moocs/fast.ai/deep-learning-2018/experiment3-home-credit-default-risk.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Experiment 3: Home Credit Default Risk\n",
    "\n",
    "This notebook attempts to apply the lessons taught in Fast.ai Lesson 4 and 5 to another, similar dataset.\n",
    "\n",
    "## Contents\n",
    "\n",
    "1. Explore and visualise dataset.\n",
    "2. Prepare dataset.\n",
    "3. Build and train model.\n",
    "4. Evaluation.\n",
    "5. Ideas for improvements."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline\n",
    "\n",
    "from pathlib import Path\n",
    "import os\n",
    "from IPython.display import FileLink\n",
    "\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from fastai.fastai.structured import proc_df, apply_cats\n",
    "from fastai.fastai.dataset import get_cv_idxs\n",
    "from fastai.fastai.column_data import ColumnarModelData\n",
    "from sklearn.model_selection import train_test_split\n",
    "from fastai.fastai.metrics import accuracy\n",
    "from sklearn.metrics import roc_auc_score\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "sns.set(color_codes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "PATH = Path('./data/home-credit-default-risk')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "PATH.mkdir(exist_ok=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## 1. Explore and visualise dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true,
    "hidden": true
   },
   "source": [
    "### Download and extract dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sample_submission.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "application_test.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "application_train.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "bureau.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "bureau_balance.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "previous_application.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "credit_card_balance.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "POS_CASH_balance.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "installments_payments.csv.zip: Skipping, found more recently modified local copy (use --force to force download)\n",
      "HomeCredit_columns_description.csv: Downloaded 37KB of 37KB to data/home-credit-default-risk\n"
     ]
    }
   ],
   "source": [
    "# Get dataset\n",
    "!kaggle competitions download -c home-credit-default-risk --path={PATH}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "for file in os.listdir(PATH):\n",
    "    if not file.endswith('zip'):\n",
    "        continue\n",
    "    \n",
    "    file_path = PATH / file\n",
    "\n",
    "    !unzip -q -d {PATH} {file_path}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "POS_CASH_balance.csv          bureau_balance.csv.zip\r\n",
      "POS_CASH_balance.csv.zip      credit_card_balance.csv\r\n",
      "application_test.csv          credit_card_balance.csv.zip\r\n",
      "application_test.csv.zip      installments_payments.csv\r\n",
      "application_train.csv         installments_payments.csv.zip\r\n",
      "application_train.csv.zip     previous_application.csv\r\n",
      "bureau.csv                    previous_application.csv.zip\r\n",
      "bureau.csv.zip                sample_submission.csv\r\n",
      "bureau_balance.csv            sample_submission.csv.zip\r\n"
     ]
    }
   ],
   "source": [
    "!ls {PATH}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true,
    "hidden": true
   },
   "source": [
    "### application_train\n",
    "\n",
    "From Kaggle data page:\n",
    "\n",
    "*\"This is the main table, broken into two files for Train (with TARGET) and Test (without TARGET). Static data for all applications. One row represents one loan in our data sample.\"*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "application_train = pd.read_csv(PATH / 'application_train.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th>TARGET</th>\n",
       "      <th>NAME_CONTRACT_TYPE</th>\n",
       "      <th>CODE_GENDER</th>\n",
       "      <th>FLAG_OWN_CAR</th>\n",
       "      <th>FLAG_OWN_REALTY</th>\n",
       "      <th>CNT_CHILDREN</th>\n",
       "      <th>AMT_INCOME_TOTAL</th>\n",
       "      <th>AMT_CREDIT</th>\n",
       "      <th>AMT_ANNUITY</th>\n",
       "      <th>...</th>\n",
       "      <th>FLAG_DOCUMENT_18</th>\n",
       "      <th>FLAG_DOCUMENT_19</th>\n",
       "      <th>FLAG_DOCUMENT_20</th>\n",
       "      <th>FLAG_DOCUMENT_21</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_HOUR</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_DAY</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_WEEK</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_MON</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_QRT</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_YEAR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100002</td>\n",
       "      <td>1</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>M</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>202500.0</td>\n",
       "      <td>406597.5</td>\n",
       "      <td>24700.5</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100003</td>\n",
       "      <td>0</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>F</td>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "      <td>270000.0</td>\n",
       "      <td>1293502.5</td>\n",
       "      <td>35698.5</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100004</td>\n",
       "      <td>0</td>\n",
       "      <td>Revolving loans</td>\n",
       "      <td>M</td>\n",
       "      <td>Y</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>67500.0</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>6750.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100006</td>\n",
       "      <td>0</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>F</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>312682.5</td>\n",
       "      <td>29686.5</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100007</td>\n",
       "      <td>0</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>M</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>121500.0</td>\n",
       "      <td>513000.0</td>\n",
       "      <td>21865.5</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 122 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   SK_ID_CURR  TARGET NAME_CONTRACT_TYPE CODE_GENDER FLAG_OWN_CAR  \\\n",
       "0      100002       1         Cash loans           M            N   \n",
       "1      100003       0         Cash loans           F            N   \n",
       "2      100004       0    Revolving loans           M            Y   \n",
       "3      100006       0         Cash loans           F            N   \n",
       "4      100007       0         Cash loans           M            N   \n",
       "\n",
       "  FLAG_OWN_REALTY  CNT_CHILDREN  AMT_INCOME_TOTAL  AMT_CREDIT  AMT_ANNUITY  \\\n",
       "0               Y             0          202500.0    406597.5      24700.5   \n",
       "1               N             0          270000.0   1293502.5      35698.5   \n",
       "2               Y             0           67500.0    135000.0       6750.0   \n",
       "3               Y             0          135000.0    312682.5      29686.5   \n",
       "4               Y             0          121500.0    513000.0      21865.5   \n",
       "\n",
       "              ...              FLAG_DOCUMENT_18 FLAG_DOCUMENT_19  \\\n",
       "0             ...                             0                0   \n",
       "1             ...                             0                0   \n",
       "2             ...                             0                0   \n",
       "3             ...                             0                0   \n",
       "4             ...                             0                0   \n",
       "\n",
       "  FLAG_DOCUMENT_20 FLAG_DOCUMENT_21 AMT_REQ_CREDIT_BUREAU_HOUR  \\\n",
       "0                0                0                        0.0   \n",
       "1                0                0                        0.0   \n",
       "2                0                0                        0.0   \n",
       "3                0                0                        NaN   \n",
       "4                0                0                        0.0   \n",
       "\n",
       "  AMT_REQ_CREDIT_BUREAU_DAY  AMT_REQ_CREDIT_BUREAU_WEEK  \\\n",
       "0                       0.0                         0.0   \n",
       "1                       0.0                         0.0   \n",
       "2                       0.0                         0.0   \n",
       "3                       NaN                         NaN   \n",
       "4                       0.0                         0.0   \n",
       "\n",
       "   AMT_REQ_CREDIT_BUREAU_MON  AMT_REQ_CREDIT_BUREAU_QRT  \\\n",
       "0                        0.0                        0.0   \n",
       "1                        0.0                        0.0   \n",
       "2                        0.0                        0.0   \n",
       "3                        NaN                        NaN   \n",
       "4                        0.0                        0.0   \n",
       "\n",
       "   AMT_REQ_CREDIT_BUREAU_YEAR  \n",
       "0                         1.0  \n",
       "1                         0.0  \n",
       "2                         0.0  \n",
       "3                         NaN  \n",
       "4                         0.0  \n",
       "\n",
       "[5 rows x 122 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "application_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "307511"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(application_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Number of rows where `TARGET = 1`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num target = 1:  24825\n",
      "Num target = 0:  282686\n"
     ]
    }
   ],
   "source": [
    "print(\"Num target = 1: \", len(application_train[application_train['TARGET'] == 1]))\n",
    "print(\"Num target = 0: \", len(application_train[application_train['TARGET'] == 0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "application_test = pd.read_csv(PATH / 'application_test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th>NAME_CONTRACT_TYPE</th>\n",
       "      <th>CODE_GENDER</th>\n",
       "      <th>FLAG_OWN_CAR</th>\n",
       "      <th>FLAG_OWN_REALTY</th>\n",
       "      <th>CNT_CHILDREN</th>\n",
       "      <th>AMT_INCOME_TOTAL</th>\n",
       "      <th>AMT_CREDIT</th>\n",
       "      <th>AMT_ANNUITY</th>\n",
       "      <th>AMT_GOODS_PRICE</th>\n",
       "      <th>...</th>\n",
       "      <th>FLAG_DOCUMENT_18</th>\n",
       "      <th>FLAG_DOCUMENT_19</th>\n",
       "      <th>FLAG_DOCUMENT_20</th>\n",
       "      <th>FLAG_DOCUMENT_21</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_HOUR</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_DAY</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_WEEK</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_MON</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_QRT</th>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_YEAR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100001</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>F</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>568800.0</td>\n",
       "      <td>20560.5</td>\n",
       "      <td>450000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100005</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>M</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>99000.0</td>\n",
       "      <td>222768.0</td>\n",
       "      <td>17370.0</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100013</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>M</td>\n",
       "      <td>Y</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>202500.0</td>\n",
       "      <td>663264.0</td>\n",
       "      <td>69777.0</td>\n",
       "      <td>630000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100028</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>F</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>2</td>\n",
       "      <td>315000.0</td>\n",
       "      <td>1575000.0</td>\n",
       "      <td>49018.5</td>\n",
       "      <td>1575000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100038</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>M</td>\n",
       "      <td>Y</td>\n",
       "      <td>N</td>\n",
       "      <td>1</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>625500.0</td>\n",
       "      <td>32067.0</td>\n",
       "      <td>625500.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 121 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   SK_ID_CURR NAME_CONTRACT_TYPE CODE_GENDER FLAG_OWN_CAR FLAG_OWN_REALTY  \\\n",
       "0      100001         Cash loans           F            N               Y   \n",
       "1      100005         Cash loans           M            N               Y   \n",
       "2      100013         Cash loans           M            Y               Y   \n",
       "3      100028         Cash loans           F            N               Y   \n",
       "4      100038         Cash loans           M            Y               N   \n",
       "\n",
       "   CNT_CHILDREN  AMT_INCOME_TOTAL  AMT_CREDIT  AMT_ANNUITY  AMT_GOODS_PRICE  \\\n",
       "0             0          135000.0    568800.0      20560.5         450000.0   \n",
       "1             0           99000.0    222768.0      17370.0         180000.0   \n",
       "2             0          202500.0    663264.0      69777.0         630000.0   \n",
       "3             2          315000.0   1575000.0      49018.5        1575000.0   \n",
       "4             1          180000.0    625500.0      32067.0         625500.0   \n",
       "\n",
       "              ...             FLAG_DOCUMENT_18 FLAG_DOCUMENT_19  \\\n",
       "0             ...                            0                0   \n",
       "1             ...                            0                0   \n",
       "2             ...                            0                0   \n",
       "3             ...                            0                0   \n",
       "4             ...                            0                0   \n",
       "\n",
       "  FLAG_DOCUMENT_20 FLAG_DOCUMENT_21 AMT_REQ_CREDIT_BUREAU_HOUR  \\\n",
       "0                0                0                        0.0   \n",
       "1                0                0                        0.0   \n",
       "2                0                0                        0.0   \n",
       "3                0                0                        0.0   \n",
       "4                0                0                        NaN   \n",
       "\n",
       "   AMT_REQ_CREDIT_BUREAU_DAY  AMT_REQ_CREDIT_BUREAU_WEEK  \\\n",
       "0                        0.0                         0.0   \n",
       "1                        0.0                         0.0   \n",
       "2                        0.0                         0.0   \n",
       "3                        0.0                         0.0   \n",
       "4                        NaN                         NaN   \n",
       "\n",
       "   AMT_REQ_CREDIT_BUREAU_MON  AMT_REQ_CREDIT_BUREAU_QRT  \\\n",
       "0                        0.0                        0.0   \n",
       "1                        0.0                        0.0   \n",
       "2                        0.0                        1.0   \n",
       "3                        0.0                        0.0   \n",
       "4                        NaN                        NaN   \n",
       "\n",
       "   AMT_REQ_CREDIT_BUREAU_YEAR  \n",
       "0                         0.0  \n",
       "1                         3.0  \n",
       "2                         4.0  \n",
       "3                         3.0  \n",
       "4                         NaN  \n",
       "\n",
       "[5 rows x 121 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "application_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "There's a file called `HomeCredit_columns_description.csv` that has a description of each column. Let's take a look at that."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "column_desc = pd.read_csv(PATH / 'HomeCredit_columns_description.csv', encoding='latin_1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>Table</th>\n",
       "      <th>Row</th>\n",
       "      <th>Description</th>\n",
       "      <th>Special</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>SK_ID_CURR</td>\n",
       "      <td>ID of loan in our sample</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>TARGET</td>\n",
       "      <td>Target variable (1 - client with payment diffi...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>NAME_CONTRACT_TYPE</td>\n",
       "      <td>Identification if loan is cash or revolving</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>CODE_GENDER</td>\n",
       "      <td>Gender of the client</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>7</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_OWN_CAR</td>\n",
       "      <td>Flag if the client owns a car</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0                         Table                 Row  \\\n",
       "0           1  application_{train|test}.csv          SK_ID_CURR   \n",
       "1           2  application_{train|test}.csv              TARGET   \n",
       "2           5  application_{train|test}.csv  NAME_CONTRACT_TYPE   \n",
       "3           6  application_{train|test}.csv         CODE_GENDER   \n",
       "4           7  application_{train|test}.csv        FLAG_OWN_CAR   \n",
       "\n",
       "                                         Description Special  \n",
       "0                           ID of loan in our sample     NaN  \n",
       "1  Target variable (1 - client with payment diffi...     NaN  \n",
       "2        Identification if loan is cash or revolving     NaN  \n",
       "3                               Gender of the client     NaN  \n",
       "4                      Flag if the client owns a car     NaN  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "column_desc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>Table</th>\n",
       "      <th>Row</th>\n",
       "      <th>Description</th>\n",
       "      <th>Special</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>SK_ID_CURR</td>\n",
       "      <td>ID of loan in our sample</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>TARGET</td>\n",
       "      <td>Target variable (1 - client with payment diffi...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>NAME_CONTRACT_TYPE</td>\n",
       "      <td>Identification if loan is cash or revolving</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>6</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>CODE_GENDER</td>\n",
       "      <td>Gender of the client</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>7</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_OWN_CAR</td>\n",
       "      <td>Flag if the client owns a car</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>8</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_OWN_REALTY</td>\n",
       "      <td>Flag if client owns a house or flat</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>9</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>CNT_CHILDREN</td>\n",
       "      <td>Number of children the client has</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>10</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_INCOME_TOTAL</td>\n",
       "      <td>Income of the client</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>11</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_CREDIT</td>\n",
       "      <td>Credit amount of the loan</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>12</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_ANNUITY</td>\n",
       "      <td>Loan annuity</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>13</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_GOODS_PRICE</td>\n",
       "      <td>For consumer loans it is the price of the good...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>14</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>NAME_TYPE_SUITE</td>\n",
       "      <td>Who was accompanying client when he was applyi...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>15</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>NAME_INCOME_TYPE</td>\n",
       "      <td>Clients income type (businessman, working, mat...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>16</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>NAME_EDUCATION_TYPE</td>\n",
       "      <td>Level of highest education the client achieved</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>17</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>NAME_FAMILY_STATUS</td>\n",
       "      <td>Family status of the client</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>18</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>NAME_HOUSING_TYPE</td>\n",
       "      <td>What is the housing situation of the client (r...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>19</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>REGION_POPULATION_RELATIVE</td>\n",
       "      <td>Normalized population of region where client l...</td>\n",
       "      <td>normalized</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>20</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>DAYS_BIRTH</td>\n",
       "      <td>Client's age in days at the time of application</td>\n",
       "      <td>time only relative to the application</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>21</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>DAYS_EMPLOYED</td>\n",
       "      <td>How many days before the application the perso...</td>\n",
       "      <td>time only relative to the application</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>22</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>DAYS_REGISTRATION</td>\n",
       "      <td>How many days before the application did clien...</td>\n",
       "      <td>time only relative to the application</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>23</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>DAYS_ID_PUBLISH</td>\n",
       "      <td>How many days before the application did clien...</td>\n",
       "      <td>time only relative to the application</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>24</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>OWN_CAR_AGE</td>\n",
       "      <td>Age of client's car</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>25</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_MOBIL</td>\n",
       "      <td>Did client provide mobile phone (1=YES, 0=NO)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>26</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_EMP_PHONE</td>\n",
       "      <td>Did client provide work phone (1=YES, 0=NO)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>27</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_WORK_PHONE</td>\n",
       "      <td>Did client provide home phone (1=YES, 0=NO)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>28</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_CONT_MOBILE</td>\n",
       "      <td>Was mobile phone reachable (1=YES, 0=NO)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>29</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_PHONE</td>\n",
       "      <td>Did client provide home phone (1=YES, 0=NO)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>30</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_EMAIL</td>\n",
       "      <td>Did client provide email (1=YES, 0=NO)</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>31</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>OCCUPATION_TYPE</td>\n",
