",
       "113           max_rech_data_8        18238\n",
       "116           count_rech_2g_8        18238\n",
       "152                 fb_user_8        18238\n",
       "119           count_rech_3g_8        18238\n",
       "122     total_rech_amt_data_8        18238\n",
       "134                 arpu_2g_8        18238\n",
       "131                 arpu_3g_8        18238\n",
       "137          night_pck_user_8        18238\n",
       "..                        ...          ...\n",
       "6                offnet_mou_6          316\n",
       "19           loc_og_t2m_mou_7          303\n",
       "16           loc_og_t2t_mou_7          303\n",
       "22           loc_og_t2f_mou_7          303\n",
       "25           loc_og_t2c_mou_7          303\n",
       "13              roam_og_mou_7          303\n",
       "28               loc_og_mou_7          303\n",
       "10              roam_ic_mou_7          303\n",
       "7                offnet_mou_7          303\n",
       "31           std_og_t2t_mou_7          303\n",
       "70           std_ic_t2m_mou_7          303\n",
       "34           std_og_t2m_mou_7          303\n",
       "58           loc_ic_t2m_mou_7          303\n",
       "67           std_ic_t2t_mou_7          303\n",
       "73           std_ic_t2f_mou_7          303\n",
       "64               loc_ic_mou_7          303\n",
       "76               std_ic_mou_7          303\n",
       "4                 onnet_mou_7          303\n",
       "61           loc_ic_t2f_mou_7          303\n",
       "82               spl_ic_mou_7          303\n",
       "85               isd_ic_mou_7          303\n",
       "37           std_og_t2f_mou_7          303\n",
       "55           loc_ic_t2t_mou_7          303\n",
       "88                ic_others_7          303\n",
       "49                og_others_7          303\n",
       "46               spl_og_mou_7          303\n",
       "43               isd_og_mou_7          303\n",
       "40               std_og_mou_7          303\n",
       "100       date_of_last_rech_7          114\n",
       "99        date_of_last_rech_6           62\n",
       "\n",
       "[114 rows x 2 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = data.isnull().sum().reset_index(name='missing_cnt')\n",
    "df.loc[df['missing_cnt']>0].sort_values('missing_cnt', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "jun_cols, jul_cols, aug_cols, sep_cols, common_cols, date_cols = get_cols_split(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "jun_call_usage_cols, jun_recharge_cols, jun_ic_usage_cols, jun_og_usage_cols = get_cols_sub_split(jun_cols)\n",
    "jul_call_usage_cols, jul_recharge_cols, jul_ic_usage_cols, jul_og_usage_cols = get_cols_sub_split(jul_cols)\n",
    "aug_call_usage_cols, aug_recharge_cols, aug_ic_usage_cols, aug_og_usage_cols = get_cols_sub_split(aug_cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_6 = ['fb_user_6','night_pck_user_6']\n",
    "cols_7 = ['fb_user_7','night_pck_user_7']\n",
    "cols_8 = ['fb_user_8','night_pck_user_8']\n",
    "\n",
    "data[cols_6] = data[cols_6].fillna(2)\n",
    "data[cols_7] = data[cols_7].fillna(2)\n",
    "data[cols_8] = data[cols_8].fillna(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_6 = ['count_rech_3g_6','max_rech_data_6','total_rech_amt_data_6','arpu_3g_6','total_rech_data_6','count_rech_2g_6','arpu_2g_6']\n",
    "cols_7 = ['count_rech_3g_7','max_rech_data_7','total_rech_amt_data_7','arpu_3g_7','total_rech_data_7','count_rech_2g_7','arpu_2g_7']\n",
    "cols_8 = ['count_rech_3g_8','max_rech_data_8','total_rech_amt_data_8','arpu_3g_8','total_rech_data_8','count_rech_2g_8','arpu_2g_8']\n",
    "\n",
    "data[cols_6] = data[cols_6].fillna(0)\n",
    "data[cols_7] = data[cols_7].fillna(0)\n",
    "data[cols_8] = data[cols_8].fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "data[jun_call_usage_cols] = data[jun_call_usage_cols].fillna(0)\n",
    "data[jul_call_usage_cols] = data[jul_call_usage_cols].fillna(0)\n",
    "data[aug_call_usage_cols] = data[aug_call_usage_cols].fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>index</th>\n",
       "      <th>missing_cnt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>date_of_last_rech_data_6</td>\n",
       "      <td>18592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>date_of_last_rech_data_7</td>\n",
       "      <td>18327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>date_of_last_rech_data_8</td>\n",
       "      <td>18238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>date_of_last_rech_8</td>\n",
       "      <td>594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>date_of_last_rech_7</td>\n",
       "      <td>114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>date_of_last_rech_6</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        index  missing_cnt\n",
       "105  date_of_last_rech_data_6        18592\n",
       "106  date_of_last_rech_data_7        18327\n",
       "107  date_of_last_rech_data_8        18238\n",
       "101       date_of_last_rech_8          594\n",
       "100       date_of_last_rech_7          114\n",
       "99        date_of_last_rech_6           62"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = data.isnull().sum().reset_index(name='missing_cnt')\n",
    "df.loc[df['missing_cnt']>0].sort_values('missing_cnt', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f9870693390>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "sns.countplot(x='churn', data=data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f9858c33450>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (ax1, ax2) = plt.subplots(1,2,figsize=(20, 6))\n",
    "\n",
    "# distribution plot for aon\n",
    "sns.distplot(data['aon'], ax=ax1)\n",
    "\n",
    "# bin the aon column with yearwise segments and plot the counts for each segments\n",
    "bins = [0, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]\n",
    "labels = [0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]\n",
    "#sns.countplot(pd.cut(round(((data['aon']/30)/12),1), bins = bins, labels = labels ), ax =ax1)\n",
    "pd.crosstab(pd.cut(round(((data['aon']/30)/12),1), bins = bins, labels = labels ), data['churn']).plot(kind='bar', stacked=True, ax = ax2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Incoming Calls Usage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": false
   },
   "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>loc_ic_mou_6</th>\n",
       "      <th>std_ic_mou_6</th>\n",
       "      <th>isd_ic_mou_6</th>\n",
       "      <th>roam_ic_mou_6</th>\n",
       "      <th>spl_ic_mou_6</th>\n",
       "      <th>ic_others_6</th>\n",
       "      <th>total_ic_mou_6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.00000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>244.064853</td>\n",
       "      <td>51.149589</td>\n",
       "      <td>11.14435</td>\n",
       "      <td>17.428073</td>\n",
       "      <td>0.066736</td>\n",
       "      <td>1.189804</td>\n",
       "      <td>307.624011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>312.931377</td>\n",
       "      <td>140.564801</td>\n",
       "      <td>67.21659</td>\n",
       "      <td>79.192826</td>\n",
       "      <td>0.194321</td>\n",
       "      <td>13.994391</td>\n",
       "      <td>361.286101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</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>56.710000</td>\n",
       "      <td>2.630000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>90.035000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>151.080000</td>\n",
       "      <td>15.060000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>205.340000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>315.620000</td>\n",
       "      <td>47.560000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.060000</td>\n",
       "      <td>393.835000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>6496.110000</td>\n",
       "      <td>5459.630000</td>\n",
       "      <td>3965.69000</td>\n",
       "      <td>2613.310000</td>\n",
       "      <td>19.760000</td>\n",
       "      <td>1344.140000</td>\n",
       "      <td>6798.640000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_ic_mou_6  std_ic_mou_6  isd_ic_mou_6  roam_ic_mou_6  spl_ic_mou_6  \\\n",
       "count  29979.000000  29979.000000   29979.00000   29979.000000  29979.000000   \n",
       "mean     244.064853     51.149589      11.14435      17.428073      0.066736   \n",
       "std      312.931377    140.564801      67.21659      79.192826      0.194321   \n",
       "min        0.000000      0.000000       0.00000       0.000000      0.000000   \n",
       "25%       56.710000      2.630000       0.00000       0.000000      0.000000   \n",
       "50%      151.080000     15.060000       0.00000       0.000000      0.000000   \n",
       "75%      315.620000     47.560000       0.00000       0.000000      0.000000   \n",
       "max     6496.110000   5459.630000    3965.69000    2613.310000     19.760000   \n",
       "\n",
       "        ic_others_6  total_ic_mou_6  \n",
       "count  29979.000000    29979.000000  \n",
       "mean       1.189804      307.624011  \n",
       "std       13.994391      361.286101  \n",
       "min        0.000000        0.000000  \n",
       "25%        0.000000       90.035000  \n",
       "50%        0.000000      205.340000  \n",
       "75%        0.060000      393.835000  \n",
       "max     1344.140000     6798.640000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>loc_ic_mou_7</th>\n",
       "      <th>std_ic_mou_7</th>\n",
       "      <th>isd_ic_mou_7</th>\n",
       "      <th>roam_ic_mou_7</th>\n",
       "      <th>spl_ic_mou_7</th>\n",
       "      <th>ic_others_7</th>\n",
       "      <th>total_ic_mou_7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.00000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>247.713761</td>\n",
       "      <td>53.394721</td>\n",
       "      <td>12.34135</td>\n",
       "      <td>13.535270</td>\n",
       "      <td>0.018064</td>\n",
       "      <td>1.478162</td>\n",
       "      <td>314.954126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>315.580778</td>\n",
       "      <td>149.236851</td>\n",
       "      <td>76.96167</td>\n",
       "      <td>76.343222</td>\n",
       "      <td>0.182005</td>\n",
       "      <td>15.414578</td>\n",
       "      <td>369.768358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</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>63.550000</td>\n",
       "      <td>2.785000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>98.840000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>154.860000</td>\n",
       "      <td>16.130000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>211.240000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>316.785000</td>\n",
       "      <td>50.390000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>396.915000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>6466.740000</td>\n",
       "      <td>6745.760000</td>\n",
       "      <td>4747.91000</td>\n",
       "      <td>3813.290000</td>\n",
       "      <td>21.330000</td>\n",
       "      <td>1495.940000</td>\n",
       "      <td>7279.080000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_ic_mou_7  std_ic_mou_7  isd_ic_mou_7  roam_ic_mou_7  spl_ic_mou_7  \\\n",
       "count  29979.000000  29979.000000   29979.00000   29979.000000  29979.000000   \n",
       "mean     247.713761     53.394721      12.34135      13.535270      0.018064   \n",
       "std      315.580778    149.236851      76.96167      76.343222      0.182005   \n",
       "min        0.000000      0.000000       0.00000       0.000000      0.000000   \n",
       "25%       63.550000      2.785000       0.00000       0.000000      0.000000   \n",
       "50%      154.860000     16.130000       0.00000       0.000000      0.000000   \n",
       "75%      316.785000     50.390000       0.00000       0.000000      0.000000   \n",
       "max     6466.740000   6745.760000    4747.91000    3813.290000     21.330000   \n",
       "\n",
       "        ic_others_7  total_ic_mou_7  \n",
       "count  29979.000000    29979.000000  \n",
       "mean       1.478162      314.954126  \n",
       "std       15.414578      369.768358  \n",
       "min        0.000000        0.000000  \n",
       "25%        0.000000       98.840000  \n",
       "50%        0.000000      211.240000  \n",
       "75%        0.000000      396.915000  \n",
       "max     1495.940000     7279.080000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>loc_ic_mou_8</th>\n",
       "      <th>std_ic_mou_8</th>\n",
       "      <th>isd_ic_mou_8</th>\n",
       "      <th>roam_ic_mou_8</th>\n",
       "      <th>spl_ic_mou_8</th>\n",
       "      <th>ic_others_8</th>\n",
       "      <th>total_ic_mou_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>234.331024</td>\n",
       "      <td>48.180997</td>\n",
       "      <td>11.688108</td>\n",
       "      <td>13.263074</td>\n",
       "      <td>0.027609</td>\n",
       "      <td>1.237346</td>\n",
       "      <td>295.472794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>307.121461</td>\n",
       "      <td>141.009723</td>\n",
       "      <td>74.924529</td>\n",
       "      <td>74.588105</td>\n",
       "      <td>0.116422</td>\n",
       "      <td>12.894187</td>\n",
       "      <td>360.411762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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>49.990000</td>\n",
       "      <td>1.430000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>78.930000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>142.880000</td>\n",
       "      <td>12.560000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>193.530000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>302.110000</td>\n",
       "      <td>43.435000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.060000</td>\n",
       "      <td>380.450000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5748.810000</td>\n",
       "      <td>5957.140000</td>\n",
       "      <td>4100.380000</td>\n",
       "      <td>4169.810000</td>\n",
       "      <td>6.230000</td>\n",
       "      <td>1209.860000</td>\n",
       "      <td>5990.710000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_ic_mou_8  std_ic_mou_8  isd_ic_mou_8  roam_ic_mou_8  spl_ic_mou_8  \\\n",
       "count  29979.000000  29979.000000  29979.000000   29979.000000  29979.000000   \n",
       "mean     234.331024     48.180997     11.688108      13.263074      0.027609   \n",
       "std      307.121461    141.009723     74.924529      74.588105      0.116422   \n",
       "min        0.000000      0.000000      0.000000       0.000000      0.000000   \n",
       "25%       49.990000      1.430000      0.000000       0.000000      0.000000   \n",
       "50%      142.880000     12.560000      0.000000       0.000000      0.000000   \n",
       "75%      302.110000     43.435000      0.000000       0.000000      0.000000   \n",
       "max     5748.810000   5957.140000   4100.380000    4169.810000      6.230000   \n",
       "\n",
       "        ic_others_8  total_ic_mou_8  \n",
       "count  29979.000000    29979.000000  \n",
       "mean       1.237346      295.472794  \n",
       "std       12.894187      360.411762  \n",
       "min        0.000000        0.000000  \n",
       "25%        0.000000       78.930000  \n",
       "50%        0.000000      193.530000  \n",
       "75%        0.060000      380.450000  \n",
       "max     1209.860000     5990.710000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols = [['loc_ic_mou_6','std_ic_mou_6','isd_ic_mou_6','roam_ic_mou_6','spl_ic_mou_6','ic_others_6','total_ic_mou_6'],\n",
    "        ['loc_ic_mou_7','std_ic_mou_7','isd_ic_mou_7','roam_ic_mou_7','spl_ic_mou_7','ic_others_7','total_ic_mou_7'],\n",
    "        ['loc_ic_mou_8','std_ic_mou_8','isd_ic_mou_8','roam_ic_mou_8','spl_ic_mou_8','ic_others_8','total_ic_mou_8']]\n",
    "\n",
    "# column description stats\n",
    "for i in range(0,3):\n",
    "    display(data[cols[i]].describe())\n",
    "\n",
    "# plot for the incoming calls usage\n",
    "plt.figure(figsize=(18, 5))\n",
    "for i in range(0,3):\n",
    "    plt.subplot(1,3,i+1)\n",
    "    X = pd.concat([data[cols[i]], data['churn']], axis=1)\n",
    "    X = pd.melt(X,id_vars=\"churn\",var_name=\"features\",value_name='value')\n",
    "    sns.boxplot(x=\"features\", y=\"value\", hue=\"churn\", data=X)\n",
    "    plt.xticks(rotation=90)    \n",
    "    plt.suptitle('Incoming Calls Usage')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Outgoing Calls Usage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": false
   },
   "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>loc_og_mou_6</th>\n",
       "      <th>std_og_mou_6</th>\n",
       "      <th>isd_og_mou_6</th>\n",
       "      <th>roam_og_mou_6</th>\n",
       "      <th>spl_og_mou_6</th>\n",
       "      <th>og_others_6</th>\n",
       "      <th>total_og_mou_6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>283.058317</td>\n",
       "      <td>395.120001</td>\n",
       "      <td>2.266842</td>\n",
       "      <td>29.348362</td>\n",
       "      <td>5.920726</td>\n",
       "      <td>0.693200</td>\n",
       "      <td>687.068925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>380.140057</td>\n",
       "      <td>606.718739</td>\n",
       "      <td>45.942529</td>\n",
       "      <td>118.629898</td>\n",
       "      <td>18.630406</td>\n",
       "      <td>2.282437</td>\n",
       "      <td>660.557039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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>51.010000</td>\n",
       "      <td>5.960000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>266.480000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>166.330000</td>\n",
