Prediction.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import pyplot\n",
    "import seaborn as sns\n",
    "\n",
    "%matplotlib inline\n",
    "%run ../Veon_Tasks/functions_book.ipynb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"final.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop('Unnamed: 0', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "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>პოლისის ნომერი (hash)</th>\n",
       "      <th>წინა პოლისის ნომერი (hash)</th>\n",
       "      <th>დამზღ პირადი ნომერი (hash)</th>\n",
       "      <th>სედანი/ჯიპი</th>\n",
       "      <th>Client Category (hash)</th>\n",
       "      <th>პრემია</th>\n",
       "      <th>გაუქმებული პრემია</th>\n",
       "      <th>სტატუსი</th>\n",
       "      <th>ვადაგადაცილებული დღე</th>\n",
       "      <th>დამზღვევის ასაკი</th>\n",
       "      <th>დამზღვევი ქალია</th>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th>გასული_პერიოდი_გაუქმებამდე</th>\n",
       "      <th>Motor TPL Package</th>\n",
       "      <th>MotorPackage</th>\n",
       "      <th>PropertyPackage</th>\n",
       "      <th>TravelPackage</th>\n",
       "      <th>1990.0</th>\n",
       "      <th>2000.0</th>\n",
       "      <th>2010.0</th>\n",
       "      <th>თანამშრომლი ქალია</th>\n",
       "      <th>პასუხისმგებლის ასაკი</th>\n",
       "      <th>დამზღ არის მესაკუთრე</th>\n",
       "      <th>ჯგუფი 1</th>\n",
       "      <th>ჯგუფი 2</th>\n",
       "      <th>ჯგუფი 3</th>\n",
       "      <th>მოთხოვნილი</th>\n",
       "      <th>ანაზღაურებული</th>\n",
       "      <th>დაკმაყოფილებული</th>\n",
       "      <th>დაუკმაყოფილებელი</th>\n",
       "      <th>ნაკლებად_დაკმაყოფილებული</th>\n",
       "      <th>ზედმეტად_დაკმაყოფილებული</th>\n",
       "      <th>no_status</th>\n",
       "      <th>დასრულებულია</th>\n",
       "      <th>პროცესშია</th>\n",
       "      <th>გარეცხვის რაოდენობა</th>\n",
       "      <th>in_abandon</th>\n",
       "      <th>out_abandon</th>\n",
       "      <th>in_answered</th>\n",
       "      <th>out_answered</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d...</td>\n",
       "      <td>0</td>\n",
       "      <td>5c05f5ab96ae3736b288a7121c222b86a82c221e782b4d...</td>\n",
       "      <td>ჰეტჩბეკი</td>\n",
       "      <td>0</td>\n",
       "      <td>1682.57</td>\n",
       "      <td>-1397.54</td>\n",
       "      <td>Cancelled</td>\n",
       "      <td>0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>61</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>200</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4...</td>\n",
       "      <td>0</td>\n",
       "      <td>70720759877097d63afb1e59d5f535413076b92a48e616...</td>\n",
       "      <td>ჰეტჩბეკი</td>\n",
       "      <td>0</td>\n",
       "      <td>670.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>194</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>634cc57a6f4ba919e81c0f01b7530de1f660b67183f597...</td>\n",
       "      <td>0</td>\n",
       "      <td>21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff3...</td>\n",
       "      <td>სედანი</td>\n",
       "      <td>0</td>\n",
       "      <td>1863.34</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Active</td>\n",
       "      <td>11</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>194</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>450</td>\n",
       "      <td>450.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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",
       "      <td>3</td>\n",
       "      <td>6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb...</td>\n",
       "      <td>0</td>\n",
       "      <td>ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a198...</td>\n",
       "      <td>სედანი</td>\n",
       "      <td>0</td>\n",
       "      <td>1289.04</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>194</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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",
       "      <td>4</td>\n",
       "      <td>e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3...</td>\n",
       "      <td>0</td>\n",
       "      <td>be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f...</td>\n",
       "      <td>მაღალი გამავლობის</td>\n",
       "      <td>0</td>\n",
       "      <td>1351.61</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>194</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>23</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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": [
       "                               პოლისის ნომერი (hash)  \\\n",
       "0  a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d...   \n",
       "1  1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4...   \n",
       "2  634cc57a6f4ba919e81c0f01b7530de1f660b67183f597...   \n",
       "3  6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb...   \n",
       "4  e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3...   \n",
       "\n",
       "   წინა პოლისის ნომერი (hash)  \\\n",
       "0                           0   \n",
       "1                           0   \n",
       "2                           0   \n",
       "3                           0   \n",
       "4                           0   \n",
       "\n",
       "                          დამზღ პირადი ნომერი (hash)        სედანი/ჯიპი  \\\n",
       "0  5c05f5ab96ae3736b288a7121c222b86a82c221e782b4d...           ჰეტჩბეკი   \n",
       "1  70720759877097d63afb1e59d5f535413076b92a48e616...           ჰეტჩბეკი   \n",
       "2  21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff3...             სედანი   \n",
       "3  ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a198...             სედანი   \n",
       "4  be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f...  მაღალი გამავლობის   \n",
       "\n",
       "   Client Category (hash)   პრემია  გაუქმებული პრემია    სტატუსი  \\\n",
       "0                       0  1682.57           -1397.54  Cancelled   \n",
       "1                       0   670.00               0.00     Active   \n",
       "2                       0  1863.34               0.00     Active   \n",
       "3                       0  1289.04               0.00     Active   \n",
       "4                       0  1351.61               0.00     Active   \n",
       "\n",
       "   ვადაგადაცილებული დღე  დამზღვევის ასაკი  დამზღვევი ქალია  \\\n",
       "0                     0              39.0                1   \n",
       "1                     0              51.0                0   \n",
       "2                    11              25.0                0   \n",
       "3                     0              35.0                0   \n",
       "4                     0              34.0                1   \n",
       "\n",
       "   ბოლომდე არ გადაიხადა  გასული_პერიოდი_გაუქმებამდე  Motor TPL Package  \\\n",
       "0                     1                          61                  0   \n",
       "1                     0                         194                  0   \n",
       "2                     0                         194                  0   \n",
       "3                     0                         194                  0   \n",
       "4                     0                         194                  0   \n",
       "\n",
       "   MotorPackage  PropertyPackage  TravelPackage  1990.0  2000.0  2010.0  \\\n",
       "0             1                0              0       0       0       1   \n",
       "1             1                0              0       1       0       0   \n",
       "2             1                0              0       0       1       0   \n",
       "3             1                0              0       0       0       1   \n",
       "4             1                0              0       0       0       1   \n",
       "\n",
       "   თანამშრომლი ქალია  პასუხისმგებლის ასაკი  დამზღ არის მესაკუთრე  ჯგუფი 1  \\\n",
       "0                  1                    30                     1        1   \n",
       "1                  0                    22                     1        1   \n",
       "2                  1                    28                     0        1   \n",
       "3                  0                    21                     0        1   \n",
       "4                  0                    23                     1        1   \n",
       "\n",
       "   ჯგუფი 2  ჯგუფი 3  მოთხოვნილი  ანაზღაურებული  დაკმაყოფილებული  \\\n",
       "0        0        0         200            0.0                0   \n",
       "1        0        0           0            0.0                0   \n",
       "2        0        0         450          450.0                1   \n",
       "3        0        0           0            0.0                0   \n",
       "4        0        0           0            0.0                0   \n",
       "\n",
       "   დაუკმაყოფილებელი  ნაკლებად_დაკმაყოფილებული  ზედმეტად_დაკმაყოფილებული  \\\n",
       "0                 1                         0                         0   \n",
       "1                 0                         0                         0   \n",
       "2                 0                         0                         0   \n",
       "3                 0                         0                         0   \n",
       "4                 0                         0                         0   \n",