       "      <td>What kind of occupation does the client have</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>32</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>CNT_FAM_MEMBERS</td>\n",
       "      <td>How many family members does client have</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>95</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>DEF_30_CNT_SOCIAL_CIRCLE</td>\n",
       "      <td>How many observation of client's social surrou...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>96</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>OBS_60_CNT_SOCIAL_CIRCLE</td>\n",
       "      <td>How many observation of client's social surrou...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>97</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>DEF_60_CNT_SOCIAL_CIRCLE</td>\n",
       "      <td>How many observation of client's social surrou...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>98</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>DAYS_LAST_PHONE_CHANGE</td>\n",
       "      <td>How many days before application did client ch...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>99</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_2</td>\n",
       "      <td>Did client provide document 2</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>100</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_3</td>\n",
       "      <td>Did client provide document 3</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>101</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_4</td>\n",
       "      <td>Did client provide document 4</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>102</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_5</td>\n",
       "      <td>Did client provide document 5</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>103</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_6</td>\n",
       "      <td>Did client provide document 6</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>104</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_7</td>\n",
       "      <td>Did client provide document 7</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>105</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_8</td>\n",
       "      <td>Did client provide document 8</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>106</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_9</td>\n",
       "      <td>Did client provide document 9</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>107</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_10</td>\n",
       "      <td>Did client provide document 10</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>108</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_11</td>\n",
       "      <td>Did client provide document 11</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>109</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_12</td>\n",
       "      <td>Did client provide document 12</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>110</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_13</td>\n",
       "      <td>Did client provide document 13</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>111</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_14</td>\n",
       "      <td>Did client provide document 14</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>112</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_15</td>\n",
       "      <td>Did client provide document 15</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>113</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_16</td>\n",
       "      <td>Did client provide document 16</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>114</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_17</td>\n",
       "      <td>Did client provide document 17</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>112</th>\n",
       "      <td>115</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_18</td>\n",
       "      <td>Did client provide document 18</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113</th>\n",
       "      <td>116</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_19</td>\n",
       "      <td>Did client provide document 19</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>117</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_20</td>\n",
       "      <td>Did client provide document 20</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>118</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>FLAG_DOCUMENT_21</td>\n",
       "      <td>Did client provide document 21</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>119</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_REQ_CREDIT_BUREAU_HOUR</td>\n",
       "      <td>Number of enquiries to Credit Bureau about the...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>117</th>\n",
       "      <td>120</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_REQ_CREDIT_BUREAU_DAY</td>\n",
       "      <td>Number of enquiries to Credit Bureau about the...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>121</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_REQ_CREDIT_BUREAU_WEEK</td>\n",
       "      <td>Number of enquiries to Credit Bureau about the...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>122</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_REQ_CREDIT_BUREAU_MON</td>\n",
       "      <td>Number of enquiries to Credit Bureau about the...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>123</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_REQ_CREDIT_BUREAU_QRT</td>\n",
       "      <td>Number of enquiries to Credit Bureau about the...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>124</td>\n",
       "      <td>application_{train|test}.csv</td>\n",
       "      <td>AMT_REQ_CREDIT_BUREAU_YEAR</td>\n",
       "      <td>Number of enquiries to Credit Bureau about the...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>122 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Unnamed: 0                         Table                         Row  \\\n",
       "0             1  application_{train|test}.csv                  SK_ID_CURR   \n",
       "1             2  application_{train|test}.csv                      TARGET   \n",
       "2             5  application_{train|test}.csv          NAME_CONTRACT_TYPE   \n",
       "3             6  application_{train|test}.csv                 CODE_GENDER   \n",
       "4             7  application_{train|test}.csv                FLAG_OWN_CAR   \n",
       "5             8  application_{train|test}.csv             FLAG_OWN_REALTY   \n",
       "6             9  application_{train|test}.csv                CNT_CHILDREN   \n",
       "7            10  application_{train|test}.csv            AMT_INCOME_TOTAL   \n",
       "8            11  application_{train|test}.csv                  AMT_CREDIT   \n",
       "9            12  application_{train|test}.csv                 AMT_ANNUITY   \n",
       "10           13  application_{train|test}.csv             AMT_GOODS_PRICE   \n",
       "11           14  application_{train|test}.csv             NAME_TYPE_SUITE   \n",
       "12           15  application_{train|test}.csv            NAME_INCOME_TYPE   \n",
       "13           16  application_{train|test}.csv         NAME_EDUCATION_TYPE   \n",
       "14           17  application_{train|test}.csv          NAME_FAMILY_STATUS   \n",
       "15           18  application_{train|test}.csv           NAME_HOUSING_TYPE   \n",
       "16           19  application_{train|test}.csv  REGION_POPULATION_RELATIVE   \n",
       "17           20  application_{train|test}.csv                  DAYS_BIRTH   \n",
       "18           21  application_{train|test}.csv               DAYS_EMPLOYED   \n",
       "19           22  application_{train|test}.csv           DAYS_REGISTRATION   \n",
       "20           23  application_{train|test}.csv             DAYS_ID_PUBLISH   \n",
       "21           24  application_{train|test}.csv                 OWN_CAR_AGE   \n",
       "22           25  application_{train|test}.csv                  FLAG_MOBIL   \n",
       "23           26  application_{train|test}.csv              FLAG_EMP_PHONE   \n",
       "24           27  application_{train|test}.csv             FLAG_WORK_PHONE   \n",
       "25           28  application_{train|test}.csv            FLAG_CONT_MOBILE   \n",
       "26           29  application_{train|test}.csv                  FLAG_PHONE   \n",
       "27           30  application_{train|test}.csv                  FLAG_EMAIL   \n",
       "28           31  application_{train|test}.csv             OCCUPATION_TYPE   \n",
       "29           32  application_{train|test}.csv             CNT_FAM_MEMBERS   \n",
       "..          ...                           ...                         ...   \n",
       "92           95  application_{train|test}.csv    DEF_30_CNT_SOCIAL_CIRCLE   \n",
       "93           96  application_{train|test}.csv    OBS_60_CNT_SOCIAL_CIRCLE   \n",
       "94           97  application_{train|test}.csv    DEF_60_CNT_SOCIAL_CIRCLE   \n",
       "95           98  application_{train|test}.csv      DAYS_LAST_PHONE_CHANGE   \n",
       "96           99  application_{train|test}.csv             FLAG_DOCUMENT_2   \n",
       "97          100  application_{train|test}.csv             FLAG_DOCUMENT_3   \n",
       "98          101  application_{train|test}.csv             FLAG_DOCUMENT_4   \n",
       "99          102  application_{train|test}.csv             FLAG_DOCUMENT_5   \n",
       "100         103  application_{train|test}.csv             FLAG_DOCUMENT_6   \n",
       "101         104  application_{train|test}.csv             FLAG_DOCUMENT_7   \n",
       "102         105  application_{train|test}.csv             FLAG_DOCUMENT_8   \n",
       "103         106  application_{train|test}.csv             FLAG_DOCUMENT_9   \n",
       "104         107  application_{train|test}.csv            FLAG_DOCUMENT_10   \n",
       "105         108  application_{train|test}.csv            FLAG_DOCUMENT_11   \n",
       "106         109  application_{train|test}.csv            FLAG_DOCUMENT_12   \n",
       "107         110  application_{train|test}.csv            FLAG_DOCUMENT_13   \n",
       "108         111  application_{train|test}.csv            FLAG_DOCUMENT_14   \n",
       "109         112  application_{train|test}.csv            FLAG_DOCUMENT_15   \n",
       "110         113  application_{train|test}.csv            FLAG_DOCUMENT_16   \n",
       "111         114  application_{train|test}.csv            FLAG_DOCUMENT_17   \n",
       "112         115  application_{train|test}.csv            FLAG_DOCUMENT_18   \n",
       "113         116  application_{train|test}.csv            FLAG_DOCUMENT_19   \n",
       "114         117  application_{train|test}.csv            FLAG_DOCUMENT_20   \n",
       "115         118  application_{train|test}.csv            FLAG_DOCUMENT_21   \n",
       "116         119  application_{train|test}.csv  AMT_REQ_CREDIT_BUREAU_HOUR   \n",
       "117         120  application_{train|test}.csv   AMT_REQ_CREDIT_BUREAU_DAY   \n",
       "118         121  application_{train|test}.csv  AMT_REQ_CREDIT_BUREAU_WEEK   \n",
       "119         122  application_{train|test}.csv   AMT_REQ_CREDIT_BUREAU_MON   \n",
       "120         123  application_{train|test}.csv   AMT_REQ_CREDIT_BUREAU_QRT   \n",
       "121         124  application_{train|test}.csv  AMT_REQ_CREDIT_BUREAU_YEAR   \n",
       "\n",
       "                                           Description  \\\n",
       "0                             ID of loan in our sample   \n",
       "1    Target variable (1 - client with payment diffi...   \n",
       "2          Identification if loan is cash or revolving   \n",
       "3                                 Gender of the client   \n",
       "4                        Flag if the client owns a car   \n",
       "5                  Flag if client owns a house or flat   \n",
       "6                    Number of children the client has   \n",
       "7                                 Income of the client   \n",
       "8                            Credit amount of the loan   \n",
       "9                                         Loan annuity   \n",
       "10   For consumer loans it is the price of the good...   \n",
       "11   Who was accompanying client when he was applyi...   \n",
       "12   Clients income type (businessman, working, mat...   \n",
       "13      Level of highest education the client achieved   \n",
       "14                         Family status of the client   \n",
       "15   What is the housing situation of the client (r...   \n",
       "16   Normalized population of region where client l...   \n",
       "17     Client's age in days at the time of application   \n",
       "18   How many days before the application the perso...   \n",
       "19   How many days before the application did clien...   \n",
       "20   How many days before the application did clien...   \n",
       "21                                 Age of client's car   \n",
       "22       Did client provide mobile phone (1=YES, 0=NO)   \n",
       "23         Did client provide work phone (1=YES, 0=NO)   \n",
       "24         Did client provide home phone (1=YES, 0=NO)   \n",
       "25            Was mobile phone reachable (1=YES, 0=NO)   \n",
       "26         Did client provide home phone (1=YES, 0=NO)   \n",
       "27              Did client provide email (1=YES, 0=NO)   \n",
       "28        What kind of occupation does the client have   \n",
       "29            How many family members does client have   \n",
       "..                                                 ...   \n",
       "92   How many observation of client's social surrou...   \n",
       "93   How many observation of client's social surrou...   \n",
       "94   How many observation of client's social surrou...   \n",
       "95   How many days before application did client ch...   \n",
       "96                       Did client provide document 2   \n",
       "97                       Did client provide document 3   \n",
       "98                       Did client provide document 4   \n",
       "99                       Did client provide document 5   \n",
       "100                      Did client provide document 6   \n",
       "101                      Did client provide document 7   \n",
       "102                      Did client provide document 8   \n",
       "103                      Did client provide document 9   \n",
       "104                     Did client provide document 10   \n",
       "105                     Did client provide document 11   \n",
       "106                     Did client provide document 12   \n",
       "107                     Did client provide document 13   \n",
       "108                     Did client provide document 14   \n",
       "109                     Did client provide document 15   \n",
       "110                     Did client provide document 16   \n",
       "111                     Did client provide document 17   \n",
       "112                     Did client provide document 18   \n",
       "113                     Did client provide document 19   \n",
       "114                     Did client provide document 20   \n",
       "115                     Did client provide document 21   \n",
       "116  Number of enquiries to Credit Bureau about the...   \n",
       "117  Number of enquiries to Credit Bureau about the...   \n",
       "118  Number of enquiries to Credit Bureau about the...   \n",
       "119  Number of enquiries to Credit Bureau about the...   \n",
       "120  Number of enquiries to Credit Bureau about the...   \n",
       "121  Number of enquiries to Credit Bureau about the...   \n",
       "\n",
       "                                   Special  \n",
       "0                                      NaN  \n",
       "1                                      NaN  \n",
       "2                                      NaN  \n",
       "3                                      NaN  \n",
       "4                                      NaN  \n",
       "5                                      NaN  \n",
       "6                                      NaN  \n",
       "7                                      NaN  \n",
       "8                                      NaN  \n",
       "9                                      NaN  \n",
       "10                                     NaN  \n",
       "11                                     NaN  \n",
       "12                                     NaN  \n",
       "13                                     NaN  \n",
       "14                                     NaN  \n",
       "15                                     NaN  \n",
       "16                             normalized   \n",
       "17   time only relative to the application  \n",
       "18   time only relative to the application  \n",
       "19   time only relative to the application  \n",
       "20   time only relative to the application  \n",
       "21                                     NaN  \n",
       "22                                     NaN  \n",
       "23                                     NaN  \n",
       "24                                     NaN  \n",
       "25                                     NaN  \n",
       "26                                     NaN  \n",
       "27                                     NaN  \n",
       "28                                     NaN  \n",
       "29                                     NaN  \n",
       "..                                     ...  \n",
       "92                                     NaN  \n",
       "93                                     NaN  \n",
       "94                                     NaN  \n",
       "95                                     NaN  \n",
       "96                                     NaN  \n",
       "97                                     NaN  \n",
       "98                                     NaN  \n",
       "99                                     NaN  \n",
       "100                                    NaN  \n",
       "101                                    NaN  \n",
       "102                                    NaN  \n",
       "103                                    NaN  \n",
       "104                                    NaN  \n",
       "105                                    NaN  \n",
       "106                                    NaN  \n",
       "107                                    NaN  \n",
       "108                                    NaN  \n",
       "109                                    NaN  \n",
       "110                                    NaN  \n",
       "111                                    NaN  \n",
       "112                                    NaN  \n",
       "113                                    NaN  \n",
       "114                                    NaN  \n",
       "115                                    NaN  \n",
       "116                                    NaN  \n",
       "117                                    NaN  \n",
       "118                                    NaN  \n",
       "119                                    NaN  \n",
       "120                                    NaN  \n",
       "121                                    NaN  \n",
       "\n",
       "[122 rows x 5 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "column_desc[column_desc['Table'] == 'application_{train|test}.csv']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Target variable (1 - client with payment difficulties: he/she had late payment more than X days on at least one of the first Y installments of the loan in our sample, 0 - all other cases)'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "column_desc[column_desc['Table'] == 'application_{train|test}.csv'].iloc[1]['Description']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### Distribution of income\n",
    "\n",
    "I'm removing outliers, since the large values can drown out the histogram."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "income_series = application_train['AMT_INCOME_TOTAL']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Largest income: 117000000.0\n",
      "Smallest income: 25650.0\n",
      "Mean income: 168797.9192969845\n",
      "Std income: 237123.14627885626\n"
     ]
    }
   ],
   "source": [
    "print(f'Largest income: {max(income_series)}')\n",
    "print(f'Smallest income: {min(income_series)}')\n",
    "print(f'Mean income: {income_series.mean()}')\n",
    "print(f'Std income: {income_series.std()}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "income_mean = income_series.mean()\n",
    "income_series_cleaned = income_series[(income_series - income_mean).abs() <= 3 * income_series.std()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lex/anaconda3/envs/fastai-cpu/lib/python3.6/site-packages/matplotlib/axes/_axes.py:6462: UserWarning: The 'normed' kwarg is deprecated, and has been replaced by the 'density' kwarg.\n",
      "  warnings.warn(\"The 'normed' kwarg is deprecated, and has been \"\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,5))\n",
    "plt.title(\"Distribution of AMT_INCOME_TOTAL\")\n",
    "ax = sns.distplot(income_series_cleaned)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### Distribution of credit amount of loans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lex/anaconda3/envs/fastai-cpu/lib/python3.6/site-packages/matplotlib/axes/_axes.py:6462: UserWarning: The 'normed' kwarg is deprecated, and has been replaced by the 'density' kwarg.\n",