       "      <td>126.360000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.210000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>510.760000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>374.600000</td>\n",
       "      <td>573.375000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.170000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>899.960000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>10643.380000</td>\n",
       "      <td>8432.990000</td>\n",
       "      <td>5900.660000</td>\n",
       "      <td>3775.110000</td>\n",
       "      <td>1023.210000</td>\n",
       "      <td>100.610000</td>\n",
       "      <td>10674.030000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_og_mou_6  std_og_mou_6  isd_og_mou_6  roam_og_mou_6  spl_og_mou_6  \\\n",
       "count  29979.000000  29979.000000  29979.000000   29979.000000  29979.000000   \n",
       "mean     283.058317    395.120001      2.266842      29.348362      5.920726   \n",
       "std      380.140057    606.718739     45.942529     118.629898     18.630406   \n",
       "min        0.000000      0.000000      0.000000       0.000000      0.000000   \n",
       "25%       51.010000      5.960000      0.000000       0.000000      0.000000   \n",
       "50%      166.330000    126.360000      0.000000       0.000000      0.210000   \n",
       "75%      374.600000    573.375000      0.000000       0.000000      5.170000   \n",
       "max    10643.380000   8432.990000   5900.660000    3775.110000   1023.210000   \n",
       "\n",
       "        og_others_6  total_og_mou_6  \n",
       "count  29979.000000    29979.000000  \n",
       "mean       0.693200      687.068925  \n",
       "std        2.282437      660.557039  \n",
       "min        0.000000        0.000000  \n",
       "25%        0.000000      266.480000  \n",
       "50%        0.000000      510.760000  \n",
       "75%        0.000000      899.960000  \n",
       "max      100.610000    10674.030000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>loc_og_mou_7</th>\n",
       "      <th>std_og_mou_7</th>\n",
       "      <th>isd_og_mou_7</th>\n",
       "      <th>roam_og_mou_7</th>\n",
       "      <th>spl_og_mou_7</th>\n",
       "      <th>og_others_7</th>\n",
       "      <th>total_og_mou_7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>284.220111</td>\n",
       "      <td>415.617820</td>\n",
       "      <td>2.207826</td>\n",
       "      <td>22.056425</td>\n",
       "      <td>7.429822</td>\n",
       "      <td>0.047651</td>\n",
       "      <td>709.531995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>375.993720</td>\n",
       "      <td>637.673317</td>\n",
       "      <td>45.642568</td>\n",
       "      <td>97.974803</td>\n",
       "      <td>23.076593</td>\n",
       "      <td>2.743248</td>\n",
       "      <td>685.283264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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>56.710000</td>\n",
       "      <td>5.570000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>275.275000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>170.460000</td>\n",
       "      <td>132.240000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.780000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>526.130000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>376.035000</td>\n",
       "      <td>615.440000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.130000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>931.305000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7674.780000</td>\n",
       "      <td>10936.730000</td>\n",
       "      <td>5490.280000</td>\n",
       "      <td>2812.040000</td>\n",
       "      <td>1265.790000</td>\n",
       "      <td>370.130000</td>\n",
       "      <td>11365.310000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_og_mou_7  std_og_mou_7  isd_og_mou_7  roam_og_mou_7  spl_og_mou_7  \\\n",
       "count  29979.000000  29979.000000  29979.000000   29979.000000  29979.000000   \n",
       "mean     284.220111    415.617820      2.207826      22.056425      7.429822   \n",
       "std      375.993720    637.673317     45.642568      97.974803     23.076593   \n",
       "min        0.000000      0.000000      0.000000       0.000000      0.000000   \n",
       "25%       56.710000      5.570000      0.000000       0.000000      0.000000   \n",
       "50%      170.460000    132.240000      0.000000       0.000000      0.780000   \n",
       "75%      376.035000    615.440000      0.000000       0.000000      7.130000   \n",
       "max     7674.780000  10936.730000   5490.280000    2812.040000   1265.790000   \n",
       "\n",
       "        og_others_7  total_og_mou_7  \n",
       "count  29979.000000    29979.000000  \n",
       "mean       0.047651      709.531995  \n",
       "std        2.743248      685.283264  \n",
       "min        0.000000        0.000000  \n",
       "25%        0.000000      275.275000  \n",
       "50%        0.000000      526.130000  \n",
       "75%        0.000000      931.305000  \n",
       "max      370.130000    11365.310000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>loc_og_mou_8</th>\n",
       "      <th>std_og_mou_8</th>\n",
       "      <th>isd_og_mou_8</th>\n",
       "      <th>roam_og_mou_8</th>\n",
       "      <th>spl_og_mou_8</th>\n",
       "      <th>og_others_8</th>\n",
       "      <th>total_og_mou_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.00000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>261.305826</td>\n",
       "      <td>353.81534</td>\n",
       "      <td>2.029675</td>\n",
       "      <td>21.483542</td>\n",
       "      <td>6.889372</td>\n",
       "      <td>0.059194</td>\n",
       "      <td>624.107691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>366.663849</td>\n",
       "      <td>616.45287</td>\n",
       "      <td>44.817776</td>\n",
       "      <td>106.294911</td>\n",
       "      <td>22.903352</td>\n",
       "      <td>3.322091</td>\n",
       "      <td>686.211375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>38.280000</td>\n",
       "      <td>1.78000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>188.965000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>148.280000</td>\n",
       "      <td>73.16000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>435.810000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>348.340000</td>\n",
       "      <td>481.41000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6.380000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>833.435000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>11039.910000</td>\n",
       "      <td>13980.06000</td>\n",
       "      <td>5681.540000</td>\n",
       "      <td>5337.040000</td>\n",
       "      <td>1390.880000</td>\n",
       "      <td>394.930000</td>\n",
       "      <td>14043.060000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_og_mou_8  std_og_mou_8  isd_og_mou_8  roam_og_mou_8  spl_og_mou_8  \\\n",
       "count  29979.000000   29979.00000  29979.000000   29979.000000  29979.000000   \n",
       "mean     261.305826     353.81534      2.029675      21.483542      6.889372   \n",
       "std      366.663849     616.45287     44.817776     106.294911     22.903352   \n",
       "min        0.000000       0.00000      0.000000       0.000000      0.000000   \n",
       "25%       38.280000       1.78000      0.000000       0.000000      0.000000   \n",
       "50%      148.280000      73.16000      0.000000       0.000000      0.500000   \n",
       "75%      348.340000     481.41000      0.000000       0.000000      6.380000   \n",
       "max    11039.910000   13980.06000   5681.540000    5337.040000   1390.880000   \n",
       "\n",
       "        og_others_8  total_og_mou_8  \n",
       "count  29979.000000    29979.000000  \n",
       "mean       0.059194      624.107691  \n",
       "std        3.322091      686.211375  \n",
       "min        0.000000        0.000000  \n",
       "25%        0.000000      188.965000  \n",
       "50%        0.000000      435.810000  \n",
       "75%        0.000000      833.435000  \n",
       "max      394.930000    14043.060000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols = [['loc_og_mou_6','std_og_mou_6','isd_og_mou_6','roam_og_mou_6','spl_og_mou_6','og_others_6','total_og_mou_6'],\n",
    "        ['loc_og_mou_7','std_og_mou_7','isd_og_mou_7','roam_og_mou_7','spl_og_mou_7','og_others_7','total_og_mou_7'],\n",
    "        ['loc_og_mou_8','std_og_mou_8','isd_og_mou_8','roam_og_mou_8','spl_og_mou_8','og_others_8','total_og_mou_8']]\n",
    "\n",
    "# column description stats\n",
    "for i in range(0,3):\n",
    "    display(data[cols[i]].describe())\n",
    "\n",
    "# plot for the outgoing calls usage\n",
    "plt.figure(figsize=(18, 5))\n",
    "#ic call usage\n",
    "for i in range(0,3):\n",
    "    plt.subplot(1,3,i+1)\n",
    "    X = pd.concat([data[cols[i]], data['churn']], axis=1)\n",
    "    X = pd.melt(X,id_vars=\"churn\",var_name=\"features\",value_name='value')\n",
    "    sns.boxplot(x=\"features\", y=\"value\", hue=\"churn\", data=X)\n",
    "    plt.xticks(rotation=90)    \n",
    "    plt.suptitle('Outgoing Calls Usage')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Operatorwise Calls Usage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "scrolled": false
   },
   "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>onnet_mou_6</th>\n",
       "      <th>offnet_mou_6</th>\n",
       "      <th>loc_ic_t2t_mou_6</th>\n",
       "      <th>loc_ic_t2m_mou_6</th>\n",
       "      <th>loc_ic_t2f_mou_6</th>\n",
       "      <th>std_ic_t2t_mou_6</th>\n",
       "      <th>std_ic_t2m_mou_6</th>\n",
       "      <th>std_ic_t2f_mou_6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.00000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>296.23634</td>\n",
       "      <td>418.127743</td>\n",
       "      <td>68.771712</td>\n",
       "      <td>159.678675</td>\n",
       "      <td>15.604617</td>\n",
       "      <td>16.238311</td>\n",
       "      <td>32.030383</td>\n",
       "      <td>2.876920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>460.94123</td>\n",
       "      <td>470.773248</td>\n",
       "      <td>158.703140</td>\n",
       "      <td>222.093299</td>\n",
       "      <td>45.847694</td>\n",
       "      <td>78.899714</td>\n",
       "      <td>101.126970</td>\n",
       "      <td>19.938857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.00000</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>41.17000</td>\n",
       "      <td>137.340000</td>\n",
       "      <td>8.290000</td>\n",
       "      <td>33.460000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.455000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>125.89000</td>\n",
       "      <td>282.260000</td>\n",
       "      <td>29.140000</td>\n",
       "      <td>93.980000</td>\n",
       "      <td>1.960000</td>\n",
       "      <td>1.060000</td>\n",
       "      <td>7.090000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>353.50000</td>\n",
       "      <td>523.625000</td>\n",
       "      <td>73.670000</td>\n",
       "      <td>202.920000</td>\n",
       "      <td>12.440000</td>\n",
       "      <td>10.285000</td>\n",
       "      <td>27.560000</td>\n",
       "      <td>0.180000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7376.71000</td>\n",
       "      <td>8362.360000</td>\n",
       "      <td>6351.440000</td>\n",
       "      <td>4693.860000</td>\n",
       "      <td>1678.410000</td>\n",
       "      <td>5459.560000</td>\n",
       "      <td>4630.230000</td>\n",
       "      <td>1351.110000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       onnet_mou_6  offnet_mou_6  loc_ic_t2t_mou_6  loc_ic_t2m_mou_6  \\\n",
       "count  29979.00000  29979.000000      29979.000000      29979.000000   \n",
       "mean     296.23634    418.127743         68.771712        159.678675   \n",
       "std      460.94123    470.773248        158.703140        222.093299   \n",
       "min        0.00000      0.000000          0.000000          0.000000   \n",
       "25%       41.17000    137.340000          8.290000         33.460000   \n",
       "50%      125.89000    282.260000         29.140000         93.980000   \n",
       "75%      353.50000    523.625000         73.670000        202.920000   \n",
       "max     7376.71000   8362.360000       6351.440000       4693.860000   \n",
       "\n",
       "       loc_ic_t2f_mou_6  std_ic_t2t_mou_6  std_ic_t2m_mou_6  std_ic_t2f_mou_6  \n",
       "count      29979.000000      29979.000000      29979.000000      29979.000000  \n",
       "mean          15.604617         16.238311         32.030383          2.876920  \n",
       "std           45.847694         78.899714        101.126970         19.938857  \n",
       "min            0.000000          0.000000          0.000000          0.000000  \n",
       "25%            0.000000          0.000000          0.455000          0.000000  \n",
       "50%            1.960000          1.060000          7.090000          0.000000  \n",
       "75%           12.440000         10.285000         27.560000          0.180000  \n",
       "max         1678.410000       5459.560000       4630.230000       1351.110000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>loc_og_t2t_mou_6</th>\n",
       "      <th>loc_og_t2m_mou_6</th>\n",
       "      <th>loc_og_t2f_mou_6</th>\n",
       "      <th>loc_og_t2c_mou_6</th>\n",
       "      <th>std_og_t2t_mou_6</th>\n",
       "      <th>std_og_t2m_mou_6</th>\n",
       "      <th>std_og_t2f_mou_6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>94.733562</td>\n",
       "      <td>181.336751</td>\n",
       "      <td>6.979198</td>\n",
       "      <td>1.567723</td>\n",
       "      <td>189.892036</td>\n",
       "      <td>203.211564</td>\n",
       "      <td>2.012068</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>236.960935</td>\n",
       "      <td>250.231261</td>\n",
       "      <td>22.671604</td>\n",
       "      <td>6.892348</td>\n",
       "      <td>409.871710</td>\n",
       "      <td>413.660478</td>\n",
       "      <td>12.463715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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>8.330000</td>\n",
       "      <td>30.290000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.600000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>32.610000</td>\n",
       "      <td>101.280000</td>\n",
       "      <td>0.330000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>12.880000</td>\n",
       "      <td>37.760000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>91.550000</td>\n",
       "      <td>240.230000</td>\n",
       "      <td>5.090000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>178.355000</td>\n",
       "      <td>211.210000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>6431.330000</td>\n",
       "      <td>4729.740000</td>\n",
       "      <td>1466.030000</td>\n",
       "      <td>342.860000</td>\n",
       "      <td>7366.580000</td>\n",
       "      <td>8314.760000</td>\n",
       "      <td>628.560000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_og_t2t_mou_6  loc_og_t2m_mou_6  loc_og_t2f_mou_6  loc_og_t2c_mou_6  \\\n",
       "count      29979.000000      29979.000000      29979.000000      29979.000000   \n",
       "mean          94.733562        181.336751          6.979198          1.567723   \n",
       "std          236.960935        250.231261         22.671604          6.892348   \n",
       "min            0.000000          0.000000          0.000000          0.000000   \n",
       "25%            8.330000         30.290000          0.000000          0.000000   \n",
       "50%           32.610000        101.280000          0.330000          0.000000   \n",
       "75%           91.550000        240.230000          5.090000          0.000000   \n",
       "max         6431.330000       4729.740000       1466.030000        342.860000   \n",
       "\n",
       "       std_og_t2t_mou_6  std_og_t2m_mou_6  std_og_t2f_mou_6  \n",
       "count      29979.000000      29979.000000      29979.000000  \n",
       "mean         189.892036        203.211564          2.012068  \n",
       "std          409.871710        413.660478         12.463715  \n",
       "min            0.000000          0.000000          0.000000  \n",
       "25%            0.000000          1.600000          0.000000  \n",
       "50%           12.880000         37.760000          0.000000  \n",
       "75%          178.355000        211.210000          0.000000  \n",
       "max         7366.580000       8314.760000        628.560000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>onnet_mou_7</th>\n",
       "      <th>offnet_mou_7</th>\n",
       "      <th>loc_ic_t2t_mou_7</th>\n",
       "      <th>loc_ic_t2m_mou_7</th>\n",
       "      <th>loc_ic_t2f_mou_7</th>\n",
       "      <th>std_ic_t2t_mou_7</th>\n",
       "      <th>std_ic_t2m_mou_7</th>\n",
       "      <th>std_ic_t2f_mou_7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>304.591922</td>\n",
       "      <td>424.103446</td>\n",
       "      <td>70.328939</td>\n",
       "      <td>160.855906</td>\n",
       "      <td>16.518919</td>\n",
       "      <td>16.906042</td>\n",
       "      <td>33.489321</td>\n",
       "      <td>2.995316</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>481.966821</td>\n",
       "      <td>486.710075</td>\n",
       "      <td>167.371389</td>\n",
       "      <td>219.513008</td>\n",
       "      <td>49.500859</td>\n",
       "      <td>84.734917</td>\n",
       "      <td>105.834745</td>\n",
       "      <td>20.521935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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>41.045000</td>\n",
