Investigation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 529,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import KBinsDiscretizer\n",
    "from sklearn.cluster import KMeans\n",
    "from scipy.spatial.distance import cdist\n",
    "from functools import reduce\n",
    "from dateutil.relativedelta import relativedelta\n",
    "from datetime import datetime, timedelta\n",
    "import re\n",
    "#from fastai.tabular import add_datepart"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_b = pd.read_excel('Train/train_bureau.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_d = pd.read_excel('Train/train_Data.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_b = pd.read_excel('Test/test_bureau.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_d = pd.read_excel('Test/test_Data.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "foo = test_b.head(1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_des = pd.read_excel('Train/data_dict.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "No Top-up Service    106677\n",
       " > 48 Months           8366\n",
       "36-48 Months           3656\n",
       "24-30 Months           3492\n",
       "30-36 Months           3062\n",
       "18-24 Months           2368\n",
       "12-18 Months           1034\n",
       "Name: Top-up Month, dtype: int64"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_d['Top-up Month'].value_counts() #7 classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>SELF-INDICATOR</th>\n",
       "      <th>MATCH-TYPE</th>\n",
       "      <th>ACCT-TYPE</th>\n",
       "      <th>CONTRIBUTOR-TYPE</th>\n",
       "      <th>DATE-REPORTED</th>\n",
       "      <th>OWNERSHIP-IND</th>\n",
       "      <th>ACCOUNT-STATUS</th>\n",
       "      <th>DISBURSED-DT</th>\n",
       "      <th>CLOSE-DT</th>\n",
       "      <th>...</th>\n",
       "      <th>INSTALLMENT-FREQUENCY</th>\n",
       "      <th>OVERDUE-AMT</th>\n",
       "      <th>WRITE-OFF-AMT</th>\n",
       "      <th>ASSET_CLASS</th>\n",
       "      <th>REPORTED DATE - HIST</th>\n",
       "      <th>DPD - HIST</th>\n",
       "      <th>CUR BAL - HIST</th>\n",
       "      <th>AMT OVERDUE - HIST</th>\n",
       "      <th>AMT PAID - HIST</th>\n",
       "      <th>TENURE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Credit Card</td>\n",
       "      <td>CCC</td>\n",
       "      <td>2019-11-17</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Active</td>\n",
       "      <td>2019-09-26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20191117,</td>\n",
       "      <td>000</td>\n",
       "      <td>589,</td>\n",
       "      <td>0,</td>\n",
       "      <td>0,</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Housing Loan</td>\n",
       "      <td>HFC</td>\n",
       "      <td>2019-10-31</td>\n",
       "      <td>Joint</td>\n",
       "      <td>Active</td>\n",
       "      <td>2017-06-21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>F03</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20191031,20190930,20190831,20190731,20190630,2...</td>\n",
       "      <td>0000000000000000000000000000000000000000000000...</td>\n",
       "      <td>2107227,2122196,2136877,2151448,2165910,218026...</td>\n",
       "      <td>0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...</td>\n",
       "      <td>,,,,,,,,,,,,,,,,,,,,,,,,,,,,,</td>\n",
       "      <td>120.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Property Loan</td>\n",
       "      <td>HFC</td>\n",
       "      <td>2019-10-31</td>\n",
       "      <td>Joint</td>\n",
       "      <td>Active</td>\n",
       "      <td>2017-06-27</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>F03</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20191031,20190930,20190831,20190731,20190630,2...</td>\n",
       "      <td>0000000000000000000000000000000000000000000000...</td>\n",
       "      <td>83458,84051,84632,85209,85781,86349,86913,8747...</td>\n",
       "      <td>0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...</td>\n",
       "      <td>,,,,,,,,,,,,,,,,,,,,,,,,,,,,,</td>\n",
       "      <td>120.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>True</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Tractor Loan</td>\n",
       "      <td>NBF</td>\n",
       "      <td>2015-10-31</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Closed</td>\n",
       "      <td>2011-12-29</td>\n",
       "      <td>2015-10-13 00:00:00</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20151031,20150901,20150831,20150731,20150630,2...</td>\n",
       "      <td>000DDD0270000000000000000000000000000000000000...</td>\n",
       "      <td>0,,21548,32097,42500,52758,62874,72850,82688,9...</td>\n",
       "      <td>0,,11000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0...</td>\n",
       "      <td>,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Gold Loan</td>\n",
       "      <td>PRB</td>\n",
       "      <td>2017-08-31</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Closed</td>\n",
       "      <td>2016-05-04</td>\n",
       "      <td>2017-06-15 00:00:00</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20170630,20170501,20170401,20170331,20170201,2...</td>\n",
       "      <td>000DDDDDD000DDDDDD000DDDDDD000DDDDDD000</td>\n",
       "      <td>0,,,25846,,,25846,,,25846,,,25846,</td>\n",
       "      <td>0,,,0,,,0,,,0,,,0</td>\n",
       "      <td>,,,,,,,,,,,,,,</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ID  SELF-INDICATOR MATCH-TYPE      ACCT-TYPE CONTRIBUTOR-TYPE  \\\n",
       "0   4           False    PRIMARY    Credit Card              CCC   \n",
       "1   4           False    PRIMARY   Housing Loan              HFC   \n",
       "2   4           False    PRIMARY  Property Loan              HFC   \n",
       "3   4            True    PRIMARY   Tractor Loan              NBF   \n",
       "4   5           False    PRIMARY      Gold Loan              PRB   \n",
       "\n",
       "  DATE-REPORTED OWNERSHIP-IND ACCOUNT-STATUS DISBURSED-DT  \\\n",
       "0    2019-11-17    Individual         Active   2019-09-26   \n",
       "1    2019-10-31         Joint         Active   2017-06-21   \n",
       "2    2019-10-31         Joint         Active   2017-06-27   \n",
       "3    2015-10-31    Individual         Closed   2011-12-29   \n",
       "4    2017-08-31    Individual         Closed   2016-05-04   \n",
       "\n",
       "              CLOSE-DT  ... INSTALLMENT-FREQUENCY OVERDUE-AMT WRITE-OFF-AMT  \\\n",
       "0                  NaN  ...                   NaN           0           0.0   \n",
       "1                  NaN  ...                   F03           0           0.0   \n",
       "2                  NaN  ...                   F03           0           0.0   \n",
       "3  2015-10-13 00:00:00  ...                   NaN           0           0.0   \n",
       "4  2017-06-15 00:00:00  ...                   NaN           0           0.0   \n",
       "\n",
       "  ASSET_CLASS                               REPORTED DATE - HIST  \\\n",
       "0         NaN                                          20191117,   \n",
       "1         NaN  20191031,20190930,20190831,20190731,20190630,2...   \n",
       "2         NaN  20191031,20190930,20190831,20190731,20190630,2...   \n",
       "3         NaN  20151031,20150901,20150831,20150731,20150630,2...   \n",
       "4         NaN  20170630,20170501,20170401,20170331,20170201,2...   \n",
       "\n",
       "                                          DPD - HIST  \\\n",
       "0                                                000   \n",
       "1  0000000000000000000000000000000000000000000000...   \n",
       "2  0000000000000000000000000000000000000000000000...   \n",
       "3  000DDD0270000000000000000000000000000000000000...   \n",
       "4            000DDDDDD000DDDDDD000DDDDDD000DDDDDD000   \n",
       "\n",
       "                                      CUR BAL - HIST  \\\n",
       "0                                               589,   \n",
       "1  2107227,2122196,2136877,2151448,2165910,218026...   \n",
       "2  83458,84051,84632,85209,85781,86349,86913,8747...   \n",
       "3  0,,21548,32097,42500,52758,62874,72850,82688,9...   \n",
       "4                 0,,,25846,,,25846,,,25846,,,25846,   \n",
       "\n",
       "                                  AMT OVERDUE - HIST  \\\n",
       "0                                                 0,   \n",
       "1  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...   \n",
       "2  0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...   \n",
       "3  0,,11000,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0...   \n",
       "4                                  0,,,0,,,0,,,0,,,0   \n",
       "\n",
       "                         AMT PAID - HIST TENURE  \n",
       "0                                     0,    NaN  \n",
       "1          ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,  120.0  \n",
       "2          ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,  120.0  \n",
       "3  ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,    NaN  \n",
       "4                         ,,,,,,,,,,,,,,    NaN  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "foo.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ID', 'SELF-INDICATOR', 'MATCH-TYPE', 'ACCT-TYPE', 'CONTRIBUTOR-TYPE',\n",
       "       'DATE-REPORTED', 'OWNERSHIP-IND', 'ACCOUNT-STATUS', 'DISBURSED-DT',\n",
       "       'CLOSE-DT', 'LAST-PAYMENT-DATE', 'CREDIT-LIMIT/SANC AMT',\n",
       "       'DISBURSED-AMT/HIGH CREDIT', 'INSTALLMENT-AMT', 'CURRENT-BAL',\n",
       "       'INSTALLMENT-FREQUENCY', 'OVERDUE-AMT', 'WRITE-OFF-AMT', 'ASSET_CLASS',\n",
       "       'REPORTED DATE - HIST', 'DPD - HIST', 'CUR BAL - HIST',\n",
       "       'AMT OVERDUE - HIST', 'AMT PAID - HIST', 'TENURE'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "foo.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>ID</th>\n",
       "      <th>SELF-INDICATOR</th>\n",
       "      <th>MATCH-TYPE</th>\n",
       "      <th>ACCT-TYPE</th>\n",
       "      <th>CONTRIBUTOR-TYPE</th>\n",
       "      <th>OWNERSHIP-IND</th>\n",
       "      <th>ACCOUNT-STATUS</th>\n",
       "      <th>INSTALLMENT-FREQUENCY</th>\n",
       "      <th>ASSET_CLASS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Credit Card</td>\n",
       "      <td>CCC</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Active</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Housing Loan</td>\n",
       "      <td>HFC</td>\n",
       "      <td>Joint</td>\n",
       "      <td>Active</td>\n",
       "      <td>F03</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Property Loan</td>\n",
       "      <td>HFC</td>\n",
       "      <td>Joint</td>\n",
       "      <td>Active</td>\n",
       "      <td>F03</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>True</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Tractor Loan</td>\n",
       "      <td>NBF</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Closed</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Gold Loan</td>\n",
       "      <td>PRB</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Closed</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>995</th>\n",
       "      <td>1910</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Credit Card</td>\n",
       "      <td>PRB</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Active</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>996</th>\n",
       "      <td>1910</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Credit Card</td>\n",
       "      <td>PRB</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Active</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>997</th>\n",
       "      <td>1910</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Credit Card</td>\n",
       "      <td>PRB</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Active</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>1910</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Auto Loan (Personal)</td>\n",
       "      <td>NAB</td>\n",
       "      <td>Guarantor</td>\n",
       "      <td>Active</td>\n",
       "      <td>F03</td>\n",
       "      <td>Standard</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999</th>\n",
       "      <td>1910</td>\n",
       "      <td>False</td>\n",
       "      <td>PRIMARY</td>\n",
       "      <td>Housing Loan</td>\n",
       "      <td>NAB</td>\n",
       "      <td>Individual</td>\n",
       "      <td>Delinquent</td>\n",