       "\n",
       "   no_status  დასრულებულია  პროცესშია  გარეცხვის რაოდენობა  in_abandon  \\\n",
       "0          0             1          0                  0.0         0.0   \n",
       "1          1             0          0                  0.0         0.0   \n",
       "2          0             1          0                  0.0         0.0   \n",
       "3          1             0          0                  0.0         0.0   \n",
       "4          1             0          0                  0.0         0.0   \n",
       "\n",
       "   out_abandon  in_answered  out_answered  \n",
       "0          0.0          1.0           0.0  \n",
       "1          0.0          1.0           0.0  \n",
       "2          0.0          0.0           0.0  \n",
       "3          0.0          0.0           0.0  \n",
       "4          0.0          0.0           0.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display_all(df.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6400, 40)"
      ]
     },
     "execution_count": 198,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.set_index(['პოლისის ნომერი (hash)', 'დამზღ პირადი ნომერი (hash)'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop(['სედანი/ჯიპი', 'სტატუსი'],axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.drop(['გაუქმებული პრემია', 'გასული_პერიოდი_გაუქმებამდე'],axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "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></th>\n",
       "      <th>წინა პოლისის ნომერი (hash)</th>\n",
       "      <th>Client Category (hash)</th>\n",
       "      <th>პრემია</th>\n",
       "      <th>ვადაგადაცილებული დღე</th>\n",
       "      <th>დამზღვევის ასაკი</th>\n",
       "      <th>დამზღვევი ქალია</th>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th>Motor TPL Package</th>\n",
       "      <th>MotorPackage</th>\n",
       "      <th>PropertyPackage</th>\n",
       "      <th>TravelPackage</th>\n",
       "      <th>1990.0</th>\n",
       "      <th>2000.0</th>\n",
       "      <th>2010.0</th>\n",
       "      <th>თანამშრომლი ქალია</th>\n",
       "      <th>პასუხისმგებლის ასაკი</th>\n",
       "      <th>დამზღ არის მესაკუთრე</th>\n",
       "      <th>ჯგუფი 1</th>\n",
       "      <th>ჯგუფი 2</th>\n",
       "      <th>ჯგუფი 3</th>\n",
       "      <th>მოთხოვნილი</th>\n",
       "      <th>ანაზღაურებული</th>\n",
       "      <th>დაკმაყოფილებული</th>\n",
       "      <th>დაუკმაყოფილებელი</th>\n",
       "      <th>ნაკლებად_დაკმაყოფილებული</th>\n",
       "      <th>ზედმეტად_დაკმაყოფილებული</th>\n",
       "      <th>no_status</th>\n",
       "      <th>დასრულებულია</th>\n",
       "      <th>პროცესშია</th>\n",
       "      <th>გარეცხვის რაოდენობა</th>\n",
       "      <th>in_abandon</th>\n",
       "      <th>out_abandon</th>\n",
       "      <th>in_answered</th>\n",
       "      <th>out_answered</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>პოლისის ნომერი (hash)</th>\n",
       "      <th>დამზღ პირადი ნომერი (hash)</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0b682bb61afa0128b8</td>\n",
       "      <td>5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dda5d8a8738d88adb96</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1682.57</td>\n",
       "      <td>0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>200</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e059b36d34c7b11036</td>\n",
       "      <td>70720759877097d63afb1e59d5f535413076b92a48e616f64363c680560f4028</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>670.00</td>\n",
       "      <td>0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a1736922575791306f</td>\n",
       "      <td>21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff3284c49d6b377906fd8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1863.34</td>\n",
       "      <td>11</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>28</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>450</td>\n",
       "      <td>450.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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",
       "      <td>6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9a05f75666df1571c7</td>\n",
       "      <td>ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a19811ddf65f7c18536cfd</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1289.04</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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",
       "      <td>e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f4e1d9f142b33b8e99</td>\n",
       "      <td>be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4194d5501a91bf3c11</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1351.61</td>\n",
       "      <td>0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>23</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</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": [
       "                                                                                                       წინა პოლისის ნომერი (hash)  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                       \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                           0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                           0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                           0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                           0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                           0   \n",
       "\n",
       "                                                                                                       Client Category (hash)  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                   \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                       0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                       0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                       0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                       0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                       0   \n",
       "\n",
       "                                                                                                        პრემია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                    \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...  1682.57   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...   670.00   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...  1863.34   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...  1289.04   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...  1351.61   \n",
       "\n",
       "                                                                                                       ვადაგადაცილებული დღე  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                 \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                     0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                     0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                    11   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                     0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                     0   \n",
       "\n",
       "                                                                                                       დამზღვევის ასაკი  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                             \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...              39.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...              51.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...              25.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...              35.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...              34.0   \n",
       "\n",
       "                                                                                                       დამზღვევი ქალია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                            \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                1   \n",
       "\n",
       "                                                                                                       ბოლომდე არ გადაიხადა  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                 \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                     1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                     0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                     0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                     0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                     0   \n",
       "\n",
       "                                                                                                       Motor TPL Package  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                              \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                  0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                  0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                  0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                  0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                  0   \n",
       "\n",