       "      <td>135.690000</td>\n",
       "      <td>9.470000</td>\n",
       "      <td>38.130000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.480000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>125.560000</td>\n",
       "      <td>282.040000</td>\n",
       "      <td>30.130000</td>\n",
       "      <td>96.840000</td>\n",
       "      <td>2.210000</td>\n",
       "      <td>1.200000</td>\n",
       "      <td>7.480000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>360.360000</td>\n",
       "      <td>532.905000</td>\n",
       "      <td>74.735000</td>\n",
       "      <td>203.535000</td>\n",
       "      <td>13.040000</td>\n",
       "      <td>10.990000</td>\n",
       "      <td>29.250000</td>\n",
       "      <td>0.260000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>8157.780000</td>\n",
       "      <td>9667.130000</td>\n",
       "      <td>5709.590000</td>\n",
       "      <td>4388.730000</td>\n",
       "      <td>1983.010000</td>\n",
       "      <td>5800.930000</td>\n",
       "      <td>3470.380000</td>\n",
       "      <td>1136.080000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        onnet_mou_7  offnet_mou_7  loc_ic_t2t_mou_7  loc_ic_t2m_mou_7  \\\n",
       "count  29979.000000  29979.000000      29979.000000      29979.000000   \n",
       "mean     304.591922    424.103446         70.328939        160.855906   \n",
       "std      481.966821    486.710075        167.371389        219.513008   \n",
       "min        0.000000      0.000000          0.000000          0.000000   \n",
       "25%       41.045000    135.690000          9.470000         38.130000   \n",
       "50%      125.560000    282.040000         30.130000         96.840000   \n",
       "75%      360.360000    532.905000         74.735000        203.535000   \n",
       "max     8157.780000   9667.130000       5709.590000       4388.730000   \n",
       "\n",
       "       loc_ic_t2f_mou_7  std_ic_t2t_mou_7  std_ic_t2m_mou_7  std_ic_t2f_mou_7  \n",
       "count      29979.000000      29979.000000      29979.000000      29979.000000  \n",
       "mean          16.518919         16.906042         33.489321          2.995316  \n",
       "std           49.500859         84.734917        105.834745         20.521935  \n",
       "min            0.000000          0.000000          0.000000          0.000000  \n",
       "25%            0.000000          0.000000          0.480000          0.000000  \n",
       "50%            2.210000          1.200000          7.480000          0.000000  \n",
       "75%           13.040000         10.990000         29.250000          0.260000  \n",
       "max         1983.010000       5800.930000       3470.380000       1136.080000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>loc_og_t2t_mou_7</th>\n",
       "      <th>loc_og_t2m_mou_7</th>\n",
       "      <th>loc_og_t2f_mou_7</th>\n",
       "      <th>loc_og_t2c_mou_7</th>\n",
       "      <th>std_og_t2t_mou_7</th>\n",
       "      <th>std_og_t2m_mou_7</th>\n",
       "      <th>std_og_t2f_mou_7</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.00000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>95.787420</td>\n",
       "      <td>181.323574</td>\n",
       "      <td>7.100147</td>\n",
       "      <td>1.86379</td>\n",
       "      <td>200.060865</td>\n",
       "      <td>213.516733</td>\n",
       "      <td>2.035893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>248.250149</td>\n",
       "      <td>240.815912</td>\n",
       "      <td>22.599153</td>\n",
       "      <td>9.26021</td>\n",
       "      <td>428.303583</td>\n",
       "      <td>438.106106</td>\n",
       "      <td>13.357273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</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>9.130000</td>\n",
       "      <td>33.580000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.350000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>33.190000</td>\n",
       "      <td>104.360000</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>13.440000</td>\n",
       "      <td>37.560000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>91.535000</td>\n",
       "      <td>239.485000</td>\n",
       "      <td>5.260000</td>\n",
       "      <td>0.10000</td>\n",
       "      <td>191.660000</td>\n",
       "      <td>223.020000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7400.660000</td>\n",
       "      <td>4557.140000</td>\n",
       "      <td>1196.430000</td>\n",
       "      <td>569.71000</td>\n",
       "      <td>8133.660000</td>\n",
       "      <td>9284.740000</td>\n",
       "      <td>544.630000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_og_t2t_mou_7  loc_og_t2m_mou_7  loc_og_t2f_mou_7  loc_og_t2c_mou_7  \\\n",
       "count      29979.000000      29979.000000      29979.000000       29979.00000   \n",
       "mean          95.787420        181.323574          7.100147           1.86379   \n",
       "std          248.250149        240.815912         22.599153           9.26021   \n",
       "min            0.000000          0.000000          0.000000           0.00000   \n",
       "25%            9.130000         33.580000          0.000000           0.00000   \n",
       "50%           33.190000        104.360000          0.400000           0.00000   \n",
       "75%           91.535000        239.485000          5.260000           0.10000   \n",
       "max         7400.660000       4557.140000       1196.430000         569.71000   \n",
       "\n",
       "       std_og_t2t_mou_7  std_og_t2m_mou_7  std_og_t2f_mou_7  \n",
       "count      29979.000000      29979.000000      29979.000000  \n",
       "mean         200.060865        213.516733          2.035893  \n",
       "std          428.303583        438.106106         13.357273  \n",
       "min            0.000000          0.000000          0.000000  \n",
       "25%            0.000000          1.350000          0.000000  \n",
       "50%           13.440000         37.560000          0.000000  \n",
       "75%          191.660000        223.020000          0.000000  \n",
       "max         8133.660000       9284.740000        544.630000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>onnet_mou_8</th>\n",
       "      <th>offnet_mou_8</th>\n",
       "      <th>loc_ic_t2t_mou_8</th>\n",
       "      <th>loc_ic_t2m_mou_8</th>\n",
       "      <th>loc_ic_t2f_mou_8</th>\n",
       "      <th>std_ic_t2t_mou_8</th>\n",
       "      <th>std_ic_t2m_mou_8</th>\n",
       "      <th>std_ic_t2f_mou_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>267.809147</td>\n",
       "      <td>375.159926</td>\n",
       "      <td>65.959031</td>\n",
       "      <td>153.649476</td>\n",
       "      <td>14.712935</td>\n",
       "      <td>15.061359</td>\n",
       "      <td>30.432842</td>\n",
       "      <td>2.683053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>466.750205</td>\n",
       "      <td>477.656991</td>\n",
       "      <td>155.769286</td>\n",
       "      <td>217.059251</td>\n",
       "      <td>43.729954</td>\n",
       "      <td>72.469566</td>\n",
       "      <td>105.328660</td>\n",
       "      <td>20.280187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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>27.070000</td>\n",
       "      <td>95.730000</td>\n",
       "      <td>6.810000</td>\n",
       "      <td>29.660000</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>99.590000</td>\n",
       "      <td>241.040000</td>\n",
       "      <td>26.840000</td>\n",
       "      <td>89.830000</td>\n",
       "      <td>1.850000</td>\n",
       "      <td>0.560000</td>\n",
       "      <td>5.710000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>298.215000</td>\n",
       "      <td>483.000000</td>\n",
       "      <td>70.360000</td>\n",
       "      <td>197.015000</td>\n",
       "      <td>11.610000</td>\n",
       "      <td>8.860000</td>\n",
       "      <td>25.330000</td>\n",
       "      <td>0.130000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>10752.560000</td>\n",
       "      <td>14007.340000</td>\n",
       "      <td>4003.210000</td>\n",
       "      <td>5738.460000</td>\n",
       "      <td>1588.530000</td>\n",
       "      <td>4309.290000</td>\n",
       "      <td>5645.860000</td>\n",
       "      <td>1394.890000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        onnet_mou_8  offnet_mou_8  loc_ic_t2t_mou_8  loc_ic_t2m_mou_8  \\\n",
       "count  29979.000000  29979.000000      29979.000000      29979.000000   \n",
       "mean     267.809147    375.159926         65.959031        153.649476   \n",
       "std      466.750205    477.656991        155.769286        217.059251   \n",
       "min        0.000000      0.000000          0.000000          0.000000   \n",
       "25%       27.070000     95.730000          6.810000         29.660000   \n",
       "50%       99.590000    241.040000         26.840000         89.830000   \n",
       "75%      298.215000    483.000000         70.360000        197.015000   \n",
       "max    10752.560000  14007.340000       4003.210000       5738.460000   \n",
       "\n",
       "       loc_ic_t2f_mou_8  std_ic_t2t_mou_8  std_ic_t2m_mou_8  std_ic_t2f_mou_8  \n",
       "count      29979.000000      29979.000000      29979.000000      29979.000000  \n",
       "mean          14.712935         15.061359         30.432842          2.683053  \n",
       "std           43.729954         72.469566        105.328660         20.280187  \n",
       "min            0.000000          0.000000          0.000000          0.000000  \n",
       "25%            0.000000          0.000000          0.000000          0.000000  \n",
       "50%            1.850000          0.560000          5.710000          0.000000  \n",
       "75%           11.610000          8.860000         25.330000          0.130000  \n",
       "max         1588.530000       4309.290000       5645.860000       1394.890000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>loc_og_t2t_mou_8</th>\n",
       "      <th>loc_og_t2m_mou_8</th>\n",
       "      <th>loc_og_t2f_mou_8</th>\n",
       "      <th>loc_og_t2c_mou_8</th>\n",
       "      <th>std_og_t2t_mou_8</th>\n",
       "      <th>std_og_t2m_mou_8</th>\n",
       "      <th>std_og_t2f_mou_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>87.184563</td>\n",
       "      <td>167.616746</td>\n",
       "      <td>6.496088</td>\n",
       "      <td>1.714052</td>\n",
       "      <td>172.355233</td>\n",
       "      <td>179.665270</td>\n",
       "      <td>1.790935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>234.833442</td>\n",
       "      <td>234.896621</td>\n",
       "      <td>20.228358</td>\n",
       "      <td>7.401177</td>\n",
       "      <td>410.213116</td>\n",
       "      <td>416.924325</td>\n",
       "      <td>11.706431</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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>5.790000</td>\n",
       "      <td>22.420000</td>\n",
       "      <td>0.000000</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>28.640000</td>\n",
       "      <td>89.880000</td>\n",
       "      <td>0.160000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.960000</td>\n",
       "      <td>23.730000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>84.700000</td>\n",
       "      <td>223.615000</td>\n",
       "      <td>4.680000</td>\n",
       "      <td>0.050000</td>\n",
       "      <td>133.210000</td>\n",
       "      <td>164.785000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>10752.560000</td>\n",
       "      <td>4961.330000</td>\n",
       "      <td>928.490000</td>\n",
       "      <td>351.830000</td>\n",
       "      <td>8014.430000</td>\n",
       "      <td>13950.040000</td>\n",
       "      <td>516.910000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       loc_og_t2t_mou_8  loc_og_t2m_mou_8  loc_og_t2f_mou_8  loc_og_t2c_mou_8  \\\n",
       "count      29979.000000      29979.000000      29979.000000      29979.000000   \n",
       "mean          87.184563        167.616746          6.496088          1.714052   \n",
       "std          234.833442        234.896621         20.228358          7.401177   \n",
       "min            0.000000          0.000000          0.000000          0.000000   \n",
       "25%            5.790000         22.420000          0.000000          0.000000   \n",
       "50%           28.640000         89.880000          0.160000          0.000000   \n",
       "75%           84.700000        223.615000          4.680000          0.050000   \n",
       "max        10752.560000       4961.330000        928.490000        351.830000   \n",
       "\n",
       "       std_og_t2t_mou_8  std_og_t2m_mou_8  std_og_t2f_mou_8  \n",
       "count      29979.000000      29979.000000      29979.000000  \n",
       "mean         172.355233        179.665270          1.790935  \n",
       "std          410.213116        416.924325         11.706431  \n",
       "min            0.000000          0.000000          0.000000  \n",
       "25%            0.000000          0.000000          0.000000  \n",
       "50%            5.960000         23.730000          0.000000  \n",
       "75%          133.210000        164.785000          0.000000  \n",
       "max         8014.430000      13950.040000        516.910000  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x1296 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols = [['onnet_mou_6','offnet_mou_6','loc_ic_t2t_mou_6','loc_ic_t2m_mou_6','loc_ic_t2f_mou_6','std_ic_t2t_mou_6','std_ic_t2m_mou_6','std_ic_t2f_mou_6'],\n",
    "        ['loc_og_t2t_mou_6','loc_og_t2m_mou_6','loc_og_t2f_mou_6','loc_og_t2c_mou_6','std_og_t2t_mou_6','std_og_t2m_mou_6','std_og_t2f_mou_6'],\n",
    "        ['onnet_mou_7','offnet_mou_7','loc_ic_t2t_mou_7','loc_ic_t2m_mou_7','loc_ic_t2f_mou_7','std_ic_t2t_mou_7','std_ic_t2m_mou_7','std_ic_t2f_mou_7'],\n",
    "        ['loc_og_t2t_mou_7','loc_og_t2m_mou_7','loc_og_t2f_mou_7','loc_og_t2c_mou_7','std_og_t2t_mou_7','std_og_t2m_mou_7','std_og_t2f_mou_7'],\n",
    "        ['onnet_mou_8','offnet_mou_8','loc_ic_t2t_mou_8','loc_ic_t2m_mou_8','loc_ic_t2f_mou_8','std_ic_t2t_mou_8','std_ic_t2m_mou_8','std_ic_t2f_mou_8'],\n",
    "        ['loc_og_t2t_mou_8','loc_og_t2m_mou_8','loc_og_t2f_mou_8','loc_og_t2c_mou_8','std_og_t2t_mou_8','std_og_t2m_mou_8','std_og_t2f_mou_8']]\n",
    "\n",
    "# column description stats\n",
    "for i in range(0,6):\n",
    "    display(data[cols[i]].describe())\n",
    "\n",
    "# plot for the operatorwise calls usage\n",
    "plt.figure(figsize=(18, 18))\n",
    "plt.subplots_adjust(hspace=0.5)\n",
    "for i in range(0,6):\n",
    "    plt.subplot(3,2,i+1)\n",
    "    X = pd.concat([data[cols[i]], data['churn']], axis=1)\n",
    "    X = pd.melt(X,id_vars=\"churn\",var_name=\"features\",value_name='value')\n",
    "    sns.boxplot(x=\"features\", y=\"value\", hue=\"churn\", data=X)\n",
    "    plt.xticks(rotation=90)    \n",
    "    plt.suptitle('Operatorwise Calls Usage')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Recharge Amount"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": false
   },
   "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_amt_6</th>\n",
       "      <th>total_rech_amt_7</th>\n",
       "      <th>total_rech_amt_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>697.720204</td>\n",
       "      <td>696.309083</td>\n",
       "      <td>613.908403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>539.488337</td>\n",
       "      <td>562.327538</td>\n",
       "      <td>602.047678</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>433.000000</td>\n",
       "      <td>427.000000</td>\n",
       "      <td>309.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>585.000000</td>\n",
       "      <td>581.000000</td>\n",
       "      <td>520.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>838.000000</td>\n",
       "      <td>836.000000</td>\n",
       "      <td>790.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>35190.000000</td>\n",
       "      <td>40335.000000</td>\n",
       "      <td>45320.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       total_rech_amt_6  total_rech_amt_7  total_rech_amt_8\n",
       "count      29979.000000      29979.000000      29979.000000\n",
       "mean         697.720204        696.309083        613.908403\n",
       "std          539.488337        562.327538        602.047678\n",
       "min            0.000000          0.000000          0.000000\n",
       "25%          433.000000        427.000000        309.000000\n",
       "50%          585.000000        581.000000        520.000000\n",
       "75%          838.000000        836.000000        790.000000\n",
       "max        35190.000000      40335.000000      45320.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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_amt_voice_6</th>\n",
       "      <th>total_rech_amt_voice_7</th>\n",
       "      <th>total_rech_amt_voice_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>608.614859</td>\n",
       "      <td>600.873285</td>\n",
       "      <td>526.838474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>527.653094</td>\n",
       "      <td>547.550096</td>\n",
       "      <td>577.651964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>352.000000</td>\n",
       "      <td>345.000000</td>\n",
       "      <td>232.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>516.000000</td>\n",
       "      <td>506.000000</td>\n",
       "      <td>441.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>755.000000</td>\n",
       "      <td>750.000000</td>\n",
       "      <td>700.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>35190.000000</td>\n",