      "  warnings.warn(\"The 'normed' kwarg is deprecated, and has been \"\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAucAAAFKCAYAAAC+bfFlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3XmcW+V99/2P9mVGs3q8Gy8sF8ZgGzAEs8UhNAk0EJqlT9q0dyAJpWl7p0led5+bNL1J0oc86dNsLVm6pEuapjQ7TUIKIQkYgh3CasDGvsD7Ml7Gnn00o9Fynj8kjeWxZkYzkkYa6ft+vfyCOec651zSHMxPP/2u33E5joOIiIiIiFSeu9ITEBERERGRNAXnIiIiIiJVQsG5iIiIiEiVUHAuIiIiIlIlFJyLiIiIiFQJBeciIiIiIlVCwbmIiIiISJXwVnoCIiLVxBizAtgDvJzZ5AYGgb+x1n4nM+Yvgd3W2m9Mcp57gBettT/Ms2/seGOMA3RYa09OY45XAO+31v6hMWYDcLe19p2FHj8TxhgP8ANgNXCftfbLecbcAvwIeLe19ts52z8JfAJ4n7X2X3O2NwDHgc3A72X+CdAILAFs5uefWWv/bIr5rQbuBc4HHKAX+Li19snM/v1ADBjO7A8AKeB/WWsfzvN7z3UlcDXwUM6cPEAX8P9aa3+eucbtwDuttW81xmwFwoAfMDnn3WGtfc9kr0VE6puCcxGRsw1ba9dnfzDGLAd+YYxJWmu/b629p4Bz3AC8km9HgcdPZg2wNHOuZ4GyBuYZS4A3Aw3W2uQEY/4I+A/gI8C3x+07CPw+8K85295B+oMP1tpeYD2AMWYT8OXc38FkjDEG+AVwh7X2p5ltbwQeNMZcY63dkRn6nsz7lT3unZn5LMpsGp7omulLsGfcfbEO+Kkx5m3W2l/njrfWXp0ZswLYXuhrERFRcC4iMgVr7YFMJvzPgO8bY75OOuD6nDHmU8BvAaPAKeB24O3ABuCzxpgk8DagDTgXeBBYkD0+c4lPZ7LhbuAvrLUP5mZh4XRWFvgg8JdAszHmX4F/Ix3IXmyMaQa+QjrIdUhnev/cWpswxowAfwW8iXQw+tfW2r8b/1qNMdcBnyWd9R0F/gLYAjwM+IDnjDHvsNbuGXfcKmATsBzYaYy5ylr7VM6Qh4HbjDFLrbWHM9veC3wTuHCq38EU7gb+NRuYA1hrf2GM+R3SmfKzGGNcwEqge6YXtda+aIy5j/SHkXfP9DwiIrlUcy4iUpgXgUtyNxhjlgEfBq6w1m4AHgFeZ639CvAs8GfW2gcyw8PW2jXW2v+d59x7rbWXkS7t+DdjTMdEk7DWHgLuAX5prb1j3O77SH9AuIT0h4N1wP/K7AsAJzMZ3XcCXzTGBMe9nnbge8CfWmvXcjp4ngfcTCazPD4wz/hD4CfW2hPAt0gHrLniwHeA92SudQ4QAbZP9FqnYQPpDxBnsNY+ZK3dm7PpP4wxLxpjDgOHSH8ouCVnf8gYs23cn69Mce2z7gsRkWIoOBcRKYwDRMdtO0I6OHveGPM5YJu19r8mOP7JSc799wDW2u2kS2E2znCON5HOojvW2ljmvDfl7M/Wvz9POlhvGHf860jXwv86M58dpIPeTZNd1BgTAO4gncUn88+3Zz685PoG6Q8gkC5xmbBmf5pSFPb/s/dYa9cB15GuP982LnjPfvjI/fPHU5wz330hIjJjCs5FRApzBeMWC1prU8DrSZeynCKdjf7rCY4fnOTcuTXcbtJZZgdw5Wz3FzBHd+a43J99OT8PZ+adHZN7fkgvcnTGbRt/jnx+G2gFvpxZePmdzHn+Z+4ga+0zgMcYsx74v4D7pzhvoZ4Crhq/0RhzjzHmrMWX1tp9pD8cfM4Yc2WR1z7rvhARKYaCcxGRKRhjLgD+D/D5cdvXkS7L2Gmt/QzwRdLBGkCCqYParNsz57sMOA/4NelOIBcbY4LGGB9nLvqc6Nw/Bf7EGOPKZLP/APhZgXMA+BVwYTZgNcasAa7ndBeViXwQ+LS1drm1doW1dgXpMpc7Mx1Zcv076ffpVWvtjOu9x/ls5lpvym4wxrwF+FPS32ycxVq7lXTm/qvGmBn9vzDzPn0Q+NuZHC8iko8WhIqInC1kjNmW+fcUMAJ8zFr7k9xBmQWB3wGeNcYMks5Mfyiz+0fAZ4wxhWS8VxljXiCdbX63tbbbGPMI8DiwCzgKPAaszYx/CviEMeYHpOvMsz4EfIl0JtdPehHmpwt90dbak8aYdwFfMsaEM6/9Dmvtq5muI2fJfEBZD9w6btc3SC8mvX3c9m+Sbnn4tkLnVcC8dxtj3kp6Ye3nSH8DcAK4JVMqNJGPkW6NeCfpDza5v/dct2f+ee64+6IP+F1rbd4PACIiM+FynPHfYIqIiIiISCUocy4iIlXNGPMG0qUw+TxmrR3fGUZEZM5S5lxEREREpEpoQaiIiIiISJVQcC4iIiIiUiVUc56jq2ugKmt8WlvD9PToGRcyc7qHpFi6h6QYun+kWLV2D3V0RMY/Z2KMMudzgNfrqfQUZI7TPSTF0j0kxdD9I8Wqp3tIwbmIiIiISJVQcC4iIiIiUiUUnIuIiIiIVAkF5yIiIiIiVULBuYiIiIhIlVBwLiIiIiJSJRSci4iIiIhUCQXnIiIiIiJVQsG5iIiIiEiVUHAuIiIiIlIlFJyLiIiIiFQJb6UnINVr87Yj0xq/af2SMs1EREREpD4ocy4iIiIiUiUUnIuIiIiIVAkF5yIiIiIiVULBuYiIiIhIlVBwLiIiIiJSJRSci4iIiIhUCQXnIiIiIiJVQsG5iIiIiEiVUHAuIiIiIlIlFJyLiIiIiFQJBeciIiIiIlVCwbmIiIiISJVQcC4iIiIiUiW8Uw0wxriBrwLrgBjwAWvt7pz9dwJ3AQngXmvtg8aYecD9QAjoBO6w1kaLHZtzzQ8DC621dxtjFgLfypnyeuBu4B+Aw8Brme2/stZ+bJrvj4iIiIjIrJkyOAduA4LW2o3GmKuAzwNvA8gExh8CNgBB4EljzM+Ae4D7rbVfN8bcDdxljPnPEox1A18DXgd8H8BaewzYlJnPRuDTmTHnAs9ba28p5g0SEREREZkthZS1XAs8DGCtfYp0wJx1JbDFWhuz1vYBu4G1uccADwE3lmhsEPgG6QD8DMYYF/Al4IPW2iRwObDEGPOYMea/jTGmkDdERERERKRSCsmcNwF9OT8njTFea20iz74BoHnc9nzbZjTWWtsDPGKMuT3PPG8Bdlhrbebno8BnrLXfNcZcC3wTuGKyF9raGsbr9Uw2pGI6OiKzfs1IY3Ba4ysxRymcfj9SLN1DUgzdP1KsermHCgnO+4Hcd8OdCczz7YsAvTnbh/NsK2bsZH4P+Nucn58lXa+OtfZJY8wSY4zLWutMdIKenugUl6iMjo4IXV0Ds37dgcGRSfc7jsMLr54kkUxx5UULKjJHKUyl7iGpHbqHpBi6f6RYtXYPTfZBo5Cyli3AzQCZmvOXc/Y9DVxnjAkaY5qB1cD23GOAm4BflmjsZC4Htub8/Angw5l5rwMOThaYy/TtOtDL9n3d2EO9OI7eWhEREZFiFRKcPwCMGGO2Al8EPmKM+agx5tbMYsz7SAfUjwIft9aOAPcC7zbGbAE2Al8u0di8jDEdwMC44PuvgNcbYx4HvgDcXthbIoU4dirKs/YEAI4DsXiywjMSERERmftcynie1tU1UJVvRqW+ytm87Uje7YPDcX6y9QCjiSTtTUFO9o1wyzUr+K3rVs3yDKVQtfZ1oMw+3UNSDN0/Uqxau4c6OiKuifbpIUQyLYlkisdfOEIsnuSK1fNZ2tEAwHAsMcWRIiIiIjIVBedSMMdxeGrHcU71xzhvSTNmWQvBQHpNsYJzERERkeIpOJeC7TnSz97OfuY1B3ndRfNxuVyEssH5qGrORURERIql4FwKdqRrEIBr1y7C40nfOmPB+Ygy5yIiIiLFUnAuBRscSeB2uYiEfWPbQoH0Q5uGRxWci4iIiBRLwbkUbGg4TkPIi8t1eoFx0J/OnI/EVNYiIiIiUiwF51KQRDLFyGiShqDvjO0et4uAz6MFoSIiIiIloOBcChLN1JQ3hLxn7QsFPCprERERESkBBedSkMHhOACNId9Z+4IBL6PxFPFEaranJSIiIlJTFJxLQYYywfn4shaAkD+9KLR/aHRW5yQiIiJSaxScS0EGM2Ut+TLn2XaKvUOxWZ2TiIiISK1RcC4FGcuc5605T2/rH1TmXERERKQYCs6lINngPJyvrCUTnPeprEVERESkKArOpSCDw3HCAS8et+usfdkHESk4FxERESmOgnOZUirlEI0l8pa0AIT8ypyLiIiIlIKCc5lSNJbAcfJ3aoF0K0WAvkEtCBUREREphoJzmdLpxaD5g/OAz43bpVaKIiIiIsVScC5TGhpro5i/rMXlchEMeFXWIiIiIlIkBecypaky55Du2NI7OIrjOLM1LREREZGao+BcpjSYCc4bJ6g5h/RTQhPJFMOxxGxNS0RERKTmKDiXKQ2NFJY5B3VsERERESmGgnOZ0tBwAr/Pjc878e0yFpzrKaEiIiIiM6bgXCblOA6Dw3EaJ8maAwT1ICIRERGRoik4l0nF4kmSKWfCHudZehCRiIiISPEUnMukBofTCzwnejpo1umacz2ISERERGSmFJzLpIYK6NQCEMqWtajmXERERGTGFJzLpArpcQ7q1iIiIiJSCpPXKgDGGDfwVWAdEAM+YK3dnbP/TuAuIAHca6190BgzD7gfCAGdwB3W2mixY3Ou+WFgobX27szPHwXeD3RlhtwFHAS+CcwHBoD3Wmuz+6VA2aeDThWcez1uQgGPMuciIiIiRSgkc34bELTWbgTuBj6f3WGMWQh8CLgGeDPwGWNMALgHuN9aex3wAnBXKcYaY0LGmG8CfzxujpcB/8NauynzxwIfBF7OnPcbwF9M+92R0w8gmqLmHKCpIUC/as5FREREZqyQ4Pxa4GEAa+1TwIacfVcCW6y1MWttH7AbWJt7DPAQcGOJxgZJB9qfHjfHy4GPGWOeNMZ8bPy8c84r0zQ0EsfjdhHweaYc29zgZyAaJ5lKzcLMRERERGrP1OlQaAL6cn5OGmO81tpEnn0DQPO47fm2zWistbYHeMQYc/u4OX4L+ArQDzxgjHnrBOedVGtrGK936iC0Ejo6IrN+zUhjkKGRBE0NfpoioSnHz28L8+qhXvyhAG1NwVmYoUxHJe4hqS26h6QYun+kWPVyDxUSnPcDue+GOxOY59sXAXpztg/n2VbM2LMYY1zA32Qy7BhjfgJcOu4cEx6fq6cnOtWQiujoiNDVNTDr1+3ujRIbTdLeFGBgcGTK8cHME0T3HugmubA+/gOaKyp1D0nt0D0kxdD9I8WqtXtosg8ahZS1bAFuBjDGXAW8nLPvaeA6Y0zQGNMMrAa25x4D3AT8skRj82kCthtjGjOB+g3AcxOcV6ZhrFPLFG0Us5ob/YA6toiIiIjMVCHB+QPAiDFmK/BF4CPGmI8aY2611h4D7iMd+D4KfNxaOwLcC7zbGLMF2Ah8uURjz5LJmP858Fhm/A5r7X8DfwesMcY8CfwB8Knpvjn1bnCksDaKWc0NAQD6BrUoVERERGQmXI7jVHoOVaOra6Aq34xKfZXzDz/azq9fOcG1axeyavGUJfu0NwX54nde5O3Xr+KtV68o/wSlYLX2daDMPt1DUgzdP1KsWruHOjoiron26SFEMqHB4UyP80LLWhpU1iIiIiJSDAXnMqFCnw6apeBcREREpDgKzmVCQyNxXC4IBwpp6gORsB+XC/pVcy4iIiIyIwrOZUJDwwnCAS9u94RlUWdwu11Ewn5lzkVERERmSMG55JVIpojGEjQWWNKS1dyg4FxERERkphScS17dA+nSlELrzbOaG/yMjCaJjSbLMS0RERGRmqbgXPI61ZduKz/t4HzsQUSqOxcRERGZLgXnktdYcB4sbDFo1tiDiFTaIiIiIjJtCs4lr6PdQwA0ZdojFmqsneKggnMRERGR6VJwLnkdPRkFoKVxmsF5o3qdi4iIiMyUgnPJq/PUEAGfh6B/umUtCs5FREREZkrBuZxlNJ6kq3d42llzOF0G068FoSIiIiLTNr20qNSFY91RHOd0icp0ZBeE9pa45nzztiMFj920fklJry0iIiIyW5Q5l7N0nkovBs0G2tMRCnjwed30q6xFREREZNoUnMtZsotBZ5I5d7lcekqoiIiIyAwpOJezZDPnLY3Tz5xDuu68f2gUx3FKOS0RERGRmqfgXM7SeXKIUMBLKOCZ0fHNDX6SKYehkUSJZyYiIiJS2xScyxkSyRQneoZZ3B7G5XLN6BxNaqcoIiIiMiMKzuUMJ3qGSaYcFs1rmPE5msfaKSo4FxEREZkOBedyhs6T6Xrzxe3FB+d96nUuIiIiMi0KzuUM2cWgi4vInJ9+EFG8JHMSERERqRcKzuUMpzPn4RmfI9sfXZlzERERkelRcC5nOHoqit/npq05OONzNDX4AOgv8VNCRURERGqdgnMZk0o5HD0VZVF7A+4ZdmqBnG4tUQXnIiIiItOh4FzGdPUNk0imiloMChD0ewn4PMqci4iIiEyTgnMZM1ZvPm/m9eZZzQ1+Zc5FREREpsk71QBjjBv4KrAOiAEfsNbuztl/J3AXkADutdY+aIyZB9wPhIBO4A5rbbTYsTnX/DCw0Fp7d+bn3wE+DCSBl4A/stamjDEvAH2Zw/ZZa++Y2dtUH46eigLFtVHMamr0s/dIP6mUg9s98xIZERERkXpSSOb8NiBord0I3A18PrvDGLMQ+BBwDfBm4DPGmABwD3C/tfY64AXgrlKMNcaEjDHfBP44Zw4h4F7gDdbaq4Fm4K3GmCCAtXZT5o8C8ymczpwXH5w3h/2kHIfBYbVTFBERESlUIcH5tcDDANbap4ANOfuuBLZYa2PW2j5gN7A29xjgIeDGEo0NAt8APp0zhxhwtbU2mvnZC4yQzvSHjTGPGGMeNcZcVcgbUs86Tw7h9biZ1zLzTi1ZTY16SqiIiIjIdE1Z1gI0cbo0BCBpjPFaaxN59g2Qzlznbs+3bUZjrbU9wCPGmNuzO6y1KeA4gDHmfwKNwM+Ai4HPAf8EnA88ZIwxmXnn1doaxuv1TPZeVExHR6Ss50+lHI51R1k6v5GFC5oBiDROL0jPneOi7L97PSWZ+3TmUu73aq7S+yLF0j0kxdD9I8Wql3uokOC8H8h9N9w5Ae74fRGgN2f7cJ5txYzNK1MX/9fABcA7rLWOMeZVYLe11gFeNcacAhYBhyY6T09PdKJdFdXREaGra6Cs1zjVN8LIaJL5LcGxaw0MjkzrHLlz9LocAA519rG0LVT0/KYzl3K/V3PRbNxDUtt0D0kxdP9IsWrtHprsg0YhZS1bgJsBMqUhL+fsexq4zhgTNMY0A6uB7bnHADcBvyzR2In8A+mSl9tyylveR6Y+3hizmHQ2/mgBr7cudZ7KPhm0+HpzSNecA/SprEVERESkYIUE5w8AI8aYrcAXgY8YYz5qjLnVWnsMuI90QP0o8HFr7QjpBZrvNsZsATYCXy7R2LMYYy4D3g9cAjxqjNlsjPkt4J+BFmPMk8C3gfdNVtJS70q5GBRUcy4iIiIyE1OWtWRquv9w3OZdOfu/Bnxt3DHHgbfkOVdRY3P2fT3n359n4g8ZvzvBdhknG5wvKlFw3px9SuhQrCTnExEREakHegiRAOke526XiwWtxdeHQ25wrsy5iIiISKEUnAuO49B5cogFbSG8ntLcEj6vh1DAOytlLaPxJL/afoy9nf04jlP264mIiIiUSyHdWqTG9Q2NEo0lWL28taTnbWrwz0rmfMe+bl473Mdrh/vYsa+bjuYQa1a24XLpyaQiIiIytyhzLhw7lW5ws7A9XNLzNjf4GYzGSaZSJT1vrlg8ya4DvQT9HlYtbqJnIMYXvvMin/vWNvYd7S/bdUVERETKQZnzOrJ525G82187nG4h3z0Qm3DMTDQ3+HGAgWiclsZAyc6ba9eBHuLJFJec18HFK9tYs7KVA8cGeWnPKT79jef45PuuYGlHY1muLSIiIlJqypwLA9E4AJGwr6TnbcouCh0sT2nLaCLJzgM9BHwezLIWAFojQT78rnX89hvOI+U47DnSN8VZRERERKqHgnMZC86bShycZzu29EfLE5zbg72MxlOsXtGKz3vmrbxqcRMAJ3qHy3JtERERkXJQcC4MREfxuF2EAqWtcipn5jyeSPHKvh58XjcXntNy1v6OlnRLyK7evM+uEhEREalKCs7rnOM4DETjNIZ9Je9uUs7M+WuHe4nFk6xe3orf5zn72o1+fF43Xcqci4iIyByi4LzOxeIp4okUkVBpS1qgfJnzZDLFjn3deD0uLpyg/aPb5WJec5CTCs5FRERkDlFwXucGMlntSNhf8nOXK3P+2uE+hmNJzDmtBP1nZ82zOlpCDI0kGBqJl/T6IiIiIuWi4LzOlatTC+RmzmMlO2cq5bB9Xzcet4uLVkz+0KTTdefKnouIiMjcoOC8zg2WMXPu9bhpCHpL+pTQY91RoiMJzl3SNOUCVi0KFRERkblGwXmdK2fmHKC5MUB/CYPz7FM/Vy5qmnLsfGXORUREZI5RcF7n+qNxXC5oLMOCUEj3Th8aSZBIpoo+VzKV4uDxQcJBL/NbQ1OO72gJAgrORUREZO5QcF7nBodHaQj6cLtL20Yxq7kxAFCS7HnnySjxRIoVCyMFtX2cp8y5iIiIzDEKzutYPJFiOJYsW0kLQFOmlr0UdefZkpYVBZS0AAR8Hpob/JzoUXAuIiIic4OC8zo2OFzeenNIPwwIis+cxxMpDp8YJBL20d4UKPi4jpYQ3f2xkpTViIiIiJSbgvM6Vs4e51nZXufFZs4Pdw2SSDqsWNQ0rSeZdrSESDkO3QOla+coIiIiUi4KzutYuTu1QE6v8yKD8/1HBwBYuTAyreO0KFRERETmEgXndWw2M+fFlLVER+Ic6RqipdFPS6TwkhbQg4hERERkbpn8KS5S07KZ81K1Udy87chZ26IjCQB2H+k7Y/+m9UsKPu/zr54k5TgF9TYfT8G5iIiIzCXKnNexgWicUMCDz1u+2yDo9wAwEkvM+BxP7zwOwIpF0ytpgZzgXB1bREREZA5QcF6nUimHoZE4jaHylbQAuN0ugn4PwzMMzvujo7yyv4d5zcEZld+0NPrxed109Y7M6PoiIiIis0llLXVqcDiO46Sf4FluQb+HoZGZBefP7TpBynGmlTUfX14TDno5emoob9kNTK/ERkRERKSclDmvU7PRqSUrFPAST6Rm1Gv8mV0ncAErptmlJVck5GM0kSIWT874HCIiIiKzQcF5ncp2amksY6eWrFAg/QXNSGx6wbHjOOw7OsDijgbCwZl/iGjMfAAZzHwgEREREalWU5a1GGPcwFeBdUAM+IC1dnfO/juBu4AEcK+19kFjzDzgfiAEdAJ3WGujxY7NueaHgYXW2rszP98C3JMZ+y/W2q8ZY0LAN4H5wADwXmtt14zfqRozm5nz7KLQ4dHEWKBciO7+GLF4kiXzGoq6fiRTVz8QHaW9OVjUuURERETKqZDM+W1A0Fq7Ebgb+Hx2hzFmIfAh4BrgzcBnjDEB0oHy/dba64AXgLtKMdYYEzLGfBP445w5+IAvAm8CXg/8Qeb4DwIvZ877DeAvZvQO1aiB4WxwPnuZ8+kuCu08NQTA4vYig/PMB4LsaxYRERGpVoUE59cCDwNYa58CNuTsuxLYYq2NWWv7gN3A2txjgIeAG0s0Nkg60P50zhxWA7uttT3W2lHgSeC6Cc4rGQPRUXxeNwFf+SubQoFMO8XR6ZW1dJ7MBOdFZs5V1iIiIiJzRSHdWpqAvpyfk8YYr7U2kWffANA8bnu+bTMaa63tAR4xxtw+yfwmO++kWlvDeL2eqYZVREfHzBdEZkUa0yUdjuMwGI3T1hSkKRIq+rxTaWtOB8VJ5/QcCnk93YPpuvg153fwyr7uGV8/GEx/OzA8mhy7fq5SvLdzQb28Tikf3UNSDN0/Uqx6uYcKCc77gdx3w50JzPPtiwC9OduH82wrZmwh88t3jsmOH9PTE51qSEV0dETo6hoo+jwDg+le39GROMmUQzjoHdtWTm4cAE71Do9dr5DXs/dILx63Cx9O0fMMBTz0DsTynqcU7221K9U9JPVL95AUQ/ePFKvW7qHJPmgUUtOwBbgZwBhzFfByzr6ngeuMMUFjTDPpEpPtuccANwG/LNHYfHYC5xtj2owxfuB64FcTnFeY3cWgkC4rcbtd9A3GCj7GcRw6T0aZ3xrC6ym+9KYx5GdoJE4q5RR9LhEREZFyKSTqeQAYMcZsJb3w8iPGmI8aY2611h4D7iMd+D4KfNxaOwLcC7zbGLMF2Ah8uURjz2KtjQMfBX5KOij/F2vtEeDvgDXGmCeBPwA+Nd03p1b1z3Jw7na5aG7w0zc0iuMUFhz3Do4yHEsUvRg0KxL24TgwNKK6cxEREaleU5a1WGtTwB+O27wrZ//XgK+NO+Y48JY85ypqbM6+r4/7+cfAj8dtiwLvynd8vRvM9DjPthicDc0NfnoGYgwNF9ZOMdupZVGRi0Gzxjq2ROOz0qFGREREZCb0EKI6NFbW0jA7mXOAlsZ0QNw3VFhpy9GxTi3hkly/MaSOLSIiIlL9FJzXoYFoHLfbRThQyHrg0mhuDADpcpVCdJ5KL84tZVkLqNe5iIiIVDcF53VoIDpKJOTD5XLN2jWbs5nzQoPzk0O4XLCwrVSZ89NPCRURERGpVgrO68xoIsloIjVW5jFbmsJ+XC7oLbBjS+fJITpaQvh9pek7Hwp48LhdDCpzLiIiIlVMwXmdiQ6nW9Q3hGavpAXA7XbRFC6sY0t/dJTB4XjJSloAXC4XkbCPgWi84I4xIiIiIrNNwXmdybYSbAjObuYc0qUt8USK4Vhi0nGnF4OWLjiH9KLQeCJFLJ4s6XlFRERESkXBeZ0ZGkkHxuHg7GbOAVoKXBTamQnOF7WXpt48K7sotdC6dxEREZHZpuC8zgxlaq4bZrnmHApfFNp5MtOppcSZ82w7x0Lr3kWA8DmvAAAgAElEQVRERERmm4LzOpPNnDdUJHNeWK/zsQcQlThzXmjmXkRERKRSFJzXmWzNeSXKWprCflwUUNZyaoj2piBBf2nnmM3c9w4ocy4iIiLVScF5nYmOJDJtBWf/V+/xuGkM++gdjE3YMWVoJE7f4GjJS1oAvB43kbCP3sGpO8aIiIiIVMLsp0+lYhzHYWg4QVtToGJzaGkMcOjEIAPROE0N/rP2Hx2rNy9tScv464+MJgllnpC6eduRaZ1j0/ol5ZiaiIiIiDLn9WRkNEnKcSpS0pLVnAnIsx1ZxsvWm5eyx3kuLQoVERGRaqbgvI5Ussd5VrbuOxuEj9dZph7nWS2RzKLQAS0KFRERkeqj4LyODFXo6aC5sh1TJsycj/U4L1fmPNuxpfDM+chogidfOkqfsu0iIiJSZgrO60g1ZM6bCihraWn0l630pqnBj8s1veB814Fe9nb2s+tgb1nmJCIiIpKl4LyORCvY4zzL53XTGPLReSp61r7hWILu/ljZSloAPG4XTWF/wR1bUo7D7iN9ABzLM2cRERGRUlJwXkcq+XTQXM2NfvqHRhnMzCfrWHemU0uZSlqyWiIB4onU2IeVyRw9OTQ2rm9otKBjRERERGZKwXkdGRpJ4HZB0O+p6DyyHVuOjlsUWu7FoFnT6djy2uF01nz5wghw+gOEiIiISDkoOK8jQyNxwkEfLperovNonmBR6KETg8BsBOfZRaGTd2wZjiU4dGKQ1kiANSvbAAXnIiIiUl4KzutEIpliOJasaL15VjZz3XnydKD74u6T/OK5w4QCXpbNbyzz9bPtFCfPnO850ofjwPlLm2lrCuD3ulV3LiIiImWl4LxO9GQC0UrXm0POg4gyZS07D/TwlQe243G7+NN3rh17cme5RMI+3C7XpJlzx3F47XAfHreLlYubcLtcLGgLMzgc52TvcFnnJyIiIvVLwXmd6O4fASrbqSXL7/PQGgnQeXKIPUf6uO97LwEOf/KOS7hgWUvZr+92u2hu9NM3FJuwY8vxnmEGonGWL4wQ8KVr9Be2hQHYebCn7HMUERGR+qTgvE5092cy5xXscZ5rcXuYnoEYX/jOi8QTKe669WIuXtk+a9dvafSTSDpndYzJ2p1ZCHr+0uaxbQvb08H5rgPqdy4iIiLloeC8TnQPpDPn4Qo+HTTXosyiz5FYgvf/5mouNx2zev3JFoXG4kkOHBugKexjfmso5xg/Qb+HXQd7CuqRLiIiIjJdCs7rxKkqy5xfeE4rHreL33+zYePFC2f9+i2RiReF7uvsJ5lyOG9ZyxmdbVyZuvOegRjHe1R3LiIiIqU3ZRrVGOMGvgqsA2LAB6y1u3P23wncBSSAe621Dxpj5gH3AyGgE7jDWhstx1jgAuBvcqZ8FXAb8DTwKrA9s/0Ba+3fTvsdqhHVVHMOcNkFHXz1o9fj81am5/pEvc6zC0FdLjh3cdNZxy1qC3Pg2AA7D/SM1aCLiIiIlEohmfPbgKC1diNwN/D57A5jzELgQ8A1wJuBzxhjAsA9wP3W2uuAF4C7yjXWWrvNWrvJWrsJ+ArwA2vtw8BlwH9m99VzYA7p4NzndeP3VfYBRLkqFZgDNIZ8eD1nd2x59VAfPQMxzlkQyds15nTduRaFioiISOkVEpxfCzwMYK19CtiQs+9KYIu1Nmat7QN2A2tzjwEeAm4s41gAjDENwKdIB/UAlwOXGWMeN8Z81xizqIDXWrO6+2NVkzWvBi6Xi+aGAH2Do6RS6frx7v4Rntl1goDPw4YL89fAR8I+WiMBdh3sIaW6cxERESmxQqK1JqAv5+ekMcZrrU3k2TcANI/bnm9bKcdmvR/4rrX2ZObnXcBz1tqfG2PeA3wJeOdkL7S1NYy3gtncyXR0RGZ8bHQkTjSWYF5rhEhjsISzmrnpvp5yzLujNcSp/hFSuAgFfTz55H5SKYcbN57DwnkTz2/9BR089txhhpOwYtHMfy+zrZh7SAR0D0lxdP9IserlHiokOO8Hct8NdyYwz7cvAvTmbB/Os63UY7Pew5nB96NA9nGODwB/OdUL7empzqc/dnRE6OoamPHxR06mH/YT8LoZGBwp1bSKMt3XU455Z79JOHJigEMnBukdjHHRilbaI/5Jr7dyQYTHgK3bDtPgXVbyeZVDsfeQiO4hKYbuHylWrd1Dk33QKKSsZQtwM4Ax5irg5Zx9TwPXGWOCxphmYDXpBZhjxwA3Ab8s41gyYwLW2kM5c/sn4B2Zf38j8FwBr7UmjS0GrYKng1aT7KLQl/acYm9nP/Oag1x6wdQtHS9cnn5QkurORUREpNQKyZw/APyGMWYr4ALuMMZ8FNhtrf2RMeY+0kGyG/i4tXbEGHMv8G+Zjisngd+11g6VY2xmjhcA+8fN+27gX4wxfwQMAR+YyRtUC6qtU0u1yPY67xmI4fe6uX7dYjxu1xRHwbzmEB0tQXYd7CWVcnBPcEzKcbjvey9x6MQgC1pDzG8Ns7AtzIK2EBctbyPgr84SKhEREakclx6mclpX10BVvhnFfpXzgyf28uDW/bzpimVj3UYqbdP6JdMav3nbkZLPwXEcvvWL3cQTKTZduphzFhRWy7Zp/RK+/tBOnnjxKP/371zKhctb8457ZtcJ/u6/thPwexgdTZJ7c12/bhG337S6BK+iMLX2daDMPt1DUgzdP1KsWruHOjoiE2YDlUqtAz1jZS36dedyuVxcceF8kqlUwYF51jWXLOKJF4/y3c17+Pj/uBy368z/xpKpFA88sRe3y8Un77iC1sYAJ3qHOd49zL89vIuX93bjOM4ZDzkSERER0RNC68CpTHAeVlnLWc5b2ow5J3/mezLnL23hytXz2Xe0n19tP3bW/l9tP86x7ijXrVvEgtYwfp+HpR2NXG46uHB5Kz0DMU7oKaMiIiIyjoLzOtA9EKOpwY/HrV93Kb1r03n4vW6+t3kPw7HE2PZ4IsUPn9yH1+PmlqtXnHXc6kwZzE4tKBUREZFxlEqtcY7j0N0fY9n8hkpPpea0Nwe56arl/PDJffzkVwd456ZzAXjixU5O9Y/wpiuW8dLeU2cdNxBNP5X08Rc700usc0y3Fl9ERERqi1KpNW4gGieRTNEWqY6HD9Wat7zuHNqaAjzyzEFO9ESJjSb58db9BPwebt64PO8xkbCPUMDL8e4oWpAtIiIiuRSc17hsvXlbk4Lzcgj4PPz2G84jkXT49qO7+flzh+gfGuVNG5bRFPbnPcblcrGoPczIaJLewdFZnrGIiIhUMwXnNa67PwZAW1OgwjOpXVdcOJ8Lljbzwmsn+fGW/TQEvbz5ynMmPWZhW7ql5bFT1flUWhEREakMBec1LvsAonZlzsvG5XLxOzdegAsYTaS4+arlU3bGyfabP9at4FxEREROU3Be47oH0sF5qzLnZbV8YYSbNy5n5aImbrh86ZTjG0M+GkM+jnVHSanuXERERDIUnNe4U5myFmXOy+8drz+X//PeDQR8noLGL2wPE0+k6Mn8jkREREQUnNe47v4RPG4XTQ35FydK5WTrzo+qtEVEREQyFJzXuJO9w8xrDp71eHmpPC0KFRERkfEUnNewkdEE/dE4HS2hSk9F8ggHvTQ3+DnREyWZUt25iIiIKDivaSd704tBFZxXr4XtYRJJh1N9w5WeioiIiFQBBec1rKs3HfApOK9eY6Ut3QrORURERMF5TTsdnKtTS7VaoLpzERERyaHgvIZ1qayl6gX9HlojAU70DpNIpio9HREREakwBec1rCtTxzyvWcF5NVvUHiaVcsa+6RAREZH6peC8hnX1DtMQ9E75KHmprMXzGgB47VBfhWciIiIilaaorUalHIeu3hGWdjRUeio1Z/O2IyU936L2MG1NAfYfG+BI1yBLOhpLen4RERGZO5Q5r1F9g6MkkinVm88BLpeL9efNA+CHW/ZXdjIiIiJSUQrOa5TaKM4tSzoaaG8O8uyuExw6MVjp6YiIiEiFKDivUWqjOLeks+ftAPzoyX0Vno2IiIhUioLzGqXM+dyzeF4D5y5u4rlXuzhwbOCs/Ue6Btl9RItGRUREapmC8xqlHudzj8vl4rbrVgHww5zseSKZ4r9+uZdP/Msz/H//8Tx9g7FKTVFERETKTN1aalRX3zBul4u2pkClp5JXqTue1IqLVrRy3tJmtu0+yb6j/fi9bv7pwZ0cOD6A1+MikXR47tUubrhsaaWnKiIiImWgzHmN6uodpr05gMetX/Fc4nK5+K1rVwLwjz9+hU99/VkOHB/gmksW8ok7rgTg2V0nKjlFERERKaMpM+fGGDfwVWAdEAM+YK3dnbP/TuAuIAHca6190BgzD7gfCAGdwB3W2mgZx94HXANkC3XfBvjyjZ3xOzWHxOJJ+gZHuWhFa6WnItO0edsRHMdhQWuI491Rgn4P161bwrL5jbx2uJeOlhC7Dvby0K8PcNPrlld6uiIiIlJihaRVbwOC1tqNwN3A57M7jDELgQ+RDozfDHzGGBMA7gHut9ZeB7wA3FWusZmpXAa82Vq7KfOnb5KxNe9kn+rN5zKXy8XVlyxk/Xnt3HrtCpbNP/1QohULIwAcOH72glERERGZ+woJzq8FHgaw1j4FbMjZdyWwxVobywTEu4G1uccADwE3lmtsJrN/PvCPxpgtxpj3jZ93znnrgjq1zH2RsJ+1580j6D/zy63lC9OB+oGjCs5FRERqUSELQpuA3P5tSWOM11qbyLNvAGgetz3ftlKObQC+BHwB8ACPGWOenWDspFpbw3i9nqmGVURHR6TgscOZmuRVy1rPOC7SqJ7nc12kMcii9gaOnhrCE/DR1lT473Q695BIPrqHpBi6f6RY9XIPFRKc9wO574Y7E5jn2xcBenO2D+fZVuqxUeBvs/XkxphHSdfH5xs7qZ6e6ixJ7+iI0NVVeKZ03+H0Sw24OeO4gcGRks9NZt/S+eng/JGt+3jj5YV1benoiPDdn+0q+Bqb1i+Z6fSkRk337yGRXLp/pFi1dg9N9kGjkLKWLcDNAMaYq4CXc/Y9DVxnjAkaY5qB1cD23GOAm4BflnHsBcCTxhiPMcZHupzl+QnG1oWT6nFe05YvSP8H/Yy6toiIiNScQoLzB4ARY8xW4IvAR4wxHzXG3GqtPQbcRzrwfRT4uLV2BLgXeLcxZguwEfhyGcfuBP4DeAp4HPiGtXZHvrHFvVVzR1fvMKGAl4ag2tjXonDQy/zWEK8d6qVXDyQSERGpKS7HcSo9h6rR1TVQlW/GdL7KcRyHD37+cRa2h/lkpi92lh78Uzt2Hejh6Z0neM9vXFBQaYvKWqRYtfaVsswu3T9SrFq7hzo6Iq6J9ukJNTWmf2iU0URKJS017pwFEVyotEVERKTWKDivMV2qN68L4aCX85e1qLRFRESkxig4rzHqcV4/rrhwPg7wnO2q9FRERESkRBSc15jTwbl6mte6DaYDF/DUK8cqPRUREREpEQXnNUaZ8/rR3BjgknPb2XOkn/3H+is9HRERESkBBec1pqt3GJcL2qfx5EiZu27ckO7U8rNnDhd1nlTK4eDxAfYc6SORTJViaiIiIjIDaoQ9x41vj3j45BDhgJcnXz5aoRnJbFqzoo1F7WGe3nmcd73hXFoaA9M6fjSe5LXDfew60MPQSPrBv8/ZLuLxFJsuXUIooL8iREREZpMy5zUkmUwRHUkQCfsrPRWZJS6Xi9/YsIxkyuGx5wvvYx8difPrV47zvc17eM52EYsnMee0sGZlG8mkw3c37+HPvrqV//rlXoZG4mV8BSIiIpJLabEaMjicznw2hnwVnonMpo0XL+T7j+9h87YjvPXq5fi8nknHHzw+wNbtxxiNpwgHvaw9r5XzlzYT8KWPu2RVG7F4ip89c4gfbdnPjn3d/PnvX47LNeHzEkRERKRElDmvIYPDowA0hhWc15OAz8P16xczEI3z1CvHJxyXTKb49SvH2fxCJ8mkw5Wr5/P261dx8cq2scAcwO/zcMvVK/jsB69m7bnt7Ons58U9p2bjpYiIiNQ9Bec1ZCCaLj+IKHNed9542VLcLhc/e+YwjuOctf/Q8QH++6mD2IO9NDf6uXnjci5c3orbPXE2POD38M5N5+ICfvjkvrznFRERkdJScF5DssG5Muf1p60pyIYLOzjcNciug71j2+OJFA89dYCP/M3j9AzEuGBZM7+5cTmtkcIWji7taGTDhfM5cGxA2XMREZFZoOC8hmQf497coAWh9ejGDcsA+PmzhwB4ac9J7vnnX/PdzXsI+Dy8fv1irlqzEK9nev/Z33LNCmXPRUREZokWhNaQvsFRwkEvft/kCwKlNp27uImVi5rY9tpJPv+tF9ixvwe3y8WNly/l/bddwkNb9s7ovNns+TO7TvDi7lOsP39eiWcuIiIiWcqc14hYPEk0lqB1mn2upXak2youxQF27O9h9fJWPvm+K/jd37iAxiLba96q7LmIiMisUOa8RoyVtDSqpKWeXbF6Pid6hlnS0cBlF3SUrP3hko5Grlg9n6d3nmDb7pNcen5HSc4rIiIiZ1LmvEb0DqTbKBa60E9qk8ft5tZrV3K5mV/yvuS3XLNS2XMREZEyU3BeI7KZ8+k+vl2kUEvmNXDF6vkcPD7Itt0nKz0dERGRmqTgvEaorEVmwy1XrwBg8wudlZ2IiIhIjVJwXgMcx6F3YJRI2DftNnki07Gko5HlCyO8sr+bgehopacjIiJSc7QgtAaMjCaJxZPMbw1VeipSIzZvOzLhvnnNQQ4cG+A/f/4aF5zTAsCm9UtKdv7xpntuERGRuUxp1hpwut5cJS1SfisWRQDYd7S/5OdOJFMc74mSTKZKfm4REZG5QJnzGpDt1NKiTi0yCxqCPua3hjjeM0x0JE446CvqfP1DoxzpGuLIySGOd0dJphyWzW9k06WLS95xRkREpNopOK8B6tRSn6ZTGhJpDJb02isXRTjRM8z+YwNctKJtRueIJ1L8/NnDdPUOj21rjQRwHIdDJwbZdbCX1ctbSzVlERGROUHBeQ3oHYzhckFTg8paZHYsXxjh6Z0n2Hd05sH5068cp6t3mIVtYVYuirC4o4GGoI/oSIIHt+7nuV1dzG/ROgoREakvqjmf47KdWpoa/HjcKgGQ2RH0e1nUHuZU3wj9Q9Pv2rK3s489nf20Nwd544alnL+shYZMeUw46OWaSxaRchyeeLGT4Vii1NMXERGpWlNmzo0xbuCrwDogBnzAWrs7Z/+dwF1AArjXWvugMWYecD8QAjqBO6y10TKO/Qjw7syU/tta+yljjAs4DLyW2f4ra+3HZvpGVauhkQTxZEolLTLrVi5qovNklP3HBqZ13PHuKE/tOI7P4+b6dYvyfqhc0tHAmpWt7NjXw7//1HLnLRep/lxEROpCIZnz24CgtXYjcDfw+ewOY8xC4EPANcCbgc8YYwLAPcD91trrgBeAu8o4dhXwHuBqYCPwJmPMWuBc4Hlr7abMn5oLzOF0vXmrOrXILFs2vxG328X+aXRtSSRT/P2PdpBIOrxuzQIi4Ynv20vP72Bec5CnXjnOky8dLcWURUREql4hwfm1wMMA1tqngA05+64EtlhrY9baPmA3sDb3GOAh4MYyjj0EvMVam7TWpgAfMAJcDiwxxjxmjPlvY4yZxvsyZ/QOpksKmpU5l1nm93lY2tFA7+Aoh08MFnTM9zbv4cCxAc5d0sSqxU2TjnW7XVy/bjGhgJf/+NmrHO+JlmLaIiIiVa2QBaFNQF/Oz0ljjNdam8izbwBoHrc937aSjbXWxoGTmTKWzwIvWGtfzWTfP2Ot/a4x5lrgm8AVk73Q1tYwXq9nsiEV09ERybt9aCRdj7t0QRMRBegyiVJ3bAFYvbKdg8cHeflAD5euWTTp2Gd3HueRZw6xpKORN15xDr4C/luLNAb54DvW8oX7n+fJ7cf5w7evLdXUZQYm+ntIpBC6f6RY9XIPFRKc9wO574Y7E5jn2xcBenO2D+fZVuqxGGOCwL+QDtj/KHPMs6Tr1bHWPmmMWWKMcVlrnYleaE+VZuY6OiJ0deWv6+3qieJ2u3CRYmBwZJZnJnNFpDFYlvujPeLH63Gx+blDvGXD0gnrwo+eGuKz//4cXo+LO9+6mr1H+xkhXtA11q9dRGskwC+eOcjNVy4jFFCTqUqY7O8hkano/pFi1do9NNkHjULKWrYANwMYY64CXs7Z9zRwnTEmaIxpBlYD23OPAW4CflmusZmM+Q+BF621d1lrk5n9nwA+nJn3OuDgZIH5XJRKOfQNjtLS6MetxXJSAV6Pm2XzG+nqHeHF3afyjumPjvLF77xINJbgvW+5kHMWTC/z4XG7ef36xYyMJnlqx7FSTFtERKRqFRKcPwCMGGO2Al8EPmKM+agx5lZr7THgPtIB9aPAx621I8C9wLuNMVtIL9L8crnGkl6w+nrgJmPM5syfjcBfAa83xjwOfAG4vbi3qvp09Q6TTDnq1CIVZc5pxeN28ZUHXuaJFzvP2DcaT/Kl77/Eyb4Rbr1mBddcMnnpy0Rev24xHreLR58/guPU1GdsERGRM7j0P7rTuroGqvLNmOirnOdf7eLLP3iZyy6Yx8Wr2iswM5krylXWkrWoLcxXHtjO4HCcN12xjN9+w3nggr//4Q6e3XWCq9Ys4M63nm6HOJ2nm25avwSAv//hdp7eeYL//buXYs7Rk0NnW619pSyzS/ePFKvW7qGOjsiEJQ96CNEcdqQr3SFDmXOpNHNOK3/x3g0sntfAI88c4r7vv8S3f7GbZ3ed4IJlLdxx0+qi+5TfcNlSAH7xfOGBvYiIyFyj4HwOO3JyCICWiIJzqbz5LSH+/Pcu5+JVbby05xQ/e/YQC9rC/MnbL8HnLf6vmvOXNrO0o4EXXu2iZyBWghmLiIhUHwXnc9iRriG8HhcNQXWvkOoQDnr503eu5S1XnsOy+Y18+F1raQz5SnJul8vFDZctJZlyzqptFxERqRUKzueoRDLFse4oLY0BPdZcqorH7ea3bziPT73vSha0hkt67qvWLCAU8LB52xESyVRJzy0iIlINFJzPUce7o+lOLSppkToS9Hu5+uJF9A2O8sJrJys9HRERkZJTPcQcdehEdjGov8IzEZle95Vi3XDZEn7x3GEefe4wV1w4f9auKyIiMhsUnM9Rr+zvAWB+icsGRKpNvsB/YXsYe6iX7z+xh/am4Bn7sq0XRURE5iKVtcxBjuPw8r5TNIZ8tDeprEXqz5oVbQC8vCf/U0lFRETmKgXnc9DhriH6Bke5eFWbFoNKXVo8L0x7U5CDxwfpVVtFERGpIQrO56Dt+9LZwotXtlV4JiKV4XK5WHte+qm4L+1V9lxERGqHgvM5aPvebgDWrGyv8ExEKmdpRwOtkQAHjg7QPzRa0nP3DMR49VAvqZRT0vOKiIhMRQtC55iR0QSvHupl+YIIzQ3q1CL1y+Vyccm57TyxrZOX957imksWFX1Ox3F4fFsn335sN7HRJPOag9xw2VKuX7eIcLA0D1MSERGZjILzOWbXwV6SKYeLV6mkRWT5gkaaG/zs7exn7bntRMLT+8Ca2wlmaDjO1u3HOHoqis/rZuWiCAePD/Kdx3bzgyf2cO6SZt5382o6WkKlfhkiIiJjVNYyx2zfq3pzkaxs9txxYMe+7hmdw3Ecdh/u40db9nP0VJQl8xp427UruG7dYt656Vwuu2Aefp8He7CX/+ffnqV3UAtQRUSkfBSczzHb93YT9Hs4d0lzpaciUhVWLIwQCfvYfbifoZH4tI491T/CI08fYuv2Y+DAxosXcMPlS8ZKWAJ+Dxevauft169i/fnzGByO87Ufv0LKUS26iIiUh4LzOeR4T5QTvcOsXt6K16NfnQiA2+3i4lXtpByHF3efKmgRZ9/QKF9/aCc/2XqA4z3DLO1o4JZrV3D+0pa87UndbheXrGpj/Xnz2Hmgh4d/fbAcL0VERETB+VyS7dJyySp1aRHJde7iJhpDPnYf7uP//POveWrHsbxB+sneYR7cup+P/cOveOLFo7Q0+rlxw1JuuHwpjaHJF3y6XC7uuPlCWhr9PPDEXvZ09pXr5YiISB3TgtA5JFtTq3pzkTO53S7edOUyXtp9ir2d/fzjj1/hR1v2c8s1Kwj6PezY182Ofd0c7xkGoDHk4/ffdC5O5thCRcJ+7rxlDZ/7zxf4hx/u4JN3XEk4qL9GRUSkdPR/lTkinkix80APC9vCzFO3CJGzNIZ8XH3JQu685SJ+8qv9bHn5GF/78Stj+4N+D5eeP481K9t43UULaAj6zujWUqjVy1v5zauX8+DWA3zjp7u469Y1elKviIiUjILzOWL34V5i8aRaKIpMYcf+blYsamJecwh7qAeP282ieWE6mkNjWfJndp0o6hq3XrOSnQd6eHrnCS5Y1sINly0tet65HxTiiRS7D/fx2uFeHAfCQW/mj4/GkI8VCyP8xoZlRV9TRESqj4LzOWL7PtWbi0xHY9jH5WZ+Wc7t9bi565Y1fOrrz/DNR15lIBrn1mtWFJ1BH44l2HXg/2/v3oPjrM47jn/3otXNkizJkoXvxo4fgxljsInJhAQ1gVBKSzJAS4cWYjJpLs0fYZrm1nBLCnTSC80kIWlLJnVICKXgQiAzBlII4ZJQBnMxBHiMUSLfsJFk635d7ds/3ldiLetqWdau+X1mPO++55x337O7x9rnPXvecw7hu1vpH8iQiMdIJuK0jVgB9fXGQ6xfVUNVedG0ziciIrlHwXmeeLnhIMlEnFWL5852VUTetUYOg/nwhkU8tm0vP3vqd7zc0ML7TqsjkTWGvX7dwkk97+63O/n1K/tp2NtOJggoLEhw+spqbMlcilJJ0oMZunvTdPemaXirnZ172rjpjuf4/GWns7Su7Ji+RhERmV0KzvNAw9429jR1smZ5FYUFidmujohE5s4p5MKzl/DY83tp2NdOd2+a+jMWkJrE/+btYvwAAAy1SURBVNNMJuCFN5p5dNtuXt/VCkBZSQGnLqtkxcKKw6ZLTSbilJemKC9NMb+qmIrSFM97E/9w5zY+c/FprHvPvBl7jSIicnwpOM9xnT0D3PzjbQCct37641pF5NgqLkxywXsX89T2t9h1oJOtz+zilKWVVMxJ0dkzMDxF42Amw1st3ew60MGuA51s8yZa2nsBOHVZJfOrSlhYU0p8gqExsViMNcurOPvUOm5/8Ld8Z8t2Lv/QSs4/a7FuTBUROQEoOM9hg5kM37//Fd4+2M3F71/G6SvVOyaSi5KJOB9ct4BtrzfxWuMhnnn1AAAPP7ubolSC4sIk7V39DGbNvZ5MxFi1uILVSyqZW1Y45XOutxqqys/k2/du578e28lvXj3ApeeezJplVVMO0jOZgMYDHbzeeIiGt9rZ29xFOp0hPZghPRgQi0FddSmVc1LUVhYzt6zwsIuIyQ7fERGRiSk4z2H3Pv4mrzUeYuOaOi4+Z/lsV0dExhGPxTjrlFpWLqrgYHsvrZ39tHX20drZT3tXPxVzUlSVF1FVVkhVeSGVZUUUJKe3Dtzyk8q57uMb+O9f7uTZ197m1rtfwhbP5dL6FaxcWDHqMUEQ0NrZz97mTva83cWO3a347kP09A2OeD2QTMZJJuKk0xne2N06nFeQiFNbVcyS+WUsri2d1msQEZHDxYJg/KWuzSwOfA84HegDPunuO7Py/wr4NJAGbnL3n5vZPOCnQDGwD7ja3btzoex4r7WpqWPidb+Pk2d+u5//ePBVTqou4Vt/U09XR++o5Y5mnmZ59ymbU0RH5+htSPLTyN7qxv0d3PdkA9vfbAFgaV0ZJYVJEokYBYkwyG7r7GNvcxddvenDjq2tLGb1kkpWL53LqkVzeXFn82GLMwVBQIY4DXtbaTrUw9utPbRHM8jEYrB6SSXrrYZTllZSW1lMIn7sFp/OBAEH23rZf7Cb7r40mSAgyITpAHXVJSydX3bYGH3JPTU1ZTQ1dcx2NSSPnWhtqKambMyfOCcTnF8CXOzum8zsbOCr7v7RKK8O+AWwASgCnooe/xPwvLtvNrOvEAb1d812WXf/1/Fea64E5437O7jlJ9tIJmJce9UG1q6uG7NBKjiXyVBwfuIZayjJjt2t3PdEAzuiOdKzxWIwvzIc275wXikLa+awYkH5EVMyjvZ3ZWQb6ujup/FAJ7v2d9Dc9k56MhGnLho/v6C6hLLSFEWpBEWpJEUFCQpTCTJBQCYT/hvMBMOz0XT1punuS9PdO8DB9j72H+xmX3PXYcOBRhOPx6guL+TMVTWsWFDBSdUl1FaWTPuXibEMZt6ZPaerN03fwCBBEBAE0YVMAMWFCeZVFFMxJzXhfQTvBidaYCXH34nWhsYLziczrOUc4CEAd3/GzDZk5b0XeNrd+4A+M9sJrI2OuSUqszV6/GYOlB03OJ8N6cEMv3urncb9HTQe6KBxfyf7mrsIgoDPfmwtJ1XrJ2MROdJ4F+Yb18xn45r5w0HwYBQIp5JxElk9zF29A2xvaDmq85eVpDhteRWnLa9i7cnVvPBGM7+Pxqvva+liT1PnUT1vtlRBnIo54Sw1FaUpCgsSxGIxYrHwxtggCDjU0UdTaw/Nbb08/OxuYDcQXohUlxdRV1VCdUURyUScZCKcNz4RjxGPxcL3JXjn/clkAgZHvGc9fWEA3tTaQ9/AIP0Dg6QHJ9+Pk0zEmVdRxLyKIuaWFVJRmqK8JEVZaQHlJSmSiTjxeGy4ThB+L/SnMwykMwykB6Pt4WkQDjsqSMSHtwXRMKTh7fDj2Dtlx/iFYcxXNEYH3ljlx+rvi6eSHOroG6X88e0Tm85N0xMdOuVnnkJdpv7cM1I0LD9LF5uFXf109gwc8+ctLUrm3M30kwnOy4G2rP1BM0u6e3qUvA6gYkT6aGmzVTbn/OSRHTzx0r7h/VRBnJMXlHPuugWs0w2gIjIN8ViMeCJGcoZnYN3e0EIiEWPFogpWLKogCAI6ewZo7+qnbyAKKAczwzeZDgXY8aFtPEZhMkFBQZzCggSpZJyiwuSUvjQH0hla2nppbu+lvaufjq5+2rv7hxdwm66CZFi3spLwIiFVECeVDLfJRJx4DIjFwkAnBnWVJTS19dIcXTjsPzjuqEoRmSXnrV/EFeevmu1qHGYywXk7kL3KRTwKzEfLKwNas9J7RkmbzbLjGu8nhpnyxavO4ouTKFdTM/pCI396/upjWyERERERmTWTGZD3NPBHANGY85ez8p4FPmBmRWZWAZwCvJJ9DHAh8GSOlBURERERyVlTma1lLeHQpKsJg96d7v5ANFPKpwgD/VvcfYuZzQd+RNhj3Qxc4e5duVD2mLxrIiIiIiIzYMLgXEREREREjg9NDCsiIiIikiMUnIuIiIiI5AgF5yIiIiIiOWIyUynKLMm6Gfd0wtVQP+nuO2e3VnK8mdlG4JvuXm9mK4HNhGuAvAJ8zt0zZnYDcBGQBq5x92dnquxxe+EybWZWAPwQWAYUAjcBr6I2JJNgZgngdsCAQcIJIWKo/cgUmVktsA04n/Cz3Iza0JjUc57bPgYUufv7gK8A/zLL9ZHjzMy+BPwAGFpf/VbgWnf/AOGX5EfN7EzgXGAj8OfAbTNcVvLHXwIt0ed6IfBd1IZk8v4EwN3fD1xP+Bmr/ciURJ0E/0647gyoDU1IwXluOwd4CMDdnwE2zG51ZBa8CVyStb8e+FX0eCtwHmE7ecTdA3ffBSTNrGYGy0r+uAe4Lms/jdqQTJK73084TTHAUuAAaj8ydf8M/BswtBy62tAEFJzntnKgLWt/0Mw0FOldxN23AANZSTF3H5r/tAOo4Mh2MpQ+U2UlT7h7p7t3mFkZcC9wLWpDMgXunjazHwHfIWxDaj8yaWa2CWhy94ezktWGJqDgPLe1Ey6iNCTu7unZqozkhEzW4zKglSPbyVD6TJWVPGJmi4FfAj9295+iNiRT5O4fB1YRjj8vzspS+5GJfAI438weB9YBdwC1WflqQ6NQcJ7bniZcjRUzOxt4eXarIzngBTOrjx5fCDxJ2E4uMLO4mS0hvIhrnsGykieilZIfAb7s7j+MktWGZFLM7Eoz+2q0200Y/Dyn9iOT5e4fdPdz3b0eeBG4CtiqNjQ+DZHIbfcRXnH+mvDmhqtnuT4y+74A3G5mKeA14F53HzSzJ4HfEF5wf26Gy0r++DugErjOzIbGnn8e+LbakEzC/wD/aWZPAAXANYSfrf4GyXToe2wCsSAIJi4lIiIiIiIzTsNaRERERERyhIJzEREREZEcoeBcRERERCRHKDgXEREREckRCs5FRERERHKEgnMRERERkRyhec5FRPKYmZ1GuEDZZe6+JUp7HDgDqHX3vqyyLxKujncp8GiUXBdt90fbD7t7yzjnu4hw/vQ5QIJwPYYb3D1jZpuBDwEHo+KlQAuwyd1fHyV/yHrgSuBWYBfhug6FwF3ATdFcxfXAje5eb2b/F+VXRfXYFT3Ple6uxdpEJK8pOBcRyW+fAO4BPg1syUpvBz4CPAhgZgYsAFqj4HtdlH4jgLvfONGJzOwPge8CF7j7DjMrBu4Gvg4MLXJ0vbtvzjrmW1H+5aPlZ5UDeMDdN0X7pcDPgBuznpuorhujMpuA+qFjREROBBrWIiKSp8ysAPgL4FrgTDNbkZW9Bbgsa/9y4N5pnvJrwM3uvgPA3XuAvwZ+NUb9UsBJHNlTPiF37yLsof+smcWOusYiInlGwbmISP66CGiMguX7gU9l5W0F6qMAHuCPgZ9P83xnAM9nJ7j7Hnf/36ykb5jZS2a2B3iVcMjJl0fkv5j177ZxzvcKUA3UTLPeIiJ5Q8NaRETy19WE47IhHF5yp5kNDQHpA54EzjOz3UAD0D3N82WA3gnKXO/um6NhNL8AHnb39pH5kzxfEG17plZNEZH8pZ5zEZE8ZGa1wIXAF8zs98APgErgkqxi9xAObfkzwuB9up4DNoyoxyozu2NkQXd3wh7zO8ys4ijPtxbY4+4dR3m8iEjeUXAuIpKfrgQedfdF7r7M3ZcCNwOfySrzEPAHhEH81mNwzn8EbjCz9wCY2RzemWHlCO5+F9BIOCZ+SqKA/u+B8Ya9iIiccDSsRUQkP20ivGEy223AlwhnasHd+8zs6ehxbzQjylFz94fM7GvA3WaWAAoIe+e/Ps5hfws8ambfj/a/YWbXjChzRbS9OJruMSD8ftoCfHNalRYRyTOxIAgmLiUiIiIiIjNOPeciIjLMzO4E1oyS9YC7X3+86yMi8m6jnnMRERERkRyhG0JFRERERHKEgnMRERERkRyh4FxEREREJEcoOBcRERERyREKzkVEREREcsT/A/dZ0EQL6mMSAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,5))\n",