       "                                                                                                       MotorPackage  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                         \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...             1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...             1   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...             1   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...             1   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...             1   \n",
       "\n",
       "                                                                                                       PropertyPackage  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                            \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                0   \n",
       "\n",
       "                                                                                                       TravelPackage  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                          \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...              0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...              0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...              0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...              0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...              0   \n",
       "\n",
       "                                                                                                       1990.0  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                   \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...       0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...       1   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...       0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...       0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...       0   \n",
       "\n",
       "                                                                                                       2000.0  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                   \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...       0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...       0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...       1   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...       0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...       0   \n",
       "\n",
       "                                                                                                       2010.0  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                   \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...       1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...       0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...       0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...       1   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...       1   \n",
       "\n",
       "                                                                                                       თანამშრომლი ქალია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                              \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                  1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                  0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                  1   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                  0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                  0   \n",
       "\n",
       "                                                                                                       პასუხისმგებლის ასაკი  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                 \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                    30   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                    22   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                    28   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                    21   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                    23   \n",
       "\n",
       "                                                                                                       დამზღ არის მესაკუთრე  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                 \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                     1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                     1   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                     0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                     0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                     1   \n",
       "\n",
       "                                                                                                       ჯგუფი 1  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                    \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...        1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...        1   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...        1   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...        1   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...        1   \n",
       "\n",
       "                                                                                                       ჯგუფი 2  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                    \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...        0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...        0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...        0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...        0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...        0   \n",
       "\n",
       "                                                                                                       ჯგუფი 3  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                    \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...        0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...        0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...        0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...        0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...        0   \n",
       "\n",
       "                                                                                                       მოთხოვნილი  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                       \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...         200   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...           0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...         450   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...           0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...           0   \n",
       "\n",
       "                                                                                                       ანაზღაურებული  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                          \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...            0.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...            0.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          450.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...            0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...            0.0   \n",
       "\n",
       "                                                                                                       დაკმაყოფილებული  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                            \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                1   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                0   \n",
       "\n",
       "                                                                                                       დაუკმაყოფილებელი  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                             \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                 1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                 0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                 0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                 0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                 0   \n",
       "\n",
       "                                                                                                       ნაკლებად_დაკმაყოფილებული  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                     \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                         0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                         0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                         0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                         0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                         0   \n",
       "\n",
       "                                                                                                       ზედმეტად_დაკმაყოფილებული  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                     \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                         0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                         0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                         0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                         0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                         0   \n",