       "      <td>40335.000000</td>\n",
       "      <td>45320.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       total_rech_amt_voice_6  total_rech_amt_voice_7  total_rech_amt_voice_8\n",
       "count            29979.000000            29979.000000            29979.000000\n",
       "mean               608.614859              600.873285              526.838474\n",
       "std                527.653094              547.550096              577.651964\n",
       "min                  0.000000                0.000000                0.000000\n",
       "25%                352.000000              345.000000              232.000000\n",
       "50%                516.000000              506.000000              441.000000\n",
       "75%                755.000000              750.000000              700.000000\n",
       "max              35190.000000            40335.000000            45320.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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_amt_data_6</th>\n",
       "      <th>total_rech_amt_data_7</th>\n",
       "      <th>total_rech_amt_data_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>90.788857</td>\n",
       "      <td>97.605184</td>\n",
       "      <td>89.803108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>191.272837</td>\n",
       "      <td>198.435609</td>\n",
       "      <td>185.886834</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>154.000000</td>\n",
       "      <td>154.000000</td>\n",
       "      <td>145.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>7546.000000</td>\n",
       "      <td>4365.000000</td>\n",
       "      <td>4061.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       total_rech_amt_data_6  total_rech_amt_data_7  total_rech_amt_data_8\n",
       "count           29979.000000           29979.000000           29979.000000\n",
       "mean               90.788857              97.605184              89.803108\n",
       "std               191.272837             198.435609             185.886834\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%               154.000000             154.000000             145.000000\n",
       "max              7546.000000            4365.000000            4061.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>max_rech_amt_6</th>\n",
       "      <th>max_rech_amt_7</th>\n",
       "      <th>max_rech_amt_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>171.471130</td>\n",
       "      <td>175.723006</td>\n",
       "      <td>162.929351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>174.762549</td>\n",
       "      <td>181.606909</td>\n",
       "      <td>172.671442</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>110.000000</td>\n",
       "      <td>110.000000</td>\n",
       "      <td>67.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>120.000000</td>\n",
       "      <td>128.000000</td>\n",
       "      <td>130.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>200.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>198.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>4010.000000</td>\n",
       "      <td>4010.000000</td>\n",
       "      <td>4449.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       max_rech_amt_6  max_rech_amt_7  max_rech_amt_8\n",
       "count    29979.000000    29979.000000    29979.000000\n",
       "mean       171.471130      175.723006      162.929351\n",
       "std        174.762549      181.606909      172.671442\n",
       "min          0.000000        0.000000        0.000000\n",
       "25%        110.000000      110.000000       67.000000\n",
       "50%        120.000000      128.000000      130.000000\n",
       "75%        200.000000      200.000000      198.000000\n",
       "max       4010.000000     4010.000000     4449.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data['total_rech_amt_voice_6'] = np.where((data['total_rech_amt_6'] >= data['total_rech_amt_data_6']), (data['total_rech_amt_6'] - data['total_rech_amt_data_6']), 0)\n",
    "data['total_rech_amt_voice_7'] = np.where((data['total_rech_amt_7'] >= data['total_rech_amt_data_7']), (data['total_rech_amt_7'] - data['total_rech_amt_data_7']), 0)\n",
    "data['total_rech_amt_voice_8'] = np.where((data['total_rech_amt_8'] >= data['total_rech_amt_data_8']), (data['total_rech_amt_8'] - data['total_rech_amt_data_8']), 0)\n",
    "\n",
    "cols = [\n",
    "        ['total_rech_amt_6','total_rech_amt_7','total_rech_amt_8'],\n",
    "        ['total_rech_amt_voice_6','total_rech_amt_voice_7','total_rech_amt_voice_8'],\n",
    "        ['total_rech_amt_data_6','total_rech_amt_data_7','total_rech_amt_data_8'],\n",
    "        ['max_rech_amt_6','max_rech_amt_7','max_rech_amt_8']\n",
    "       ]\n",
    "\n",
    "\n",
    "for i in range(0,4):\n",
    "    display(data[cols[i]].describe())\n",
    "\n",
    "\n",
    "plt.figure(figsize=(18, 10))\n",
    "plt.subplots_adjust(hspace=0.5)\n",
    "for i in range(0,4):\n",
    "    plt.subplot(2,2,i+1)\n",
    "    X = pd.concat([data[cols[i]], data['churn']], axis=1)\n",
    "    X = pd.melt(X,id_vars=\"churn\",var_name=\"features\",value_name='value')\n",
    "    sns.boxplot(x=\"features\", y=\"value\", hue=\"churn\", data=X)\n",
    "    plt.xticks(rotation=90)    \n",
    "    plt.suptitle('Recharge Amount')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Recharge Count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": false
   },
   "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_num_6</th>\n",
       "      <th>total_rech_num_7</th>\n",
       "      <th>total_rech_num_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>12.123253</td>\n",
       "      <td>11.914207</td>\n",
       "      <td>10.226092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>9.546424</td>\n",
       "      <td>9.605757</td>\n",
       "      <td>9.480012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>4.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>8.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>15.000000</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>13.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>307.000000</td>\n",
       "      <td>138.000000</td>\n",
       "      <td>196.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       total_rech_num_6  total_rech_num_7  total_rech_num_8\n",
       "count      29979.000000      29979.000000      29979.000000\n",
       "mean          12.123253         11.914207         10.226092\n",
       "std            9.546424          9.605757          9.480012\n",
       "min            0.000000          0.000000          0.000000\n",
       "25%            6.000000          6.000000          4.000000\n",
       "50%            9.000000          9.000000          8.000000\n",
       "75%           15.000000         15.000000         13.000000\n",
       "max          307.000000        138.000000        196.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.936889</td>\n",
       "      <td>1.053938</td>\n",
       "      <td>1.035425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>2.176824</td>\n",
       "      <td>2.412950</td>\n",
       "      <td>2.405824</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</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",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       total_rech_data_6  total_rech_data_7  total_rech_data_8\n",
       "count       29979.000000       29979.000000       29979.000000\n",
       "mean            0.936889           1.053938           1.035425\n",
       "std             2.176824           2.412950           2.405824\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"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>max_rech_data_6</th>\n",
       "      <th>max_rech_data_7</th>\n",
       "      <th>max_rech_data_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>56.949331</td>\n",
       "      <td>58.716702</td>\n",
       "      <td>55.276594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>105.849744</td>\n",
       "      <td>108.176034</td>\n",
       "      <td>104.814186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>98.000000</td>\n",
       "      <td>98.000000</td>\n",
       "      <td>56.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1555.000000</td>\n",
       "      <td>1555.000000</td>\n",
       "      <td>1555.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       max_rech_data_6  max_rech_data_7  max_rech_data_8\n",
       "count     29979.000000     29979.000000     29979.000000\n",
       "mean         56.949331        58.716702        55.276594\n",
       "std         105.849744       108.176034       104.814186\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%          98.000000        98.000000        56.000000\n",
       "max        1555.000000      1555.000000      1555.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>count_rech_2g_6</th>\n",
       "      <th>count_rech_2g_7</th>\n",
       "      <th>count_rech_2g_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.624771</td>\n",
       "      <td>0.718303</td>\n",
       "      <td>0.721605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.730741</td>\n",
       "      <td>1.930804</td>\n",
       "      <td>1.871110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>39.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>44.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       count_rech_2g_6  count_rech_2g_7  count_rech_2g_8\n",
       "count     29979.000000     29979.000000     29979.000000\n",
       "mean          0.624771         0.718303         0.721605\n",
       "std           1.730741         1.930804         1.871110\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          39.000000        48.000000        44.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>count_rech_3g_6</th>\n",
       "      <th>count_rech_3g_7</th>\n",
       "      <th>count_rech_3g_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.312118</td>\n",
       "      <td>0.335635</td>\n",
       "      <td>0.313820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.066232</td>\n",
       "      <td>1.176567</td>\n",
       "      <td>1.162114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>29.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>45.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       count_rech_3g_6  count_rech_3g_7  count_rech_3g_8\n",
       "count     29979.000000     29979.000000     29979.000000\n",
       "mean          0.312118         0.335635         0.313820\n",
       "std           1.066232         1.176567         1.162114\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          29.000000        35.000000        45.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x720 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols = [\n",
    "        ['total_rech_num_6','total_rech_num_7','total_rech_num_8'],\n",
    "        ['total_rech_data_6','total_rech_data_7','total_rech_data_8'],\n",
    "        ['max_rech_data_6','max_rech_data_7','max_rech_data_8'],\n",
    "        ['count_rech_2g_6','count_rech_2g_7','count_rech_2g_8'],\n",
    "        ['count_rech_3g_6','count_rech_3g_7','count_rech_3g_8'] \n",
    "       ]\n",
    "\n",
    "\n",
    "for i in range(0,5):\n",
    "    display(data[cols[i]].describe())\n",
    "\n",
    "\n",
    "plt.figure(figsize=(18, 10))\n",
    "plt.subplots_adjust(hspace=0.5)\n",
    "for i in range(0,5):\n",
    "    plt.subplot(2,3,i+1)\n",
    "    X = pd.concat([data[cols[i]], data['churn']], axis=1)\n",
    "    X = pd.melt(X,id_vars=\"churn\",var_name=\"features\",value_name='value')\n",
    "    sns.boxplot(x=\"features\", y=\"value\", hue=\"churn\", data=X)\n",
    "    plt.xticks(rotation=90)    \n",
    "    plt.suptitle('Recharge Count')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Average Revenue per User (Arpu)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "scrolled": false
   },
   "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>arpu_6</th>\n",
       "      <th>arpu_7</th>\n",
       "      <th>arpu_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>587.587624</td>\n",
       "      <td>589.424681</td>\n",
       "      <td>535.067748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>442.848876</td>\n",
       "      <td>463.049067</td>\n",
       "      <td>492.447887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-2258.709000</td>\n",
       "      <td>-2014.045000</td>\n",
       "      <td>-945.808000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>364.352500</td>\n",
       "      <td>365.235000</td>\n",
       "      <td>289.782000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>495.951000</td>\n",
       "      <td>493.781000</td>\n",
       "      <td>452.232000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>704.184500</td>\n",
       "      <td>700.986000</td>\n",
       "      <td>671.349500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>27731.088000</td>\n",
       "      <td>35145.834000</td>\n",
       "      <td>33543.624000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             arpu_6        arpu_7        arpu_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean     587.587624    589.424681    535.067748\n",
       "std      442.848876    463.049067    492.447887\n",
       "min    -2258.709000  -2014.045000   -945.808000\n",
       "25%      364.352500    365.235000    289.782000\n",
       "50%      495.951000    493.781000    452.232000\n",
       "75%      704.184500    700.986000    671.349500\n",
       "max    27731.088000  35145.834000  33543.624000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>arpu_2g_6</th>\n",
       "      <th>arpu_2g_7</th>\n",
       "      <th>arpu_2g_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.280483</td>\n",
       "      <td>51.890691</td>\n",
       "      <td>47.866114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>153.212614</td>\n",
       "      <td>157.959498</td>\n",
       "      <td>145.869033</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-15.320000</td>\n",
       "      <td>-15.480000</td>\n",
       "      <td>-24.430000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2.215000</td>\n",
       "      <td>2.375000</td>\n",
       "      <td>0.605000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>6433.760000</td>\n",
       "      <td>4809.360000</td>\n",
       "      <td>3467.170000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          arpu_2g_6     arpu_2g_7     arpu_2g_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean      50.280483     51.890691     47.866114\n",
       "std      153.212614    157.959498    145.869033\n",
       "min      -15.320000    -15.480000    -24.430000\n",
       "25%        0.000000      0.000000      0.000000\n",
       "50%        0.000000      0.000000      0.000000\n",
       "75%        2.215000      2.375000      0.605000\n",
       "max     6433.760000   4809.360000   3467.170000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>arpu_3g_6</th>\n",
       "      <th>arpu_3g_7</th>\n",
       "      <th>arpu_3g_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>54.043360</td>\n",
       "      <td>55.934679</td>\n",
       "      <td>52.160455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>171.872571</td>\n",
       "      <td>176.375885</td>\n",
       "      <td>163.905193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-30.280000</td>\n",
       "      <td>-26.040000</td>\n",
       "      <td>-20.040000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\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>6362.280000</td>\n",
       "      <td>4980.900000</td>\n",
       "      <td>3716.900000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          arpu_3g_6     arpu_3g_7     arpu_3g_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean      54.043360     55.934679     52.160455\n",
       "std      171.872571    176.375885    163.905193\n",
       "min      -30.280000    -26.040000    -20.040000\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     6362.280000   4980.900000   3716.900000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols = [\n",
    "        ['arpu_6','arpu_7','arpu_8'],\n",
    "        ['arpu_2g_6','arpu_2g_7','arpu_2g_8'],\n",
    "        ['arpu_3g_6','arpu_3g_7','arpu_3g_8']\n",
    "       ]\n",
    "\n",
    "\n",
    "for i in range(0,3):\n",
    "    display(data[cols[i]].describe())\n",
    "\n",
    "\n",
    "plt.figure(figsize=(18, 5))\n",
    "for i in range(0,3):\n",
    "    plt.subplot(1,3,i+1)\n",
    "    X = pd.concat([data[cols[i]], data['churn']], axis=1)\n",
    "    X = pd.melt(X,id_vars=\"churn\",var_name=\"features\",value_name='value')\n",
    "    sns.boxplot(x=\"features\", y=\"value\", hue=\"churn\", data=X)\n",
    "    plt.xticks(rotation=90)    \n",
    "    plt.suptitle('Arpu')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2G and 3G"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": false
   },
   "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>vol_2g_mb_6</th>\n",
       "      <th>vol_2g_mb_7</th>\n",
       "      <th>vol_2g_mb_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>78.875263</td>\n",
       "      <td>78.196804</td>\n",
       "      <td>69.248761</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>277.525473</td>\n",
       "      <td>280.430249</td>\n",
       "      <td>268.616268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>14.470000</td>\n",
       "      <td>14.960000</td>\n",
       "      <td>9.625000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>10285.900000</td>\n",
       "      <td>7873.550000</td>\n",
       "      <td>11117.610000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        vol_2g_mb_6   vol_2g_mb_7   vol_2g_mb_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean      78.875263     78.196804     69.248761\n",