       "      <td>F03</td>\n",
       "      <td>Standard</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       ID  SELF-INDICATOR MATCH-TYPE             ACCT-TYPE CONTRIBUTOR-TYPE  \\\n",
       "0       4           False    PRIMARY           Credit Card              CCC   \n",
       "1       4           False    PRIMARY          Housing Loan              HFC   \n",
       "2       4           False    PRIMARY         Property Loan              HFC   \n",
       "3       4            True    PRIMARY          Tractor Loan              NBF   \n",
       "4       5           False    PRIMARY             Gold Loan              PRB   \n",
       "..    ...             ...        ...                   ...              ...   \n",
       "995  1910           False    PRIMARY           Credit Card              PRB   \n",
       "996  1910           False    PRIMARY           Credit Card              PRB   \n",
       "997  1910           False    PRIMARY           Credit Card              PRB   \n",
       "998  1910           False    PRIMARY  Auto Loan (Personal)              NAB   \n",
       "999  1910           False    PRIMARY          Housing Loan              NAB   \n",
       "\n",
       "    OWNERSHIP-IND ACCOUNT-STATUS INSTALLMENT-FREQUENCY ASSET_CLASS  \n",
       "0      Individual         Active                   NaN         NaN  \n",
       "1           Joint         Active                   F03         NaN  \n",
       "2           Joint         Active                   F03         NaN  \n",
       "3      Individual         Closed                   NaN         NaN  \n",
       "4      Individual         Closed                   NaN         NaN  \n",
       "..            ...            ...                   ...         ...  \n",
       "995    Individual         Active                   NaN         NaN  \n",
       "996    Individual         Active                   NaN         NaN  \n",
       "997    Individual         Active                   NaN         NaN  \n",
       "998     Guarantor         Active                   F03    Standard  \n",
       "999    Individual     Delinquent                   F03    Standard  \n",
       "\n",
       "[1000 rows x 9 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "foo[['ID','SELF-INDICATOR','MATCH-TYPE', 'ACCT-TYPE', 'CONTRIBUTOR-TYPE','OWNERSHIP-IND', 'ACCOUNT-STATUS','INSTALLMENT-FREQUENCY','ASSET_CLASS']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0         4\n",
       "1         4\n",
       "2         4\n",
       "3         4\n",
       "4         5\n",
       "       ... \n",
       "995    1910\n",
       "996    1910\n",
       "997    1910\n",
       "998    1910\n",
       "999    1910\n",
       "Name: ID, Length: 1000, dtype: int64"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "foo.ID"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>ASSET_CLASS</th>\n",
       "      <th>01</th>\n",
       "      <th>1</th>\n",
       "      <th>Doubtful</th>\n",
       "      <th>Loss</th>\n",
       "      <th>Special Mention Account</th>\n",
       "      <th>Standard</th>\n",
       "      <th>SubStandard</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>361</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>393</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143396</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143397</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143398</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143399</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143400</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8112 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "ASSET_CLASS  01  1  Doubtful  Loss  Special Mention Account  Standard  \\\n",
       "ID                                                                      \n",
       "6             0  0         0     0                        0         1   \n",
       "25            0  0         0     0                        0         1   \n",
       "127           0  0         0     0                        0         4   \n",
       "361           0  0         0     0                        0         1   \n",
       "393           0  0         0     0                        0         2   \n",
       "...          .. ..       ...   ...                      ...       ...   \n",
       "143396        0  0         0     0                        0         5   \n",
       "143397        0  0         0     0                        0         1   \n",
       "143398        0  0         0     0                        0         1   \n",
       "143399        0  0         0     0                        0         8   \n",
       "143400        0  0         0     0                        0         8   \n",
       "\n",
       "ASSET_CLASS  SubStandard  \n",
       "ID                        \n",
       "6                      0  \n",
       "25                     0  \n",
       "127                    0  \n",
       "361                    0  \n",
       "393                    0  \n",
       "...                  ...  \n",
       "143396                 0  \n",
       "143397                 0  \n",
       "143398                 0  \n",
       "143399                 0  \n",
       "143400                 0  \n",
       "\n",
       "[8112 rows x 7 columns]"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(foo.ID,foo['MATCH-TYPE'],dropna= False)\n",
    "pd.crosstab(foo.ID,foo['ACCT-TYPE'],dropna= False) #CONTRIBUTOR-TYPE\n",
    "pd.crosstab(train_b.ID,train_b['INSTALLMENT-FREQUENCY'],dropna= False)\n",
    "#pd.crosstab(foo.ID,foo['CONTRIBUTOR-TYPE'],dropna= False)\n",
    "#pd.crosstab(foo.ID,foo['ACCOUNT-STATUS'],dropna= False)\n",
    "#pd.crosstab(foo.ID,foo['OWNERSHIP-IND'],dropna= False)\n",
    "pd.crosstab(test_b.ID,test_b['ASSET_CLASS'],dropna= False) #test has one less category \n",
    "#pd.crosstab(train_b.ID,train_b['TENURE'],dropna= False) #324 kinds of tenure "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = ['SELF-INDICATOR','MATCH-TYPE', 'ACCT-TYPE', 'CONTRIBUTOR-TYPE','OWNERSHIP-IND', 'ACCOUNT-STATUS',\n",
    "        'INSTALLMENT-FREQUENCY','ASSET_CLASS']\n",
    "# for i in cols:\n",
    "#     train_data_+i = pd.crosstab(train.ID,train[i],dropna= False)\n",
    "#      train_cols_name.append(train_data_+i)\n",
    "#     test_data_+i = pd.crosstab(test.ID,test[i],dropna= False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Column name ID\n",
      "         index        ID\n",
      "0        71060  0.000749\n",
      "1       141732  0.000294\n",
      "2        51786  0.000271\n",
      "3         1167  0.000246\n",
      "4        97794  0.000221\n",
      "...        ...       ...\n",
      "128650  107223  0.000002\n",
      "128651  109270  0.000002\n",
      "128652  116205  0.000002\n",
      "128653  103121  0.000002\n",
      "128654    9436  0.000002\n",
      "\n",
      "[128655 rows x 2 columns]         index        ID\n",
      "0      114111  0.001453\n",
      "1      114063  0.000937\n",
      "2       69953  0.000922\n",
      "3       26305  0.000922\n",
      "4      114125  0.000875\n",
      "...       ...       ...\n",
      "14740   78473  0.000016\n",
      "14741  141962  0.000016\n",
      "14742   62097  0.000016\n",
      "14743  141994  0.000016\n",
      "14744   98272  0.000016\n",
      "\n",
      "[14745 rows x 2 columns]\n",
      "Column name SELF-INDICATOR\n",
      "   index  SELF-INDICATOR\n",
      "0  False        0.707011\n",
      "1   True        0.292989    index  SELF-INDICATOR\n",
      "0  False         0.72366\n",
      "1   True         0.27634\n",
      "Column name MATCH-TYPE\n",
      "       index  MATCH-TYPE\n",
      "0    PRIMARY    0.999649\n",
      "1  SECONDARY    0.000351        index  MATCH-TYPE\n",
      "0    PRIMARY    0.999563\n",
      "1  SECONDARY    0.000437\n",
      "Column name ACCT-TYPE\n",
      "                                                index  ACCT-TYPE\n",
      "0                                        Tractor Loan   0.332075\n",
      "1                                           Gold Loan   0.162298\n",
      "2          Business Loan Priority Sector  Agriculture   0.142792\n",
      "3                                   Kisan Credit Card   0.059498\n",
      "4                                Auto Loan (Personal)   0.052733\n",
      "5                                               Other   0.048545\n",
      "6                             Commercial Vehicle Loan   0.034808\n",
      "7                                    Two-Wheeler Loan   0.028181\n",
      "8                                         Credit Card   0.022648\n",
      "9                                       Consumer Loan   0.021617\n",
      "10                                          Overdraft   0.016176\n",
      "11                                      Personal Loan   0.015389\n",
      "12                                       Housing Loan   0.009646\n",
      "13                        Construction Equipment Loan   0.008239\n",
      "14                                     JLG Individual   0.007498\n",
      "15                              Business Loan General   0.007114\n",
      "16                         Loan Against Bank Deposits   0.006852\n",
      "17                                      Used Car Loan   0.004643\n",
      "18      Business Loan Priority Sector  Small Business   0.003529\n",
      "19                   Loan Against Shares / Securities   0.002774\n",
      "20                                      Property Loan   0.002368\n",
      "21              Business Loan Priority Sector  Others   0.002040\n",
      "22              Mudra Loans   Shishu / Kishor / Tarun   0.002013\n",
      "23  Business Non-Funded Credit Facility-Priority S...   0.000956\n",
      "24                         Non-Funded Credit Facility   0.000804\n",
      "25                                     Education Loan   0.000799\n",
      "26                            Business Loan - Secured   0.000620\n",
      "27                Business Loan Against Bank Deposits   0.000551\n",
      "28                                         Individual   0.000528\n",
      "29                            Business Loan Unsecured   0.000399\n",
      "30                         Microfinance Business Loan   0.000383\n",
      "31                               Loan to Professional   0.000325\n",
      "32  Business Non-Funded Credit Facility-Priority S...   0.000226\n",
      "33                                Microfinance Others   0.000219\n",
      "34                                Secured Credit Card   0.000148\n",
      "35        Business Non-Funded Credit Facility General   0.000130\n",
      "36        Prime Minister Jaan Dhan Yojana - Overdraft   0.000103\n",
      "37                                          JLG Group   0.000061\n",
      "38                  Pradhan Mantri Awas Yojana - CLSS   0.000057\n",
      "39                         Microfinance Personal Loan   0.000046\n",
      "40                                     SHG Individual   0.000045\n",
      "41                                         Fleet Card   0.000041\n",
      "42                          Commercial Equipment Loan   0.000025\n",
      "43                          Microfinance Housing Loan   0.000025\n",
      "44                                Loan on Credit Card   0.000009\n",
      "45                              Corporate Credit Card   0.000009\n",
      "46                                            Leasing   0.000005\n",
      "47  Business Non-Funded Credit Facility-Priority S...   0.000005\n",
      "48                                          SHG Group   0.000002\n",
      "49                                     Telco Landline   0.000002                                                 index  ACCT-TYPE\n",
      "0                                        Tractor Loan   0.340664\n",
      "1          Business Loan Priority Sector  Agriculture   0.127931\n",
      "2                                           Gold Loan   0.118371\n",
      "3                                               Other   0.070338\n",
      "4                                   Kisan Credit Card   0.066418\n",
      "5                                Auto Loan (Personal)   0.061372\n",
      "6                                    Two-Wheeler Loan   0.028913\n",
      "7                             Commercial Vehicle Loan   0.027711\n",
      "8                                         Credit Card   0.026477\n",
      "9                                       Consumer Loan   0.025695\n",