    "plt.title(\"Distribution of AMT_CREDIT\")\n",
    "ax = sns.distplot(application_train['AMT_CREDIT'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### Distribution of annuity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lex/anaconda3/envs/fastai-cpu/lib/python3.6/site-packages/matplotlib/axes/_axes.py:6462: UserWarning: The 'normed' kwarg is deprecated, and has been replaced by the 'density' kwarg.\n",
      "  warnings.warn(\"The 'normed' kwarg is deprecated, and has been \"\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,5))\n",
    "plt.title(\"Distribution of AMT_ANNUITY\")\n",
    "ax = sns.distplot(application_train['AMT_ANNUITY'].dropna())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### Balance of dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x12edc70f0>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "application_train['CODE_GENDER'].value_counts().plot(kind='pie', title='Gender balance')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x12eebccf8>"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "application_train['FLAG_OWN_CAR'].value_counts().plot(kind='pie', title='Own car balance')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x12f191e48>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "application_train['FLAG_OWN_REALTY'].value_counts().plot(kind='pie', title='Own realty balance')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x130983e80>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "application_train['CNT_CHILDREN'].value_counts().plot(kind='pie', title='Num children')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x12ee236d8>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "application_train['NAME_EDUCATION_TYPE'].value_counts().plot(kind='pie', title='Education')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Now let's check out the other DFs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "dfs = []\n",
    "\n",
    "for file in [\n",
    "    'POS_CASH_balance.csv', 'bureau.csv', 'bureau_balance.csv', 'credit_card_balance.csv',\n",
    "    'installments_payments.csv', 'previous_application.csv'\n",
    "]:\n",
    "    dfs.append(pd.read_csv(PATH / file))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "pos, bureau, bureau_balance, credit_card_balance, installments_payments, previous_application = dfs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### POS"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Monthly balance snapshots of previous POS (point of sales) and cash loans that the applicant had with Home Credit.\n",
    "\n",
    "This table has one row for each month of history of every previous credit in Home Credit (consumer credit and cash loans) related to loans in our sample – i.e. the table has (#loans in sample * # of relative previous credits * # of months in which we have some history observable for the previous credits) rows."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_PREV</th>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th>MONTHS_BALANCE</th>\n",
       "      <th>CNT_INSTALMENT</th>\n",
       "      <th>CNT_INSTALMENT_FUTURE</th>\n",
       "      <th>NAME_CONTRACT_STATUS</th>\n",
       "      <th>SK_DPD</th>\n",
       "      <th>SK_DPD_DEF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1803195</td>\n",
       "      <td>182943</td>\n",
       "      <td>-31</td>\n",
       "      <td>48.0</td>\n",
       "      <td>45.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1715348</td>\n",
       "      <td>367990</td>\n",
       "      <td>-33</td>\n",
       "      <td>36.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1784872</td>\n",
       "      <td>397406</td>\n",
       "      <td>-32</td>\n",
       "      <td>12.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1903291</td>\n",
       "      <td>269225</td>\n",
       "      <td>-35</td>\n",
       "      <td>48.0</td>\n",
       "      <td>42.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2341044</td>\n",
       "      <td>334279</td>\n",
       "      <td>-35</td>\n",
       "      <td>36.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SK_ID_PREV  SK_ID_CURR  MONTHS_BALANCE  CNT_INSTALMENT  \\\n",
       "0     1803195      182943             -31            48.0   \n",
       "1     1715348      367990             -33            36.0   \n",
       "2     1784872      397406             -32            12.0   \n",
       "3     1903291      269225             -35            48.0   \n",
       "4     2341044      334279             -35            36.0   \n",
       "\n",
       "   CNT_INSTALMENT_FUTURE NAME_CONTRACT_STATUS  SK_DPD  SK_DPD_DEF  \n",
       "0                   45.0               Active       0           0  \n",
       "1                   35.0               Active       0           0  \n",
       "2                    9.0               Active       0           0  \n",
       "3                   42.0               Active       0           0  \n",
       "4                   35.0               Active       0           0  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pos.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "pos_col_desc = column_desc[column_desc['Table'] == 'POS_CASH_balance.csv']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* SK_ID_PREV: ID of previous credit in Home Credit related to loan in our sample. (One loan in our sample can have 0,1,2 or more previous loans in Home Credit)\n",
      "* SK_ID_CURR: ID of loan in our sample\n",
      "* MONTHS_BALANCE: Month of balance relative to application date (-1 means the information to the freshest monthly snapshot, 0 means the information at application - often it will be the same as -1 as many banks are not updating the information to Credit Bureau regularly )\n",
      "* CNT_INSTALMENT: Term of previous credit (can change over time)\n",
      "* CNT_INSTALMENT_FUTURE: Installments left to pay on the previous credit\n",
      "* NAME_CONTRACT_STATUS: Contract status during the month\n",
      "* SK_DPD: DPD (days past due) during the month of previous credit\n",
      "* SK_DPD_DEF: DPD during the month with tolerance (debts with low loan amounts are ignored) of the previous credit\n"
     ]
    }
   ],
   "source": [
    "for idx, row in pos_col_desc.iterrows():\n",
    "    print(f\"* {row['Row'].strip()}: {row['Description']}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_PREV</th>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th>MONTHS_BALANCE</th>\n",
       "      <th>CNT_INSTALMENT</th>\n",
       "      <th>CNT_INSTALMENT_FUTURE</th>\n",
       "      <th>NAME_CONTRACT_STATUS</th>\n",
       "      <th>SK_DPD</th>\n",
       "      <th>SK_DPD_DEF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2197888</th>\n",
       "      <td>1369693</td>\n",
       "      <td>100001</td>\n",
       "      <td>-53</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Completed</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4704415</th>\n",
       "      <td>1369693</td>\n",
       "      <td>100001</td>\n",
       "      <td>-54</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7823681</th>\n",
       "      <td>1369693</td>\n",
       "      <td>100001</td>\n",
       "      <td>-55</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8789081</th>\n",
       "      <td>1369693</td>\n",
       "      <td>100001</td>\n",
       "      <td>-56</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7167007</th>\n",
       "      <td>1369693</td>\n",
       "      <td>100001</td>\n",
       "      <td>-57</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4928574</th>\n",
       "      <td>1851984</td>\n",
       "      <td>100001</td>\n",
       "      <td>-93</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Completed</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8531326</th>\n",
       "      <td>1851984</td>\n",
       "      <td>100001</td>\n",
       "      <td>-94</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1891462</th>\n",
       "      <td>1851984</td>\n",
       "      <td>100001</td>\n",
       "      <td>-95</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>7</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1261679</th>\n",
       "      <td>1851984</td>\n",
       "      <td>100001</td>\n",
       "      <td>-96</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         SK_ID_PREV  SK_ID_CURR  MONTHS_BALANCE  CNT_INSTALMENT  \\\n",
       "2197888     1369693      100001             -53             4.0   \n",
       "4704415     1369693      100001             -54             4.0   \n",
       "7823681     1369693      100001             -55             4.0   \n",
       "8789081     1369693      100001             -56             4.0   \n",
       "7167007     1369693      100001             -57             4.0   \n",
       "4928574     1851984      100001             -93             4.0   \n",
       "8531326     1851984      100001             -94             4.0   \n",
       "1891462     1851984      100001             -95             4.0   \n",
       "1261679     1851984      100001             -96             4.0   \n",
       "\n",
       "         CNT_INSTALMENT_FUTURE NAME_CONTRACT_STATUS  SK_DPD  SK_DPD_DEF  \n",
       "2197888                    0.0            Completed       0           0  \n",
       "4704415                    1.0               Active       0           0  \n",
       "7823681                    2.0               Active       0           0  \n",
       "8789081                    3.0               Active       0           0  \n",
       "7167007                    4.0               Active       0           0  \n",
       "4928574                    0.0            Completed       0           0  \n",
       "8531326                    0.0               Active       0           0  \n",
       "1891462                    1.0               Active       7           7  \n",
       "1261679                    2.0               Active       0           0  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pos[pos['SK_ID_CURR'] == 100001].sort_values(by='MONTHS_BALANCE', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "* I could take an approach of including each months balance values as a column:\n",
    "  \n",
    "  ``MONTHS_BALANCER_-96_SK_DPD``\n",
    "  \n",
    "  Going to create a lot of columns though.\n",
    "  \n",
    "* Another alternative would be to sum the ``SK_DPD`` and ``SK_DPD_DEF`` values, but that doesn't capture the other information that could be useful.\n",
    "\n",
    "* I could also create previous loan columns:\n",
    "  loan_1_installment: 4\n",
    "  loan_1_days_past_due: 7\n",
    "  \n",
    "* Could consider average DPD per loan."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### Bureau"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "All client's previous credits provided by other financial institutions that were reported to Credit Bureau (for clients who have a loan in our sample).\n",
    "\n",
    "For every loan in our sample, there are as many rows as number of credits the client had in Credit Bureau before the application date."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "pos_col_desc = column_desc[column_desc['Table'] == 'bureau.csv']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* SK_ID_CURR: ID of loan in our sample - one loan in our sample can have 0,1,2 or more related previous credits in credit bureau \n",
      "* SK_BUREAU_ID: Recoded ID of previous Credit Bureau credit related to our loan (unique coding for each loan application)\n",
      "* CREDIT_ACTIVE: Status of the Credit Bureau (CB) reported credits\n",
      "* CREDIT_CURRENCY: Recoded currency of the Credit Bureau credit\n",
      "* DAYS_CREDIT: How many days before current application did client apply for Credit Bureau credit\n",
      "* CREDIT_DAY_OVERDUE: Number of days past due on CB credit at the time of application for related loan in our sample\n",
      "* DAYS_CREDIT_ENDDATE: Remaining duration of CB credit (in days) at the time of application in Home Credit\n",
      "* DAYS_ENDDATE_FACT: Days since CB credit ended at the time of application in Home Credit (only for closed credit)\n",
      "* AMT_CREDIT_MAX_OVERDUE: Maximal amount overdue on the Credit Bureau credit so far (at application date of loan in our sample)\n",
      "* CNT_CREDIT_PROLONG: How many times was the Credit Bureau credit prolonged\n",
      "* AMT_CREDIT_SUM: Current credit amount for the Credit Bureau credit\n",
      "* AMT_CREDIT_SUM_DEBT: Current debt on Credit Bureau credit\n",
      "* AMT_CREDIT_SUM_LIMIT: Current credit limit of credit card reported in Credit Bureau\n",
      "* AMT_CREDIT_SUM_OVERDUE: Current amount overdue on Credit Bureau credit\n",
      "* CREDIT_TYPE: Type of Credit Bureau credit (Car, cash,...)\n",
      "* DAYS_CREDIT_UPDATE: How many days before loan application did last information about the Credit Bureau credit come\n",
      "* AMT_ANNUITY: Annuity of the Credit Bureau credit\n"
     ]
    }
   ],
   "source": [
    "for idx, row in pos_col_desc.iterrows():\n",
    "    print(f\"* {row['Row'].strip()}: {row['Description']}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th>SK_ID_BUREAU</th>\n",
       "      <th>CREDIT_ACTIVE</th>\n",
       "      <th>CREDIT_CURRENCY</th>\n",
       "      <th>DAYS_CREDIT</th>\n",
       "      <th>CREDIT_DAY_OVERDUE</th>\n",
       "      <th>DAYS_CREDIT_ENDDATE</th>\n",
       "      <th>DAYS_ENDDATE_FACT</th>\n",
       "      <th>AMT_CREDIT_MAX_OVERDUE</th>\n",
       "      <th>CNT_CREDIT_PROLONG</th>\n",
       "      <th>AMT_CREDIT_SUM</th>\n",
       "      <th>AMT_CREDIT_SUM_DEBT</th>\n",
       "      <th>AMT_CREDIT_SUM_LIMIT</th>\n",
       "      <th>AMT_CREDIT_SUM_OVERDUE</th>\n",
       "      <th>CREDIT_TYPE</th>\n",
       "      <th>DAYS_CREDIT_UPDATE</th>\n",
       "      <th>AMT_ANNUITY</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>248484</th>\n",
       "      <td>100001</td>\n",
       "      <td>5896630</td>\n",
       "      <td>Closed</td>\n",
       "      <td>currency 1</td>\n",
       "      <td>-857</td>\n",
       "      <td>0</td>\n",
       "      <td>-492.0</td>\n",
       "      <td>-553.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>112500.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Consumer credit</td>\n",
       "      <td>-155</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248485</th>\n",
       "      <td>100001</td>\n",
       "      <td>5896631</td>\n",
       "      <td>Closed</td>\n",
       "      <td>currency 1</td>\n",
       "      <td>-909</td>\n",
       "      <td>0</td>\n",
       "      <td>-179.0</td>\n",
       "      <td>-877.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>279720.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Consumer credit</td>\n",
       "      <td>-155</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248486</th>\n",
       "      <td>100001</td>\n",
       "      <td>5896632</td>\n",
       "      <td>Closed</td>\n",
       "      <td>currency 1</td>\n",
       "      <td>-879</td>\n",
       "      <td>0</td>\n",
       "      <td>-514.0</td>\n",
       "      <td>-544.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>91620.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Consumer credit</td>\n",
       "      <td>-155</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248487</th>\n",
       "      <td>100001</td>\n",
       "      <td>5896633</td>\n",
       "      <td>Closed</td>\n",
       "      <td>currency 1</td>\n",
       "      <td>-1572</td>\n",
       "      <td>0</td>\n",
       "      <td>-1329.0</td>\n",
       "      <td>-1328.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>85500.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Consumer credit</td>\n",
       "      <td>-155</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248488</th>\n",
       "      <td>100001</td>\n",
       "      <td>5896634</td>\n",
       "      <td>Active</td>\n",
       "      <td>currency 1</td>\n",
       "      <td>-559</td>\n",
       "      <td>0</td>\n",
       "      <td>902.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>337680.0</td>\n",
       "      <td>113166.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Consumer credit</td>\n",
       "      <td>-6</td>\n",
       "      <td>4630.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248489</th>\n",
       "      <td>100001</td>\n",
       "      <td>5896635</td>\n",
       "      <td>Active</td>\n",
       "      <td>currency 1</td>\n",
       "      <td>-49</td>\n",
       "      <td>0</td>\n",
       "      <td>1778.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>378000.0</td>\n",
       "      <td>373239.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Consumer credit</td>\n",
       "      <td>-16</td>\n",
       "      <td>10822.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248490</th>\n",
       "      <td>100001</td>\n",
       "      <td>5896636</td>\n",
       "      <td>Active</td>\n",
       "      <td>currency 1</td>\n",
       "      <td>-320</td>\n",
       "      <td>0</td>\n",
       "      <td>411.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>168345.0</td>\n",
       "      <td>110281.5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Consumer credit</td>\n",
       "      <td>-10</td>\n",
       "      <td>9364.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        SK_ID_CURR  SK_ID_BUREAU CREDIT_ACTIVE CREDIT_CURRENCY  DAYS_CREDIT  \\\n",
       "248484      100001       5896630        Closed      currency 1         -857   \n",
       "248485      100001       5896631        Closed      currency 1         -909   \n",
       "248486      100001       5896632        Closed      currency 1         -879   \n",
       "248487      100001       5896633        Closed      currency 1        -1572   \n",
       "248488      100001       5896634        Active      currency 1         -559   \n",
       "248489      100001       5896635        Active      currency 1          -49   \n",
       "248490      100001       5896636        Active      currency 1         -320   \n",
       "\n",
       "        CREDIT_DAY_OVERDUE  DAYS_CREDIT_ENDDATE  DAYS_ENDDATE_FACT  \\\n",
       "248484                   0               -492.0             -553.0   \n",
       "248485                   0               -179.0             -877.0   \n",
       "248486                   0               -514.0             -544.0   \n",
       "248487                   0              -1329.0            -1328.0   \n",
       "248488                   0                902.0                NaN   \n",
       "248489                   0               1778.0                NaN   \n",
       "248490                   0                411.0                NaN   \n",
       "\n",
       "        AMT_CREDIT_MAX_OVERDUE  CNT_CREDIT_PROLONG  AMT_CREDIT_SUM  \\\n",
       "248484                     NaN                   0        112500.0   \n",
       "248485                     NaN                   0        279720.0   \n",
       "248486                     NaN                   0         91620.0   \n",
       "248487                     NaN                   0         85500.0   \n",
       "248488                     NaN                   0        337680.0   \n",
       "248489                     NaN                   0        378000.0   \n",
       "248490                     NaN                   0        168345.0   \n",
       "\n",
       "        AMT_CREDIT_SUM_DEBT  AMT_CREDIT_SUM_LIMIT  AMT_CREDIT_SUM_OVERDUE  \\\n",
       "248484                  0.0                   0.0                     0.0   \n",
       "248485                  0.0                   0.0                     0.0   \n",
       "248486                  0.0                   0.0                     0.0   \n",
       "248487                  0.0                   0.0                     0.0   \n",
       "248488             113166.0                   0.0                     0.0   \n",
       "248489             373239.0                   0.0                     0.0   \n",
       "248490             110281.5                   NaN                     0.0   \n",
       "\n",
       "            CREDIT_TYPE  DAYS_CREDIT_UPDATE  AMT_ANNUITY  \n",
       "248484  Consumer credit                -155          0.0  \n",
       "248485  Consumer credit                -155          0.0  \n",
       "248486  Consumer credit                -155          0.0  \n",
       "248487  Consumer credit                -155          0.0  \n",
       "248488  Consumer credit                  -6       4630.5  \n",
       "248489  Consumer credit                 -16      10822.5  \n",
       "248490  Consumer credit                 -10       9364.5  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bureau[bureau['SK_ID_CURR'] == 100001]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "All client's previous credits provided by other financial institutions that were reported to Credit Bureau (for clients who have a loan in our sample).\n",
    "\n",
    "For every loan in our sample, there are as many rows as number of credits the client had in Credit Bureau before the application date."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### Bureau Balance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Monthly balances of previous credits in Credit Bureau.\n",
    "\n",
    "This table has one row for each month of history of every previous credit reported to Credit Bureau – i.e the table has (#loans in sample * # of relative previous credits * # of months where we have some history observable for the previous credits) rows."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "bureau_balance_col_desc = column_desc[column_desc['Table'] == 'bureau_balance.csv']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* SK_BUREAU_ID: Recoded ID of Credit Bureau credit (unique coding for each application) - use this to join to CREDIT_BUREAU table \n",
      "* MONTHS_BALANCE: Month of balance relative to application date (-1 means the freshest balance date)\n",
      "* STATUS: Status of Credit Bureau loan during the month (active, closed, DPD0-30,",
      " [C means closed, X means status unknown, 0 means no DPD, 1 means maximal did during month between 1-30, 2 means DPD 31-60,",
      " 5 means DPD 120+ or sold or written off ] )\n"
     ]
    }
   ],
   "source": [
    "for idx, row in bureau_balance_col_desc.iterrows():\n",
    "    print(f\"* {row['Row'].strip()}: {row['Description']}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_BUREAU</th>\n",
       "      <th>MONTHS_BALANCE</th>\n",
       "      <th>STATUS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4040588</th>\n",
       "      <td>5896630</td>\n",
       "      <td>0</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4040589</th>\n",
       "      <td>5896630</td>\n",
       "      <td>-1</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4040590</th>\n",
       "      <td>5896630</td>\n",
       "      <td>-2</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4040591</th>\n",
       "      <td>5896630</td>\n",
       "      <td>-3</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4040592</th>\n",
       "      <td>5896630</td>\n",
       "      <td>-4</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         SK_ID_BUREAU  MONTHS_BALANCE STATUS\n",
       "4040588       5896630               0      C\n",
       "4040589       5896630              -1      C\n",
       "4040590       5896630              -2      C\n",
       "4040591       5896630              -3      C\n",
       "4040592       5896630              -4      C"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bureau_balance[bureau_balance['SK_ID_BUREAU'] == 5896630].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### Credit card balance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Monthly balance snapshots of previous credit cards that the applicant has with Home Credit.\n",