       "\n",
       "                                                                                                       no_status  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                      \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...          0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...          1   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...          1   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...          1   \n",
       "\n",
       "                                                                                                       დასრულებულია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                         \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...             1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...             0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...             1   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...             0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...             0   \n",
       "\n",
       "                                                                                                       პროცესშია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                      \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...          0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...          0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...          0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...          0   \n",
       "\n",
       "                                                                                                       გარეცხვის რაოდენობა  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                  0.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                  0.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                  0.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                  0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                  0.0   \n",
       "\n",
       "                                                                                                       in_abandon  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                       \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...         0.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...         0.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...         0.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...         0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...         0.0   \n",
       "\n",
       "                                                                                                       out_abandon  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                        \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...          0.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...          0.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          0.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...          0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...          0.0   \n",
       "\n",
       "                                                                                                       in_answered  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                        \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...          1.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...          1.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          0.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...          0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...          0.0   \n",
       "\n",
       "                                                                                                       out_answered  \n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                        \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...           0.0  \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...           0.0  \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...           0.0  \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...           0.0  \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...           0.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display_all(df.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Now let's split the dataset and start preparing the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [],
   "source": [
    "new_df=df.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from imblearn.over_sampling import SMOTE\n",
    "from sklearn.utils import resample\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "import pickle\n",
    "import time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = new_df.drop(labels = [\"ბოლომდე არ გადაიხადა\"], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "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></th>\n",
       "      <th>წინა პოლისის ნომერი (hash)</th>\n",
       "      <th>Client Category (hash)</th>\n",
       "      <th>პრემია</th>\n",
       "      <th>ვადაგადაცილებული დღე</th>\n",
       "      <th>დამზღვევის ასაკი</th>\n",
       "      <th>დამზღვევი ქალია</th>\n",
       "      <th>Motor TPL Package</th>\n",
       "      <th>MotorPackage</th>\n",
       "      <th>PropertyPackage</th>\n",
       "      <th>TravelPackage</th>\n",
       "      <th>...</th>\n",
       "      <th>ნაკლებად_დაკმაყოფილებული</th>\n",
       "      <th>ზედმეტად_დაკმაყოფილებული</th>\n",
       "      <th>no_status</th>\n",
       "      <th>დასრულებულია</th>\n",
       "      <th>პროცესშია</th>\n",
       "      <th>გარეცხვის რაოდენობა</th>\n",
       "      <th>in_abandon</th>\n",
       "      <th>out_abandon</th>\n",
       "      <th>in_answered</th>\n",
       "      <th>out_answered</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>პოლისის ნომერი (hash)</th>\n",
       "      <th>დამზღ პირადი ნომერი (hash)</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0b682bb61afa0128b8</td>\n",
       "      <td>5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dda5d8a8738d88adb96</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1682.57</td>\n",
       "      <td>0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e059b36d34c7b11036</td>\n",
       "      <td>70720759877097d63afb1e59d5f535413076b92a48e616f64363c680560f4028</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>670.00</td>\n",
       "      <td>0</td>\n",
       "      <td>51.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a1736922575791306f</td>\n",
       "      <td>21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff3284c49d6b377906fd8</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1863.34</td>\n",
       "      <td>11</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>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",
       "      <td>6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9a05f75666df1571c7</td>\n",
       "      <td>ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a19811ddf65f7c18536cfd</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1289.04</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>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",
       "      <td>e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f4e1d9f142b33b8e99</td>\n",
       "      <td>be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4194d5501a91bf3c11</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1351.61</td>\n",
       "      <td>0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>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",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                       წინა პოლისის ნომერი (hash)  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                       \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                           0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                           0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                           0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                           0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                           0   \n",
       "\n",
       "                                                                                                       Client Category (hash)  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                   \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                       0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                       0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                       0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                       0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                       0   \n",
       "\n",
       "                                                                                                        პრემია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                    \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...  1682.57   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...   670.00   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...  1863.34   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...  1289.04   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...  1351.61   \n",