       "std      277.525473    280.430249    268.616268\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%       14.470000     14.960000      9.625000\n",
       "max    10285.900000   7873.550000  11117.610000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>vol_3g_mb_6</th>\n",
       "      <th>vol_3g_mb_7</th>\n",
       "      <th>vol_3g_mb_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>258.623297</td>\n",
       "      <td>278.341301</td>\n",
       "      <td>270.061116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>866.619871</td>\n",
       "      <td>856.096362</td>\n",
       "      <td>859.715968</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>0.000000</td>\n",
       "      <td>2.275000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>45735.400000</td>\n",
       "      <td>28144.120000</td>\n",
       "      <td>30036.060000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        vol_3g_mb_6   vol_3g_mb_7   vol_3g_mb_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean     258.623297    278.341301    270.061116\n",
       "std      866.619871    856.096362    859.715968\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      2.275000      0.000000\n",
       "max    45735.400000  28144.120000  30036.060000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>night_pck_user_6</th>\n",
       "      <th>night_pck_user_7</th>\n",
       "      <th>night_pck_user_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.250709</td>\n",
       "      <td>1.232529</td>\n",
       "      <td>1.226025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.962706</td>\n",
       "      <td>0.967516</td>\n",
       "      <td>0.969349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       night_pck_user_6  night_pck_user_7  night_pck_user_8\n",
       "count      29979.000000      29979.000000      29979.000000\n",
       "mean           1.250709          1.232529          1.226025\n",
       "std            0.962706          0.967516          0.969349\n",
       "min            0.000000          0.000000          0.000000\n",
       "25%            0.000000          0.000000          0.000000\n",
       "50%            2.000000          2.000000          2.000000\n",
       "75%            2.000000          2.000000          2.000000\n",
       "max            2.000000          2.000000          2.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>fb_user_6</th>\n",
       "      <th>fb_user_7</th>\n",
       "      <th>fb_user_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.581274</td>\n",
       "      <td>1.568498</td>\n",
       "      <td>1.544648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.566739</td>\n",
       "      <td>0.575308</td>\n",
       "      <td>0.612733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          fb_user_6     fb_user_7     fb_user_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean       1.581274      1.568498      1.544648\n",
       "std        0.566739      0.575308      0.612733\n",
       "min        0.000000      0.000000      0.000000\n",
       "25%        1.000000      1.000000      1.000000\n",
       "50%        2.000000      2.000000      2.000000\n",
       "75%        2.000000      2.000000      2.000000\n",
       "max        2.000000      2.000000      2.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>monthly_2g_6</th>\n",
       "      <th>monthly_2g_7</th>\n",
       "      <th>monthly_2g_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.124954</td>\n",
       "      <td>0.131425</td>\n",
       "      <td>0.114080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.374656</td>\n",
       "      <td>0.387793</td>\n",
       "      <td>0.357335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>4.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       monthly_2g_6  monthly_2g_7  monthly_2g_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean       0.124954      0.131425      0.114080\n",
       "std        0.374656      0.387793      0.357335\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        4.000000      5.000000      5.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>monthly_3g_6</th>\n",
       "      <th>monthly_3g_7</th>\n",
       "      <th>monthly_3g_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.174055</td>\n",
       "      <td>0.185296</td>\n",
       "      <td>0.173321</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.574545</td>\n",
       "      <td>0.620703</td>\n",
       "      <td>0.583178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>14.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>16.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       monthly_3g_6  monthly_3g_7  monthly_3g_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean       0.174055      0.185296      0.173321\n",
       "std        0.574545      0.620703      0.583178\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       14.000000     16.000000     16.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>sachet_2g_6</th>\n",
       "      <th>sachet_2g_7</th>\n",
       "      <th>sachet_2g_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.499817</td>\n",
       "      <td>0.586877</td>\n",
       "      <td>0.607525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.698845</td>\n",
       "      <td>1.900280</td>\n",
       "      <td>1.844605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>39.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>44.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        sachet_2g_6   sachet_2g_7   sachet_2g_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean       0.499817      0.586877      0.607525\n",
       "std        1.698845      1.900280      1.844605\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       39.000000     48.000000     44.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>sachet_3g_6</th>\n",
       "      <th>sachet_3g_7</th>\n",
       "      <th>sachet_3g_8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.138063</td>\n",
       "      <td>0.150339</td>\n",
       "      <td>0.140498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.875984</td>\n",
       "      <td>0.969779</td>\n",
       "      <td>0.975212</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>29.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>41.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        sachet_3g_6   sachet_3g_7   sachet_3g_8\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean       0.138063      0.150339      0.140498\n",
       "std        0.875984      0.969779      0.975212\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       29.000000     35.000000     41.000000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>jun_vbc_3g</th>\n",
       "      <th>jul_vbc_3g</th>\n",
       "      <th>aug_vbc_3g</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "      <td>29979.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>121.450911</td>\n",
       "      <td>135.225368</td>\n",
       "      <td>129.482436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>389.910814</td>\n",
       "      <td>408.218570</td>\n",
       "      <td>390.625027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\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>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\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>0.000000</td>\n",
       "      <td>2.025000</td>\n",
       "      <td>1.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>11166.210000</td>\n",
       "      <td>9165.600000</td>\n",
       "      <td>12916.220000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         jun_vbc_3g    jul_vbc_3g    aug_vbc_3g\n",
       "count  29979.000000  29979.000000  29979.000000\n",
       "mean     121.450911    135.225368    129.482436\n",
       "std      389.910814    408.218570    390.625027\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      2.025000      1.600000\n",
       "max    11166.210000   9165.600000  12916.220000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x1080 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cols = [\n",
    "        ['vol_2g_mb_6','vol_2g_mb_7','vol_2g_mb_8'],\n",
    "        ['vol_3g_mb_6','vol_3g_mb_7','vol_3g_mb_8'],\n",
    "        ['night_pck_user_6','night_pck_user_7','night_pck_user_8'],\n",
    "        ['fb_user_6','fb_user_7','fb_user_8'],\n",
    "        ['monthly_2g_6','monthly_2g_7','monthly_2g_8'],\n",
    "        ['monthly_3g_6','monthly_3g_7','monthly_3g_8'],\n",
    "        ['sachet_2g_6','sachet_2g_7','sachet_2g_8'],\n",
    "        ['sachet_3g_6','sachet_3g_7','sachet_3g_8'],\n",
    "        ['jun_vbc_3g','jul_vbc_3g','aug_vbc_3g']\n",
    "       ]\n",
    "\n",
    "\n",
    "for i in range(0,9):\n",
    "    display(data[cols[i]].describe())\n",
    "\n",
    "\n",
    "plt.figure(figsize=(18, 15))\n",
    "plt.subplots_adjust(hspace=0.5)\n",
    "for i in range(0,9):\n",
    "    plt.subplot(3,3,i+1)\n",
    "    X = pd.concat([data[cols[i]], data['churn']], axis=1)\n",
    "    X = pd.melt(X,id_vars=\"churn\",var_name=\"features\",value_name='value')\n",
    "    sns.boxplot(x=\"features\", y=\"value\", hue=\"churn\", data=X)\n",
    "    plt.xticks(rotation=90)    \n",
    "    plt.suptitle('2G-3G Volume')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "cols_to_exclude = ['night_pck_user_6','night_pck_user_7','night_pck_user_8',\n",
    "                   'fb_user_6','fb_user_7','fb_user_8',\n",
    "                   'monthly_2g_6','monthly_2g_7','monthly_2g_8',\n",
    "                   'monthly_3g_6','monthly_3g_7','monthly_3g_8',\n",
    "                   'sachet_2g_6','sachet_2g_7','sachet_2g_8',\n",
    "                   'sachet_3g_6','sachet_3g_7','sachet_3g_8',\n",
    "                   'date_of_last_rech_6','date_of_last_rech_7','date_of_last_rech_8','date_of_last_rech_data_6','date_of_last_rech_data_7','date_of_last_rech_data_8',\n",
    "                   'spl_ic_mou_6','spl_ic_mou_7','spl_ic_mou_8','spl_og_mou_6','og_others_6','spl_og_mou_7','og_others_7','spl_og_mou_8','og_others_8',\n",
    "                   'loc_og_t2c_mou_6','std_og_t2f_mou_6','std_ic_t2f_mou_6','loc_ic_t2f_mou_6',\n",
    "                   'loc_og_t2c_mou_7','std_og_t2f_mou_7','std_ic_t2f_mou_7','loc_ic_t2f_mou_7',\n",
    "                   'loc_og_t2c_mou_8','std_og_t2f_mou_8','std_ic_t2f_mou_8','loc_ic_t2f_mou_8',\n",
    "                   'aon','churn'\n",
    "                  ]\n",
    "cols = list(set(data.columns).difference(set(cols_to_exclude)))\n",
    "\n",
    "\n",
    "for col in cols:\n",
    "    percentiles = data[col].quantile([0.01,0.99]).values\n",
    "    #data[col][data[col] <= percentiles[0]] = percentiles[0]\n",
    "    data[col][data[col] >= percentiles[1]] = percentiles[1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Remove the outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "data = data.loc[~(\n",
    "                    ((data['roam_og_mou_8'] > 2200) & (data['churn'] == 1)) |\n",
    "                    ((data['arpu_7'] > 10000) & (data['churn'] == 1)) |\n",
    "                    ((data['loc_og_mou_8'] > 2000) & (data['churn'] == 1)) |\n",
    "                    ((data['loc_ic_mou_7'] > 4000) & (data['churn'] == 1)) |\n",
    "                    ((data['std_og_mou_7'] > 7000) & (data['churn'] == 1)) |\n",
    "                    ((data['vol_2g_mb_8'] > 2500) & (data['churn'] == 1)) \n",
    "                 )\n",
    "               ]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature Engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Convert date columns to date format\n",
    "for col in date_cols:\n",
    "    data[col] = pd.to_datetime(data[col], format='%m/%d/%Y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "cols = ['date_of_last_rech_6','date_of_last_rech_7','date_of_last_rech_8']\n",
    "# get the recent date of recharge in the last 3 months\n",
    "data['last_rech_date'] = data[cols].max(axis=1)\n",
    "# get the number of days from the recent recharge date till the last date of august month\n",
    "data['days_since_last_rech'] = np.floor(( pd.to_datetime('2014-08-31', format='%Y-%m-%d') - data['last_rech_date'] ).astype('timedelta64[D]'))\n",
    "# fill the null values as 0\n",
    "data['days_since_last_rech'] = data['days_since_last_rech'].fillna(0)\n",
    "\n",
    "# subtract it from 3 to add higher weightage for values present in all the columns. \n",
    "# len(cols) = 3,  means present in all columns, 0 means not present in any column\n",
    "data['rech_weightage'] = len(cols) - (data[cols].isnull().sum(axis=1))\n",
    "data.drop(['last_rech_date','date_of_last_rech_6','date_of_last_rech_7','date_of_last_rech_8'], axis=1, inplace=True)\n",
    "\n",
    "\n",
    "cols = ['date_of_last_rech_data_6','date_of_last_rech_data_7','date_of_last_rech_data_8']\n",
    "# get the recent date of recharge data in the last 3 months\n",
    "data['last_rech_data_date'] = data[cols].max(axis=1)\n",
    "# get the number of days from the recent recharge data date till the last date of august month\n",
    "data['days_since_last_data_rech'] = np.floor(( pd.to_datetime('2014-08-31', format='%Y-%m-%d') - data['last_rech_data_date'] ).astype('timedelta64[D]'))\n",
    "# fill the null values as 0\n",
    "data['days_since_last_data_rech'] = data['days_since_last_data_rech'].fillna(0)\n",
    "\n",
    "# subtract it from 3 to add higher weightage for values present in all the columns. \n",
    "# len(cols) = 3, means present in all columns, 0 means not present in any column\n",
    "data['rech_data_weightage'] = len(cols) - (data[cols].isnull().sum(axis=1))\n",
    "\n",
    "# drop the unwanted columns\n",
    "data.drop(['last_rech_data_date','date_of_last_rech_data_6','date_of_last_rech_data_7','date_of_last_rech_data_8'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Numeric Columns - call usage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# network columns\n",
    "# get the mean of onnet mou in the last 3 months\n",
    "cols = ['onnet_mou_6','onnet_mou_7','onnet_mou_8']\n",
    "data['mean_onnet_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "# get the mean of offnet mou in the last 3 months\n",
    "cols = ['offnet_mou_6','offnet_mou_7','offnet_mou_8']\n",
    "data['mean_offnet_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "# get the mean total of both onnet and offnet mou in the last 3 months\n",
    "data['mean_onnet_offnet_mou'] = data['mean_onnet_mou'] + data['mean_offnet_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Roaming columns\n",
    "# get the mean of roam ic mou in the last 3 months\n",
    "cols = ['roam_ic_mou_6','roam_ic_mou_7','roam_ic_mou_8']\n",
    "data['mean_roam_ic_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "# get the mean of roam og mou in the last 3 months\n",
    "cols = ['roam_og_mou_6','roam_og_mou_7','roam_og_mou_8']\n",
    "data['mean_roam_og_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "# get the mean total of both roam ic and og mou in the last 3 months\n",
    "data['mean_roam_mou'] = data['mean_roam_ic_mou'] + data['mean_roam_og_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# loc-t2t columns\n",
    "cols = ['loc_ic_t2t_mou_6','loc_ic_t2t_mou_7','loc_ic_t2t_mou_8']\n",
    "data['mean_loc_ic_t2t_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['loc_og_t2t_mou_6','loc_og_t2t_mou_7','loc_og_t2t_mou_8']\n",
    "data['mean_loc_og_t2t_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_loc_t2t_mou'] = data['mean_loc_ic_t2t_mou'] + data['mean_loc_og_t2t_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# loc-t2m columns\n",
    "cols = ['loc_ic_t2m_mou_6','loc_ic_t2m_mou_7','loc_ic_t2m_mou_8']\n",
    "data['mean_loc_ic_t2m_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['loc_og_t2m_mou_6','loc_og_t2m_mou_7','loc_og_t2m_mou_8']\n",
    "data['mean_loc_og_t2m_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_loc_t2m_mou'] = data['mean_loc_ic_t2m_mou'] + data['mean_loc_og_t2m_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# loc-t2f columns\n",
    "cols = ['loc_ic_t2f_mou_6','loc_ic_t2f_mou_7','loc_ic_t2f_mou_8']\n",
    "data['mean_loc_ic_t2f_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['loc_og_t2f_mou_6','loc_og_t2f_mou_7','loc_og_t2f_mou_8']\n",
    "data['mean_loc_og_t2f_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_loc_t2f_mou'] = data['mean_loc_ic_t2f_mou'] + data['mean_loc_og_t2f_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# loc-t2c columns\n",
    "cols = ['loc_og_t2c_mou_6','loc_og_t2c_mou_7','loc_og_t2c_mou_8']\n",
    "data['mean_loc_og_t2c_mou'] = round(data[cols].mean(axis=1),2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# std-t2t columns\n",
    "cols = ['std_ic_t2t_mou_6','std_ic_t2t_mou_7','std_ic_t2t_mou_8']\n",
    "data['mean_std_ic_t2t_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['std_og_t2t_mou_6','std_og_t2t_mou_7','std_og_t2t_mou_8']\n",
    "data['mean_std_og_t2t_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_std_t2t_mou'] = data['mean_std_ic_t2t_mou'] + data['mean_std_og_t2t_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# std-t2m columns\n",
    "cols = ['std_ic_t2m_mou_6','std_ic_t2m_mou_7','std_ic_t2m_mou_8']\n",
    "data['mean_std_ic_t2m_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['std_og_t2m_mou_6','std_og_t2m_mou_7','std_og_t2m_mou_8']\n",
    "data['mean_std_og_t2m_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_std_t2m_mou'] = data['mean_std_ic_t2m_mou'] + data['mean_std_og_t2m_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "# std-t2f columns\n",
    "cols = ['std_ic_t2f_mou_6','std_ic_t2f_mou_7','std_ic_t2f_mou_8']\n",