    "\n",
    "This table has one row for each month of history of every previous credit in Home Credit (consumer credit and cash loans) related to loans in our sample – i.e. the table has (#loans in sample * # of relative previous credit cards * # of months where we have some history observable for the previous credit card) rows."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "credit_card_balance_col_desc = column_desc[column_desc['Table'] == 'credit_card_balance.csv']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* SK_ID_PREV: ID of previous credit in Home credit related to loan in our sample. (One loan in our sample can have 0,1,2 or more previous loans in Home Credit)\n",
      "* SK_ID_CURR: ID of loan in our sample\n",
      "* MONTHS_BALANCE: Month of balance relative to application date (-1 means the freshest balance date)\n",
      "* AMT_BALANCE: Balance during the month of previous credit\n",
      "* AMT_CREDIT_LIMIT_ACTUAL: Credit card limit during the month of the previous credit\n",
      "* AMT_DRAWINGS_ATM_CURRENT: Amount drawing at ATM during the month of the previous credit\n",
      "* AMT_DRAWINGS_CURRENT: Amount drawing during the month of the previous credit\n",
      "* AMT_DRAWINGS_OTHER_CURRENT: Amount of other drawings during the month of the previous credit\n",
      "* AMT_DRAWINGS_POS_CURRENT: Amount drawing or buying goods during the month of the previous credit\n",
      "* AMT_INST_MIN_REGULARITY: Minimal installment for this month of the previous credit\n",
      "* AMT_PAYMENT_CURRENT: How much did the client pay during the month on the previous credit\n",
      "* AMT_PAYMENT_TOTAL_CURRENT: How much did the client pay during the month in total on the previous credit\n",
      "* AMT_RECEIVABLE_PRINCIPAL: Amount receivable for principal on the previous credit\n",
      "* AMT_RECIVABLE: Amount receivable on the previous credit\n",
      "* AMT_TOTAL_RECEIVABLE: Total amount receivable on the previous credit\n",
      "* CNT_DRAWINGS_ATM_CURRENT: Number of drawings at ATM during this month on the previous credit\n",
      "* CNT_DRAWINGS_CURRENT: Number of drawings during this month on the previous credit\n",
      "* CNT_DRAWINGS_OTHER_CURRENT: Number of other drawings during this month on the previous credit\n",
      "* CNT_DRAWINGS_POS_CURRENT: Number of drawings for goods during this month on the previous credit\n",
      "* CNT_INSTALMENT_MATURE_CUM: Number of paid installments on the previous credit\n",
      "* NAME_CONTRACT_STATUS: Contract status (active signed,...) on the previous credit\n",
      "* SK_DPD: DPD (Days past due) during the month on the previous credit\n",
      "* SK_DPD_DEF: DPD (Days past due) during the month with tolerance (debts with low loan amounts are ignored) of the previous credit\n"
     ]
    }
   ],
   "source": [
    "for idx, row in credit_card_balance_col_desc.iterrows():\n",
    "    print(f\"* {row['Row'].strip()}: {row['Description']}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_PREV</th>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th>MONTHS_BALANCE</th>\n",
       "      <th>AMT_BALANCE</th>\n",
       "      <th>AMT_CREDIT_LIMIT_ACTUAL</th>\n",
       "      <th>AMT_DRAWINGS_ATM_CURRENT</th>\n",
       "      <th>AMT_DRAWINGS_CURRENT</th>\n",
       "      <th>AMT_DRAWINGS_OTHER_CURRENT</th>\n",
       "      <th>AMT_DRAWINGS_POS_CURRENT</th>\n",
       "      <th>AMT_INST_MIN_REGULARITY</th>\n",
       "      <th>...</th>\n",
       "      <th>AMT_RECIVABLE</th>\n",
       "      <th>AMT_TOTAL_RECEIVABLE</th>\n",
       "      <th>CNT_DRAWINGS_ATM_CURRENT</th>\n",
       "      <th>CNT_DRAWINGS_CURRENT</th>\n",
       "      <th>CNT_DRAWINGS_OTHER_CURRENT</th>\n",
       "      <th>CNT_DRAWINGS_POS_CURRENT</th>\n",
       "      <th>CNT_INSTALMENT_MATURE_CUM</th>\n",
       "      <th>NAME_CONTRACT_STATUS</th>\n",
       "      <th>SK_DPD</th>\n",
       "      <th>SK_DPD_DEF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-6</td>\n",
       "      <td>56.970</td>\n",
       "      <td>135000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>877.500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>877.500</td>\n",
       "      <td>1700.325</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122799</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-30</td>\n",
       "      <td>45328.320</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2332.035</td>\n",
       "      <td>...</td>\n",
       "      <td>45670.275</td>\n",
       "      <td>45670.275</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153474</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-8</td>\n",
       "      <td>129.240</td>\n",
       "      <td>135000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>22005.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>22005.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>33.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>330701</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-2</td>\n",
       "      <td>28.575</td>\n",
       "      <td>9000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>1916.910</td>\n",
       "      <td>...</td>\n",
       "      <td>28.575</td>\n",
       "      <td>28.575</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>419198</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-12</td>\n",
       "      <td>0.000</td>\n",
       "      <td>45000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>561468</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-39</td>\n",
       "      <td>56570.940</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2761.740</td>\n",
       "      <td>...</td>\n",
       "      <td>56570.940</td>\n",
       "      <td>56570.940</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>577279</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-27</td>\n",
       "      <td>43030.845</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>43360.290</td>\n",
       "      <td>43360.290</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>596084</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-19</td>\n",
       "      <td>0.000</td>\n",
       "      <td>45000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>281.475</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>722949</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-13</td>\n",
       "      <td>0.000</td>\n",
       "      <td>45000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>889286</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-17</td>\n",
       "      <td>0.000</td>\n",
       "      <td>45000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>917059</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-7</td>\n",
       "      <td>1687.365</td>\n",
       "      <td>135000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1790.145</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1790.145</td>\n",
       "      <td>129.240</td>\n",
       "      <td>...</td>\n",
       "      <td>1687.365</td>\n",
       "      <td>1687.365</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>986730</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-21</td>\n",
       "      <td>14332.905</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>14452.470</td>\n",
       "      <td>14452.470</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1007062</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-29</td>\n",
       "      <td>44749.395</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2292.345</td>\n",
       "      <td>...</td>\n",
       "      <td>45089.460</td>\n",
       "      <td>45089.460</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1039133</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-44</td>\n",
       "      <td>60736.680</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3291.390</td>\n",
       "      <td>...</td>\n",
       "      <td>61224.345</td>\n",
       "      <td>61224.345</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1210264</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-23</td>\n",
       "      <td>17579.925</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>17723.115</td>\n",
       "      <td>17723.115</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1448236</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-37</td>\n",
       "      <td>49298.310</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2667.825</td>\n",
       "      <td>...</td>\n",
       "      <td>49693.590</td>\n",
       "      <td>49693.590</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1548947</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-4</td>\n",
       "      <td>4036.860</td>\n",
       "      <td>9000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>900.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>900.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>4036.860</td>\n",
       "      <td>4036.860</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1614906</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-25</td>\n",
       "      <td>20609.010</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>20770.875</td>\n",
       "      <td>20770.875</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1663555</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-32</td>\n",
       "      <td>46864.575</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2420.010</td>\n",
       "      <td>...</td>\n",
       "      <td>47206.530</td>\n",
       "      <td>47206.530</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1679171</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-22</td>\n",
       "      <td>15988.005</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>16119.675</td>\n",
       "      <td>16119.675</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1764744</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-42</td>\n",
       "      <td>58743.675</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3194.190</td>\n",
       "      <td>...</td>\n",
       "      <td>59216.985</td>\n",
       "      <td>59216.985</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1811769</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-48</td>\n",
       "      <td>69186.690</td>\n",
       "      <td>270000</td>\n",
       "      <td>67500.0</td>\n",
       "      <td>67500.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>67841.190</td>\n",
       "      <td>67841.190</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1821907</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-5</td>\n",
       "      <td>5228.280</td>\n",
       "      <td>9000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5229.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5229.000</td>\n",
       "      <td>840.780</td>\n",
       "      <td>...</td>\n",
       "      <td>5228.280</td>\n",
       "      <td>5228.280</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>36.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1845734</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-31</td>\n",
       "      <td>46119.600</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2369.295</td>\n",
       "      <td>...</td>\n",
       "      <td>46461.555</td>\n",
       "      <td>46461.555</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021610</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-43</td>\n",
       "      <td>59786.865</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3242.250</td>\n",
       "      <td>...</td>\n",
       "      <td>60267.735</td>\n",
       "      <td>60267.735</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2029171</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-14</td>\n",
       "      <td>0.000</td>\n",
       "      <td>45000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2070436</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-3</td>\n",
       "      <td>1897.875</td>\n",
       "      <td>9000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>1897.875</td>\n",
       "      <td>1897.875</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2239489</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-41</td>\n",
       "      <td>57725.595</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3141.450</td>\n",
       "      <td>...</td>\n",
       "      <td>58191.525</td>\n",
       "      <td>58191.525</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2257989</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-10</td>\n",
       "      <td>11112.525</td>\n",
       "      <td>135000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>11112.525</td>\n",
       "      <td>11112.525</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>31.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2286944</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-36</td>\n",
       "      <td>44956.125</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2629.620</td>\n",
       "      <td>...</td>\n",
       "      <td>45346.185</td>\n",
       "      <td>45346.185</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2339745</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-9</td>\n",
       "      <td>9169.380</td>\n",
       "      <td>135000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>9169.380</td>\n",
       "      <td>9169.380</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2377322</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-35</td>\n",
       "      <td>45877.140</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2501.010</td>\n",
       "      <td>...</td>\n",
       "      <td>46247.445</td>\n",
       "      <td>46247.445</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2533249</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-20</td>\n",
       "      <td>281.475</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>35.055</td>\n",
       "      <td>35.055</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2584127</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-47</td>\n",
       "      <td>62727.075</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3498.975</td>\n",
       "      <td>...</td>\n",
       "      <td>63246.825</td>\n",
       "      <td>63246.825</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2648695</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-40</td>\n",
       "      <td>51462.360</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3089.925</td>\n",
       "      <td>...</td>\n",
       "      <td>51920.145</td>\n",
       "      <td>51920.145</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2653597</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-45</td>\n",
       "      <td>61484.400</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3346.695</td>\n",
       "      <td>...</td>\n",
       "      <td>61980.255</td>\n",
       "      <td>61980.255</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2736406</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-34</td>\n",
       "      <td>46585.530</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2456.640</td>\n",
       "      <td>...</td>\n",
       "      <td>46949.895</td>\n",
       "      <td>46949.895</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2853861</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-24</td>\n",
       "      <td>19131.840</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>19286.325</td>\n",
       "      <td>19286.325</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2873870</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-15</td>\n",
       "      <td>0.000</td>\n",
       "      <td>45000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3064061</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-26</td>\n",
       "      <td>22064.355</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>22387.725</td>\n",
       "      <td>22387.725</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3123527</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-16</td>\n",
       "      <td>0.000</td>\n",
       "      <td>45000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3205656</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-28</td>\n",
       "      <td>43969.230</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2263.050</td>\n",
       "      <td>...</td>\n",
       "      <td>44304.570</td>\n",
       "      <td>44304.570</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3284876</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-46</td>\n",
       "      <td>62577.900</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>3399.795</td>\n",
       "      <td>...</td>\n",
       "      <td>63082.080</td>\n",
       "      <td>63082.080</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3486594</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-38</td>\n",
       "      <td>52754.445</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2713.005</td>\n",
       "      <td>...</td>\n",
       "      <td>53157.240</td>\n",
       "      <td>53157.240</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3568086</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-18</td>\n",
       "      <td>0.000</td>\n",
       "      <td>45000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3657049</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-33</td>\n",
       "      <td>47879.325</td>\n",
       "      <td>270000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>2355.615</td>\n",
       "      <td>...</td>\n",
       "      <td>48224.745</td>\n",
       "      <td>48224.745</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3748710</th>\n",
       "      <td>2562384</td>\n",
       "      <td>378907</td>\n",
       "      <td>-11</td>\n",
       "      <td>13002.075</td>\n",
       "      <td>135000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>26640.000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>26640.000</td>\n",
       "      <td>2250.000</td>\n",
       "      <td>...</td>\n",
       "      <td>13002.075</td>\n",
       "      <td>13002.075</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>Active</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>47 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         SK_ID_PREV  SK_ID_CURR  MONTHS_BALANCE  AMT_BALANCE  \\\n",
       "0           2562384      378907              -6       56.970   \n",
       "122799      2562384      378907             -30    45328.320   \n",
       "153474      2562384      378907              -8      129.240   \n",
       "330701      2562384      378907              -2       28.575   \n",
       "419198      2562384      378907             -12        0.000   \n",
       "561468      2562384      378907             -39    56570.940   \n",
       "577279      2562384      378907             -27    43030.845   \n",
       "596084      2562384      378907             -19        0.000   \n",
       "722949      2562384      378907             -13        0.000   \n",
       "889286      2562384      378907             -17        0.000   \n",
       "917059      2562384      378907              -7     1687.365   \n",
       "986730      2562384      378907             -21    14332.905   \n",
       "1007062     2562384      378907             -29    44749.395   \n",
       "1039133     2562384      378907             -44    60736.680   \n",
       "1210264     2562384      378907             -23    17579.925   \n",
       "1448236     2562384      378907             -37    49298.310   \n",
       "1548947     2562384      378907              -4     4036.860   \n",
       "1614906     2562384      378907             -25    20609.010   \n",
       "1663555     2562384      378907             -32    46864.575   \n",
       "1679171     2562384      378907             -22    15988.005   \n",
       "1764744     2562384      378907             -42    58743.675   \n",
       "1811769     2562384      378907             -48    69186.690   \n",
       "1821907     2562384      378907              -5     5228.280   \n",
       "1845734     2562384      378907             -31    46119.600   \n",
       "2021610     2562384      378907             -43    59786.865   \n",
       "2029171     2562384      378907             -14        0.000   \n",
       "2070436     2562384      378907              -3     1897.875   \n",
       "2239489     2562384      378907             -41    57725.595   \n",
       "2257989     2562384      378907             -10    11112.525   \n",
       "2286944     2562384      378907             -36    44956.125   \n",
       "2339745     2562384      378907              -9     9169.380   \n",
       "2377322     2562384      378907             -35    45877.140   \n",
       "2533249     2562384      378907             -20      281.475   \n",
       "2584127     2562384      378907             -47    62727.075   \n",
       "2648695     2562384      378907             -40    51462.360   \n",
       "2653597     2562384      378907             -45    61484.400   \n",
       "2736406     2562384      378907             -34    46585.530   \n",
       "2853861     2562384      378907             -24    19131.840   \n",
       "2873870     2562384      378907             -15        0.000   \n",
       "3064061     2562384      378907             -26    22064.355   \n",
       "3123527     2562384      378907             -16        0.000   \n",
       "3205656     2562384      378907             -28    43969.230   \n",
       "3284876     2562384      378907             -46    62577.900   \n",
       "3486594     2562384      378907             -38    52754.445   \n",
       "3568086     2562384      378907             -18        0.000   \n",
       "3657049     2562384      378907             -33    47879.325   \n",
       "3748710     2562384      378907             -11    13002.075   \n",
       "\n",
       "         AMT_CREDIT_LIMIT_ACTUAL  AMT_DRAWINGS_ATM_CURRENT  \\\n",
       "0                         135000                       0.0   \n",
       "122799                    270000                       0.0   \n",
       "153474                    135000                       0.0   \n",
       "330701                      9000                       0.0   \n",
       "419198                     45000                       0.0   \n",
       "561468                    270000                       0.0   \n",
       "577279                    270000                       0.0   \n",
       "596084                     45000                       0.0   \n",
       "722949                     45000                       0.0   \n",
       "889286                     45000                       0.0   \n",
       "917059                    135000                       0.0   \n",
       "986730                    270000                       0.0   \n",
       "1007062                   270000                       0.0   \n",
       "1039133                   270000                       0.0   \n",
       "1210264                   270000                       0.0   \n",
       "1448236                   270000                       0.0   \n",
       "1548947                     9000                       0.0   \n",
       "1614906                   270000                       0.0   \n",
       "1663555                   270000                       0.0   \n",
       "1679171                   270000                       0.0   \n",
       "1764744                   270000                       0.0   \n",
       "1811769                   270000                   67500.0   \n",
       "1821907                     9000                       0.0   \n",
       "1845734                   270000                       0.0   \n",
       "2021610                   270000                       0.0   \n",
       "2029171                    45000                       0.0   \n",