       "\n",
       "                                                                                                       ვადაგადაცილებული დღე  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                 \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                     0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                     0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                    11   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                     0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                     0   \n",
       "\n",
       "                                                                                                       დამზღვევის ასაკი  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                             \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...              39.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...              51.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...              25.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...              35.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...              34.0   \n",
       "\n",
       "                                                                                                       დამზღვევი ქალია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                            \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                1   \n",
       "\n",
       "                                                                                                       Motor TPL Package  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                              \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                  0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                  0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                  0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                  0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                  0   \n",
       "\n",
       "                                                                                                       MotorPackage  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                         \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...             1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...             1   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...             1   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...             1   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...             1   \n",
       "\n",
       "                                                                                                       PropertyPackage  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                            \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                0   \n",
       "\n",
       "                                                                                                       TravelPackage  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                          \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...              0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...              0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...              0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...              0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...              0   \n",
       "\n",
       "                                                                                                       ...  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                          ...   \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...  ...   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...  ...   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...  ...   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...  ...   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...  ...   \n",
       "\n",
       "                                                                                                       ნაკლებად_დაკმაყოფილებული  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                     \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                         0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                         0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                         0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                         0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                         0   \n",
       "\n",
       "                                                                                                       ზედმეტად_დაკმაყოფილებული  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                     \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                         0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                         0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                         0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                         0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                         0   \n",
       "\n",
       "                                                                                                       no_status  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                      \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...          0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...          1   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...          1   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...          1   \n",
       "\n",
       "                                                                                                       დასრულებულია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                         \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...             1   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...             0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...             1   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...             0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...             0   \n",
       "\n",
       "                                                                                                       პროცესშია  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                      \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...          0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...          0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...          0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...          0   \n",
       "\n",
       "                                                                                                       გარეცხვის რაოდენობა  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                                \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...                  0.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...                  0.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...                  0.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...                  0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...                  0.0   \n",
       "\n",
       "                                                                                                       in_abandon  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                       \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...         0.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...         0.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...         0.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...         0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...         0.0   \n",
       "\n",
       "                                                                                                       out_abandon  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                        \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...          0.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...          0.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          0.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...          0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...          0.0   \n",
       "\n",
       "                                                                                                       in_answered  \\\n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                        \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...          1.0   \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...          1.0   \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...          0.0   \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...          0.0   \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...          0.0   \n",