      "10                                          Overdraft   0.024727\n",
      "11                                      Personal Loan   0.018416\n",
      "12                                       Housing Loan   0.012371\n",
      "13                              Business Loan General   0.007404\n",
      "14                                      Used Car Loan   0.007217\n",
      "15                         Loan Against Bank Deposits   0.007045\n",
      "16                        Construction Equipment Loan   0.006810\n",
      "17                                     JLG Individual   0.006233\n",
      "18      Business Loan Priority Sector  Small Business   0.002812\n",
      "19              Mudra Loans   Shishu / Kishor / Tarun   0.002734\n",
      "20                                      Property Loan   0.002421\n",
      "21                   Loan Against Shares / Securities   0.001984\n",
      "22              Business Loan Priority Sector  Others   0.001234\n",
      "23                                     Education Loan   0.000594\n",
      "24                         Non-Funded Credit Facility   0.000578\n",
      "25                                         Individual   0.000547\n",
      "26                         Microfinance Business Loan   0.000422\n",
      "27                            Business Loan - Secured   0.000422\n",
      "28                            Business Loan Unsecured   0.000391\n",
      "29                Business Loan Against Bank Deposits   0.000359\n",
      "30                               Loan to Professional   0.000312\n",
      "31                                Microfinance Others   0.000172\n",
      "32        Prime Minister Jaan Dhan Yojana - Overdraft   0.000172\n",
      "33  Business Non-Funded Credit Facility-Priority S...   0.000172\n",
      "34                                     SHG Individual   0.000141\n",
      "35  Business Non-Funded Credit Facility-Priority S...   0.000078\n",
      "36                         Microfinance Personal Loan   0.000062\n",
      "37        Business Non-Funded Credit Facility General   0.000062\n",
      "38                  Pradhan Mantri Awas Yojana - CLSS   0.000047\n",
      "39                          Commercial Equipment Loan   0.000047\n",
      "40                          Microfinance Housing Loan   0.000047\n",
      "41                                Secured Credit Card   0.000031\n",
      "42                                         Staff Loan   0.000016\n",
      "43                              Corporate Credit Card   0.000016\n",
      "44                                         Fleet Card   0.000016\n",
      "Column name CONTRIBUTOR-TYPE\n",
      "   index  CONTRIBUTOR-TYPE\n",
      "0    NBF          0.457944\n",
      "1    NAB          0.310059\n",
      "2    PRB          0.134356\n",
      "3    RRB          0.042834\n",
      "4    COP          0.039585\n",
      "5    MFI          0.006009\n",
      "6    HFC          0.004893\n",
      "7    CCC          0.002621\n",
      "8    FRB          0.001346\n",
      "9    SFB          0.000333\n",
      "10   ARC          0.000014\n",
      "11   OFI          0.000005    index  CONTRIBUTOR-TYPE\n",
      "0    NBF          0.460426\n",
      "1    NAB          0.272935\n",
      "2    PRB          0.141192\n",
      "3    COP          0.072666\n",
      "4    RRB          0.036723\n",
      "5    HFC          0.006717\n",
      "6    MFI          0.004780\n",
      "7    CCC          0.002827\n",
      "8    FRB          0.000906\n",
      "9    SFB          0.000812\n",
      "10   OFI          0.000016\n",
      "Column name DATE-REPORTED\n",
      "          index  DATE-REPORTED\n",
      "0    2020-01-31       0.211000\n",
      "1    2019-12-31       0.125788\n",
      "2    2016-12-31       0.057027\n",
      "3    2019-11-30       0.036952\n",
      "4    2019-06-30       0.028220\n",
      "...         ...            ...\n",
      "2647 2019-05-11       0.000002\n",
      "2648 2013-02-04       0.000002\n",
      "2649 2015-08-26       0.000002\n",
      "2650 2014-06-25       0.000002\n",
      "2651 2010-05-27       0.000002\n",
      "\n",
      "[2652 rows x 2 columns]           index  DATE-REPORTED\n",
      "0    2019-10-31       0.244631\n",
      "1    2019-11-30       0.134023\n",
      "2    2016-12-31       0.050891\n",
      "3    2019-08-31       0.047548\n",
      "4    2019-06-30       0.032600\n",
      "...         ...            ...\n",
      "1126 2016-12-23       0.000016\n",
      "1127 2015-04-23       0.000016\n",
      "1128 2014-09-20       0.000016\n",
      "1129 2016-06-04       0.000016\n",
      "1130 2015-09-02       0.000016\n",
      "\n",
      "[1131 rows x 2 columns]\n",
      "Column name OWNERSHIP-IND\n",
      "              index  OWNERSHIP-IND\n",
      "0        Individual       0.862707\n",
      "1             Joint       0.066730\n",
      "2         Guarantor       0.062185\n",
      "3           Primary       0.008132\n",
      "4  Supl Card Holder       0.000246               index  OWNERSHIP-IND\n",
      "0        Individual       0.857386\n",
      "1         Guarantor       0.068667\n",
      "2             Joint       0.066933\n",
      "3           Primary       0.006920\n",
      "4  Supl Card Holder       0.000094\n",
      "Column name ACCOUNT-STATUS\n",
      "                          index  ACCOUNT-STATUS\n",
      "0                        Closed        0.571023\n",
      "1                        Active        0.359988\n",
      "2                    Delinquent        0.057872\n",
      "3                   Written Off        0.005237\n",
      "4                    Suit Filed        0.003677\n",
      "5                       Settled        0.001116\n",
      "6                  Restructured        0.000911\n",
      "7   SUIT FILED (WILFUL DEFAULT)        0.000125\n",
      "8                WILFUL DEFAULT        0.000048\n",
      "9                Sold/Purchased        0.000002\n",
      "10                    Cancelled        0.000002                          index  ACCOUNT-STATUS\n",
      "0                       Closed        0.501289\n",
      "1                       Active        0.420906\n",
      "2                   Delinquent        0.070057\n",
      "3                   Suit Filed        0.003343\n",
      "4                  Written Off        0.002640\n",
      "5                      Settled        0.000968\n",
      "6                 Restructured        0.000625\n",
      "7  SUIT FILED (WILFUL DEFAULT)        0.000141\n",
      "8               WILFUL DEFAULT        0.000031\n",
      "Column name DISBURSED-DT\n",
      "          index  DISBURSED-DT\n",
      "0    2017-10-31      0.003361\n",
      "1    2017-06-30      0.003333\n",
      "2    2017-05-31      0.002105\n",
      "3    2018-03-31      0.002065\n",
      "4    2017-03-31      0.002064\n",
      "...         ...           ...\n",
      "6274 2001-12-04      0.000002\n",
      "6275 2008-03-09      0.000002\n",
      "6276 2002-12-30      0.000002\n",
      "6277 2004-03-04      0.000002\n",
      "6278 2002-04-06      0.000002\n",
      "\n",
      "[6279 rows x 2 columns]           index  DISBURSED-DT\n",
      "0    2017-03-10      0.004377\n",
      "1    2017-03-07      0.004048\n",
      "2    2017-10-31      0.003408\n",
      "3    2017-02-08      0.003314\n",
      "4    2017-06-30      0.002845\n",
      "...         ...           ...\n",
      "4614 2006-11-29      0.000016\n",
      "4615 2008-02-03      0.000016\n",
      "4616 2003-07-11      0.000016\n",
      "4617 2006-06-09      0.000016\n",
      "4618 2009-12-22      0.000016\n",
      "\n",
      "[4619 rows x 2 columns]\n",
      "Column name CLOSE-DT\n",
      "                    index  CLOSE-DT\n",
      "0     2019-06-01 00:00:00  0.002268\n",
      "1     2017-07-31 00:00:00  0.002223\n",
      "2     2018-03-31 00:00:00  0.002122\n",
      "3     2018-01-31 00:00:00  0.002016\n",
      "4     2017-10-31 00:00:00  0.001870\n",
      "...                   ...       ...\n",
      "4679  2007-08-14 00:00:00  0.000003\n",
      "4680  2008-05-01 00:00:00  0.000003\n",
      "4681  2004-12-25 00:00:00  0.000003\n",
      "4682  2006-12-01 00:00:00  0.000003\n",
      "4683  2006-01-17 00:00:00  0.000003\n",
      "\n",
      "[4684 rows x 2 columns]                     index  CLOSE-DT\n",
      "0     2015-07-01 00:00:00  0.004207\n",
      "1     2019-06-01 00:00:00  0.004084\n",
      "2     2017-07-31 00:00:00  0.003132\n",
      "3     2015-01-01 00:00:00  0.002702\n",
      "4     2017-06-30 00:00:00  0.002610\n",
      "...                   ...       ...\n",
      "3325  2012-02-27 00:00:00  0.000031\n",
      "3326  2012-12-06 00:00:00  0.000031\n",
      "3327  2009-06-26 00:00:00  0.000031\n",
      "3328  2010-03-18 00:00:00  0.000031\n",
      "3329  2012-08-24 00:00:00  0.000031\n",
      "\n",
      "[3330 rows x 2 columns]\n",
      "Column name LAST-PAYMENT-DATE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "          index  LAST-PAYMENT-DATE\n",
      "0    2019-09-30           0.014829\n",
      "1    2019-12-31           0.011070\n",
      "2    2019-12-05           0.010267\n",
      "3    2019-12-02           0.006851\n",
      "4    2019-12-27           0.006839\n",
      "...         ...                ...\n",
      "4080 2012-01-15           0.000004\n",
      "4081 2008-03-29           0.000004\n",
      "4082 2011-11-04           0.000004\n",
      "4083 2011-01-16           0.000004\n",
      "4084 2009-07-03           0.000004\n",
      "\n",
      "[4085 rows x 2 columns]           index  LAST-PAYMENT-DATE\n",
      "0    2019-09-30           0.017605\n",
      "1    2019-10-05           0.014554\n",
      "2    2019-10-31           0.009120\n",
      "3    2019-11-02           0.007944\n",
      "4    2019-03-31           0.007309\n",
      "...         ...                ...\n",
      "2790 2016-10-30           0.000032\n",
      "2791 2011-06-02           0.000032\n",
      "2792 2013-09-19           0.000032\n",
      "2793 2006-04-07           0.000032\n",
      "2794 2013-06-18           0.000032\n",
      "\n",
      "[2795 rows x 2 columns]\n",
      "Column name CREDIT-LIMIT/SANC AMT\n",
      "        index  CREDIT-LIMIT/SANC AMT\n",
      "0           0               0.305033\n",
      "1      30,000               0.060492\n",
      "2      25,000               0.058117\n",
      "3      15,000               0.046573\n",
      "4      10,000               0.037733\n",
      "..        ...                    ...\n",
      "872     4,700               0.000066\n",
      "873    37,289               0.000066\n",
      "874  3,73,000               0.000066\n",
      "875    43,199               0.000066\n",
      "876    16,667               0.000066\n",
      "\n",
      "[877 rows x 2 columns]       index  CREDIT-LIMIT/SANC AMT\n",
      "0         0               0.280350\n",
      "1    25,000               0.078223\n",
      "2    30,000               0.061327\n",
      "3    15,000               0.053817\n",
      "4    50,000               0.030038\n",
      "..      ...                    ...\n",
      "230  91,000               0.000626\n",
      "231  38,623               0.000626\n",
      "232  27,500               0.000626\n",
      "233  77,900               0.000626\n",
      "234  10,047               0.000626\n",
      "\n",
      "[235 rows x 2 columns]\n",
      "Column name DISBURSED-AMT/HIGH CREDIT\n",
      "          index  DISBURSED-AMT/HIGH CREDIT\n",
      "0             0                   0.057144\n",
      "1      3,00,000                   0.044843\n",
      "2      2,00,000                   0.028143\n",
      "3      2,50,000                   0.023342\n",
      "4      4,00,000                   0.023285\n",
      "...         ...                        ...\n",
      "72421  4,25,713                   0.000002\n",
      "72422  3,59,060                   0.000002\n",
      "72423    79,247                   0.000002\n",
      "72424  21,2,170                   0.000002\n",
      "72425   3,4,572                   0.000002\n",
      "\n",
      "[72426 rows x 2 columns]           index  DISBURSED-AMT/HIGH CREDIT\n",
      "0             0                   0.082743\n",
      "1      3,00,000                   0.045404\n",
      "2      4,00,000                   0.033729\n",
      "3      2,00,000                   0.031791\n",
      "4      1,00,000                   0.025586\n",
      "...         ...                        ...\n",
      "13839     9,900                   0.000016\n",