       "2070436                     9000                       0.0   \n",
       "2239489                   270000                       0.0   \n",
       "2257989                   135000                       0.0   \n",
       "2286944                   270000                       0.0   \n",
       "2339745                   135000                       0.0   \n",
       "2377322                   270000                       0.0   \n",
       "2533249                   270000                       0.0   \n",
       "2584127                   270000                       0.0   \n",
       "2648695                   270000                       0.0   \n",
       "2653597                   270000                       0.0   \n",
       "2736406                   270000                       0.0   \n",
       "2853861                   270000                       0.0   \n",
       "2873870                    45000                       0.0   \n",
       "3064061                   270000                       0.0   \n",
       "3123527                    45000                       0.0   \n",
       "3205656                   270000                       0.0   \n",
       "3284876                   270000                       0.0   \n",
       "3486594                   270000                       0.0   \n",
       "3568086                    45000                       0.0   \n",
       "3657049                   270000                       0.0   \n",
       "3748710                   135000                       0.0   \n",
       "\n",
       "         AMT_DRAWINGS_CURRENT  AMT_DRAWINGS_OTHER_CURRENT  \\\n",
       "0                     877.500                         0.0   \n",
       "122799                  0.000                         0.0   \n",
       "153474              22005.000                         0.0   \n",
       "330701                  0.000                         0.0   \n",
       "419198                  0.000                         0.0   \n",
       "561468                  0.000                         0.0   \n",
       "577279                  0.000                         0.0   \n",
       "596084                  0.000                         0.0   \n",
       "722949                  0.000                         0.0   \n",
       "889286                  0.000                         0.0   \n",
       "917059               1790.145                         0.0   \n",
       "986730                  0.000                         0.0   \n",
       "1007062                 0.000                         0.0   \n",
       "1039133                 0.000                         0.0   \n",
       "1210264                 0.000                         0.0   \n",
       "1448236                 0.000                         0.0   \n",
       "1548947               900.000                         0.0   \n",
       "1614906                 0.000                         0.0   \n",
       "1663555                 0.000                         0.0   \n",
       "1679171                 0.000                         0.0   \n",
       "1764744                 0.000                         0.0   \n",
       "1811769             67500.000                         0.0   \n",
       "1821907              5229.000                         0.0   \n",
       "1845734                 0.000                         0.0   \n",
       "2021610                 0.000                         0.0   \n",
       "2029171                 0.000                         0.0   \n",
       "2070436                 0.000                         0.0   \n",
       "2239489                 0.000                         0.0   \n",
       "2257989                 0.000                         0.0   \n",
       "2286944                 0.000                         0.0   \n",
       "2339745                 0.000                         0.0   \n",
       "2377322                 0.000                         0.0   \n",
       "2533249                 0.000                         0.0   \n",
       "2584127                 0.000                         0.0   \n",
       "2648695                 0.000                         0.0   \n",
       "2653597                 0.000                         0.0   \n",
       "2736406                 0.000                         0.0   \n",
       "2853861                 0.000                         0.0   \n",
       "2873870                 0.000                         0.0   \n",
       "3064061                 0.000                         0.0   \n",
       "3123527                 0.000                         0.0   \n",
       "3205656                 0.000                         0.0   \n",
       "3284876                 0.000                         0.0   \n",
       "3486594                 0.000                         0.0   \n",
       "3568086                 0.000                         0.0   \n",
       "3657049                 0.000                         0.0   \n",
       "3748710             26640.000                         0.0   \n",
       "\n",
       "         AMT_DRAWINGS_POS_CURRENT  AMT_INST_MIN_REGULARITY     ...      \\\n",
       "0                         877.500                 1700.325     ...       \n",
       "122799                      0.000                 2332.035     ...       \n",
       "153474                  22005.000                 2250.000     ...       \n",
       "330701                      0.000                 1916.910     ...       \n",
       "419198                      0.000                    0.000     ...       \n",
       "561468                      0.000                 2761.740     ...       \n",
       "577279                      0.000                 2250.000     ...       \n",
       "596084                      0.000                  281.475     ...       \n",
       "722949                      0.000                    0.000     ...       \n",
       "889286                      0.000                    0.000     ...       \n",
       "917059                   1790.145                  129.240     ...       \n",
       "986730                      0.000                 2250.000     ...       \n",
       "1007062                     0.000                 2292.345     ...       \n",
       "1039133                     0.000                 3291.390     ...       \n",
       "1210264                     0.000                 2250.000     ...       \n",
       "1448236                     0.000                 2667.825     ...       \n",
       "1548947                   900.000                 2250.000     ...       \n",
       "1614906                     0.000                 2250.000     ...       \n",
       "1663555                     0.000                 2420.010     ...       \n",
       "1679171                     0.000                 2250.000     ...       \n",
       "1764744                     0.000                 3194.190     ...       \n",
       "1811769                     0.000                      NaN     ...       \n",
       "1821907                  5229.000                  840.780     ...       \n",
       "1845734                     0.000                 2369.295     ...       \n",
       "2021610                     0.000                 3242.250     ...       \n",
       "2029171                     0.000                    0.000     ...       \n",
       "2070436                     0.000                 2250.000     ...       \n",
       "2239489                     0.000                 3141.450     ...       \n",
       "2257989                     0.000                 2250.000     ...       \n",
       "2286944                     0.000                 2629.620     ...       \n",
       "2339745                     0.000                 2250.000     ...       \n",
       "2377322                     0.000                 2501.010     ...       \n",
       "2533249                     0.000                 2250.000     ...       \n",
       "2584127                     0.000                 3498.975     ...       \n",
       "2648695                     0.000                 3089.925     ...       \n",
       "2653597                     0.000                 3346.695     ...       \n",
       "2736406                     0.000                 2456.640     ...       \n",
       "2853861                     0.000                 2250.000     ...       \n",
       "2873870                     0.000                    0.000     ...       \n",
       "3064061                     0.000                 2250.000     ...       \n",
       "3123527                     0.000                    0.000     ...       \n",
       "3205656                     0.000                 2263.050     ...       \n",
       "3284876                     0.000                 3399.795     ...       \n",
       "3486594                     0.000                 2713.005     ...       \n",
       "3568086                     0.000                    0.000     ...       \n",
       "3657049                     0.000                 2355.615     ...       \n",
       "3748710                 26640.000                 2250.000     ...       \n",
       "\n",
       "         AMT_RECIVABLE  AMT_TOTAL_RECEIVABLE  CNT_DRAWINGS_ATM_CURRENT  \\\n",
       "0                0.000                 0.000                       0.0   \n",
       "122799       45670.275             45670.275                       0.0   \n",
       "153474           0.000                 0.000                       0.0   \n",
       "330701          28.575                28.575                       0.0   \n",
       "419198           0.000                 0.000                       0.0   \n",
       "561468       56570.940             56570.940                       0.0   \n",
       "577279       43360.290             43360.290                       0.0   \n",
       "596084           0.000                 0.000                       0.0   \n",
       "722949           0.000                 0.000                       0.0   \n",
       "889286           0.000                 0.000                       0.0   \n",
       "917059        1687.365              1687.365                       0.0   \n",
       "986730       14452.470             14452.470                       0.0   \n",
       "1007062      45089.460             45089.460                       0.0   \n",
       "1039133      61224.345             61224.345                       0.0   \n",
       "1210264      17723.115             17723.115                       0.0   \n",
       "1448236      49693.590             49693.590                       0.0   \n",
       "1548947       4036.860              4036.860                       0.0   \n",
       "1614906      20770.875             20770.875                       0.0   \n",
       "1663555      47206.530             47206.530                       0.0   \n",
       "1679171      16119.675             16119.675                       0.0   \n",
       "1764744      59216.985             59216.985                       0.0   \n",
       "1811769      67841.190             67841.190                       1.0   \n",
       "1821907       5228.280              5228.280                       0.0   \n",
       "1845734      46461.555             46461.555                       0.0   \n",
       "2021610      60267.735             60267.735                       0.0   \n",
       "2029171          0.000                 0.000                       0.0   \n",
       "2070436       1897.875              1897.875                       0.0   \n",
       "2239489      58191.525             58191.525                       0.0   \n",
       "2257989      11112.525             11112.525                       0.0   \n",
       "2286944      45346.185             45346.185                       0.0   \n",
       "2339745       9169.380              9169.380                       0.0   \n",
       "2377322      46247.445             46247.445                       0.0   \n",
       "2533249         35.055                35.055                       0.0   \n",
       "2584127      63246.825             63246.825                       0.0   \n",
       "2648695      51920.145             51920.145                       0.0   \n",
       "2653597      61980.255             61980.255                       0.0   \n",
       "2736406      46949.895             46949.895                       0.0   \n",
       "2853861      19286.325             19286.325                       0.0   \n",
       "2873870          0.000                 0.000                       0.0   \n",
       "3064061      22387.725             22387.725                       0.0   \n",
       "3123527          0.000                 0.000                       0.0   \n",
       "3205656      44304.570             44304.570                       0.0   \n",
       "3284876      63082.080             63082.080                       0.0   \n",
       "3486594      53157.240             53157.240                       0.0   \n",
       "3568086          0.000                 0.000                       0.0   \n",
       "3657049      48224.745             48224.745                       0.0   \n",
       "3748710      13002.075             13002.075                       0.0   \n",
       "\n",
       "         CNT_DRAWINGS_CURRENT  CNT_DRAWINGS_OTHER_CURRENT  \\\n",
       "0                           1                         0.0   \n",
       "122799                      0                         0.0   \n",
       "153474                      1                         0.0   \n",
       "330701                      0                         0.0   \n",
       "419198                      0                         0.0   \n",
       "561468                      0                         0.0   \n",
       "577279                      0                         0.0   \n",
       "596084                      0                         0.0   \n",
       "722949                      0                         0.0   \n",
       "889286                      0                         0.0   \n",
       "917059                      1                         0.0   \n",
       "986730                      0                         0.0   \n",
       "1007062                     0                         0.0   \n",
       "1039133                     0                         0.0   \n",
       "1210264                     0                         0.0   \n",
       "1448236                     0                         0.0   \n",
       "1548947                     1                         0.0   \n",
       "1614906                     0                         0.0   \n",
       "1663555                     0                         0.0   \n",
       "1679171                     0                         0.0   \n",
       "1764744                     0                         0.0   \n",
       "1811769                     1                         0.0   \n",
       "1821907                     2                         0.0   \n",
       "1845734                     0                         0.0   \n",
       "2021610                     0                         0.0   \n",
       "2029171                     0                         0.0   \n",
       "2070436                     0                         0.0   \n",
       "2239489                     0                         0.0   \n",
       "2257989                     0                         0.0   \n",
       "2286944                     0                         0.0   \n",
       "2339745                     0                         0.0   \n",
       "2377322                     0                         0.0   \n",
       "2533249                     0                         0.0   \n",
       "2584127                     0                         0.0   \n",
       "2648695                     0                         0.0   \n",
       "2653597                     0                         0.0   \n",
       "2736406                     0                         0.0   \n",
       "2853861                     0                         0.0   \n",
       "2873870                     0                         0.0   \n",
       "3064061                     0                         0.0   \n",
       "3123527                     0                         0.0   \n",
       "3205656                     0                         0.0   \n",
       "3284876                     0                         0.0   \n",
       "3486594                     0                         0.0   \n",
       "3568086                     0                         0.0   \n",
       "3657049                     0                         0.0   \n",
       "3748710                     1                         0.0   \n",
       "\n",
       "         CNT_DRAWINGS_POS_CURRENT  CNT_INSTALMENT_MATURE_CUM  \\\n",
       "0                             1.0                       35.0   \n",
       "122799                        0.0                       18.0   \n",
       "153474                        1.0                       33.0   \n",
       "330701                        0.0                       39.0   \n",
       "419198                        0.0                       29.0   \n",
       "561468                        0.0                        9.0   \n",
       "577279                        0.0                       21.0   \n",
       "596084                        0.0                       29.0   \n",
       "722949                        0.0                       29.0   \n",
       "889286                        0.0                       29.0   \n",
       "917059                        1.0                       34.0   \n",
       "986730                        0.0                       27.0   \n",
       "1007062                       0.0                       19.0   \n",
       "1039133                       0.0                        4.0   \n",
       "1210264                       0.0                       25.0   \n",
       "1448236                       0.0                       11.0   \n",
       "1548947                       1.0                       37.0   \n",
       "1614906                       0.0                       23.0   \n",
       "1663555                       0.0                       16.0   \n",
       "1679171                       0.0                       26.0   \n",
       "1764744                       0.0                        6.0   \n",
       "1811769                       0.0                        NaN   \n",
       "1821907                       2.0                       36.0   \n",
       "1845734                       0.0                       17.0   \n",
       "2021610                       0.0                        5.0   \n",
       "2029171                       0.0                       29.0   \n",
       "2070436                       0.0                       38.0   \n",
       "2239489                       0.0                        7.0   \n",
       "2257989                       0.0                       31.0   \n",
       "2286944                       0.0                       12.0   \n",
       "2339745                       0.0                       32.0   \n",
       "2377322                       0.0                       13.0   \n",
       "2533249                       0.0                       28.0   \n",
       "2584127                       0.0                        1.0   \n",
       "2648695                       0.0                        8.0   \n",
       "2653597                       0.0                        3.0   \n",
       "2736406                       0.0                       14.0   \n",
       "2853861                       0.0                       24.0   \n",
       "2873870                       0.0                       29.0   \n",
       "3064061                       0.0                       22.0   \n",
       "3123527                       0.0                       29.0   \n",
       "3205656                       0.0                       20.0   \n",
       "3284876                       0.0                        2.0   \n",
       "3486594                       0.0                       10.0   \n",
       "3568086                       0.0                       29.0   \n",
       "3657049                       0.0                       15.0   \n",
       "3748710                       1.0                       30.0   \n",
       "\n",
       "         NAME_CONTRACT_STATUS  SK_DPD  SK_DPD_DEF  \n",
       "0                      Active       0           0  \n",
       "122799                 Active       0           0  \n",
       "153474                 Active       0           0  \n",
       "330701                 Active       0           0  \n",
       "419198                 Active       0           0  \n",
       "561468                 Active       6           6  \n",
       "577279                 Active       0           0  \n",
       "596084                 Active       0           0  \n",
       "722949                 Active       0           0  \n",
       "889286                 Active       0           0  \n",
       "917059                 Active       0           0  \n",
       "986730                 Active       0           0  \n",
       "1007062                Active       0           0  \n",
       "1039133                Active       0           0  \n",
       "1210264                Active       0           0  \n",
       "1448236                Active       0           0  \n",
       "1548947                Active       0           0  \n",
       "1614906                Active       0           0  \n",
       "1663555                Active       0           0  \n",
       "1679171                Active       0           0  \n",
       "1764744                Active       0           0  \n",
       "1811769                Active       0           0  \n",
       "1821907                Active       0           0  \n",
       "1845734                Active       0           0  \n",
       "2021610                Active       0           0  \n",
       "2029171                Active       0           0  \n",
       "2070436                Active       0           0  \n",
       "2239489                Active       0           0  \n",
       "2257989                Active       0           0  \n",
       "2286944                Active       0           0  \n",
       "2339745                Active       0           0  \n",
       "2377322                Active       0           0  \n",
       "2533249                Active       0           0  \n",
       "2584127                Active       0           0  \n",
       "2648695                Active       0           0  \n",
       "2653597                Active       0           0  \n",
       "2736406                Active       0           0  \n",
       "2853861                Active       0           0  \n",
       "2873870                Active       0           0  \n",
       "3064061                Active       0           0  \n",
       "3123527                Active       0           0  \n",
       "3205656                Active       0           0  \n",
       "3284876                Active       0           0  \n",
       "3486594                Active       0           0  \n",
       "3568086                Active       0           0  \n",
       "3657049                Active       0           0  \n",
       "3748710                Active       0           0  \n",
       "\n",
       "[47 rows x 23 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "credit_card_balance[credit_card_balance['SK_ID_CURR'] == 378907]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### Installments Payments"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Repayment history for the previously disbursed credits in Home Credit related to the loans in our sample.\n",
    "\n",
    "There is a) one row for every payment that was made plus b) one row each for missed payment.\n",