       "\n",
       "                                                                                                       out_answered  \n",
       "პოლისის ნომერი (hash)                              დამზღ პირადი ნომერი (hash)                                        \n",
       "a33fff3fbe65ff7923448554f5f32f05b52ddf1f4cf88d0... 5c05f5ab96ae3736b288a7121c222b86a82c221e782b4dd...           0.0  \n",
       "1caa70ee07ef8cb0f55f06f6bf43cb7d4853b13fe837a4e... 70720759877097d63afb1e59d5f535413076b92a48e616f...           0.0  \n",
       "634cc57a6f4ba919e81c0f01b7530de1f660b67183f597a... 21af9ba2c667dc7758c1d8fee8cf845c10a2b912077ff32...           0.0  \n",
       "6e58638e7cd1bb63394922816cdbc8a1e50cc6ae6206cb9... ce3f168dac3aa97e04b770006e29b2bc6a5a4dca34a1981...           0.0  \n",
       "e831f0776cd9fd656387eb9de5a72009f1f747ce8d90e3f... be7f92e24d1ad7c35f4b44afde3077178512c4a846da5f4...           0.0  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = new_df[\"ბოლომდე არ გადაიხადა\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "30000.0"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.max().max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 210,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.min().min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TMalishava\\AppData\\Local\\Continuum\\anaconda3\\envs\\testenv\\lib\\site-packages\\sklearn\\preprocessing\\data.py:323: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by MinMaxScaler.\n",
      "  return self.partial_fit(X, y)\n"
     ]
    }
   ],
   "source": [
    "scaler = MinMaxScaler()\n",
    "X_train = scaler.fit_transform(X_train)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.1666666666666665"
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 213,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.min()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic Regression (unbalanced)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of Logistic regression classifier on training set: 0.793359375\n",
      "Accuracy of Logistic regression classifier on test set: 0.79921875\n"
     ]
    }
   ],
   "source": [
    "logreg = LogisticRegression(solver='liblinear')\n",
    "logreg.fit(X_train, y_train)\n",
    "print(f'Accuracy of Logistic regression classifier on training set: {logreg.score(X_train, y_train)}')\n",
    "print(f'Accuracy of Logistic regression classifier on test set: {logreg.score(X_test, y_test)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = logreg.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### From the confusion matrix we can see how much mistakes has the model made on churned customers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>998</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>249</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0   1\n",
       "ბოლომდე არ გადაიხადა         \n",
       "0                     998   8\n",
       "1                     249  25"
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### This is due to unbalanced dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Let's fix it and even out the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Over sampling method v.1 (SMOTE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TMalishava\\AppData\\Local\\Continuum\\anaconda3\\envs\\testenv\\lib\\site-packages\\sklearn\\preprocessing\\data.py:323: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by MinMaxScaler.\n",
      "  return self.partial_fit(X, y)\n"
     ]
    }
   ],
   "source": [
    "scaler = MinMaxScaler()\n",
    "X_train = scaler.fit_transform(X_train)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((5120, 33), (5120,))"
      ]
     },
     "execution_count": 272,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "metadata": {},
   "outputs": [],
   "source": [
    "sm = SMOTE(random_state=27, ratio=1.0)\n",
    "X_train_sm, y_train_sm = sm.fit_sample(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((8038, 33), (8038,))"
      ]
     },
     "execution_count": 274,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_sm.shape, y_train_sm.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.02034557263056437"
      ]
     },
     "execution_count": 275,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "start_time = time.time()\n",
    "rfc_smote = RandomForestClassifier(n_estimators = 100).fit(X_train_sm, y_train_sm)\n",
    "(time.time() - start_time)/60"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 276,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc_smote_pred = rfc_smote.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>862</td>\n",
       "      <td>144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>129</td>\n",
       "      <td>145</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     862  144\n",
       "1                     129  145"
      ]
     },
     "execution_count": 277,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc_conf_matrix = pd.crosstab(y_test, rfc_smote_pred)\n",
    "rfc_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.78671875"
      ]
     },
     "execution_count": 278,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, rfc_smote_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.856858846918489\n",
      "Churn guessed correctly: 0.5291970802919708\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', rfc_conf_matrix[0][0]/(rfc_conf_matrix[0][0] + rfc_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', rfc_conf_matrix[1][1]/(rfc_conf_matrix[0][1] + rfc_conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 326,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost import XGBClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 327,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_smote = XGBClassifier().fit(X_train_down, y_train_down)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 328,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_smote_pred = xgb_smote.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 329,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>577</td>\n",
       "      <td>429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>244</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     577  429\n",
       "1                      30  244"
      ]
     },
     "execution_count": 329,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc_conf_matrix = pd.crosstab(y_test, xgb_smote_pred)\n",
    "rfc_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.78671875"
      ]
     },
     "execution_count": 330,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, rfc_smote_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.5735586481113321\n",
      "Churn guessed correctly: 0.8905109489051095\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', rfc_conf_matrix[0][0]/(rfc_conf_matrix[0][0] + rfc_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', rfc_conf_matrix[1][1]/(rfc_conf_matrix[0][1] + rfc_conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TMalishava\\AppData\\Local\\Continuum\\anaconda3\\envs\\testenv\\lib\\site-packages\\sklearn\\linear_model\\logistic.py:758: ConvergenceWarning: lbfgs failed to converge. Increase the number of iterations.\n",
      "  \"of iterations.\", ConvergenceWarning)\n"
     ]
    }
   ],
   "source": [
    "smote = LogisticRegression(solver='lbfgs').fit(X_train_sm, y_train_sm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [],
   "source": [
    "smote_pred = smote.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>600</td>\n",
       "      <td>406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>52</td>\n",
       "      <td>222</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     600  406\n",
       "1                      52  222"
      ]
     },
     "execution_count": 224,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conf_matrix = pd.crosstab(y_test, smote_pred)\n",
    "conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6421875"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, smote_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.5964214711729622\n",
      "Churn guessed correctly: 0.8102189781021898\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', conf_matrix[0][0]/(conf_matrix[0][0] + conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', conf_matrix[1][1]/(conf_matrix[0][1] + conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Linear Discriminant Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TMalishava\\AppData\\Local\\Continuum\\anaconda3\\envs\\testenv\\lib\\site-packages\\sklearn\\discriminant_analysis.py:388: UserWarning: Variables are collinear.\n",
      "  warnings.warn(\"Variables are collinear.\")\n"
     ]
    }
   ],
   "source": [
    "lda_smote = LinearDiscriminantAnalysis().fit(X_train_sm, y_train_sm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [],
   "source": [
    "lda_smote_pred = lda_smote.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>579</td>\n",
       "      <td>427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>50</td>\n",
       "      <td>224</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     579  427\n",
       "1                      50  224"
      ]
     },
     "execution_count": 229,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda_conf_matrix = pd.crosstab(y_test, lda_smote_pred)\n",
    "lda_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.62734375"
      ]
     },