      "13840  4,66,602                   0.000016\n",
      "13841  2,51,073                   0.000016\n",
      "13842  2,45,606                   0.000016\n",
      "13843  2,48,363                   0.000016\n",
      "\n",
      "[13844 rows x 2 columns]\n",
      "Column name INSTALLMENT-AMT\n",
      "                  index  INSTALLMENT-AMT\n",
      "0             0/Monthly         0.203841\n",
      "1                     0         0.013781\n",
      "2      1,00,000/Monthly         0.012363\n",
      "3        50,000/Monthly         0.005651\n",
      "4        40,000/Monthly         0.004439\n",
      "...                 ...              ...\n",
      "50128      11,000/Other         0.000007\n",
      "50129    13,221/Monthly         0.000007\n",
      "50130  1,69,067/Monthly         0.000007\n",
      "50131   1,2,334/Monthly         0.000007\n",
      "50132    13,637/Monthly         0.000007\n",
      "\n",
      "[50133 rows x 2 columns]                  index  INSTALLMENT-AMT\n",
      "0            0/Monthly         0.300931\n",
      "1                    0         0.019233\n",
      "2     1,00,000/Monthly         0.007350\n",
      "3       50,000/Monthly         0.003369\n",
      "4       10,000/Monthly         0.003185\n",
      "...                ...              ...\n",
      "7805  2,74,502/Monthly         0.000061\n",
      "7806  55,930/Quarterly         0.000061\n",
      "7807  1,36,759/Monthly         0.000061\n",
      "7808    17,850/Monthly         0.000061\n",
      "7809          1,05,424         0.000061\n",
      "\n",
      "[7810 rows x 2 columns]\n",
      "Column name CURRENT-BAL\n",
      "           index  CURRENT-BAL\n",
      "0              0     0.630289\n",
      "1       2,50,000     0.001356\n",
      "2       1,00,000     0.000762\n",
      "3       3,00,000     0.000694\n",
      "4         50,000     0.000635\n",
      "...          ...          ...\n",
      "147440  4,98,478     0.000002\n",
      "147441  5,20,229     0.000002\n",
      "147442  3,42,303     0.000002\n",
      "147443  1,79,522     0.000002\n",
      "147444  1,05,544     0.000002\n",
      "\n",
      "[147445 rows x 2 columns]            index  CURRENT-BAL\n",
      "0              0     0.574632\n",
      "1       1,00,000     0.000985\n",
      "2       2,00,000     0.000656\n",
      "3       3,00,000     0.000578\n",
      "4       2,50,000     0.000578\n",
      "...          ...          ...\n",
      "24124   1,19,719     0.000016\n",
      "24125   2,14,337     0.000016\n",
      "24126   1,22,850     0.000016\n",
      "24127  13,44,988     0.000016\n",
      "24128   3,18,600     0.000016\n",
      "\n",
      "[24129 rows x 2 columns]\n",
      "Column name INSTALLMENT-FREQUENCY\n",
      "  index  INSTALLMENT-FREQUENCY\n",
      "0   F03               0.949738\n",
      "1   F01               0.017515\n",
      "2   F05               0.017199\n",
      "3   F02               0.013102\n",
      "4   F10               0.002247\n",
      "5   F04               0.000096\n",
      "6   F07               0.000052\n",
      "7   F08               0.000044\n",
      "8   F06               0.000007   index  INSTALLMENT-FREQUENCY\n",
      "0   F03               0.945931\n",
      "1   F05               0.025735\n",
      "2   F01               0.014009\n",
      "3   F02               0.012487\n",
      "4   F10               0.001775\n",
      "5   F06               0.000063\n",
      "Column name OVERDUE-AMT\n",
      "          index  OVERDUE-AMT\n",
      "0             0     0.924155\n",
      "1           118     0.000613\n",
      "2             1     0.000328\n",
      "3           100     0.000201\n",
      "4           236     0.000138\n",
      "...         ...          ...\n",
      "22380    81,400     0.000002\n",
      "22381    13,437     0.000002\n",
      "22382    18,846     0.000002\n",
      "22383    46,435     0.000002\n",
      "22384  1,03,569     0.000002\n",
      "\n",
      "[22385 rows x 2 columns]        index  OVERDUE-AMT\n",
      "0          0     0.906744\n",
      "1          2     0.000688\n",
      "2          1     0.000472\n",
      "3        118     0.000354\n",
      "4        100     0.000314\n",
      "...      ...          ...\n",
      "4133  24,619     0.000020\n",
      "4134   4,292     0.000020\n",
      "4135  22,418     0.000020\n",
      "4136   1,286     0.000020\n",
      "4137   1,422     0.000020\n",
      "\n",
      "[4138 rows x 2 columns]\n",
      "Column name WRITE-OFF-AMT\n",
      "         index  WRITE-OFF-AMT\n",
      "0          0.0       0.998032\n",
      "1     352886.0       0.000017\n",
      "2     143614.0       0.000007\n",
      "3        238.0       0.000006\n",
      "4      53682.0       0.000006\n",
      "...        ...            ...\n",
      "1004  318935.0       0.000002\n",
      "1005   79500.0       0.000002\n",
      "1006   79326.0       0.000002\n",
      "1007  317034.0       0.000002\n",
      "1008  261512.0       0.000002\n",
      "\n",
      "[1009 rows x 2 columns]         index  WRITE-OFF-AMT\n",
      "0         0.0       0.998259\n",
      "1     61887.0       0.000033\n",
      "2      6300.0       0.000016\n",
      "3     10026.0       0.000016\n",
      "4     45989.0       0.000016\n",
      "..        ...            ...\n",
      "102    9949.0       0.000016\n",
      "103   24013.0       0.000016\n",
      "104  224948.0       0.000016\n",
      "105    2014.0       0.000016\n",
      "106   69187.0       0.000016\n",
      "\n",
      "[107 rows x 2 columns]\n",
      "Column name ASSET_CLASS\n",
      "                     index  ASSET_CLASS\n",
      "0                 Standard     0.954705\n",
      "1              SubStandard     0.020348\n",
      "2                 Doubtful     0.009898\n",
      "3  Special Mention Account     0.009617\n",
      "4                     Loss     0.003352\n",
      "5                        1     0.001390\n",
      "6                       01     0.000680\n",
      "7                        2     0.000012                      index  ASSET_CLASS\n",
      "0                 Standard     0.954975\n",
      "1              SubStandard     0.018443\n",
      "2                 Doubtful     0.012124\n",
      "3  Special Mention Account     0.009307\n",
      "4                     Loss     0.002576\n",
      "5                        1     0.001442\n",
      "6                       01     0.001133\n",
      "Column name REPORTED DATE - HIST\n",
      "                                                   index  REPORTED DATE - HIST\n",
      "0      20200131,20191231,20191130,20191031,20190930,2...              0.046830\n",
      "1      20191231,20191130,20191031,20190930,20190831,2...              0.012366\n",
      "2      20200131,20191231,20191130,20191031,20190930,2...              0.006424\n",
      "3      20200131,20191231,20191130,20191031,20190930,2...              0.005396\n",
      "4                                              20200131,              0.005318\n",
      "...                                                  ...                   ...\n",
      "57841  20160731,20160630,20160531,20160430,20160331,2...              0.000002\n",
      "57842  20170630,20170531,20170430,20170331,20170228,2...              0.000002\n",
      "57843  20170930,20170801,20170731,20170630,20170531,2...              0.000002\n",
      "57844  20190228,20190212,20190131,20181231,20181130,2...              0.000002\n",
      "57845  20191130,20191031,20190930,20190801,20190701,2...              0.000002\n",
      "\n",
      "[57846 rows x 2 columns]                                                    index  REPORTED DATE - HIST\n",
      "0      20191031,20190930,20190831,20190731,20190630,2...              0.052361\n",
      "1      20191130,20191031,20190930,20190831,20190731,2...              0.019168\n",
      "2      20191031,20190930,20190801,20190701,20190601,2...              0.010528\n",
      "3                                     20190831,20190820,              0.008429\n",
      "4                                              20190831,              0.007745\n",
      "...                                                  ...                   ...\n",
      "12393      20151231,20151130,20151031,20150930,20150901,              0.000016\n",
      "12394  20190131,20181231,20181130,20181031,20180930,2...              0.000016\n",
      "12395  20151031,20150901,20150801,20150701,20150601,2...              0.000016\n",
      "12396  20161004,20160901,20160801,20160701,20160630,2...              0.000016\n",
      "12397  20190105,20181231,20181130,20181031,20180930,2...              0.000016\n",
      "\n",
      "[12398 rows x 2 columns]\n",
      "Column name DPD - HIST\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                                    index  DPD - HIST\n",
      "0                                                       0    0.080850\n",
      "1                                                     XXX    0.072240\n",
      "2       0000000000000000000000000000000000000000000000...    0.041260\n",
      "3       XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX...    0.019915\n",
      "4                                                     000    0.013616\n",
      "...                                                   ...         ...\n",
      "134333   000068038007008005008008007007007008008038008038    0.000002\n",
      "134334  038275XXXXXXXXXXXXDDDDDDXXXXXXXXXXXXXXXDDDXXXX...    0.000002\n",
      "134335  332XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX...    0.000002\n",
      "134336  0610600290600290610300600600290300270300300290...    0.000002\n",
      "134337  000000XXXXXXXXXDDDXXXDDDXXXDDDXXXXXXXXXDDDDDDX...    0.000002\n",
      "\n",
      "[134338 rows x 2 columns]                                                    index  DPD - HIST\n",
      "0                                                    XXX    0.069890\n",
      "1      0000000000000000000000000000000000000000000000...    0.068214\n",
      "2      XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX...    0.024653\n",
      "3      000DDD0000000000000000000000000000000000000000...    0.020910\n",
      "4                                                    000    0.017672\n",
      "...                                                  ...         ...\n",
      "17207            000000244XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX    0.000016\n",
      "17208  XXX0600000000000000000000000000000000000000000...    0.000016\n",
      "17209  0000000000000000000000000000000000070060000970...    0.000016\n",
      "17210  000DDDDDDDDD0000000000000000000000000000000000...    0.000016\n",
      "17211  0000000000000000000220000000001000670390000001...    0.000016\n",
      "\n",
      "[17212 rows x 2 columns]\n",
      "Column name CUR BAL - HIST\n",
      "                                                    index  CUR BAL - HIST\n",
      "0                                                       ,        0.058752\n",
      "1                                                      0,        0.016950\n",
      "2                                                    0,0,        0.004482\n",
      "3       0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...        0.003764\n",
      "4                                    0,0,,,,,,,,,,,,,0,0,        0.000956\n",
      "...                                                   ...             ...\n",
      "447067  0,,19062,21180,21180,21180,23298,23298,23298,2...        0.000002\n",
      "447068  0,72990,71993,71010,70040,69083,68140,140199,1...        0.000002\n",
      "447069   0,,,,,,,,,,,,,,,,,,,,,,,,,,354244,354244,384327,        0.000002\n",
      "447070  0,-19,-19,-19,-39,-39,-39,702,-9,-9,-9,-9,-9,-...        0.000002\n",
      "447071  1234813,1165000,1164980,1164980,1164980,115940...        0.000002\n",
      "\n",
      "[447072 rows x 2 columns]                                                    index  CUR BAL - HIST\n",
      "0                                                      ,        0.049660\n",
      "1                                                     0,        0.018631\n",
      "2                                                   0,0,        0.006183\n",
      "3                                   0,0,,,,,,,,,,,,,0,0,        0.005028\n",
      "4      0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...        0.003710\n",
      "...                                                  ...             ...\n",