    "One row is equivalent to one payment of one installment OR one installment corresponding to one payment of one previous Home Credit credit related to loans in our sample."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "installments_payments_col_desc = column_desc[column_desc['Table'] == 'installments_payments.csv']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* SK_ID_PREV: ID of previous credit in Home credit related to loan in our sample. (One loan in our sample can have 0,1,2 or more previous loans in Home Credit)\n",
      "* SK_ID_CURR: ID of loan in our sample\n",
      "* NUM_INSTALMENT_VERSION: Version of installment calendar (0 is for credit card) of previous credit. Change of installment version from month to month signifies that some parameter of payment calendar has changed\n",
      "* NUM_INSTALMENT_NUMBER: On which installment we observe payment\n",
      "* DAYS_INSTALMENT: When the installment of previous credit was supposed to be paid (relative to application date of current loan)\n",
      "* DAYS_ENTRY_PAYMENT: When was the installments of previous credit paid actually (relative to application date of current loan)\n",
      "* AMT_INSTALMENT: What was the prescribed installment amount of previous credit on this installment\n",
      "* AMT_PAYMENT: What the client actually paid on previous credit on this installment\n"
     ]
    }
   ],
   "source": [
    "for idx, row in installments_payments_col_desc.iterrows():\n",
    "    print(f\"* {row['Row'].strip()}: {row['Description']}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_PREV</th>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th>NUM_INSTALMENT_VERSION</th>\n",
       "      <th>NUM_INSTALMENT_NUMBER</th>\n",
       "      <th>DAYS_INSTALMENT</th>\n",
       "      <th>DAYS_ENTRY_PAYMENT</th>\n",
       "      <th>AMT_INSTALMENT</th>\n",
       "      <th>AMT_PAYMENT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1006147</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10</td>\n",
       "      <td>-1854.0</td>\n",
       "      <td>-1860.0</td>\n",
       "      <td>3647.700</td>\n",
       "      <td>3647.700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2407765</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9</td>\n",
       "      <td>-1884.0</td>\n",
       "      <td>-1888.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>581223</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8</td>\n",
       "      <td>-1914.0</td>\n",
       "      <td>-1917.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>489749</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7</td>\n",
       "      <td>-1944.0</td>\n",
       "      <td>-1950.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>819151</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>-1974.0</td>\n",
       "      <td>-1980.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1745693</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5</td>\n",
       "      <td>-2004.0</td>\n",
       "      <td>-2007.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3460941</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>-2034.0</td>\n",
       "      <td>-2037.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3704788</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>-2064.0</td>\n",
       "      <td>-2070.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2456163</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>-2094.0</td>\n",
       "      <td>-2100.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1347951</th>\n",
       "      <td>2841573</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-2124.0</td>\n",
       "      <td>-2129.0</td>\n",
       "      <td>3703.275</td>\n",
       "      <td>3703.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2710642</th>\n",
       "      <td>2524148</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8</td>\n",
       "      <td>-2284.0</td>\n",
       "      <td>-2284.0</td>\n",
       "      <td>7511.130</td>\n",
       "      <td>7511.130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>585836</th>\n",
       "      <td>2524148</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7</td>\n",
       "      <td>-2314.0</td>\n",
       "      <td>-2314.0</td>\n",
       "      <td>7524.540</td>\n",
       "      <td>7524.540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>912742</th>\n",
       "      <td>2524148</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>-2344.0</td>\n",
       "      <td>-2344.0</td>\n",
       "      <td>7524.540</td>\n",
       "      <td>7524.540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2805152</th>\n",
       "      <td>2524148</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5</td>\n",
       "      <td>-2374.0</td>\n",
       "      <td>-2374.0</td>\n",
       "      <td>7524.540</td>\n",
       "      <td>7524.540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>438555</th>\n",
       "      <td>2524148</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>-2404.0</td>\n",
       "      <td>-2404.0</td>\n",
       "      <td>7524.540</td>\n",
       "      <td>7524.540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1894799</th>\n",
       "      <td>2524148</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>-2434.0</td>\n",
       "      <td>-2434.0</td>\n",
       "      <td>7524.540</td>\n",
       "      <td>7524.540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1364128</th>\n",
       "      <td>2524148</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>-2464.0</td>\n",
       "      <td>-2464.0</td>\n",
       "      <td>7524.540</td>\n",
       "      <td>7524.540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2663607</th>\n",
       "      <td>2524148</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-2494.0</td>\n",
       "      <td>-2494.0</td>\n",
       "      <td>7524.540</td>\n",
       "      <td>7524.540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2673514</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>36</td>\n",
       "      <td>-96.0</td>\n",
       "      <td>-117.0</td>\n",
       "      <td>10855.260</td>\n",
       "      <td>10855.260</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>778085</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>35</td>\n",
       "      <td>-126.0</td>\n",
       "      <td>-144.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1795608</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>34</td>\n",
       "      <td>-156.0</td>\n",
       "      <td>-172.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3739160</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>33</td>\n",
       "      <td>-186.0</td>\n",
       "      <td>-215.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3433028</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>32</td>\n",
       "      <td>-216.0</td>\n",
       "      <td>-232.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1877113</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>31</td>\n",
       "      <td>-246.0</td>\n",
       "      <td>-263.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2660327</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>30</td>\n",
       "      <td>-276.0</td>\n",
       "      <td>-299.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2479237</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>29</td>\n",
       "      <td>-306.0</td>\n",
       "      <td>-332.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2673446</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>28</td>\n",
       "      <td>-336.0</td>\n",
       "      <td>-355.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2644646</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>27</td>\n",
       "      <td>-366.0</td>\n",
       "      <td>-384.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3724901</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>26</td>\n",
       "      <td>-396.0</td>\n",
       "      <td>-396.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1092004</th>\n",
       "      <td>2452854</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>25</td>\n",
       "      <td>-426.0</td>\n",
       "      <td>-427.0</td>\n",
       "      <td>11302.605</td>\n",
       "      <td>11302.605</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1747613</th>\n",
       "      <td>1644291</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>-380.0</td>\n",
       "      <td>-384.0</td>\n",
       "      <td>22048.560</td>\n",
       "      <td>22048.560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>898625</th>\n",
       "      <td>1644291</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>-410.0</td>\n",
       "      <td>-414.0</td>\n",
       "      <td>22048.560</td>\n",
       "      <td>22048.560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3531700</th>\n",
       "      <td>1644291</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-440.0</td>\n",
       "      <td>-442.0</td>\n",
       "      <td>22048.560</td>\n",
       "      <td>22048.560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3589526</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>2.0</td>\n",
       "      <td>11</td>\n",
       "      <td>-1431.0</td>\n",
       "      <td>-1454.0</td>\n",
       "      <td>72377.280</td>\n",
       "      <td>72377.280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2835586</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10</td>\n",
       "      <td>-1461.0</td>\n",
       "      <td>-1467.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>667782</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9</td>\n",
       "      <td>-1491.0</td>\n",
       "      <td>-1496.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2536220</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8</td>\n",
       "      <td>-1521.0</td>\n",
       "      <td>-1525.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1168850</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7</td>\n",
       "      <td>-1551.0</td>\n",
       "      <td>-1558.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>708824</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>-1581.0</td>\n",
       "      <td>-1587.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1614970</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5</td>\n",
       "      <td>-1611.0</td>\n",
       "      <td>-1616.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3778071</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>-1641.0</td>\n",
       "      <td>-1645.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>951171</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>-1671.0</td>\n",
       "      <td>-1671.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2955110</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>-1701.0</td>\n",
       "      <td>-1707.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2738477</th>\n",
       "      <td>1588995</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-1731.0</td>\n",
       "      <td>-1736.0</td>\n",
       "      <td>13305.060</td>\n",
       "      <td>13305.060</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>831532</th>\n",
       "      <td>1294823</td>\n",
       "      <td>161674</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7</td>\n",
       "      <td>-2869.0</td>\n",
       "      <td>-2869.0</td>\n",
       "      <td>8591.130</td>\n",
       "      <td>8591.130</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2315826</th>\n",
       "      <td>1294823</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>-2899.0</td>\n",
       "      <td>-2926.0</td>\n",
       "      <td>4340.745</td>\n",
       "      <td>4340.745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>614294</th>\n",
       "      <td>1065354</td>\n",
       "      <td>161674</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2</td>\n",
       "      <td>-272.0</td>\n",
       "      <td>-299.0</td>\n",
       "      <td>117414.000</td>\n",
       "      <td>117414.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2671816</th>\n",
       "      <td>1065354</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-302.0</td>\n",
       "      <td>-315.0</td>\n",
       "      <td>12595.185</td>\n",
       "      <td>12595.185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>216905</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>12</td>\n",
       "      <td>-1000.0</td>\n",
       "      <td>-1033.0</td>\n",
       "      <td>6935.220</td>\n",
       "      <td>6935.220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1402152</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11</td>\n",
       "      <td>-1030.0</td>\n",
       "      <td>-1063.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1038640</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10</td>\n",
       "      <td>-1060.0</td>\n",
       "      <td>-1125.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1608829</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9</td>\n",
       "      <td>-1090.0</td>\n",
       "      <td>-1154.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138614</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8</td>\n",
       "      <td>-1120.0</td>\n",
       "      <td>-1175.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1128187</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7</td>\n",
       "      <td>-1150.0</td>\n",
       "      <td>-1175.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6</td>\n",
       "      <td>-1180.0</td>\n",
       "      <td>-1187.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1878907</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5</td>\n",
       "      <td>-1210.0</td>\n",
       "      <td>-1217.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3612113</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4</td>\n",
       "      <td>-1240.0</td>\n",
       "      <td>-1247.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>425089</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "      <td>-1270.0</td>\n",
       "      <td>-1275.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47438</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "      <td>-1300.0</td>\n",
       "      <td>-1307.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>577850</th>\n",
       "      <td>1054186</td>\n",
       "      <td>161674</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>-1330.0</td>\n",
       "      <td>-1338.0</td>\n",
       "      <td>6948.360</td>\n",
       "      <td>6948.360</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>101 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         SK_ID_PREV  SK_ID_CURR  NUM_INSTALMENT_VERSION  \\\n",
       "1006147     2841573      161674                     1.0   \n",
       "2407765     2841573      161674                     1.0   \n",
       "581223      2841573      161674                     1.0   \n",
       "489749      2841573      161674                     1.0   \n",
       "819151      2841573      161674                     1.0   \n",
       "1745693     2841573      161674                     1.0   \n",
       "3460941     2841573      161674                     1.0   \n",
       "3704788     2841573      161674                     1.0   \n",
       "2456163     2841573      161674                     1.0   \n",
       "1347951     2841573      161674                     1.0   \n",
       "2710642     2524148      161674                     1.0   \n",
       "585836      2524148      161674                     1.0   \n",
       "912742      2524148      161674                     1.0   \n",
       "2805152     2524148      161674                     1.0   \n",
       "438555      2524148      161674                     1.0   \n",
       "1894799     2524148      161674                     1.0   \n",
       "1364128     2524148      161674                     1.0   \n",
       "2663607     2524148      161674                     1.0   \n",
       "2673514     2452854      161674                     1.0   \n",
       "778085      2452854      161674                     1.0   \n",
       "1795608     2452854      161674                     1.0   \n",
       "3739160     2452854      161674                     1.0   \n",
       "3433028     2452854      161674                     1.0   \n",
       "1877113     2452854      161674                     1.0   \n",
       "2660327     2452854      161674                     1.0   \n",
       "2479237     2452854      161674                     1.0   \n",
       "2673446     2452854      161674                     1.0   \n",
       "2644646     2452854      161674                     1.0   \n",
       "3724901     2452854      161674                     1.0   \n",
       "1092004     2452854      161674                     1.0   \n",
       "...             ...         ...                     ...   \n",
       "1747613     1644291      161674                     1.0   \n",
       "898625      1644291      161674                     1.0   \n",
       "3531700     1644291      161674                     1.0   \n",
       "3589526     1588995      161674                     2.0   \n",
       "2835586     1588995      161674                     1.0   \n",
       "667782      1588995      161674                     1.0   \n",
       "2536220     1588995      161674                     1.0   \n",
       "1168850     1588995      161674                     1.0   \n",
       "708824      1588995      161674                     1.0   \n",
       "1614970     1588995      161674                     1.0   \n",
       "3778071     1588995      161674                     1.0   \n",
       "951171      1588995      161674                     1.0   \n",
       "2955110     1588995      161674                     1.0   \n",
       "2738477     1588995      161674                     1.0   \n",
       "831532      1294823      161674                     2.0   \n",
       "2315826     1294823      161674                     1.0   \n",
       "614294      1065354      161674                     2.0   \n",
       "2671816     1065354      161674                     1.0   \n",
       "216905      1054186      161674                     1.0   \n",
       "1402152     1054186      161674                     1.0   \n",
       "1038640     1054186      161674                     1.0   \n",
       "1608829     1054186      161674                     1.0   \n",
       "138614      1054186      161674                     1.0   \n",
       "1128187     1054186      161674                     1.0   \n",
       "0           1054186      161674                     1.0   \n",
       "1878907     1054186      161674                     1.0   \n",
       "3612113     1054186      161674                     1.0   \n",
       "425089      1054186      161674                     1.0   \n",
       "47438       1054186      161674                     1.0   \n",
       "577850      1054186      161674                     1.0   \n",
       "\n",
       "         NUM_INSTALMENT_NUMBER  DAYS_INSTALMENT  DAYS_ENTRY_PAYMENT  \\\n",
       "1006147                     10          -1854.0             -1860.0   \n",
       "2407765                      9          -1884.0             -1888.0   \n",
       "581223                       8          -1914.0             -1917.0   \n",
       "489749                       7          -1944.0             -1950.0   \n",
       "819151                       6          -1974.0             -1980.0   \n",
       "1745693                      5          -2004.0             -2007.0   \n",
       "3460941                      4          -2034.0             -2037.0   \n",
       "3704788                      3          -2064.0             -2070.0   \n",
       "2456163                      2          -2094.0             -2100.0   \n",
       "1347951                      1          -2124.0             -2129.0   \n",
       "2710642                      8          -2284.0             -2284.0   \n",
       "585836                       7          -2314.0             -2314.0   \n",
       "912742                       6          -2344.0             -2344.0   \n",
       "2805152                      5          -2374.0             -2374.0   \n",
       "438555                       4          -2404.0             -2404.0   \n",
       "1894799                      3          -2434.0             -2434.0   \n",
       "1364128                      2          -2464.0             -2464.0   \n",
       "2663607                      1          -2494.0             -2494.0   \n",
       "2673514                     36            -96.0              -117.0   \n",
       "778085                      35           -126.0              -144.0   \n",
       "1795608                     34           -156.0              -172.0   \n",
       "3739160                     33           -186.0              -215.0   \n",
       "3433028                     32           -216.0              -232.0   \n",
       "1877113                     31           -246.0              -263.0   \n",
       "2660327                     30           -276.0              -299.0   \n",
       "2479237                     29           -306.0              -332.0   \n",
       "2673446                     28           -336.0              -355.0   \n",
       "2644646                     27           -366.0              -384.0   \n",
       "3724901                     26           -396.0              -396.0   \n",
       "1092004                     25           -426.0              -427.0   \n",
       "...                        ...              ...                 ...   \n",
       "1747613                      3           -380.0              -384.0   \n",
       "898625                       2           -410.0              -414.0   \n",
       "3531700                      1           -440.0              -442.0   \n",
       "3589526                     11          -1431.0             -1454.0   \n",
       "2835586                     10          -1461.0             -1467.0   \n",
       "667782                       9          -1491.0             -1496.0   \n",
       "2536220                      8          -1521.0             -1525.0   \n",
       "1168850                      7          -1551.0             -1558.0   \n",
       "708824                       6          -1581.0             -1587.0   \n",
       "1614970                      5          -1611.0             -1616.0   \n",
       "3778071                      4          -1641.0             -1645.0   \n",
       "951171                       3          -1671.0             -1671.0   \n",
       "2955110                      2          -1701.0             -1707.0   \n",
       "2738477                      1          -1731.0             -1736.0   \n",
       "831532                       7          -2869.0             -2869.0   \n",
       "2315826                      6          -2899.0             -2926.0   \n",
       "614294                       2           -272.0              -299.0   \n",
       "2671816                      1           -302.0              -315.0   \n",
       "216905                      12          -1000.0             -1033.0   \n",
       "1402152                     11          -1030.0             -1063.0   \n",
       "1038640                     10          -1060.0             -1125.0   \n",
       "1608829                      9          -1090.0             -1154.0   \n",
       "138614                       8          -1120.0             -1175.0   \n",
       "1128187                      7          -1150.0             -1175.0   \n",
       "0                            6          -1180.0             -1187.0   \n",
       "1878907                      5          -1210.0             -1217.0   \n",