     "execution_count": 230,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, lda_smote_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.5755467196819085\n",
      "Churn guessed correctly: 0.8175182481751825\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', lda_conf_matrix[0][0]/(lda_conf_matrix[0][0] + lda_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', lda_conf_matrix[1][1]/(lda_conf_matrix[0][1] + lda_conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Over sampling method v.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = new_df.drop(labels = [\"ბოლომდე არ გადაიხადა\"], axis = 1)\n",
    "y = new_df[\"ბოლომდე არ გადაიხადა\"]\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_up = pd.concat([X_train, y_train], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 314,
   "metadata": {},
   "outputs": [],
   "source": [
    "non_churn = X_up[X_up['ბოლომდე არ გადაიხადა']==0]\n",
    "churn = X_up[X_up['ბოლომდე არ გადაიხადა']==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {},
   "outputs": [],
   "source": [
    "churn_upsampled = resample(churn,\n",
    "                          replace=True, # sample with replacement\n",
    "                          n_samples=len(non_churn), # match number in majority class\n",
    "                          random_state=27) # reproducible results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [],
   "source": [
    "# combine majority and upsampled minority\n",
    "upsampled = pd.concat([non_churn, churn_upsampled])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    4019\n",
       "0    4019\n",
       "Name: ბოლომდე არ გადაიხადა, dtype: int64"
      ]
     },
     "execution_count": 317,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# combine majority and upsampled minority\n",
    "upsampled = pd.concat([non_churn, churn_upsampled])\n",
    "\n",
    "# check new class counts\n",
    "upsampled['ბოლომდე არ გადაიხადა'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_up = upsampled['ბოლომდე არ გადაიხადა']\n",
    "X_train_up = upsampled.drop('ბოლომდე არ გადაიხადა', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TMalishava\\AppData\\Local\\Continuum\\anaconda3\\envs\\testenv\\lib\\site-packages\\sklearn\\preprocessing\\data.py:323: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by MinMaxScaler.\n",
      "  return self.partial_fit(X, y)\n"
     ]
    }
   ],
   "source": [
    "scaler = MinMaxScaler()\n",
    "X_train_up = scaler.fit_transform(X_train_up)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc_upsampled = RandomForestClassifier(n_estimators = 100).fit(X_train_up, y_train_up)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc_upsampled_pred = rfc_upsampled.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>875</td>\n",
       "      <td>131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>147</td>\n",
       "      <td>127</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     875  131\n",
       "1                     147  127"
      ]
     },
     "execution_count": 290,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc_conf_matrix = pd.crosstab(y_test, rfc_upsampled_pred)\n",
    "rfc_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7828125"
      ]
     },
     "execution_count": 291,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, rfc_upsampled_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.8697813121272365\n",
      "Churn guessed correctly: 0.4635036496350365\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', rfc_conf_matrix[0][0]/(rfc_conf_matrix[0][0] + rfc_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', rfc_conf_matrix[1][1]/(rfc_conf_matrix[0][1] + rfc_conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost import XGBClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_upsampled = XGBClassifier().fit(X_train_down, y_train_down)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_upsampled_pred = xgb_upsampled.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>577</td>\n",
       "      <td>429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>244</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     577  429\n",
       "1                      30  244"
      ]
     },
     "execution_count": 323,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc_conf_matrix = pd.crosstab(y_test, xgb_upsampled_pred)\n",
    "rfc_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7828125"
      ]
     },
     "execution_count": 324,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, rfc_upsampled_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.5735586481113321\n",
      "Churn guessed correctly: 0.8905109489051095\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', rfc_conf_matrix[0][0]/(rfc_conf_matrix[0][0] + rfc_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', rfc_conf_matrix[1][1]/(rfc_conf_matrix[0][1] + rfc_conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {},
   "outputs": [],
   "source": [
    "upsampled = LogisticRegression(solver='liblinear').fit(X_train_up, y_train_up)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [],
   "source": [
    "upsampled_pred = upsampled.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>583</td>\n",
       "      <td>423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>50</td>\n",
       "      <td>224</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     583  423\n",
       "1                      50  224"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conf_matrix = pd.crosstab(y_test, upsampled_pred)\n",
    "conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.63046875"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, upsampled_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.5795228628230616\n",
      "Churn guessed correctly: 0.8175182481751825\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', conf_matrix[0][0]/(conf_matrix[0][0] + conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', conf_matrix[1][1]/(conf_matrix[0][1] + conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Linear Discriminant Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TMalishava\\AppData\\Local\\Continuum\\anaconda3\\envs\\testenv\\lib\\site-packages\\sklearn\\discriminant_analysis.py:388: UserWarning: Variables are collinear.\n",
      "  warnings.warn(\"Variables are collinear.\")\n"
     ]
    }
   ],
   "source": [
    "lda_upsampled = LinearDiscriminantAnalysis().fit(X_train_up, y_train_up)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [],
   "source": [
    "lda_upsampled_pred = lda_upsampled.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>562</td>\n",
       "      <td>444</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>42</td>\n",
       "      <td>232</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     562  444\n",
       "1                      42  232"
      ]
     },
     "execution_count": 247,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda_conf_matrix = pd.crosstab(y_test, lda_upsampled_pred)\n",
    "lda_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6203125"
      ]
     },
     "execution_count": 248,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, lda_upsampled_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.558648111332008\n",
      "Churn guessed correctly: 0.8467153284671532\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', lda_conf_matrix[0][0]/(lda_conf_matrix[0][0] + lda_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', lda_conf_matrix[1][1]/(lda_conf_matrix[0][1] + lda_conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Under sampling method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = new_df.drop(labels = [\"ბოლომდე არ გადაიხადა\"], axis = 1)\n",
    "y = new_df[\"ბოლომდე არ გადაიხადა\"]\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_down = pd.concat([X_train, y_train], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {},
   "outputs": [],
   "source": [
    "non_churn = X_down[X_down['ბოლომდე არ გადაიხადა']==0]\n",
    "churn = X_down[X_down['ბოლომდე არ გადაიხადა']==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {},
   "outputs": [],
   "source": [
    "non_churn_downsampled = resample(non_churn,\n",
    "                                replace = False, # sample without replacement\n",
    "                                n_samples = len(churn), # match minority n\n",
    "                                random_state = 27) # reproducible results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "metadata": {},
   "outputs": [],
   "source": [
    "# combine minority and downsampled majority\n",
    "downsampled = pd.concat([non_churn_downsampled, churn])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    1101\n",
       "0    1101\n",
       "Name: ბოლომდე არ გადაიხადა, dtype: int64"
      ]
     },
     "execution_count": 298,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# checking counts\n",
    "downsampled['ბოლომდე არ გადაიხადა'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_down = downsampled['ბოლომდე არ გადაიხადა']\n",
    "X_train_down = downsampled.drop('ბოლომდე არ გადაიხადა', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TMalishava\\AppData\\Local\\Continuum\\anaconda3\\envs\\testenv\\lib\\site-packages\\sklearn\\preprocessing\\data.py:323: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by MinMaxScaler.\n",
      "  return self.partial_fit(X, y)\n"
     ]
    }
   ],
   "source": [
    "scaler = MinMaxScaler()\n",