      "53587  0,0,3258,6457,9601,9601,15723,18703,21631,2450...        0.000016\n",
      "53588  101159,100105,,120909,,118619,117396,116213,11...        0.000016\n",
      "53589  121888,121888,115273,114388,114388,114388,1143...        0.000016\n",
      "53590  4064,4064,4064,4064,4064,4064,4064,4064,4064,0...        0.000016\n",
      "53591  45270,88894,88894,88894,130929,130929,130929,1...        0.000016\n",
      "\n",
      "[53592 rows x 2 columns]\n",
      "Column name AMT OVERDUE - HIST\n",
      "                                                    index  AMT OVERDUE - HIST\n",
      "0                                                       ,            0.065307\n",
      "1       0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...            0.026440\n",
      "2                    ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,            0.020520\n",
      "3       0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...            0.019049\n",
      "4                                                      0,            0.017066\n",
      "...                                                   ...                 ...\n",
      "187336  ,,58455,57756,57520,57520,53863,53863,53863,52...            0.000002\n",
      "187337  0,,2083,2083,4166,2083,2083,2083,1842,1842,184...            0.000002\n",
      "187338  0,0,0,0,0,0,0,0,0,0,0,0,0,0,170,170,170,170,10...            0.000002\n",
      "187339  0,0,0,0,0,0,0,0,,0,0,0,0,0,0,,0,,0,0,0,0,0,0,0...            0.000002\n",
      "187340  0,0,0,0,0,0,0,0,0,0,,0,0,0,0,0,0,0,5005,5005,5...            0.000002\n",
      "\n",
      "[187341 rows x 2 columns]                                                    index  AMT OVERDUE - HIST\n",
      "0                                                      ,            0.056461\n",
      "1      0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...            0.040890\n",
      "2                   ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,            0.025497\n",
      "3      0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,...            0.023593\n",
      "4      0,,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0...            0.020876\n",
      "...                                                  ...                 ...\n",
      "22220        59647,0,0,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,            0.000016\n",
      "22221  0,0,0,,,,,,,,,,,202500,202500,202500,202500,20...            0.000016\n",
      "22222  0,0,9320,9320,0,4237,4237,4237,4237,4237,0,641...            0.000016\n",
      "22223  0,0,0,0,0,,0,0,0,0,,0,0,0,0,0,0,0,0,0,0,0,0,0,...            0.000016\n",
      "22224  0,,0,0,0,3138,82138,0,0,0,0,0,79940,0,0,0,0,0,...            0.000016\n",
      "\n",
      "[22225 rows x 2 columns]\n",
      "Column name AMT PAID - HIST\n",
      "                                                   index  AMT PAID - HIST\n",
      "0                  ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,         0.116924\n",
      "1                   ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,         0.082509\n",
      "2                                                      ,         0.077279\n",
      "3                                         ,,,,,,,,,,,,,,         0.025735\n",
      "4                                                     ,,         0.021162\n",
      "...                                                  ...              ...\n",
      "83412  12811,12811,12811,,12810,12811,,25811,,25811,1...         0.000002\n",
      "83413  0,32000,0,0,0,0,0,,0,0,0,101000,0,0,0,,0,,0,0,...         0.000002\n",
      "83414                 70281,0,0,0,0,0,0,0,0,0,0,0,0,640,         0.000002\n",
      "83415  162,120000,100800,70,59,20200,153,595000,,,900...         0.000002\n",
      "83416                10000,0,,0,,0,0,0,0,0,0,0,0,6000,0,         0.000002\n",
      "\n",
      "[83417 rows x 2 columns]                                                    index  AMT PAID - HIST\n",
      "0                  ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,         0.128221\n",
      "1                   ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,         0.099076\n",
      "2                                                      ,         0.067990\n",
      "3                                         ,,,,,,,,,,,,,,         0.020637\n",
      "4                                                     ,,         0.018371\n",
      "...                                                  ...              ...\n",
      "11021                       3000,3000,28500,3000,3000,0,         0.000016\n",
      "11022  51456,49312,47168,45024,42880,40736,38592,3644...         0.000016\n",
      "11023            ,,,,,,,,5163,,,,,,,643500,,,,,2478,,,,,         0.000016\n",
      "11024  ,,,,,,,10200,10200,10200,10200,10200,19990,102...         0.000016\n",
      "11025  137482,0,0,0,0,0,138911,0,0,0,0,0,0,0,0,0,0,0,...         0.000016\n",
      "\n",
      "[11026 rows x 2 columns]\n",
      "Column name TENURE\n",
      "     index    TENURE\n",
      "0     12.0  0.248478\n",
      "1     36.0  0.140616\n",
      "2      0.0  0.093656\n",
      "3     60.0  0.071828\n",
      "4     24.0  0.060352\n",
      "..     ...       ...\n",
      "319  436.0  0.000005\n",
      "320  266.0  0.000005\n",
      "321  244.0  0.000005\n",
      "322  391.0  0.000005\n",
      "323  306.0  0.000005\n",
      "\n",
      "[324 rows x 2 columns]      index    TENURE\n",
      "0     12.0  0.163836\n",
      "1     36.0  0.148839\n",
      "2      0.0  0.087666\n",
      "3     24.0  0.080475\n",
      "4     60.0  0.077542\n",
      "..     ...       ...\n",
      "188  110.0  0.000047\n",
      "189  164.0  0.000047\n",
      "190  720.0  0.000047\n",
      "191   86.0  0.000047\n",
      "192  135.0  0.000047\n",
      "\n",
      "[193 rows x 2 columns]\n"
     ]
    }
   ],
   "source": [
    "for i in test_b.columns:\n",
    "    print(\"Column name\",i)\n",
    "    print(train_b[i].value_counts(1).reset_index(),test_b[i].value_counts(1).reset_index())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>TENURE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12.0</td>\n",
       "      <td>0.248478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>36.0</td>\n",
       "      <td>0.140616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.093656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>60.0</td>\n",
       "      <td>0.071828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>24.0</td>\n",
       "      <td>0.060352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>48.0</td>\n",
       "      <td>0.037019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.026399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>30.0</td>\n",
       "      <td>0.026186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.021921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>18.0</td>\n",
       "      <td>0.017558</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index    TENURE\n",
       "0   12.0  0.248478\n",
       "1   36.0  0.140616\n",
       "2    0.0  0.093656\n",
       "3   60.0  0.071828\n",
       "4   24.0  0.060352\n",
       "5   48.0  0.037019\n",
       "6    1.0  0.026399\n",
       "7   30.0  0.026186\n",
       "8    6.0  0.021921\n",
       "9   18.0  0.017558"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_b['TENURE'].value_counts(1).reset_index().head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>TENURE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12.0</td>\n",
       "      <td>0.163836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>36.0</td>\n",
       "      <td>0.148839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.087666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>24.0</td>\n",
       "      <td>0.080475</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>60.0</td>\n",
       "      <td>0.077542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>48.0</td>\n",
       "      <td>0.044803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.024601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>84.0</td>\n",
       "      <td>0.020107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>30.0</td>\n",
       "      <td>0.019539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.018735</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index    TENURE\n",
       "0   12.0  0.163836\n",
       "1   36.0  0.148839\n",
       "2    0.0  0.087666\n",
       "3   24.0  0.080475\n",
       "4   60.0  0.077542\n",
       "5   48.0  0.044803\n",
       "6    6.0  0.024601\n",
       "7   84.0  0.020107\n",
       "8   30.0  0.019539\n",
       "9    1.0  0.018735"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_b['TENURE'].value_counts(1).reset_index().head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_bb = train_b.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_bb = test_b.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data Treatment:\n",
    "#     -categorical vars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# id -dont do it\n",
    "# self indiactor -nochange\n",
    "# match type- not req\n",
    "# acct type -required 3 columns chosen for remaning crosstab then combine cols\n",
    "    -\n",
    " #contributor type ARC is not ptesent in test so crosstab and then delete arc   \n",
    "# ownership indicator no rpoblem\n",
    "# account status not present in test Sold/Purchased ,Cancelled cross tab then delete                            \n",
    "# credit limit CREDIT-LIMIT/SANC AMT  DISBURSED-AMT/HIGH CREDIT kbindiscretization\n",
    "# installment not present in test data F04 F07  F08 crosstab then delete it\n",
    "# asset class 2 not ptresent in test crosstab and remove it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_SELF_INDICATOR=pd.crosstab(train.ID,train['SELF-INDICATOR'],dropna= True).reset_index()\n",
    "test_SELF_INDICATOR=pd.crosstab(test.ID,test['SELF-INDICATOR'],dropna= True).reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_MATCH_TYPE=pd.crosstab(train.ID,train['MATCH-TYPE'],dropna= True).reset_index()\n",
    "test_MATCH_TYPE=pd.crosstab(test.ID,test['MATCH-TYPE'],dropna= True).reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Property Loan', 'SHG Individual', 'Commercial Vehicle Loan', 'Construction Equipment Loan', 'Loan to Professional', 'Overdraft', 'Microfinance Business Loan', 'Business Non-Funded Credit Facility-Priority Sector-Agriculture', 'Two-Wheeler Loan', 'Business Loan General', 'Personal Loan', 'Loan Against Shares / Securities', 'Microfinance Personal Loan', 'Consumer Loan', 'Tractor Loan', 'Business Loan Unsecured', 'Prime Minister Jaan Dhan Yojana - Overdraft', 'Business Non-Funded Credit Facility General', 'Business Loan - Secured', 'Secured Credit Card', 'Auto Loan (Personal)', 'Gold Loan', 'Pradhan Mantri Awas Yojana - CLSS', 'Individual', 'Business Loan Against Bank Deposits', 'Microfinance Others', 'Mudra Loans   Shishu / Kishor / Tarun', 'Non-Funded Credit Facility', 'Business Loan Priority Sector  Others', 'Business Loan Priority Sector  Agriculture', 'Used Car Loan', 'Credit Card', 'Commercial Equipment Loan', 'Kisan Credit Card', 'Other', 'Loan Against Bank Deposits', 'Microfinance Housing Loan', 'Fleet Card', 'Business Loan Priority Sector  Small Business', 'Education Loan', 'JLG Individual', 'Housing Loan', 'Business Non-Funded Credit Facility-Priority Sector- Small Business', 'Corporate Credit Card']\n"
     ]
    }
   ],
   "source": [
    "x = set(train_bb['ACCT-TYPE'].unique())  \n",
    "y = set(test_bb['ACCT-TYPE'].unique()) \n",
    "common = x.intersection(y)\n",
    "print(list(common)) #now decide cols to keep "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_ACCT_TYPE=pd.crosstab(train.ID,train['ACCT-TYPE'],dropna= True).reset_index()\n",
    "test_ACCT_TYPE=pd.crosstab(test.ID,test['ACCT-TYPE'],dropna= True).reset_index()\n",
    "\n",
    "cols_to_keep =['Tractor Loan','Gold Loan','Business Loan Priority Sector  Agriculture']\n",
    "\n",
    "train_ACCT_TYPE[\"Others\"] = pd.DataFrame(train_ACCT_TYPE.drop(columns=cols_to_keep).sum(axis=1))\n",
    "test_ACCT_TYPE[\"Others\"] = pd.DataFrame(test_ACCT_TYPE.drop(columns=cols_to_keep).sum(axis=1))\n",
    "filter_cols =cols_to_keep + ['Others'] \n",
    "train_ACCT_TYPE = train_ACCT_TYPE.filter(filter_cols)\n",
    "test_ACCT_TYPE = test_ACCT_TYPE.filter(filter_cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_CONTRIBUTOR_TYPE=pd.crosstab(train.ID,train['CONTRIBUTOR-TYPE'],dropna= True).drop(columns = ['ARC']).reset_index()\n",