    "data['mean_std_ic_t2f_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['std_og_t2f_mou_6','std_og_t2f_mou_7','std_og_t2f_mou_8']\n",
    "data['mean_std_og_t2f_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_std_t2f_mou'] = data['mean_std_ic_t2f_mou'] + data['mean_std_og_t2f_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "# loc columns\n",
    "cols = ['loc_ic_mou_6','loc_ic_mou_7','loc_ic_mou_8']\n",
    "data['mean_loc_ic_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['loc_og_mou_6','loc_og_mou_7','loc_og_mou_8']\n",
    "data['mean_loc_og_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_loc_mou'] = data['mean_loc_ic_mou'] + data['mean_loc_og_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "# std columns\n",
    "cols = ['std_ic_mou_6','std_ic_mou_7','std_ic_mou_8']\n",
    "data['mean_std_ic_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['std_og_mou_6','std_og_mou_7','std_og_mou_8']\n",
    "data['mean_std_og_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_std_mou'] = data['mean_std_ic_mou'] + data['mean_std_og_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "# isd columns\n",
    "cols = ['isd_ic_mou_6','isd_ic_mou_7','isd_ic_mou_8']\n",
    "data['mean_isd_ic_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['isd_og_mou_6','isd_og_mou_7','isd_og_mou_8']\n",
    "data['mean_isd_og_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_isd_mou'] = data['mean_isd_ic_mou'] + data['mean_isd_og_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "# spl columns\n",
    "cols = ['spl_ic_mou_6','spl_ic_mou_7','spl_ic_mou_8']\n",
    "data['mean_spl_ic_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['spl_og_mou_6','spl_og_mou_7','spl_og_mou_8']\n",
    "data['mean_spl_og_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_spl_mou'] = data['mean_spl_ic_mou'] + data['mean_spl_og_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "# others columns\n",
    "cols = ['ic_others_6','ic_others_7','ic_others_8']\n",
    "data['mean_ic_others_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['og_others_6','og_others_7','og_others_8']\n",
    "data['mean_og_others_mou'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "data['mean_others_mou'] = data['mean_ic_others_mou'] + data['mean_og_others_mou']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total columns\n",
    "cols = ['total_ic_mou_6','total_ic_mou_7','total_ic_mou_8']\n",
    "data['mean_total_ic_mou'] = round(data[cols].mean(axis=1),2)\n",
    "# Weightage for ic for the last 3 months\n",
    "df = data[cols].astype(bool)\n",
    "data['total_ic_weightage'] = ( df['total_ic_mou_6'] * 1 ) + ( df['total_ic_mou_7'] * 10 ) + ( df['total_ic_mou_8'] * 100 )\n",
    "\n",
    "cols = ['total_og_mou_6','total_og_mou_7','total_og_mou_8']\n",
    "data['mean_total_og_mou'] = round(data[cols].mean(axis=1),2)\n",
    "# Weightage for og for the last 3 months\n",
    "df = data[cols].astype(bool)\n",
    "data['total_og_weightage'] = ( df['total_og_mou_6'] * 1 ) + ( df['total_og_mou_7'] * 10 ) + ( df['total_og_mou_8'] * 100 )\n",
    "\n",
    "data['mean_total_mou'] = data['mean_total_ic_mou'] + data['mean_total_og_mou']\n",
    "\n",
    "data['mean_total_mou_6'] = round(data[['total_ic_mou_6','total_og_mou_6']].mean(axis=1),2)\n",
    "data['mean_total_mou_7'] = round(data[['total_ic_mou_7','total_og_mou_7']].mean(axis=1),2)\n",
    "data['mean_total_mou_8'] = round(data[['total_ic_mou_8','total_og_mou_8']].mean(axis=1),2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Numeric Columns - Recharge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total_rech_num columns\n",
    "cols = ['total_rech_num_6','total_rech_num_7','total_rech_num_8']\n",
    "# mean of total recharge number\n",
    "data['mean_total_rech_num'] = round(data[cols].mean(axis=1),2)\n",
    "# Minimum of total recharge number\n",
    "data['min_total_rech_num'] = data[cols].min(axis=1)\n",
    "# Maximum of total recharge number\n",
    "data['max_total_rech_num'] = data[cols].max(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total_rech_amt columns\n",
    "cols = ['total_rech_amt_6','total_rech_amt_7','total_rech_amt_8']\n",
    "data['mean_total_rech_amt'] = round(data[cols].mean(axis=1),2)\n",
    "data['min_total_rech_amt'] = data[cols].min(axis=1)\n",
    "data['max_total_rech_amt'] = data[cols].max(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "# max_rech_amt columns\n",
    "cols = ['max_rech_amt_6','max_rech_amt_7','max_rech_amt_8']\n",
    "data['mean_max_rech_amt'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "# last_day_rch_amt columns\n",
    "cols = ['last_day_rch_amt_6','last_day_rch_amt_7','last_day_rch_amt_8']\n",
    "data['mean_last_day_rch_amt'] = round(data[cols].mean(axis=1),2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total_rech_data columns\n",
    "cols = ['total_rech_data_6','total_rech_data_7','total_rech_data_8']\n",
    "data['mean_total_rech_data'] = round(data[cols].mean(axis=1),2)\n",
    "data['min_total_rech_data'] = data[cols].min(axis=1)\n",
    "data['max_total_rech_data'] = data[cols].max(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total_rech_amt_data columns\n",
    "cols = ['total_rech_amt_data_6','total_rech_amt_data_7','total_rech_amt_data_8']\n",
    "data['mean_total_rech_amt_data'] = round(data[cols].mean(axis=1),2)\n",
    "data['min_total_rech_amt_data'] = data[cols].min(axis=1)\n",
    "data['max_total_rech_amt_data'] = data[cols].max(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total_rech_voice columns\n",
    "data['mean_total_rech_voice'] = data['mean_total_rech_num'] - data['mean_total_rech_data']\n",
    "data['min_total_rech_voice'] = data['min_total_rech_num'] - data['min_total_rech_data']\n",
    "data['max_total_rech_voice'] = data['max_total_rech_num'] - data['max_total_rech_data']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "# total_rech_amt_voice columns\n",
    "data['mean_total_rech_amt_voice'] = data['mean_total_rech_amt'] - data['mean_total_rech_amt_data']\n",
    "data['min_total_rech_amt_voice'] = data['min_total_rech_amt'] - data['min_total_rech_amt_data']\n",
    "data['max_total_rech_amt_voice'] = data['max_total_rech_amt'] - data['max_total_rech_amt_data']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "# max_rech_data columns\n",
    "cols = ['max_rech_data_6','max_rech_data_7','max_rech_data_8']\n",
    "data['mean_max_rech_data'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "# count_rech_2g columns\n",
    "cols = ['count_rech_2g_6','count_rech_2g_7','count_rech_2g_8']\n",
    "data['mean_count_rech_2g'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "# count_rech_3g columns\n",
    "cols = ['count_rech_3g_6','count_rech_3g_7','count_rech_3g_8']\n",
    "data['mean_count_rech_3g'] = round(data[cols].mean(axis=1),2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "#get recharge num weightage for the last three months\n",
    "cols = ['total_rech_num_6','total_rech_num_7','total_rech_num_8']\n",
    "df = data[cols].astype(bool)\n",
    "data['rech_num_weightage'] = ( df['total_rech_num_6'] * 1 ) + ( df['total_rech_num_7'] * 10 ) + ( df['total_rech_num_8'] * 100 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "#get recharge amount weightage for the last three months\n",
    "cols = ['total_rech_amt_6','total_rech_amt_7','total_rech_amt_8']\n",
    "df = data[cols].astype(bool)\n",
    "data['rech_amt_weightage'] = ( df['total_rech_amt_6'] * 1 ) + ( df['total_rech_amt_7'] * 10 ) + ( df['total_rech_amt_8'] * 100 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "# arpu columns\n",
    "# ARPU = Total Revenue / Average Subscribers\n",
    "cols = ['arpu_6','arpu_7','arpu_8']\n",
    "data['mean_arpu'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['arpu_2g_6','arpu_2g_7','arpu_2g_8']\n",
    "data['mean_arpu_2g_data'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['arpu_3g_6','arpu_3g_7','arpu_3g_8']\n",
    "data['mean_arpu_3g_data'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['vol_2g_mb_6','vol_2g_mb_7','vol_2g_mb_8']\n",
    "data['mean_vol_2g_mb_data'] = round(data[cols].mean(axis=1),2)\n",
    "\n",
    "cols = ['vol_3g_mb_6','vol_3g_mb_7','vol_3g_mb_8']\n",
    "data['mean_vol_3g_mb_data'] = round(data[cols].mean(axis=1),2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "#get night_pck_user weightage for the last three months\n",
    "cols = ['night_pck_user_6','night_pck_user_7','night_pck_user_8']\n",
    "data['night_pck_weightage'] = ( data['night_pck_user_6'] * 1 ) + ( data['night_pck_user_7'] * 10 ) + ( data['night_pck_user_8'] * 100 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "#get fb_user weightage for the last three months\n",
    "cols = ['fb_user_6','fb_user_7','fb_user_8']\n",
    "data['fb_user_weightage'] = ( data['fb_user_6'] * 1 ) + ( data['fb_user_7'] * 10 ) + ( data['fb_user_8'] * 100 )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "#get vbc mean for the last three months\n",
    "cols = ['jun_vbc_3g','jul_vbc_3g','aug_vbc_3g']\n",
    "data['mean_vbc_3g'] = round(data[cols].mean(axis=1),2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#get monthly pack weightage for the last three months\n",
    "data['monthly_2g_weightage'] = ( data['monthly_2g_6'] * 1 ) + ( data['monthly_2g_7'] * 10 ) + ( data['monthly_2g_8'] * 100 )\n",
    "data['monthly_3g_weightage'] = ( data['monthly_3g_6'] * 1 ) + ( data['monthly_3g_7'] * 10 ) + ( data['monthly_3g_8'] * 100 )\n",
    "\n",
    "#get sachet pack weightage for the last three months\n",
    "data['sachet_2g_weightage'] = ( data['sachet_2g_6'] * 1 ) + ( data['sachet_2g_7'] * 10 ) + ( data['sachet_2g_8'] * 100 )\n",
    "data['sachet_3g_weightage'] = ( data['sachet_3g_6'] * 1 ) + ( data['sachet_3g_7'] * 10 ) + ( data['sachet_3g_8'] * 100 )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model 1 - Customer Churn Prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First let us take a backup of the dataset, so we can use it later for Model 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "master_df = data.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "churn = data['churn']\n",
    "data = data.drop('churn', axis=1)\n",
    "\n",
    "#split the columns into category and numerical\n",
    "cat_cols = ['night_pck_user_6','monthly_2g_6','sachet_2g_6','monthly_3g_6','sachet_3g_6','fb_user_6',\n",
    "            'night_pck_user_7','monthly_2g_7','sachet_2g_7','monthly_3g_7','sachet_3g_7','fb_user_7',\n",
    "            'night_pck_user_8','monthly_2g_8','sachet_2g_8','monthly_3g_8','sachet_3g_8','fb_user_8'] \n",
    "\n",
    "num_cols = list(set(data.columns).difference(set(cat_cols)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### OneHot encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "# dummy encode the categorical columns\n",
    "data = pd.concat([data,pd.get_dummies(data[cat_cols], drop_first=True)], axis=1)\n",
    "\n",
    "data.drop(cat_cols, axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "# log transform with constant 10000 for real numbers\n",
    "data[num_cols] = np.log((10000 + data[num_cols]))\n",
    "data = np.log((10000 + data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Standardize the numeric values\n",
    "data[num_cols] = (( data[num_cols] - data[num_cols].mean() ) / data[num_cols].std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "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>index</th>\n",
       "      <th>missing_cnt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [index, missing_cnt]\n",
       "Index: []"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Missing values count\n",
    "#data = data.replace([np.inf, -np.inf], np.nan)\n",
    "df = data.isnull().sum().reset_index(name='missing_cnt')\n",
    "df.loc[df['missing_cnt']>0].sort_values('missing_cnt', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data imbalance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class Balance count :  Counter({1: 27629, 0: 27390})\n"
     ]
    }
   ],
   "source": [
    "X = data\n",
    "Y = churn\n",
    "\n",
    "# smote = SMOTE(random_state=RANDOM_STATE, ratio=1.0)\n",
    "# X, Y = smote.fit_sample(X, Y)\n",
    "\n",
    "adasyn = ADASYN(random_state=RANDOM_STATE)\n",
    "X, Y = adasyn.fit_sample(X, Y)\n",
    "\n",
    "print('Class Balance count : ',Counter(Y))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Split into train test "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(38513, 225) (38513,)\n",
      "(16506, 225) (16506,)\n"
     ]
    }
   ],
   "source": [
    "X_train, X_test, Y_train, Y_test = train_test_split(X,Y, test_size=0.3, random_state=RANDOM_STATE)\n",
    "print(X_train.shape, Y_train.shape)\n",
    "print(X_test.shape, Y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Baseline Model Building"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_score_list = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate_model(actual, pred):\n",
    "    # Accuracy Score\n",
    "    acc_score = round(accuracy_score(actual, pred)*100,2)\n",
    "    print('Accuracy Score : ',acc_score)\n",
    "    # ROC AUC score\n",
    "    roc_score = round(roc_auc_score(actual, pred)*100,2)\n",
    "    print('ROC AUC score : ',roc_score)\n",
    "    # Precision score\n",
    "    prec_score = round(precision_score(actual, pred)*100,2)\n",
    "    print('Precision score : ', prec_score)\n",
    "    # Recall score\n",
    "    rec_score = round(recall_score(actual, pred)*100,2)\n",
    "    print('Recall score : ', rec_score)\n",
    "\n",
    "    return acc_score, roc_score, prec_score, rec_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  83.18\n",
      "ROC AUC score :  83.18\n",
      "Precision score :  82.8\n",
      "Recall score :  83.84\n"
     ]
    }
   ],
   "source": [
    "model = LogisticRegression(class_weight='balanced', random_state=RANDOM_STATE,max_iter=10000)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'LogisticRegression', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RidgeClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  82.67\n",
      "ROC AUC score :  82.67\n",
      "Precision score :  82.48\n",
      "Recall score :  83.03\n"
     ]
    }
   ],
   "source": [
    "model = RidgeClassifier(class_weight='balanced', random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'RidgeClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SGDClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  80.2\n",
      "ROC AUC score :  80.2\n",
      "Precision score :  80.07\n",
      "Recall score :  80.52\n"
     ]
    }
   ],
   "source": [
    "model = SGDClassifier(class_weight='balanced', random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'SGDClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  87.74\n",
      "ROC AUC score :  87.72\n",
      "Precision score :  80.35\n",
      "Recall score :  99.98\n"
     ]
    }
   ],
   "source": [
    "model = KNeighborsClassifier()\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'KNeighborsClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### LinearSVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  82.89\n",
      "ROC AUC score :  82.89\n",
      "Precision score :  82.1\n",
      "Recall score :  84.2\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/nikhilkashyap/miniconda3/lib/python3.7/site-packages/sklearn/svm/_base.py:977: ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.\n",
      "  \"the number of iterations.\", ConvergenceWarning)\n"
     ]
    }
   ],
   "source": [
    "model = LinearSVC(class_weight='balanced', random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'LinearSVC', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  92.09\n",
      "ROC AUC score :  92.09\n",
      "Precision score :  89.6\n",
      "Recall score :  95.27\n"
     ]
    }
   ],
   "source": [
    "model = SVC(kernel='rbf', class_weight='balanced', random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'SVC_Kernel_rbf', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decision Tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  92.62\n",
      "ROC AUC score :  92.62\n",
      "Precision score :  91.47\n",
      "Recall score :  94.04\n"
     ]
    }
   ],
   "source": [
    "model = DecisionTreeClassifier(class_weight='balanced', random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'DecisionTreeClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### RandomForest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  96.79\n",
      "ROC AUC score :  96.79\n",
      "Precision score :  95.71\n",
      "Recall score :  97.98\n"
     ]
    }
   ],
   "source": [
    "model = RandomForestClassifier(class_weight='balanced', random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'RandomForestClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ExtraTrees"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  97.77\n",
      "ROC AUC score :  97.77\n",
      "Precision score :  96.38\n",
      "Recall score :  99.27\n"
     ]
    }
   ],
   "source": [
    "model = ExtraTreesClassifier(class_weight='balanced', random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'ExtraTreesClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ADABoost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  91.55\n",
      "ROC AUC score :  91.55\n",
      "Precision score :  91.33\n",
      "Recall score :  91.85\n"
     ]
    }
   ],
   "source": [
    "model = AdaBoostClassifier(random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'AdaBoostClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### GradientBoosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  94.44\n",
      "ROC AUC score :  94.44\n",
      "Precision score :  93.7\n",
      "Recall score :  95.32\n"
     ]
    }
   ],
   "source": [