    "X_train_down = scaler.fit_transform(X_train_down)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc_undersampled = RandomForestClassifier(n_estimators = 100).fit(X_train_down, y_train_down)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc_undersampled_pred = rfc_undersampled.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>706</td>\n",
       "      <td>300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>59</td>\n",
       "      <td>215</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     706  300\n",
       "1                      59  215"
      ]
     },
     "execution_count": 303,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc_conf_matrix = pd.crosstab(y_test, rfc_undersampled_pred)\n",
    "rfc_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.71953125"
      ]
     },
     "execution_count": 304,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, rfc_undersampled_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.7017892644135189\n",
      "Churn guessed correctly: 0.7846715328467153\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', rfc_conf_matrix[0][0]/(rfc_conf_matrix[0][0] + rfc_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', rfc_conf_matrix[1][1]/(rfc_conf_matrix[0][1] + rfc_conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Feature Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 332,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_importances = pd.DataFrame(rfc_undersampled.feature_importances_,\n",
    "                                   index = X.columns,\n",
    "                                   columns=['importance']).sort_values('importance', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 333,
   "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>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>პრემია</td>\n",
       "      <td>0.191721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>დამზღვევის ასაკი</td>\n",
       "      <td>0.185155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>პასუხისმგებლის ასაკი</td>\n",
       "      <td>0.132940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>in_answered</td>\n",
       "      <td>0.052070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>ვადაგადაცილებული დღე</td>\n",
       "      <td>0.050222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>გარეცხვის რაოდენობა</td>\n",
       "      <td>0.048236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>მოთხოვნილი</td>\n",
       "      <td>0.036136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>თანამშრომლი ქალია</td>\n",
       "      <td>0.029478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>დამზღ არის მესაკუთრე</td>\n",
       "      <td>0.028171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>დამზღვევი ქალია</td>\n",
       "      <td>0.026915</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      importance\n",
       "პრემია                  0.191721\n",
       "დამზღვევის ასაკი        0.185155\n",
       "პასუხისმგებლის ასაკი    0.132940\n",
       "in_answered             0.052070\n",
       "ვადაგადაცილებული დღე    0.050222\n",
       "გარეცხვის რაოდენობა     0.048236\n",
       "მოთხოვნილი              0.036136\n",
       "თანამშრომლი ქალია       0.029478\n",
       "დამზღ არის მესაკუთრე    0.028171\n",
       "დამზღვევი ქალია         0.026915"
      ]
     },
     "execution_count": 333,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_importances.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 340,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2683a4cc288>"
      ]
     },
     "execution_count": 340,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "feat_importances = pd.Series(rfc_undersampled.feature_importances_, index=new_df.columns.tolist()[0:-1])\n",
    "feat_importances.nlargest(15).plot(kind='barh',figsize=(12,7),fontsize=15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gradient Boosting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {},
   "outputs": [],
   "source": [
    "from xgboost import XGBClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_undersampled = XGBClassifier().fit(X_train_down, y_train_down)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [],
   "source": [
    "xgb_undersampled_pred = xgb_undersampled.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>631</td>\n",
       "      <td>375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>37</td>\n",
       "      <td>237</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     631  375\n",
       "1                      37  237"
      ]
     },
     "execution_count": 309,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfc_conf_matrix = pd.crosstab(y_test, xgb_undersampled_pred)\n",
    "rfc_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.71953125"
      ]
     },
     "execution_count": 310,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, rfc_undersampled_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.6272365805168986\n",
      "Churn guessed correctly: 0.864963503649635\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', rfc_conf_matrix[0][0]/(rfc_conf_matrix[0][0] + rfc_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', rfc_conf_matrix[1][1]/(rfc_conf_matrix[0][1] + rfc_conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {},
   "outputs": [],
   "source": [
    "undersampled = LogisticRegression(solver='liblinear').fit(X_train_down, y_train_down)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [],
   "source": [
    "undersampled_pred = undersampled.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>584</td>\n",
       "      <td>422</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>51</td>\n",
       "      <td>223</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     584  422\n",
       "1                      51  223"
      ]
     },
     "execution_count": 260,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conf_matrix = pd.crosstab(y_test, undersampled_pred)\n",
    "conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.63046875"
      ]
     },
     "execution_count": 261,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, undersampled_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.5805168986083499\n",
      "Churn guessed correctly: 0.8138686131386861\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', conf_matrix[0][0]/(conf_matrix[0][0] + conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', conf_matrix[1][1]/(conf_matrix[0][1] + conf_matrix[1][1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 263,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Relative Feature Importance')"
      ]
     },
     "execution_count": 263,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_importance = abs(undersampled.coef_[0])\n",
    "feature_importance = 100.0 * (feature_importance / feature_importance.max())\n",
    "sorted_idx = np.argsort(feature_importance)\n",
    "pos = np.arange(sorted_idx.shape[0]) + .5\n",
    "\n",
    "featfig = plt.figure(figsize=(15,15))\n",
    "featax = featfig.add_subplot(1, 1, 1)\n",
    "featax.barh(pos, feature_importance[sorted_idx], align='center')\n",
    "featax.set_yticks(pos)\n",
    "featax.set_yticklabels(np.array(X.columns)[sorted_idx], fontsize=8)\n",
    "featax.set_xlabel('Relative Feature Importance')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Linear Discriminant Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\TMalishava\\AppData\\Local\\Continuum\\anaconda3\\envs\\testenv\\lib\\site-packages\\sklearn\\discriminant_analysis.py:388: UserWarning: Variables are collinear.\n",
      "  warnings.warn(\"Variables are collinear.\")\n"
     ]
    }
   ],
   "source": [
    "lda_undersampled = LinearDiscriminantAnalysis().fit(X_train_down, y_train_down)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {},
   "outputs": [],
   "source": [
    "lda_undersampled_pred = lda_undersampled.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "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>col_0</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ბოლომდე არ გადაიხადა</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>569</td>\n",
       "      <td>437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>47</td>\n",
       "      <td>227</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "col_0                   0    1\n",
       "ბოლომდე არ გადაიხადა          \n",
       "0                     569  437\n",
       "1                      47  227"
      ]
     },
     "execution_count": 266,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda_conf_matrix = pd.crosstab(y_test, lda_undersampled_pred)\n",
    "lda_conf_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.621875"
      ]
     },
     "execution_count": 267,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy_score(y_test, lda_undersampled_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Non-churn guessed correctly: 0.5656063618290258\n",
      "Churn guessed correctly: 0.8284671532846716\n"
     ]
    }
   ],
   "source": [
    "print('Non-churn guessed correctly:', lda_conf_matrix[0][0]/(lda_conf_matrix[0][0] + lda_conf_matrix[1][0]))\n",
    "\n",
    "print('Churn guessed correctly:', lda_conf_matrix[1][1]/(lda_conf_matrix[0][1] + lda_conf_matrix[1][1]))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:testenv]",
   "language": "python",
   "name": "conda-env-testenv-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}