    "test_CONTRIBUTOR_TYPE=pd.crosstab(test.ID,test['CONTRIBUTOR-TYPE'],dropna= True).reset_index()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_OWNERSHIP_IND=pd.crosstab(train.ID,train['OWNERSHIP-IND'],dropna= True).reset_index()\n",
    "test_OWNERSHIP_IND=pd.crosstab(test.ID,test['OWNERSHIP-IND'],dropna= True).reset_index()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_ACCOUNT_STATUS=pd.crosstab(train.ID,train['ACCOUNT-STATUS'],dropna= True).drop(columns=['Sold/Purchased','Cancelled']).reset_index()\n",
    "test_ACCOUNT_STATUS=pd.crosstab(test.ID,test['ACCOUNT-STATUS'],dropna= True).reset_index()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_INSTALLMENT_FREQUENCY=pd.crosstab(train.ID,train['INSTALLMENT-FREQUENCY'],dropna= True).drop(columns=['F04','F07','F08']).reset_index()\n",
    "test_INSTALLMENT_FREQUENCY=pd.crosstab(test.ID,test['INSTALLMENT-FREQUENCY'],dropna= True).reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_ASSET_CLASS=pd.crosstab(train.ID,train['ASSET_CLASS'],dropna= True).drop(columns=['2']).reset_index()\n",
    "test_ASSET_CLASS=pd.crosstab(test.ID,test['ASSET_CLASS'],dropna= True).reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 314,
   "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>SELF-INDICATOR</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143396</th>\n",
       "      <td>16</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143397</th>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143398</th>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143399</th>\n",
       "      <td>12</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143400</th>\n",
       "      <td>12</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>14745 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "SELF-INDICATOR  False  True \n",
       "ID                          \n",
       "4                   3      1\n",
       "5                   7      1\n",
       "6                  10      2\n",
       "25                  1      1\n",
       "119                 1      1\n",
       "...               ...    ...\n",
       "143396             16     23\n",
       "143397              4      2\n",
       "143398              4      2\n",
       "143399             12      4\n",
       "143400             12      4\n",
       "\n",
       "[14745 rows x 2 columns]"
      ]
     },
     "execution_count": 314,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 376,
   "metadata": {},
   "outputs": [],
   "source": [
    "def onehotencoder(train,test,cols=[]):\n",
    "    \n",
    "    from sklearn.preprocessing import OneHotEncoder\n",
    "    #object_cols = catVar1(train) #catVar1 gives desired categorical column and not all object columns\n",
    "    object_cols=cols\n",
    "    print(object_cols)\n",
    "    \n",
    "    OH_encoder = OneHotEncoder(handle_unknown='ignore', sparse=False)\n",
    "    OH_cols_train = pd.DataFrame(OH_encoder.fit_transform(train[object_cols]))\n",
    "    OH_cols_test = pd.DataFrame(OH_encoder.transform(test[object_cols]))\n",
    "\n",
    "    # One-hot encoding removed index; put it back\n",
    "    OH_cols_train.index = train.index\n",
    "    OH_cols_test.index = test.index\n",
    "\n",
    "    ##hack for restoring columns names just like get dummies\n",
    "    column_name = OH_encoder.get_feature_names(object_cols)\n",
    "    OH_cols_train.columns = column_name\n",
    "    OH_cols_test.columns = column_name\n",
    "    \n",
    "\n",
    "    # Remove desired categorical columns (will replace with one-hot encoding)\n",
    "    num_train = train.drop(object_cols, axis=1)\n",
    "    num_test = test.drop(object_cols, axis=1)\n",
    "\n",
    "    # Add one-hot encoded columns to numerical/remaining features\n",
    "    OH_train = pd.concat([num_train, OH_cols_train], axis=1)\n",
    "    OH_test = pd.concat([num_test, OH_cols_test], axis=1)\n",
    "    \n",
    "\n",
    "    \n",
    "    print(OH_train.shape,OH_test.shape)\n",
    "    \n",
    "    \n",
    "    return OH_train,OH_test\n",
    "\n",
    "def targetencoding(train,test,y_train,drop=False,cols=[]):\n",
    "    import category_encoders as ce\n",
    "    # Create the encoder itself\n",
    "    cat_features = cols\n",
    "    print(f'targest emcoding for features {cat_features}')\n",
    "    target_enc = ce.TargetEncoder(cols=cat_features)\n",
    "\n",
    "    \n",
    "\n",
    "    # Fit the encoder using the categorical features and target\n",
    "    target_enc.fit(train[cat_features], y_train)\n",
    "    \n",
    "\n",
    "    # Transform the features, rename the columns with _target suffix, and join to dataframe\n",
    "    traintrgtenc = train.join(target_enc.transform(train[cat_features]).add_suffix('_target'))\n",
    "    testtrgtenc = test.join(target_enc.transform(test[cat_features]).add_suffix('_target'))\n",
    "\n",
    "    if drop :\n",
    "        traintrgtenc = traintrgtenc.drop(cat_features, axis =1)\n",
    "        testtrgtenc = testtrgtenc.drop(cat_features, axis =1)\n",
    "\n",
    "    \n",
    "    print(traintrgtenc.shape,testtrgtenc.shape)\n",
    "    \n",
    "    return traintrgtenc,testtrgtenc\n",
    "\n",
    "\n",
    "def ordinalencoding(train,test,y_train,mapping,drop=False,cols=[]):\n",
    "    import category_encoders as ce\n",
    "    # Create the encoder itself\n",
    "    cat_features = cols\n",
    "    print(f'ordinal encoding for features {cat_features}')\n",
    "    target_enc = ce.OrdinalEncoder(cols=cat_features,mapping = mapping)\n",
    "\n",
    "    \n",
    "\n",
    "    # Fit the encoder using the categorical features and target\n",
    "    target_enc.fit(train[cat_features], y_train)\n",
    "    \n",
    "\n",
    "    # Transform the features, rename the columns with _target suffix, and join to dataframe\n",
    "    traintrgtenc = train.join(target_enc.transform(train[cat_features]).add_suffix('_ordinal'))\n",
    "    testtrgtenc = test.join(target_enc.transform(test[cat_features]).add_suffix('_ordinal'))\n",
    "\n",
    "    if drop :\n",
    "        traintrgtenc = traintrgtenc.drop(cat_features, axis =1)\n",
    "        testtrgtenc = testtrgtenc.drop(cat_features, axis =1)\n",
    "\n",
    "    \n",
    "    print(traintrgtenc.shape,testtrgtenc.shape)\n",
    "    \n",
    "    return traintrgtenc,testtrgtenc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [],
   "source": [
    "def clusters(X,K=[2,3,4,5,6,7,8,9]):\n",
    "    inertias = [] \n",
    "    distortions=[]\n",
    "\n",
    "    for k in K:\n",
    "\n",
    "        km = KMeans(n_clusters=k, random_state=10)\n",
    "        cluster_labels = km.fit_predict(X)\n",
    "\n",
    "        distortions.append(sum(np.min(cdist(X, km.cluster_centers_, 'euclidean'),axis=1)) / X.shape[0]) \n",
    "\n",
    "        inertias.append(km.inertia_)\n",
    "\n",
    "    plt.plot(K, distortions, 'bx-') \n",
    "    plt.xlabel('Values of K') \n",
    "    plt.ylabel('Distortion') \n",
    "    plt.title('The Elbow Method using Distortion') \n",
    "    plt.show()\n",
    "\n",
    "    plt.plot(K, inertias, 'bx-') \n",
    "    plt.xlabel('Values of K') \n",
    "    plt.ylabel('Inertia') \n",
    "    plt.title('The Elbow Method using Inertia') \n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ID                                0\n",
       "SELF-INDICATOR                    0\n",
       "MATCH-TYPE                        0\n",
       "ACCT-TYPE                         0\n",
       "CONTRIBUTOR-TYPE                  0\n",
       "DATE-REPORTED                  3683\n",
       "OWNERSHIP-IND                     0\n",
       "ACCOUNT-STATUS                    0\n",
       "DISBURSED-DT                  32150\n",
       "CLOSE-DT                     251758\n",
       "LAST-PAYMENT-DATE            319283\n",
       "CREDIT-LIMIT/SANC AMT        545685\n",
       "DISBURSED-AMT/HIGH CREDIT         0\n",
       "INSTALLMENT-AMT              420509\n",
       "CURRENT-BAL                     233\n",
       "INSTALLMENT-FREQUENCY        425135\n",
       "OVERDUE-AMT                  118891\n",
       "WRITE-OFF-AMT                 19123\n",
       "ASSET_CLASS                  300376\n",
       "REPORTED DATE - HIST          19123\n",
       "DPD - HIST                    19647\n",
       "CUR BAL - HIST                19123\n",
       "AMT OVERDUE - HIST            19123\n",
       "AMT PAID - HIST               20294\n",
       "TENURE                       368107\n",
       "dtype: int64"
      ]
     },
     "execution_count": 249,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_bb.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_bb['DISBURSED-AMT/HIGH CREDIT'] = train_b['DISBURSED-AMT/HIGH CREDIT'].str.replace(',', '').astype('float')\n",
    "test_bb['DISBURSED-AMT/HIGH CREDIT'] = test_b['DISBURSED-AMT/HIGH CREDIT'].str.replace(',', '').astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_bb['CURRENT-BAL'] = train_b['CURRENT-BAL'].fillna(0)\n",
    "test_bb['CURRENT-BAL'] = test_b['CURRENT-BAL'].fillna(0)\n",
    "train_bb['CURRENT-BAL'] = train_b['CURRENT-BAL'].str.replace(',', '').astype('float')\n",
    "test_bb['CURRENT-BAL'] = test_b['CURRENT-BAL'].str.replace(',', '').astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for tenure just use drop na"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = train_bb['DISBURSED-AMT/HIGH CREDIT'].values.reshape(-1,1)\n",
    "clusters(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "X = train_bb['TENURE'].dropna().values.reshape(-1,1)\n",
    "clusters(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 335,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "train_CURRENT = train_bb[['CURRENT-BAL','ID']].dropna()\n",
    "test_CURRENT = test_bb[['CURRENT-BAL','ID']].dropna()\n",
    "\n",
    "enc = KBinsDiscretizer(n_bins=4, encode='ordinal', strategy='kmeans')\n",
    "enc.fit(train_bb['CURRENT-BAL'].dropna().values.reshape(-1,1))\n",
    "\n",
    "train_CURRENT['CURRENT-BAL'] = enc.transform(train_CURRENT['CURRENT-BAL'].values.reshape(-1,1))\n",
    "test_CURRENT['CURRENT-BAL']= enc.transform(test_CURRENT['CURRENT-BAL'].values.reshape(-1,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 336,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_CURRENT=pd.crosstab(train_CURRENT.ID,train_CURRENT['CURRENT-BAL'],dropna= True).reset_index()\n",
    "test_CURRENT=pd.crosstab(test_CURRENT.ID,test_CURRENT['CURRENT-BAL'],dropna= True).reset_index()\n",
    "test_CURRENT['3.0']=0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 341,
   "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>CURRENT-BAL</th>\n",
       "      <th>ID</th>\n",
       "      <th>0.0</th>\n",
       "      <th>1.0</th>\n",
       "      <th>2.0</th>\n",
       "      <th>3.0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>31</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128650</th>\n",
       "      <td>143390</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128651</th>\n",
       "      <td>143391</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128652</th>\n",
       "      <td>143393</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128653</th>\n",
       "      <td>143394</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128654</th>\n",
       "      <td>143395</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>128655 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "CURRENT-BAL      ID  0.0  1.0  2.0  3.0\n",
       "0                 1    0    9    0    0\n",
       "1                 2    0   13    0    0\n",
       "2                 3    0   31    0    0\n",
       "3                 7    0    4    0    0\n",
       "4                 8    0    7    0    0\n",
       "...             ...  ...  ...  ...  ...\n",
       "128650       143390    0    2    0    0\n",
       "128651       143391    0    1    0    0\n",
       "128652       143393    0    4    0    0\n",
       "128653       143394    0    1    0    0\n",
       "128654       143395    0    1    0    0\n",
       "\n",
       "[128655 rows x 5 columns]"
      ]
     },
     "execution_count": 341,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "train_DISBURSED = train_bb[['DISBURSED-AMT/HIGH CREDIT','ID']].dropna()\n",
    "test_DISBURSED = test_bb[['DISBURSED-AMT/HIGH CREDIT','ID']].dropna()\n",
    "\n",
    "enc = KBinsDiscretizer(n_bins=4, encode='ordinal', strategy='kmeans')\n",
    "enc.fit(train_bb['DISBURSED-AMT/HIGH CREDIT'].dropna().values.reshape(-1,1))\n",