    "model = GradientBoostingClassifier(random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'GradientBoostingClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### XGBoosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy Score :  97.15\n",
      "ROC AUC score :  97.15\n",
      "Precision score :  96.96\n",
      "Recall score :  97.35\n"
     ]
    }
   ],
   "source": [
    "model = XGBClassifier(random_state=RANDOM_STATE)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)\n",
    "model_score_list.append({'model_name':'XGBClassifier', 'acc_score':acc_score, 'roc_score':roc_score, 'precision_score':prec_score, 'recall_score':rec_score})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Score chart"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "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>model_name</th>\n",
       "      <th>acc_score</th>\n",
       "      <th>roc_score</th>\n",
       "      <th>precision_score</th>\n",
       "      <th>recall_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>KNeighborsClassifier</td>\n",
       "      <td>87.74</td>\n",
       "      <td>87.72</td>\n",
       "      <td>80.35</td>\n",
       "      <td>99.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>ExtraTreesClassifier</td>\n",
       "      <td>97.77</td>\n",
       "      <td>97.77</td>\n",
       "      <td>96.38</td>\n",
       "      <td>99.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>RandomForestClassifier</td>\n",
       "      <td>96.79</td>\n",
       "      <td>96.79</td>\n",
       "      <td>95.71</td>\n",
       "      <td>97.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>XGBClassifier</td>\n",
       "      <td>97.15</td>\n",
       "      <td>97.15</td>\n",
       "      <td>96.96</td>\n",
       "      <td>97.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>GradientBoostingClassifier</td>\n",
       "      <td>94.44</td>\n",
       "      <td>94.44</td>\n",
       "      <td>93.70</td>\n",
       "      <td>95.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>SVC_Kernel_rbf</td>\n",
       "      <td>92.09</td>\n",
       "      <td>92.09</td>\n",
       "      <td>89.60</td>\n",
       "      <td>95.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>DecisionTreeClassifier</td>\n",
       "      <td>92.62</td>\n",
       "      <td>92.62</td>\n",
       "      <td>91.47</td>\n",
       "      <td>94.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>AdaBoostClassifier</td>\n",
       "      <td>91.55</td>\n",
       "      <td>91.55</td>\n",
       "      <td>91.33</td>\n",
       "      <td>91.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>LinearSVC</td>\n",
       "      <td>82.89</td>\n",
       "      <td>82.89</td>\n",
       "      <td>82.10</td>\n",
       "      <td>84.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>LogisticRegression</td>\n",
       "      <td>83.18</td>\n",
       "      <td>83.18</td>\n",
       "      <td>82.80</td>\n",
       "      <td>83.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>LogisticRegression</td>\n",
       "      <td>83.10</td>\n",
       "      <td>83.10</td>\n",
       "      <td>82.87</td>\n",
       "      <td>83.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>RidgeClassifier</td>\n",
       "      <td>82.67</td>\n",
       "      <td>82.67</td>\n",
       "      <td>82.48</td>\n",
       "      <td>83.03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SGDClassifier</td>\n",
       "      <td>80.20</td>\n",
       "      <td>80.20</td>\n",
       "      <td>80.07</td>\n",
       "      <td>80.52</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    model_name  acc_score  roc_score  precision_score  \\\n",
       "4         KNeighborsClassifier      87.74      87.72            80.35   \n",
       "9         ExtraTreesClassifier      97.77      97.77            96.38   \n",
       "8       RandomForestClassifier      96.79      96.79            95.71   \n",
       "12               XGBClassifier      97.15      97.15            96.96   \n",
       "11  GradientBoostingClassifier      94.44      94.44            93.70   \n",
       "6               SVC_Kernel_rbf      92.09      92.09            89.60   \n",
       "7       DecisionTreeClassifier      92.62      92.62            91.47   \n",
       "10          AdaBoostClassifier      91.55      91.55            91.33   \n",
       "5                    LinearSVC      82.89      82.89            82.10   \n",
       "1           LogisticRegression      83.18      83.18            82.80   \n",
       "0           LogisticRegression      83.10      83.10            82.87   \n",
       "2              RidgeClassifier      82.67      82.67            82.48   \n",
       "3                SGDClassifier      80.20      80.20            80.07   \n",
       "\n",
       "    recall_score  \n",
       "4          99.98  \n",
       "9          99.27  \n",
       "8          97.98  \n",
       "12         97.35  \n",
       "11         95.32  \n",
       "6          95.27  \n",
       "7          94.04  \n",
       "10         91.85  \n",
       "5          84.20  \n",
       "1          83.84  \n",
       "0          83.53  \n",
       "2          83.03  \n",
       "3          80.52  "
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_score_df = pd.DataFrame(model_score_list,columns=['model_name', 'acc_score', 'roc_score','precision_score','recall_score'])\n",
    "\n",
    "\n",
    "model_score_df.sort_values(['recall_score','roc_score'], ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cross Validation and Model Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "# initialize Stratified KFold with 5 splits\n",
    "n_splits = 5\n",
    "skf = StratifiedKFold(n_splits=n_splits, shuffle = True, random_state = RANDOM_STATE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "KNeighborsClassifier cv_score_mean :  100.0\n",
      "KNeighborsClassifier cv_score_std :  0.01\n"
     ]
    }
   ],
   "source": [
    "# Cross validation for KNeighborsClassifier\n",
    "model = KNeighborsClassifier()\n",
    "knn_cv_scores = cross_val_score(model, X, Y, scoring='recall', cv = skf, n_jobs=-1)\n",
    "print('KNeighborsClassifier cv_score_mean : ', round(knn_cv_scores.mean() * 100, 2))\n",
    "print('KNeighborsClassifier cv_score_std : ', round(knn_cv_scores.std() * 100, 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ExtraTreesClassifier cv_score_mean :  99.43\n",
      "ExtraTreesClassifier cv_score_std :  0.1\n"
     ]
    }
   ],
   "source": [
    "# Cross validation for ExtraTreesClassifier\n",
    "model = ExtraTreesClassifier(random_state=RANDOM_STATE)\n",
    "etc_cv_scores = cross_val_score(model, X, Y, scoring='recall', cv = skf, n_jobs=-1)\n",
    "print('ExtraTreesClassifier cv_score_mean : ', round(etc_cv_scores.mean() * 100, 2))\n",
    "print('ExtraTreesClassifier cv_score_std : ', round(etc_cv_scores.std() * 100, 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForestClassifier cv_score_mean :  98.5\n",
      "RandomForestClassifier cv_score_std :  0.21\n"
     ]
    }
   ],
   "source": [
    "# Cross validation for RandomForestClassifier\n",
    "model = RandomForestClassifier(random_state=RANDOM_STATE)\n",
    "rfc_cv_scores = cross_val_score(model, X, Y, scoring='recall', cv = skf, n_jobs=-1)\n",
    "print('RandomForestClassifier cv_score_mean : ', round(rfc_cv_scores.mean() * 100, 2))\n",
    "print('RandomForestClassifier cv_score_std : ', round(rfc_cv_scores.std() * 100, 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "XGBoost cv_score_mean :  97.27\n",
      "XGBoost cv_score_std :  0.2\n"
     ]
    }
   ],
   "source": [
    "# Cross validation for XGBoost\n",
    "model = XGBClassifier(random_state=RANDOM_STATE)\n",
    "xgb_cv_scores = cross_val_score(model, X, Y, scoring='recall', cv = skf, n_jobs=-1)\n",
    "print('XGBoost cv_score_mean : ', round(xgb_cv_scores.mean() * 100, 2))\n",
    "print('XGBoost cv_score_std : ', round(xgb_cv_scores.std() * 100, 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f9858c54250>"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "# plot the cv scores for the above 4 models\n",
    "cv_scores_df = pd.DataFrame({'index':np.linspace(1,n_splits,n_splits),'knn':knn_cv_scores,'etc':etc_cv_scores,'xgb':xgb_cv_scores,'rfc':rfc_cv_scores})\n",
    "cv_scores_df.plot.line(x='index', y=['knn','etc','xgb','rfc'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hyperparameter Tuning for KNeighborsClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 1 candidates, totalling 5 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:  8.7min remaining: 13.1min\n",
      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed: 11.2min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Best estimator:\n",
      "KNeighborsClassifier(n_jobs=-1, n_neighbors=3)\n",
      "\n",
      " Best score:\n",
      "0.9998552167212829\n",
      "\n",
      " Best parameters:\n",
      "{'p': 2, 'n_neighbors': 3}\n"
     ]
    }
   ],
   "source": [
    "n_splits = 5\n",
    "\n",
    "n_neighbors =  [int(x) for x in np.linspace(start = 1, stop = 3, num = 3)]\n",
    "\n",
    "p = [int(x) for x in np.linspace(start = 1, stop = 3, num = 3)]\n",
    "\n",
    "\n",
    "params = {\n",
    "            'n_neighbors': n_neighbors,\n",
    "            'p': p\n",
    "         }\n",
    "\n",
    "\n",
    "model = KNeighborsClassifier(weights='uniform', n_jobs=-1)\n",
    "\n",
    "skf = StratifiedKFold(n_splits=n_splits, shuffle = True, random_state = RANDOM_STATE)\n",
    "\n",
    "\n",
    "#grid = GridSearchCV(estimator=model, param_grid=params, scoring='recall', n_jobs=-1, cv=skf.split(X,Y), verbose=3 )\n",
    "grid = RandomizedSearchCV(estimator=model, param_distributions=params, n_iter = 1, scoring='recall', n_jobs=-1, cv=skf.split(X,Y), verbose=3, random_state=RANDOM_STATE )\n",
    "\n",
    "\n",
    "grid.fit(X, Y)\n",
    "\n",
    "\n",
    "print('\\n Best estimator:')\n",
    "print(grid.best_estimator_)\n",
    "print('\\n Best score:')\n",
    "print(grid.best_score_ * 2 - 1)\n",
    "print('\\n Best parameters:')\n",
    "print(grid.best_params_)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "# Fitting the Best Estimator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = grid.best_estimator_\n",
    "\n",
    "\n",
    "model.fit(X_train, Y_train)\n",
    "\n",
    "\n",
    "Y_pred = model.predict(X_test)\n",
    "\n",
    "\n",
    "Y_pred_proba = model.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "def draw_roc( actual, probs ):\n",
    "    fpr, tpr, thresholds = roc_curve( actual, probs, drop_intermediate = False )\n",
    "    auc_score = roc_auc_score( actual, probs )\n",
    "    plt.figure(figsize=(6, 6))\n",
    "    plt.plot( fpr, tpr, label='ROC curve (area = %0.2f)' % auc_score )\n",
    "    plt.plot([0, 1], [0, 1], 'k--')\n",
    "    plt.xlim([0.0, 1.0])\n",
    "    plt.ylim([0.0, 1.05])\n",
    "    plt.xlabel('False Positive Rate or [1 - True Negative Rate]')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.title('ROC Curve')\n",
    "    plt.legend(loc=\"lower right\")\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Classification Report : \n",
      "\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.78      0.88      8238\n",
      "           1       0.82      1.00      0.90      8268\n",
      "\n",
      "    accuracy                           0.89     16506\n",
      "   macro avg       0.91      0.89      0.89     16506\n",
      "weighted avg       0.91      0.89      0.89     16506\n",
      "\n",
      "\n",
      "TN = 6457, FP = 1781, FN = 0, TP = 8268\n",
      "\n",
      "\n",
      "Accuracy Score :  89.21\n",
      "ROC AUC score :  89.19\n",
      "Precision score :  82.28\n",
      "Recall score :  100.0\n"
     ]
    }
   ],
   "source": [
    "print('\\nClassification Report : \\n\\n', classification_report(Y_test, Y_pred))\n",
    "\n",
    "\n",
    "tn, fp, fn, tp = confusion_matrix(Y_test, Y_pred).ravel()\n",
    "print('\\nTN = {0}, FP = {1}, FN = {2}, TP = {3}\\n\\n'.format(tn, fp, fn, tp))\n",
    "\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_roc(Y_test, Y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model 2 - Important features of churn "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature Selection using ExtraTreesClassifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's prepare the dataset for model building"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = master_df.copy()\n",
    "\n",
    "churn = data['churn']\n",
    "data = data.drop('churn', axis=1)\n",
    "\n",
    "X = data\n",
    "Y = churn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "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>index</th>\n",
       "      <th>Feature</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>199</td>\n",
       "      <td>total_ic_weightage</td>\n",
       "      <td>0.079892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>201</td>\n",
       "      <td>total_og_weightage</td>\n",
       "      <td>0.069146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>230</td>\n",
       "      <td>rech_amt_weightage</td>\n",
       "      <td>0.047513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>154</td>\n",
       "      <td>days_since_last_rech</td>\n",
       "      <td>0.013660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>210</td>\n",
       "      <td>min_total_rech_amt</td>\n",
       "      <td>0.012795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>59</td>\n",
       "      <td>loc_ic_t2m_mou_8</td>\n",
       "      <td>0.011501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>205</td>\n",
       "      <td>mean_total_mou_8</td>\n",
       "      <td>0.011105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>95</td>\n",
       "      <td>total_rech_amt_8</td>\n",
       "      <td>0.010763</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>101</td>\n",
       "      <td>last_day_rch_amt_8</td>\n",
       "      <td>0.010092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>65</td>\n",
       "      <td>loc_ic_mou_8</td>\n",
       "      <td>0.010012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>14</td>\n",
       "      <td>roam_og_mou_8</td>\n",
       "      <td>0.009914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2</td>\n",
       "      <td>arpu_8</td>\n",
       "      <td>0.009066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>53</td>\n",
       "      <td>total_og_mou_8</td>\n",
       "      <td>0.008800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>11</td>\n",
       "      <td>roam_ic_mou_8</td>\n",
       "      <td>0.008166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>229</td>\n",
       "      <td>rech_num_weightage</td>\n",
       "      <td>0.007810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>80</td>\n",
       "      <td>total_ic_mou_8</td>\n",
       "      <td>0.007423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>224</td>\n",
       "      <td>min_total_rech_amt_voice</td>\n",
       "      <td>0.007362</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>153</td>\n",
       "      <td>total_rech_amt_voice_8</td>\n",
       "      <td>0.006553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>98</td>\n",
       "      <td>max_rech_amt_8</td>\n",
       "      <td>0.006512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>20</td>\n",
       "      <td>loc_og_t2m_mou_8</td>\n",
       "      <td>0.006294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>207</td>\n",
       "      <td>min_total_rech_num</td>\n",
       "      <td>0.005882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>8</td>\n",
       "      <td>offnet_mou_8</td>\n",
       "      <td>0.005872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>5</td>\n",
       "      <td>onnet_mou_8</td>\n",
       "      <td>0.005387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>29</td>\n",
       "      <td>loc_og_mou_8</td>\n",
       "      <td>0.005350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>221</td>\n",
       "      <td>min_total_rech_voice</td>\n",
       "      <td>0.005240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>56</td>\n",
       "      <td>loc_ic_t2t_mou_8</td>\n",
       "      <td>0.005193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>169</td>\n",
       "      <td>mean_loc_t2m_mou</td>\n",
       "      <td>0.005168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>92</td>\n",
       "      <td>total_rech_num_8</td>\n",
       "      <td>0.005153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>131</td>\n",
       "      <td>night_pck_user_8</td>\n",
       "      <td>0.005028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>162</td>\n",
       "      <td>mean_roam_og_mou</td>\n",
       "      <td>0.005006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>147</td>\n",
       "      <td>aon</td>\n",
       "      <td>0.004969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>41</td>\n",
       "      <td>std_og_mou_8</td>\n",
       "      <td>0.004898</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>163</td>\n",
       "      <td>mean_roam_mou</td>\n",
       "      <td>0.004693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>40</td>\n",
       "      <td>std_og_mou_7</td>\n",
       "      <td>0.004692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>91</td>\n",
       "      <td>total_rech_num_7</td>\n",
       "      <td>0.004691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>212</td>\n",
       "      <td>mean_max_rech_amt</td>\n",
       "      <td>0.004668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>39</td>\n",
       "      <td>std_og_mou_6</td>\n",
       "      <td>0.004657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>90</td>\n",
       "      <td>total_rech_num_6</td>\n",
       "      <td>0.004637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>1</td>\n",