    "\n",
    "train_DISBURSED['DISBURSED-AMT/HIGH CREDIT'] = enc.transform(train_DISBURSED['DISBURSED-AMT/HIGH CREDIT'].values.reshape(-1,1))\n",
    "test_DISBURSED['DISBURSED-AMT/HIGH CREDIT']= enc.transform(test_DISBURSED['DISBURSED-AMT/HIGH CREDIT'].values.reshape(-1,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 347,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_DISBURSED=pd.crosstab(train_DISBURSED.ID,train_DISBURSED['DISBURSED-AMT/HIGH CREDIT'],dropna= True).reset_index()\n",
    "test_DISBURSED=pd.crosstab(test_DISBURSED.ID,test_DISBURSED['DISBURSED-AMT/HIGH CREDIT'],dropna= True).reset_index()\n",
    "test_DISBURSED['2.0']=0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {},
   "outputs": [],
   "source": [
    "#discretisation baes on clusters\n",
    "\n",
    "train_tenure = train_bb[['TENURE','ID']].dropna()\n",
    "test_tenure = test_bb[['TENURE','ID']].dropna()\n",
    "\n",
    "enc = KBinsDiscretizer(n_bins=4, encode='ordinal', strategy='kmeans')\n",
    "enc.fit(train_bb['TENURE'].dropna().values.reshape(-1,1))\n",
    "\n",
    "train_tenure['TENURE'] = enc.transform(train_tenure['TENURE'].values.reshape(-1,1))\n",
    "test_tenure['TENURE']= enc.transform(test_tenure['TENURE'].values.reshape(-1,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_tenure=pd.crosstab(train_tenure.ID,train_tenure['TENURE'],dropna= True).reset_index()\n",
    "test_tenure=pd.crosstab(test_tenure.ID,test_tenure['TENURE'],dropna= True).reset_index()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 354,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataframes = [train_b.ID,train_SELF_INDICATOR,train_INSTALLMENT_FREQUENCY,train_ACCOUNT_STATUS,train_OWNERSHIP_IND,\n",
    "              train_CONTRIBUTOR_TYPE,train_ACCT_TYPE,train_MATCH_TYPE,train_ASSET_CLASS,train_CURRENT,train_DISBURSED,\n",
    "              train_tenure]\n",
    "final_df_train_beauro = reduce(lambda left, right: pd.merge(left, right, on='ID', how='left'), dataframes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 355,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataframes = [test_b.ID,test_SELF_INDICATOR,test_INSTALLMENT_FREQUENCY,test_ACCOUNT_STATUS,test_OWNERSHIP_IND,\n",
    "              test_CONTRIBUTOR_TYPE,test_ACCT_TYPE,test_MATCH_TYPE,test_ASSET_CLASS,test_CURRENT,test_DISBURSED,test_tenure]\n",
    "final_df_test_beauro = reduce(lambda left, right: pd.merge(left, right, on='ID', how='left'), dataframes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "###### now train data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 545,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dd = train_d.copy()\n",
    "test_dd = test_d.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 546,
   "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>ID</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>InstlmentMode</th>\n",
       "      <th>LoanStatus</th>\n",
       "      <th>PaymentMode</th>\n",
       "      <th>BranchID</th>\n",
       "      <th>Area</th>\n",
       "      <th>Tenure</th>\n",
       "      <th>AssetCost</th>\n",
       "      <th>AmountFinance</th>\n",
       "      <th>...</th>\n",
       "      <th>ManufacturerID</th>\n",
       "      <th>SupplierID</th>\n",
       "      <th>LTV</th>\n",
       "      <th>SEX</th>\n",
       "      <th>AGE</th>\n",
       "      <th>MonthlyIncome</th>\n",
       "      <th>City</th>\n",
       "      <th>State</th>\n",
       "      <th>ZiPCODE</th>\n",
       "      <th>Top-up Month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>Monthly</td>\n",
       "      <td>Arrear</td>\n",
       "      <td>Closed</td>\n",
       "      <td>PDC_E</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>48</td>\n",
       "      <td>450000</td>\n",
       "      <td>275000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1568</td>\n",
       "      <td>21946</td>\n",
       "      <td>61.11</td>\n",
       "      <td>M</td>\n",
       "      <td>49.0</td>\n",
       "      <td>35833.33</td>\n",
       "      <td>RAISEN</td>\n",
       "      <td>MADHYA PRADESH</td>\n",
       "      <td>464993.0</td>\n",
       "      <td>&gt; 48 Months</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>Monthly</td>\n",
       "      <td>Advance</td>\n",
       "      <td>Closed</td>\n",
       "      <td>PDC</td>\n",
       "      <td>333</td>\n",
       "      <td>BHOPAL</td>\n",
       "      <td>47</td>\n",
       "      <td>485000</td>\n",
       "      <td>350000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1062</td>\n",
       "      <td>34802</td>\n",
       "      <td>70.00</td>\n",
       "      <td>M</td>\n",
       "      <td>23.0</td>\n",
       "      <td>666.67</td>\n",
       "      <td>SEHORE</td>\n",
       "      <td>MADHYA PRADESH</td>\n",
       "      <td>466001.0</td>\n",
       "      <td>No Top-up Service</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>Quatrly</td>\n",
       "      <td>Arrear</td>\n",
       "      <td>Active</td>\n",
       "      <td>Direct Debit</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>68</td>\n",
       "      <td>690000</td>\n",
       "      <td>519728.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1060</td>\n",
       "      <td>127335</td>\n",
       "      <td>69.77</td>\n",
       "      <td>M</td>\n",
       "      <td>39.0</td>\n",
       "      <td>45257.00</td>\n",
       "      <td>BHOPAL</td>\n",
       "      <td>MADHYA PRADESH</td>\n",
       "      <td>462030.0</td>\n",
       "      <td>12-18 Months</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>7</td>\n",
       "      <td>Monthly</td>\n",
       "      <td>Advance</td>\n",
       "      <td>Closed</td>\n",
       "      <td>Billed</td>\n",
       "      <td>125</td>\n",
       "      <td>GUNA</td>\n",
       "      <td>48</td>\n",
       "      <td>480000</td>\n",
       "      <td>400000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1060</td>\n",
       "      <td>25094</td>\n",
       "      <td>80.92</td>\n",
       "      <td>M</td>\n",
       "      <td>24.0</td>\n",
       "      <td>20833.33</td>\n",
       "      <td>ASHOK NAGAR</td>\n",
       "      <td>MADHYA PRADESH</td>\n",
       "      <td>473335.0</td>\n",
       "      <td>&gt; 48 Months</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8</td>\n",
       "      <td>Monthly</td>\n",
       "      <td>Arrear</td>\n",
       "      <td>Closed</td>\n",
       "      <td>Billed</td>\n",
       "      <td>152</td>\n",
       "      <td>BILASPUR</td>\n",
       "      <td>44</td>\n",
       "      <td>619265</td>\n",
       "      <td>440000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1046</td>\n",
       "      <td>21853</td>\n",
       "      <td>71.05</td>\n",
       "      <td>M</td>\n",
       "      <td>56.0</td>\n",
       "      <td>27313.67</td>\n",
       "      <td>BILASPUR</td>\n",
       "      <td>CHATTISGARH</td>\n",
       "      <td>495442.0</td>\n",
       "      <td>36-48 Months</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ID Frequency InstlmentMode LoanStatus   PaymentMode  BranchID      Area  \\\n",
       "0   1   Monthly        Arrear     Closed         PDC_E         1       NaN   \n",
       "1   2   Monthly       Advance     Closed           PDC       333    BHOPAL   \n",
       "2   3   Quatrly        Arrear     Active  Direct Debit         1       NaN   \n",
       "3   7   Monthly       Advance     Closed        Billed       125      GUNA   \n",
       "4   8   Monthly        Arrear     Closed        Billed       152  BILASPUR   \n",
       "\n",
       "   Tenure  AssetCost  AmountFinance  ...  ManufacturerID  SupplierID    LTV  \\\n",
       "0      48     450000       275000.0  ...            1568       21946  61.11   \n",
       "1      47     485000       350000.0  ...            1062       34802  70.00   \n",
       "2      68     690000       519728.0  ...            1060      127335  69.77   \n",
       "3      48     480000       400000.0  ...            1060       25094  80.92   \n",
       "4      44     619265       440000.0  ...            1046       21853  71.05   \n",
       "\n",
       "  SEX   AGE  MonthlyIncome         City           State   ZiPCODE  \\\n",
       "0   M  49.0       35833.33       RAISEN  MADHYA PRADESH  464993.0   \n",
       "1   M  23.0         666.67       SEHORE  MADHYA PRADESH  466001.0   \n",
       "2   M  39.0       45257.00       BHOPAL  MADHYA PRADESH  462030.0   \n",
       "3   M  24.0       20833.33  ASHOK NAGAR  MADHYA PRADESH  473335.0   \n",
       "4   M  56.0       27313.67     BILASPUR     CHATTISGARH  495442.0   \n",
       "\n",
       "        Top-up Month  \n",
       "0        > 48 Months  \n",
       "1  No Top-up Service  \n",
       "2       12-18 Months  \n",
       "3        > 48 Months  \n",
       "4       36-48 Months  \n",
       "\n",
       "[5 rows x 26 columns]"
      ]
     },
     "execution_count": 546,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 547,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([' > 48 Months', 'No Top-up Service', '12-18 Months',\n",
       "       '36-48 Months', '18-24 Months', '24-30 Months', '30-36 Months'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 547,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['Top-up Month'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 548,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ID                     0\n",
       "Frequency              0\n",
       "InstlmentMode          0\n",
       "LoanStatus             0\n",
       "PaymentMode            0\n",
       "BranchID               0\n",
       "Area               11653\n",
       "Tenure                 0\n",
       "AssetCost              0\n",
       "AmountFinance          0\n",
       "DisbursalAmount        0\n",
       "EMI                    0\n",
       "DisbursalDate          0\n",
       "MaturityDAte           1\n",
       "AuthDate               0\n",
       "AssetID                0\n",
       "ManufacturerID         0\n",
       "SupplierID             0\n",
       "LTV                    0\n",
       "SEX                   59\n",
       "AGE                   59\n",
       "MonthlyIncome        234\n",
       "City               11256\n",
       "State                  0\n",
       "ZiPCODE              372\n",
       "Top-up Month           0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 548,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_dd.isna().sum()\n",
    "train_dd.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 549,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dd['SEX'] = train_dd['SEX'].fillna(train_dd['SEX'].mode()[0])\n",
    "test_dd['SEX'] = test_dd['SEX'].fillna(test_dd['SEX'].mode()[0])\n",
    "\n",
    "train_dd['AGE'] = train_dd['AGE'].fillna(train_dd['AGE'].mode()[0])\n",
    "test_dd['AGE'] = test_dd['AGE'].fillna(test_dd['AGE'].mode()[0])\n",
    "\n",
    "train_dd['MonthlyIncome'] = train_dd['MonthlyIncome'].fillna(train_dd['MonthlyIncome'].mode()[0])\n",
    "test_dd['MonthlyIncome'] = test_dd['MonthlyIncome'].fillna(test_dd['MonthlyIncome'].mode()[0])\n",
    "\n",
    "train_dd['ZiPCODE']  =train_dd['ZiPCODE'].fillna(method = 'ffill').astype('int') \n",
    "test_dd['ZiPCODE']  =test_dd['ZiPCODE'].fillna(method = 'ffill').astype('int') \n",
    "\n",
    "train_dd =train_dd.fillna(method = 'ffill') \n",
    "test_dd =test_dd.fillna(method = 'ffill') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 550,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dd['AuthDate'] = pd.to_datetime(train_d['AuthDate'] ,errors='coerce')\n",
    "test_dd['AuthDate'] = pd.to_datetime(test_d['AuthDate'] ,errors='coerce')\n",
    "\n",
    "train_dd['DisbursalDate'] = pd.to_datetime(train_d['DisbursalDate'] ,errors='coerce')\n",
    "test_dd['DisbursalDate'] = pd.to_datetime(test_d['DisbursalDate'] ,errors='coerce')\n",
    "\n",
    "train_dd['MaturityDAte'] = pd.to_datetime(train_d['MaturityDAte'] ,errors='coerce')\n",
    "test_dd['MaturityDAte'] = pd.to_datetime(test_d['MaturityDAte'] ,errors='coerce')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 551,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dd['AuthtoMatureDays'] = train_dd['AuthDate'] - train_dd['MaturityDAte']\n",
    "test_dd['AuthtoMatureDays'] = test_dd['AuthDate'] - test_dd['MaturityDAte']\n",
    "\n",
    "train_dd['AuthtoDisDays'] = train_dd['AuthDate'] - train_dd['DisbursalDate']\n",
    "test_dd['AuthtoDisDays'] = test_dd['AuthDate'] - test_dd['DisbursalDate']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 552,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dd['AuthYear'] = train_dd['AuthDate'].dt.year \n",
    "test_dd['AuthtYear'] = test_dd['AuthDate'].dt.year \n",
    "train_dd['AuthMonth'] = train_dd['AuthDate'].dt.month \n",
    "test_dd['AuthtMonth'] = test_dd['AuthDate'].dt.month \n",
    "\n",
    "\n",
    "train_dd['MaturityYear'] = train_d['MaturityDAte'].dt.year \n",
    "test_dd['MaturityYear'] = test_d['MaturityDAte'].dt.year \n",
    "\n",
    "train_dd['MaturityMonth'] = train_d['MaturityDAte'].dt.month \n",
    "test_dd['MaturityMonth'] = test_d['MaturityDAte'].dt.month "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 553,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dd = train_dd.drop(columns=['City','Area','AuthDate','DisbursalDate','MaturityDAte'])\n",
    "test_dd = test_dd.drop(columns=['City','Area','AuthDate','DisbursalDate','MaturityDAte'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 554,
   "metadata": {},
   "outputs": [],
   "source": [
    "mapping_dict = {'No Top-up Service':0, '12-18 Months':1,'18-24 Months':2, '24-30 Months':3, \n",
    "                '30-36 Months':4,'36-48 Months':5,' > 48 Months':6}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 555,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dd['Top-up Month'] = train_dd['Top-up Month'].replace(mapping_dict) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 556,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    106677\n",
       "6      8366\n",
       "5      3656\n",
       "3      3492\n",
       "4      3062\n",
       "2      2368\n",
       "1      1034\n",
       "Name: Top-up Month, dtype: int64"
      ]
     },
     "execution_count": 556,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['Top-up Month'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 557,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Monthly', 'Quatrly', 'Half Yearly', 'BI-Monthly'], dtype=object)"
      ]
     },
     "execution_count": 557,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['Frequency'].unique()\n",
    "#test_dd['Frequency'].unique() #ordinal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 558,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Advance', 'Arrear'], dtype=object)"
      ]
     },
     "execution_count": 558,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['InstlmentMode'].unique()\n",
    "test_dd['InstlmentMode'].unique() #one hot encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 559,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Closed', 'Active'], dtype=object)"
      ]
     },
     "execution_count": 559,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['LoanStatus'].unique()\n",
    "test_dd['LoanStatus'].unique() #one hot encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 560,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['PDC_E', 'PDC', 'Billed', 'Direct Debit', 'ECS', 'ECS Reject',\n",
       "       'SI Reject', 'Cheque', 'Auto Debit', 'PDC Reject'], dtype=object)"
      ]
     },
     "execution_count": 560,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['PaymentMode'].unique()\n",
    "test_dd['PaymentMode'].unique()  # target encoding\n",
    "\n",
    "# ['PDC_E', 'PDC', 'Direct Debit', 'Billed', 'ECS', 'Auto Debit',\n",
    "#        'SI Reject', 'ECS Reject', 'Cheque', 'PDC Reject', 'Escrow']   \n",
    "\n",
    "# ['PDC_E', 'PDC', 'Billed', 'Direct Debit', 'ECS', 'ECS Reject',\n",
    "#        'SI Reject', 'Cheque', 'Auto Debit', 'PDC Reject']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 561,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  1, 333, 125, 152,   5, 154,  79,   8, 194, 316,   9,  10, 231,\n",
       "       289, 149, 229,  16,  17,  19,  20, 200,  21, 281, 110, 205, 341,\n",
       "        22, 311,  29,  24,  26, 202,  25, 243, 272,  32,  38, 309,  39,\n",
       "        40,  41, 334,  45,  46,  47, 340,  48,  50, 305,  53,  54,  55,\n",
       "       304, 196, 292, 310,  57,  59, 296,  62, 124,  67,  68,  70,  73,\n",
       "       324, 275,  71, 318, 338, 314,  72,  76,  85,  91,  93, 130, 131,\n",
       "       133, 137, 141, 283, 148, 143, 280, 263, 147, 150, 163,  56, 166,\n",
       "       199, 173, 189, 298, 327, 193, 294, 204, 206, 407, 207, 246, 273,\n",
       "       301, 376, 343, 116, 325,   6, 330, 336,   3, 328, 308, 209, 265,\n",
       "       320, 306,  11,  14,  23, 322, 271,  33, 300, 278, 145, 312, 106,\n",
       "        69, 332, 293, 164,  88,  89,  95, 120, 313, 136, 138, 139, 337,\n",
       "       302, 326, 342, 151, 162, 335, 303, 195, 379, 370, 406, 380, 323,\n",
       "       269, 270, 233, 423, 274, 276, 279, 344, 357, 359, 375, 377, 386,\n",
       "       387, 388, 389, 390, 393, 395, 397, 400, 401, 404, 405, 412, 413,\n",
       "       414, 416, 418, 421, 424, 277, 408], dtype=int64)"
      ]
     },
     "execution_count": 561,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['BranchID'].unique()\n",
    "# test_dd['BranchID'].unique()  # target or leave it as it is delete area and city "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 562,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "36     3770\n",
       "48     3597\n",
       "60     2029\n",
       "24     1416\n",
       "42      578\n",
       "       ... \n",
       "75        1\n",
       "108       1\n",
       "92        1\n",
       "86        1\n",
       "270       1\n",
       "Name: Tenure, Length: 98, dtype: int64"
      ]
     },
     "execution_count": 562,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['Tenure'].value_counts()\n",
    "test_dd['Tenure'].value_counts() #can be done as target encoding or leave it as it is"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 563,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1049.0    3524\n",
       "1046.0    3087\n",
       "1186.0    1728\n",
       "1019.0    1542\n",
       "1060.0    1533\n",
       "1062.0    1298\n",
       "1568.0    1183\n",
       "1187.0     735\n",
       "3473.0      74\n",
       "2965.0      20\n",
       "2608.0      19\n",
       "2721.0       2\n",
       "Name: ManufacturerID, dtype: int64"
      ]
     },
     "execution_count": 563,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['ManufacturerID'].value_counts()\n",
    "test_dd['ManufacturerID'].value_counts() #can be done as target encoding or leave it as it is"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 564,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "42707     1465\n",
       "29012     1301\n",
       "38832     1039\n",
       "89813     1004\n",
       "27328      735\n",
       "          ... \n",
       "125748       1\n",
       "105362       1\n",
       "41757        1\n",
       "140389       1\n",
       "137117       1\n",
       "Name: SupplierID, Length: 4539, dtype: int64"
      ]
     },
     "execution_count": 564,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['SupplierID'].value_counts()\n",
    "#test_dd['SupplierID'].value_counts() #can be done as target encoding or leave it as it is"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 565,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "14425238    1\n",
       "26856210    1\n",
       "432888      1\n",
       "25671318    1\n",
       "14094476    1\n",
       "           ..\n",
       "23303785    1\n",
       "7931765     1\n",
       "16423430    1\n",
       "1666936     1\n",
       "28319748    1\n",
       "Name: AssetID, Length: 128655, dtype: int64"
      ]
     },
     "execution_count": 565,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['AssetID'].value_counts()\n",
    "#test_dd['AssetID'].value_counts() #no repetition label enco"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 566,
   "metadata": {},
   "outputs": [],
   "source": [
    "#.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 567,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "125001    806\n",
       "125055    691\n",
       "584128    524\n",
       "125050    409\n",
       "334001    407\n",
       "         ... \n",
       "517297      1\n",
       "335022      1\n",
       "501504      1\n",
       "412804      1\n",
       "325601      1\n",
       "Name: ZiPCODE, Length: 9123, dtype: int64"
      ]
     },
     "execution_count": 567,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_dd['ZiPCODE'].value_counts() #can be again encoded as target or label that is leave as it is."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 568,
   "metadata": {},
   "outputs": [],
   "source": [
    "#sex one hot encoding male female"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 569,
   "metadata": {},
   "outputs": [],
   "source": [
    "mapping_dict_f= [\n",
    "{'col': 'Frequency', 'mapping': {'BI-Monthly':0, 'Monthly': 1, 'Quatrly': 2,  'Half Yearly': 3}}\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 570,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ordinal encoding for features ['Frequency']\n",
      "(128655, 27) (14745, 26)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\HP\\anaconda3\\lib\\site-packages\\category_encoders\\utils.py:21: FutureWarning: is_categorical is deprecated and will be removed in a future version.  Use is_categorical_dtype instead\n",
      "  elif pd.api.types.is_categorical(cols):\n"
     ]
    }
   ],
   "source": [
    "f1,f2 = ordinalencoding(train_dd,test_dd,train_dd['Top-up Month'],drop =True,cols=['Frequency'],mapping= mapping_dict_f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 571,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['InstlmentMode', 'LoanStatus', 'SEX']\n",
      "(128655, 30) (14745, 29)\n"
     ]
    }
   ],
   "source": [
    "f11,f22 = onehotencoder(f1,f2,cols=['InstlmentMode','LoanStatus','SEX'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 572,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "targest emcoding for features ['PaymentMode', 'State', 'BranchID', 'ZiPCODE', 'ManufacturerID']\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\HP\\anaconda3\\lib\\site-packages\\category_encoders\\utils.py:21: FutureWarning: is_categorical is deprecated and will be removed in a future version.  Use is_categorical_dtype instead\n",
      "  elif pd.api.types.is_categorical(cols):\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(128655, 30) (14745, 29)\n"
     ]
    }
   ],
   "source": [
    "f111,f222 = targetencoding(f11,f22,f11['Top-up Month'],drop=True,cols=['PaymentMode','State','BranchID','ZiPCODE','ManufacturerID']) #because branch id and area same"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 573,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['ID', 'Tenure', 'AssetCost', 'AmountFinance', 'DisbursalAmount', 'EMI',\n",
       "       'AssetID', 'SupplierID', 'LTV', 'AGE', 'MonthlyIncome', 'Top-up Month',\n",
       "       'AuthtoMatureDays', 'AuthtoDisDays', 'AuthYear', 'AuthMonth',\n",
       "       'MaturityYear', 'MaturityMonth', 'Frequency_ordinal',\n",
       "       'InstlmentMode_Advance', 'InstlmentMode_Arrear', 'LoanStatus_Active',\n",
       "       'LoanStatus_Closed', 'SEX_F', 'SEX_M', 'PaymentMode_target',\n",
       "       'State_target', 'BranchID_target', 'ZiPCODE_target',\n",
       "       'ManufacturerID_target'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 573,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f111.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 576,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataframes = [f111,final_df_train_beauro]\n",
    "final_df_train = reduce(lambda left, right: pd.merge(left, right, on='ID', how='left'), dataframes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 575,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataframes = [f222,final_df_test_beauro]\n",
    "final_df_test = reduce(lambda left, right: pd.merge(left, right, on='ID', how='left'), dataframes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 589,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ID                      0\n",
       "Tenure                  0\n",
       "AssetCost               0\n",
       "AmountFinance           0\n",
       "DisbursalAmount         0\n",
       "                    ...  \n",
       "3.0_y                   0\n",
       "0.0_y              145781\n",
       "1.0_y              145781\n",
       "2.0                145781\n",
       "3.0                145781\n",
       "Length: 88, dtype: int64"
      ]
     },
     "execution_count": 589,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df_train.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}