       "      <td>arpu_7</td>\n",
       "      <td>0.004611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>204</td>\n",
       "      <td>mean_total_mou_7</td>\n",
       "      <td>0.004558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>184</td>\n",
       "      <td>mean_loc_og_mou</td>\n",
       "      <td>0.004507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>213</td>\n",
       "      <td>mean_last_day_rch_amt</td>\n",
       "      <td>0.004490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>52</td>\n",
       "      <td>total_og_mou_7</td>\n",
       "      <td>0.004455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>161</td>\n",
       "      <td>mean_roam_ic_mou</td>\n",
       "      <td>0.004436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>55</td>\n",
       "      <td>loc_ic_t2t_mou_7</td>\n",
       "      <td>0.004429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>183</td>\n",
       "      <td>mean_loc_ic_mou</td>\n",
       "      <td>0.004427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>236</td>\n",
       "      <td>night_pck_weightage</td>\n",
       "      <td>0.004402</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>0</td>\n",
       "      <td>arpu_6</td>\n",
       "      <td>0.004376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>51</td>\n",
       "      <td>total_og_mou_6</td>\n",
       "      <td>0.004354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>19</td>\n",
       "      <td>loc_og_t2m_mou_7</td>\n",
       "      <td>0.004346</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    index                   Feature     Score\n",
       "0     199        total_ic_weightage  0.079892\n",
       "1     201        total_og_weightage  0.069146\n",
       "2     230        rech_amt_weightage  0.047513\n",
       "3     154      days_since_last_rech  0.013660\n",
       "4     210        min_total_rech_amt  0.012795\n",
       "5      59          loc_ic_t2m_mou_8  0.011501\n",
       "6     205          mean_total_mou_8  0.011105\n",
       "7      95          total_rech_amt_8  0.010763\n",
       "8     101        last_day_rch_amt_8  0.010092\n",
       "9      65              loc_ic_mou_8  0.010012\n",
       "10     14             roam_og_mou_8  0.009914\n",
       "11      2                    arpu_8  0.009066\n",
       "12     53            total_og_mou_8  0.008800\n",
       "13     11             roam_ic_mou_8  0.008166\n",
       "14    229        rech_num_weightage  0.007810\n",
       "15     80            total_ic_mou_8  0.007423\n",
       "16    224  min_total_rech_amt_voice  0.007362\n",
       "17    153    total_rech_amt_voice_8  0.006553\n",
       "18     98            max_rech_amt_8  0.006512\n",
       "19     20          loc_og_t2m_mou_8  0.006294\n",
       "20    207        min_total_rech_num  0.005882\n",
       "21      8              offnet_mou_8  0.005872\n",
       "22      5               onnet_mou_8  0.005387\n",
       "23     29              loc_og_mou_8  0.005350\n",
       "24    221      min_total_rech_voice  0.005240\n",
       "25     56          loc_ic_t2t_mou_8  0.005193\n",
       "26    169          mean_loc_t2m_mou  0.005168\n",
       "27     92          total_rech_num_8  0.005153\n",
       "28    131          night_pck_user_8  0.005028\n",
       "29    162          mean_roam_og_mou  0.005006\n",
       "30    147                       aon  0.004969\n",
       "31     41              std_og_mou_8  0.004898\n",
       "32    163             mean_roam_mou  0.004693\n",
       "33     40              std_og_mou_7  0.004692\n",
       "34     91          total_rech_num_7  0.004691\n",
       "35    212         mean_max_rech_amt  0.004668\n",
       "36     39              std_og_mou_6  0.004657\n",
       "37     90          total_rech_num_6  0.004637\n",
       "38      1                    arpu_7  0.004611\n",
       "39    204          mean_total_mou_7  0.004558\n",
       "40    184           mean_loc_og_mou  0.004507\n",
       "41    213     mean_last_day_rch_amt  0.004490\n",
       "42     52            total_og_mou_7  0.004455\n",
       "43    161          mean_roam_ic_mou  0.004436\n",
       "44     55          loc_ic_t2t_mou_7  0.004429\n",
       "45    183           mean_loc_ic_mou  0.004427\n",
       "46    236       night_pck_weightage  0.004402\n",
       "47      0                    arpu_6  0.004376\n",
       "48     51            total_og_mou_6  0.004354\n",
       "49     19          loc_og_t2m_mou_7  0.004346"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = ExtraTreesClassifier(n_jobs=-1, random_state=RANDOM_STATE)\n",
    "model.fit(X, Y)\n",
    "\n",
    "\n",
    "feature_imp_df = pd.DataFrame({'Feature':data.columns, 'Score':model.feature_importances_})\n",
    "\n",
    "feature_imp_df = feature_imp_df.sort_values('Score', ascending=False).reset_index()\n",
    "feature_imp_df.head(50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f9872ba8650>"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "(pd.Series(model.feature_importances_, index=data.columns)\n",
    "   .nlargest(50)\n",
    "   .plot(kind='barh', figsize=(18,12))) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/nikhilkashyap/miniconda3/lib/python3.7/site-packages/numpy/core/fromnumeric.py:2542: FutureWarning: Method .ptp is deprecated and will be removed in a future version. Use numpy.ptp instead.\n",
      "  return ptp(axis=axis, out=out, **kwargs)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Generalized Linear Model Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>         <td>churn</td>      <th>  No. Observations:  </th>  <td> 29979</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                  <td>GLM</td>       <th>  Df Residuals:      </th>  <td> 29968</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model Family:</th>        <td>Binomial</td>     <th>  Df Model:          </th>  <td>    10</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Link Function:</th>         <td>logit</td>      <th>  Scale:             </th> <td>  1.0000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>                <td>IRLS</td>       <th>  Log-Likelihood:    </th> <td> -5777.7</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>            <td>Thu, 10 Sep 2020</td> <th>  Deviance:          </th> <td>  11555.</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                <td>23:19:46</td>     <th>  Pearson chi2:      </th> <td>7.01e+04</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>No. Iterations:</th>          <td>8</td>        <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>     <td>nonrobust</td>    <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "            <td></td>              <th>coef</th>     <th>std err</th>      <th>z</th>      <th>P>|z|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>const</th>                <td>   -0.5215</td> <td>    0.130</td> <td>   -3.998</td> <td> 0.000</td> <td>   -0.777</td> <td>   -0.266</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>total_og_weightage</th>   <td>   -0.0206</td> <td>    0.001</td> <td>  -31.588</td> <td> 0.000</td> <td>   -0.022</td> <td>   -0.019</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>max_rech_amt_8</th>       <td>   -0.0015</td> <td>    0.000</td> <td>   -5.155</td> <td> 0.000</td> <td>   -0.002</td> <td>   -0.001</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>days_since_last_rech</th> <td>    0.0401</td> <td>    0.003</td> <td>   14.423</td> <td> 0.000</td> <td>    0.035</td> <td>    0.045</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>loc_og_t2m_mou_8</th>     <td>   -0.0042</td> <td>    0.000</td> <td>  -12.068</td> <td> 0.000</td> <td>   -0.005</td> <td>   -0.003</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>roam_og_mou_8</th>        <td>    0.0037</td> <td>    0.000</td> <td>   10.406</td> <td> 0.000</td> <td>    0.003</td> <td>    0.004</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>total_og_mou_8</th>       <td>-1.588e-05</td> <td> 8.16e-05</td> <td>   -0.195</td> <td> 0.846</td> <td>   -0.000</td> <td>    0.000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>total_rech_num_8</th>     <td>   -0.0159</td> <td>    0.005</td> <td>   -3.145</td> <td> 0.002</td> <td>   -0.026</td> <td>   -0.006</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>arpu_8</th>               <td>   -0.0007</td> <td>    0.000</td> <td>   -4.263</td> <td> 0.000</td> <td>   -0.001</td> <td>   -0.000</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>fb_user_weightage</th>    <td>    0.0056</td> <td>    0.001</td> <td>   10.466</td> <td> 0.000</td> <td>    0.005</td> <td>    0.007</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>aon</th>                  <td>   -0.0004</td> <td> 3.38e-05</td> <td>  -11.033</td> <td> 0.000</td> <td>   -0.000</td> <td>   -0.000</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                 Generalized Linear Model Regression Results                  \n",
       "==============================================================================\n",
       "Dep. Variable:                  churn   No. Observations:                29979\n",
       "Model:                            GLM   Df Residuals:                    29968\n",
       "Model Family:                Binomial   Df Model:                           10\n",
       "Link Function:                  logit   Scale:                          1.0000\n",
       "Method:                          IRLS   Log-Likelihood:                -5777.7\n",
       "Date:                Thu, 10 Sep 2020   Deviance:                       11555.\n",
       "Time:                        23:19:46   Pearson chi2:                 7.01e+04\n",
       "No. Iterations:                     8                                         \n",
       "Covariance Type:            nonrobust                                         \n",
       "========================================================================================\n",
       "                           coef    std err          z      P>|z|      [0.025      0.975]\n",
       "----------------------------------------------------------------------------------------\n",
       "const                   -0.5215      0.130     -3.998      0.000      -0.777      -0.266\n",
       "total_og_weightage      -0.0206      0.001    -31.588      0.000      -0.022      -0.019\n",
       "max_rech_amt_8          -0.0015      0.000     -5.155      0.000      -0.002      -0.001\n",
       "days_since_last_rech     0.0401      0.003     14.423      0.000       0.035       0.045\n",
       "loc_og_t2m_mou_8        -0.0042      0.000    -12.068      0.000      -0.005      -0.003\n",
       "roam_og_mou_8            0.0037      0.000     10.406      0.000       0.003       0.004\n",
       "total_og_mou_8       -1.588e-05   8.16e-05     -0.195      0.846      -0.000       0.000\n",
       "total_rech_num_8        -0.0159      0.005     -3.145      0.002      -0.026      -0.006\n",
       "arpu_8                  -0.0007      0.000     -4.263      0.000      -0.001      -0.000\n",
       "fb_user_weightage        0.0056      0.001     10.466      0.000       0.005       0.007\n",
       "aon                     -0.0004   3.38e-05    -11.033      0.000      -0.000      -0.000\n",
       "========================================================================================\n",
       "\"\"\""
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = ['total_og_weightage','max_rech_amt_8','days_since_last_rech','loc_og_t2m_mou_8','roam_og_mou_8',\n",
    "        'total_og_mou_8','total_rech_num_8','arpu_8','fb_user_weightage','aon']\n",
    "\n",
    "X = data[cols]\n",
    "Y = churn\n",
    "\n",
    "\n",
    "logm = sm.GLM(Y,(sm.add_constant(X)), family = sm.families.Binomial())\n",
    "modres = logm.fit()\n",
    "\n",
    "logm.fit().summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = ['total_og_weightage','max_rech_amt_8','days_since_last_rech','loc_og_t2m_mou_8','roam_og_mou_8',\n",
    "        'total_rech_num_8','arpu_8','fb_user_weightage','aon']\n",
    "data = data[cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Standardize the numeric values\n",
    "data = (( data - data.mean() ) / data.std())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ADASYN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class Balance count :  Counter({1: 28057, 0: 27390})\n"
     ]
    }
   ],
   "source": [
    "X = data\n",
    "Y = churn\n",
    "\n",
    "\n",
    "#smote = SMOTE(random_state=RANDOM_STATE, ratio=1.0)\n",
    "#X, Y = smote.fit_sample(X, Y)\n",
    "\n",
    "adasyn = ADASYN(random_state=RANDOM_STATE)\n",
    "X, Y = adasyn.fit_sample(X, Y)\n",
    "\n",
    "print('Class Balance count : ',Counter(Y))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(38812, 9) (38812,)\n",
      "(16635, 9) (16635,)\n"
     ]
    }
   ],
   "source": [
    "#Split into train test \n",
    "X_train, X_test, Y_train, Y_test = train_test_split(X,Y, test_size=0.3, random_state=RANDOM_STATE)\n",
    "print(X_train.shape, Y_train.shape)\n",
    "print(X_test.shape, Y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Y_train : Counter({1: 19648, 0: 19164})\n",
      "Y_test : Counter({1: 8409, 0: 8226})\n"
     ]
    }
   ],
   "source": [
    "print('Y_train :', Counter(Y_train))\n",
    "print('Y_test :', Counter(Y_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Building"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ExtraTrees, XGBoost, RandomForest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = ExtraTreesClassifier(bootstrap=False, class_weight='balanced', criterion='gini', max_depth=60, max_features='sqrt',\n",
    "            max_leaf_nodes=None, min_impurity_decrease=0.0, min_impurity_split=None, min_samples_leaf=1, min_samples_split=2, \n",
    "            min_weight_fraction_leaf=0.0, n_estimators=600, n_jobs=-1, oob_score=False, random_state=RANDOM_STATE, verbose=0, warm_start=False)\n",
    "model.fit(X_train, Y_train)\n",
    "Y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [],
   "source": [
    "# param_dist = {'objective':'binary:logistic','n_estimators':600, 'n_jobs':-1,'max_depth':60,'random_state':42}\n",
    "model_xgb = XGBClassifier(random_state=RANDOM_STATE)\n",
    "model_xgb.fit(X_train, Y_train)\n",
    "Y_pred_xgb = model_xgb.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_rfc = RandomForestClassifier(random_state=RANDOM_STATE)\n",
    "model_rfc.fit(X_train, Y_train)\n",
    "Y_pred_rfc = model_rfc.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Classification Report : \n",
      "\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.95      0.97      8226\n",
      "           1       0.95      0.99      0.97      8409\n",
      "\n",
      "    accuracy                           0.97     16635\n",
      "   macro avg       0.97      0.97      0.97     16635\n",
      "weighted avg       0.97      0.97      0.97     16635\n",
      "\n",
      "\n",
      "TN = 7777, FP = 449, FN = 77, TP = 8332\n",
      "\n",
      "\n",
      "Accuracy Score :  96.84\n",
      "ROC AUC score :  96.81\n",
      "Precision score :  94.89\n",
      "Recall score :  99.08\n"
     ]
    }
   ],
   "source": [
    "print('\\nClassification Report : \\n\\n', classification_report(Y_test, Y_pred))\n",
    "\n",
    "tn, fp, fn, tp = confusion_matrix(Y_test, Y_pred).ravel()\n",
    "print('\\nTN = {0}, FP = {1}, FN = {2}, TP = {3}\\n\\n'.format(tn, fp, fn, tp))\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Classification Report : \n",
      "\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.96      0.96      0.96      8226\n",
      "           1       0.96      0.96      0.96      8409\n",
      "\n",
      "    accuracy                           0.96     16635\n",
      "   macro avg       0.96      0.96      0.96     16635\n",
      "weighted avg       0.96      0.96      0.96     16635\n",
      "\n",
      "\n",
      "TN = 7916, FP = 310, FN = 303, TP = 8106\n",
      "\n",
      "\n",
      "Accuracy Score :  96.31\n",
      "ROC AUC score :  96.31\n",
      "Precision score :  96.32\n",
      "Recall score :  96.4\n"
     ]
    }
   ],
   "source": [
    "print('\\nClassification Report : \\n\\n', classification_report(Y_test, Y_pred_xgb))\n",
    "\n",
    "tn, fp, fn, tp = confusion_matrix(Y_test, Y_pred_xgb).ravel()\n",
    "print('\\nTN = {0}, FP = {1}, FN = {2}, TP = {3}\\n\\n'.format(tn, fp, fn, tp))\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred_xgb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Classification Report : \n",
      "\n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.97      0.94      0.96      8226\n",
      "           1       0.94      0.98      0.96      8409\n",
      "\n",
      "    accuracy                           0.96     16635\n",
      "   macro avg       0.96      0.96      0.96     16635\n",
      "weighted avg       0.96      0.96      0.96     16635\n",
      "\n",
      "\n",
      "TN = 7742, FP = 484, FN = 205, TP = 8204\n",
      "\n",
      "\n",
      "Accuracy Score :  95.86\n",
      "ROC AUC score :  95.84\n",
      "Precision score :  94.43\n",
      "Recall score :  97.56\n"
     ]
    }
   ],
   "source": [
    "print('\\nClassification Report : \\n\\n', classification_report(Y_test, Y_pred_rfc))\n",
    "\n",
    "tn, fp, fn, tp = confusion_matrix(Y_test, Y_pred_rfc).ravel()\n",
    "print('\\nTN = {0}, FP = {1}, FN = {2}, TP = {3}\\n\\n'.format(tn, fp, fn, tp))\n",
    "\n",
    "acc_score, roc_score, prec_score, rec_score = evaluate_model(Y_test, Y_pred_rfc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_roc(Y_test, Y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_roc(Y_test, Y_pred_xgb)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "draw_roc(Y_test, Y_pred_rfc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}