       "3612113                      4          -1240.0             -1247.0   \n",
       "425089                       3          -1270.0             -1275.0   \n",
       "47438                        2          -1300.0             -1307.0   \n",
       "577850                       1          -1330.0             -1338.0   \n",
       "\n",
       "         AMT_INSTALMENT  AMT_PAYMENT  \n",
       "1006147        3647.700     3647.700  \n",
       "2407765        3703.275     3703.275  \n",
       "581223         3703.275     3703.275  \n",
       "489749         3703.275     3703.275  \n",
       "819151         3703.275     3703.275  \n",
       "1745693        3703.275     3703.275  \n",
       "3460941        3703.275     3703.275  \n",
       "3704788        3703.275     3703.275  \n",
       "2456163        3703.275     3703.275  \n",
       "1347951        3703.275     3703.275  \n",
       "2710642        7511.130     7511.130  \n",
       "585836         7524.540     7524.540  \n",
       "912742         7524.540     7524.540  \n",
       "2805152        7524.540     7524.540  \n",
       "438555         7524.540     7524.540  \n",
       "1894799        7524.540     7524.540  \n",
       "1364128        7524.540     7524.540  \n",
       "2663607        7524.540     7524.540  \n",
       "2673514       10855.260    10855.260  \n",
       "778085        11302.605    11302.605  \n",
       "1795608       11302.605    11302.605  \n",
       "3739160       11302.605    11302.605  \n",
       "3433028       11302.605    11302.605  \n",
       "1877113       11302.605    11302.605  \n",
       "2660327       11302.605    11302.605  \n",
       "2479237       11302.605    11302.605  \n",
       "2673446       11302.605    11302.605  \n",
       "2644646       11302.605    11302.605  \n",
       "3724901       11302.605    11302.605  \n",
       "1092004       11302.605    11302.605  \n",
       "...                 ...          ...  \n",
       "1747613       22048.560    22048.560  \n",
       "898625        22048.560    22048.560  \n",
       "3531700       22048.560    22048.560  \n",
       "3589526       72377.280    72377.280  \n",
       "2835586       13305.060    13305.060  \n",
       "667782        13305.060    13305.060  \n",
       "2536220       13305.060    13305.060  \n",
       "1168850       13305.060    13305.060  \n",
       "708824        13305.060    13305.060  \n",
       "1614970       13305.060    13305.060  \n",
       "3778071       13305.060    13305.060  \n",
       "951171        13305.060    13305.060  \n",
       "2955110       13305.060    13305.060  \n",
       "2738477       13305.060    13305.060  \n",
       "831532         8591.130     8591.130  \n",
       "2315826        4340.745     4340.745  \n",
       "614294       117414.000   117414.000  \n",
       "2671816       12595.185    12595.185  \n",
       "216905         6935.220     6935.220  \n",
       "1402152        6948.360     6948.360  \n",
       "1038640        6948.360     6948.360  \n",
       "1608829        6948.360     6948.360  \n",
       "138614         6948.360     6948.360  \n",
       "1128187        6948.360     6948.360  \n",
       "0              6948.360     6948.360  \n",
       "1878907        6948.360     6948.360  \n",
       "3612113        6948.360     6948.360  \n",
       "425089         6948.360     6948.360  \n",
       "47438          6948.360     6948.360  \n",
       "577850         6948.360     6948.360  \n",
       "\n",
       "[101 rows x 8 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "installments_payments[installments_payments['SK_ID_CURR'] == 161674].sort_values(by=['SK_ID_PREV', 'DAYS_INSTALMENT'], ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### Previous Application"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "All previous applications for Home Credit loans of clients who have loans in our sample.\n",
    "\n",
    "There is one row for each previous application related to loans in our data sample."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "previous_application_col_desc = column_desc[column_desc['Table'] == 'previous_application.csv']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* SK_ID_PREV: ID of previous credit in Home credit related to loan in our sample. (One loan in our sample can have 0,1,2 or more previous loan applications in Home Credit, previous application could, but not necessarily have to lead to credit) \n",
      "* SK_ID_CURR: ID of loan in our sample\n",
      "* NAME_CONTRACT_TYPE: Contract product type (Cash loan, consumer loan [POS] ,...) of the previous application\n",
      "* AMT_ANNUITY: Annuity of previous application\n",
      "* AMT_APPLICATION: For how much credit did client ask on the previous application\n",
      "* AMT_CREDIT: Final credit amount on the previous application. This differs from AMT_APPLICATION in a way that the AMT_APPLICATION is the amount for which the client initially applied for, but during our approval process he could have received different amount - AMT_CREDIT\n",
      "* AMT_DOWN_PAYMENT: Down payment on the previous application\n",
      "* AMT_GOODS_PRICE: Goods price of good that client asked for (if applicable) on the previous application\n",
      "* WEEKDAY_APPR_PROCESS_START: On which day of the week did the client apply for previous application\n",
      "* HOUR_APPR_PROCESS_START: Approximately at what day hour did the client apply for the previous application\n",
      "* FLAG_LAST_APPL_PER_CONTRACT: Flag if it was last application for the previous contract. Sometimes by mistake of client or our clerk there could be more applications for one single contract\n",
      "* NFLAG_LAST_APPL_IN_DAY: Flag if the application was the last application per day of the client. Sometimes clients apply for more applications a day. Rarely it could also be error in our system that one application is in the database twice\n",
      "* NFLAG_MICRO_CASH: Flag Micro finance loan\n",
      "* RATE_DOWN_PAYMENT: Down payment rate normalized on previous credit\n",
      "* RATE_INTEREST_PRIMARY: Interest rate normalized on previous credit\n",
      "* RATE_INTEREST_PRIVILEGED: Interest rate normalized on previous credit\n",
      "* NAME_CASH_LOAN_PURPOSE: Purpose of the cash loan\n",
      "* NAME_CONTRACT_STATUS: Contract status (approved, cancelled, ...) of previous application\n",
      "* DAYS_DECISION: Relative to current application when was the decision about previous application made\n",
      "* NAME_PAYMENT_TYPE: Payment method that client chose to pay for the previous application\n",
      "* CODE_REJECT_REASON: Why was the previous application rejected\n",
      "* NAME_TYPE_SUITE: Who accompanied client when applying for the previous application\n",
      "* NAME_CLIENT_TYPE: Was the client old or new client when applying for the previous application\n",
      "* NAME_GOODS_CATEGORY: What kind of goods did the client apply for in the previous application\n",
      "* NAME_PORTFOLIO: Was the previous application for CASH, POS, CAR, ",
      "\n",
      "* NAME_PRODUCT_TYPE: Was the previous application x-sell o walk-in\n",
      "* CHANNEL_TYPE: Through which channel we acquired the client on the previous application\n",
      "* SELLERPLACE_AREA: Selling area of seller place of the previous application\n",
      "* NAME_SELLER_INDUSTRY: The industry of the seller\n",
      "* CNT_PAYMENT: Term of previous credit at application of the previous application\n",
      "* NAME_YIELD_GROUP: Grouped interest rate into small medium and high of the previous application\n",
      "* PRODUCT_COMBINATION: Detailed product combination of the previous application\n",
      "* DAYS_FIRST_DRAWING: Relative to application date of current application when was the first disbursement of the previous application\n",
      "* DAYS_FIRST_DUE: Relative to application date of current application when was the first due supposed to be of the previous application\n",
      "* DAYS_LAST_DUE_1ST_VERSION: Relative to application date of current application when was the first due of the previous application\n",
      "* DAYS_LAST_DUE: Relative to application date of current application when was the last due date of the previous application\n",
      "* DAYS_TERMINATION: Relative to application date of current application when was the expected termination of the previous application\n",
      "* NFLAG_INSURED_ON_APPROVAL: Did the client requested insurance during the previous application\n"
     ]
    }
   ],
   "source": [
    "for idx, row in previous_application_col_desc.iterrows():\n",
    "    print(f\"* {row['Row'].strip()}: {row['Description']}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SK_ID_PREV</th>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th>NAME_CONTRACT_TYPE</th>\n",
       "      <th>AMT_ANNUITY</th>\n",
       "      <th>AMT_APPLICATION</th>\n",
       "      <th>AMT_CREDIT</th>\n",
       "      <th>AMT_DOWN_PAYMENT</th>\n",
       "      <th>AMT_GOODS_PRICE</th>\n",
       "      <th>WEEKDAY_APPR_PROCESS_START</th>\n",
       "      <th>HOUR_APPR_PROCESS_START</th>\n",
       "      <th>...</th>\n",
       "      <th>NAME_SELLER_INDUSTRY</th>\n",
       "      <th>CNT_PAYMENT</th>\n",
       "      <th>NAME_YIELD_GROUP</th>\n",
       "      <th>PRODUCT_COMBINATION</th>\n",
       "      <th>DAYS_FIRST_DRAWING</th>\n",
       "      <th>DAYS_FIRST_DUE</th>\n",
       "      <th>DAYS_LAST_DUE_1ST_VERSION</th>\n",
       "      <th>DAYS_LAST_DUE</th>\n",
       "      <th>DAYS_TERMINATION</th>\n",
       "      <th>NFLAG_INSURED_ON_APPROVAL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2030495</td>\n",
       "      <td>271877</td>\n",
       "      <td>Consumer loans</td>\n",
       "      <td>1730.430</td>\n",
       "      <td>17145.0</td>\n",
       "      <td>17145.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17145.0</td>\n",
       "      <td>SATURDAY</td>\n",
       "      <td>15</td>\n",
       "      <td>...</td>\n",
       "      <td>Connectivity</td>\n",
       "      <td>12.0</td>\n",
       "      <td>middle</td>\n",
       "      <td>POS mobile with interest</td>\n",
       "      <td>365243.0</td>\n",
       "      <td>-42.0</td>\n",
       "      <td>300.0</td>\n",
       "      <td>-42.0</td>\n",
       "      <td>-37.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>735737</th>\n",
       "      <td>1696966</td>\n",
       "      <td>271877</td>\n",
       "      <td>Consumer loans</td>\n",
       "      <td>68258.655</td>\n",
       "      <td>1800000.0</td>\n",
       "      <td>1754721.0</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>1800000.0</td>\n",
       "      <td>SATURDAY</td>\n",
       "      <td>18</td>\n",
       "      <td>...</td>\n",
       "      <td>Furniture</td>\n",
       "      <td>36.0</td>\n",
       "      <td>low_normal</td>\n",
       "      <td>POS industry with interest</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1567518</th>\n",
       "      <td>2154916</td>\n",
       "      <td>271877</td>\n",
       "      <td>Consumer loans</td>\n",
       "      <td>12417.390</td>\n",
       "      <td>108400.5</td>\n",
       "      <td>119848.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>108400.5</td>\n",
       "      <td>SUNDAY</td>\n",
       "      <td>14</td>\n",
       "      <td>...</td>\n",
       "      <td>Furniture</td>\n",
       "      <td>12.0</td>\n",
       "      <td>middle</td>\n",
       "      <td>POS industry with interest</td>\n",
       "      <td>365243.0</td>\n",
       "      <td>-512.0</td>\n",
       "      <td>-182.0</td>\n",
       "      <td>-392.0</td>\n",
       "      <td>-387.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 37 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         SK_ID_PREV  SK_ID_CURR NAME_CONTRACT_TYPE  AMT_ANNUITY  \\\n",
       "0           2030495      271877     Consumer loans     1730.430   \n",
       "735737      1696966      271877     Consumer loans    68258.655   \n",
       "1567518     2154916      271877     Consumer loans    12417.390   \n",
       "\n",
       "         AMT_APPLICATION  AMT_CREDIT  AMT_DOWN_PAYMENT  AMT_GOODS_PRICE  \\\n",
       "0                17145.0     17145.0               0.0          17145.0   \n",
       "735737         1800000.0   1754721.0          180000.0        1800000.0   \n",
       "1567518         108400.5    119848.5               0.0         108400.5   \n",
       "\n",
       "        WEEKDAY_APPR_PROCESS_START  HOUR_APPR_PROCESS_START  \\\n",
       "0                         SATURDAY                       15   \n",
       "735737                    SATURDAY                       18   \n",
       "1567518                     SUNDAY                       14   \n",
       "\n",
       "                   ...            NAME_SELLER_INDUSTRY  CNT_PAYMENT  \\\n",
       "0                  ...                    Connectivity         12.0   \n",
       "735737             ...                       Furniture         36.0   \n",
       "1567518            ...                       Furniture         12.0   \n",
       "\n",
       "         NAME_YIELD_GROUP         PRODUCT_COMBINATION  DAYS_FIRST_DRAWING  \\\n",
       "0                  middle    POS mobile with interest            365243.0   \n",
       "735737         low_normal  POS industry with interest                 NaN   \n",
       "1567518            middle  POS industry with interest            365243.0   \n",
       "\n",
       "        DAYS_FIRST_DUE DAYS_LAST_DUE_1ST_VERSION  DAYS_LAST_DUE  \\\n",
       "0                -42.0                     300.0          -42.0   \n",
       "735737             NaN                       NaN            NaN   \n",
       "1567518         -512.0                    -182.0         -392.0   \n",
       "\n",
       "        DAYS_TERMINATION NFLAG_INSURED_ON_APPROVAL  \n",
       "0                  -37.0                       0.0  \n",
       "735737               NaN                       NaN  \n",
       "1567518           -387.0                       0.0  \n",
       "\n",
       "[3 rows x 37 columns]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "previous_application[previous_application['SK_ID_CURR'] == 271877]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Prepare dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "### Aggregate and join extra data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Below function performs aggregation and joining of additional datasets. A lot of the feature engineering ideas were taken from https://www.kaggle.com/shep312/lightgbm-with-weighted-averages-dropout-771/code"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### Previous applications"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Count the number of previous applications and get the total credit amount for each."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "prev_app_agg = previous_application.groupby('SK_ID_CURR').agg(\n",
    "    {'SK_ID_CURR': 'count', 'AMT_CREDIT': 'sum'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "prev_app_agg.columns = ['PREV APP COUNT', 'TOTAL PREV LOAN AMT']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PREV APP COUNT</th>\n",
       "      <th>TOTAL PREV LOAN AMT</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>100001</th>\n",
       "      <td>1</td>\n",
       "      <td>23787.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100002</th>\n",
       "      <td>1</td>\n",
       "      <td>179055.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100003</th>\n",
       "      <td>3</td>\n",
       "      <td>1452573.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100004</th>\n",
       "      <td>1</td>\n",
       "      <td>20106.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100005</th>\n",
       "      <td>2</td>\n",
       "      <td>40153.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            PREV APP COUNT  TOTAL PREV LOAN AMT\n",
       "SK_ID_CURR                                     \n",
       "100001                   1              23787.0\n",
       "100002                   1             179055.0\n",
       "100003                   3            1452573.0\n",
       "100004                   1              20106.0\n",
       "100005                   2              40153.5"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prev_app_agg.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_df = application_train.merge(\n",
    "    prev_app_agg, left_on='SK_ID_CURR', right_index=True, how='left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "test_df = application_test.merge(\n",
    "    prev_app_agg, left_on='SK_ID_CURR', right_index=True, how='left')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Add the mean of all previous application features to the set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "prev_app_mean = previous_application.groupby('SK_ID_CURR').mean()\n",
    "\n",
    "train_df = train_df.merge(\n",
    "    prev_app_mean, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_PAVG'])\n",
    "test_df = test_df.merge(\n",
    "    prev_app_mean, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_PAVG'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Get the mode of each categorical variable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "categorical_feats = previous_application.columns[\n",
    "    previous_application.dtypes == 'object']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "def get_mode(x):\n",
    "    x = x.fillna('NaN')\n",
    "    return x.mode().iloc[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 34min 30s, sys: 20.5 s, total: 34min 50s\n",
      "Wall time: 34min 45s\n"
     ]
    }
   ],
   "source": [
    "%time prev_apps_cat_mode = previous_application[list(categorical_feats) + ['SK_ID_CURR']].groupby('SK_ID_CURR').agg({k: get_mode for k in categorical_feats})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_df = train_df.merge(\n",
    "    prev_apps_cat_mode, left_on='SK_ID_CURR',\n",
    "    right_index=True, how='left', suffixes=['', '_BMODE'])\n",
    "test_df = test_df.merge(\n",
    "    prev_apps_cat_mode, left_on='SK_ID_CURR',\n",
    "    right_index=True, how='left', suffixes=['', '_BMODE'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### Credit card balance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Add credit card balance weighted by recency."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "def get_weighted_avg(x):\n",
    "    return np.average(x, weights=-1/credit_card_balance.loc[x.index, 'MONTHS_BALANCE'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "credit_card_avgs = credit_card_balance.groupby('SK_ID_CURR').agg(get_weighted_avg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_df = train_df.merge(\n",
    "    credit_card_avgs, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_CCAVG'])\n",
    "test_df = test_df.merge(\n",
    "    credit_card_avgs, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_CCAVG'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Add credit card categorical data (using most recent category)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "most_recent_index = credit_card_balance.groupby('SK_ID_CURR')['MONTHS_BALANCE'].idxmax()\n",
    "cat_feats = credit_card_balance.columns[credit_card_balance.dtypes == 'object'].tolist()  + ['SK_ID_CURR']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_df = train_df.merge(\n",
    "    credit_card_balance.loc[most_recent_index, cat_feats],\n",
    "    left_on='SK_ID_CURR', right_on='SK_ID_CURR',\n",
    "    how='left', suffixes=['', '_CCAVG'])\n",
    "test_df = test_df.merge(\n",
    "    credit_card_balance.loc[most_recent_index, cat_feats],\n",
    "    left_on='SK_ID_CURR', right_on='SK_ID_CURR',\n",
    "    how='left', suffixes=['', '_CCAVG'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### Credit bureau data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "credit_bureau_avgs = bureau.groupby('SK_ID_CURR').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_df = train_df.merge(\n",
    "    credit_bureau_avgs, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_BAVG'])\n",
    "test_df = test_df.merge(\n",
    "    credit_bureau_avgs, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_BAVG'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### POS cash sales"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "def get_weighted_avg(x):\n",
    "    return np.average(x, weights=-1/pos.loc[x.index, 'MONTHS_BALANCE'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "f = {\n",
    "    'CNT_INSTALMENT': get_weighted_avg,\n",
    "    'CNT_INSTALMENT_FUTURE': get_weighted_avg,\n",
    "    'SK_DPD': get_weighted_avg,\n",
    "    'SK_DPD_DEF': get_weighted_avg}\n",
    "\n",
    "cash_avg = pos.groupby('SK_ID_CURR')[\n",
    "    'CNT_INSTALMENT','CNT_INSTALMENT_FUTURE', 'SK_DPD', 'SK_DPD_DEF'].agg(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_df = train_df.merge(\n",
    "    cash_avg, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_CAVG'])\n",
    "test_df = test_df.merge(\n",
    "    cash_avg, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_CAVG'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### POS cash balance - categorical features (most recent)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "most_recent_index = pos.groupby('SK_ID_CURR')['MONTHS_BALANCE'].idxmax()\n",
    "cat_feats = pos.columns[pos.dtypes == 'object'].tolist()  + ['SK_ID_CURR']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_df = train_df.merge(\n",
    "    pos.loc[most_recent_index, cat_feats],\n",
    "    left_on='SK_ID_CURR', right_on='SK_ID_CURR',\n",
    "    how='left', suffixes=['', '_CAVG'])\n",
    "test_df = test_df.merge(\n",
    "    pos.loc[most_recent_index, cat_feats],\n",
    "    left_on='SK_ID_CURR', right_on='SK_ID_CURR',\n",
    "    how='left', suffixes=['', '_CAVG'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "#### Installments payments (average)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "installments_avg = installments_payments.groupby('SK_ID_CURR').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_df = train_df.merge(\n",
    "    installments_avg, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_IAVG'])\n",
    "test_df = test_df.merge(\n",
    "    installments_avg, left_on='SK_ID_CURR', right_index=True,\n",
    "    how='left', suffixes=['', '_IAVG'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save features and load features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df.to_feather(PATH / 'train_df_2018_may_24')\n",
    "test_df.to_feather(PATH / 'test_df_2018_may_24')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_df = pd.read_feather(PATH / 'train_df_2018_may_24')\n",
    "test_df_raw = pd.read_feather(PATH / 'test_df_2018_may_24')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "CAT_COLUMNS = [\n",
    "    'NAME_CONTRACT_TYPE', 'CODE_GENDER', 'CNT_CHILDREN',\n",
    "    'NAME_TYPE_SUITE', 'NAME_INCOME_TYPE', 'NAME_EDUCATION_TYPE',\n",
    "    'NAME_FAMILY_STATUS', 'NAME_HOUSING_TYPE', 'OWN_CAR_AGE', 'OCCUPATION_TYPE',\n",
    "    'REGION_RATING_CLIENT', 'REGION_RATING_CLIENT_W_CITY', 'WEEKDAY_APPR_PROCESS_START',\n",
    "    'HOUR_APPR_PROCESS_START', 'ORGANIZATION_TYPE', 'FONDKAPREMONT_MODE', 'HOUSETYPE_MODE',\n",
    "    'WALLSMATERIAL_MODE', 'EMERGENCYSTATE_MODE', 'NAME_CONTRACT_TYPE_BMODE', 'WEEKDAY_APPR_PROCESS_START_BMODE',\n",
    "    'FLAG_LAST_APPL_PER_CONTRACT', 'NAME_CASH_LOAN_PURPOSE', 'NAME_CONTRACT_STATUS', 'NAME_PAYMENT_TYPE',\n",
    "    'CODE_REJECT_REASON', 'NAME_TYPE_SUITE_BMODE', 'NAME_CLIENT_TYPE', 'NAME_GOODS_CATEGORY', 'NAME_PORTFOLIO',\n",
    "    'NAME_PRODUCT_TYPE', 'CHANNEL_TYPE', 'NAME_SELLER_INDUSTRY', 'NAME_YIELD_GROUP', 'PRODUCT_COMBINATION',\n",
    "    'NAME_CONTRACT_STATUS_CCAVG', 'NAME_CONTRACT_STATUS_CAVG', \n",
    "    \n",
    "]\n",
    "\n",
    "BOOL_COLS = [\n",
    "    'FLAG_OWN_CAR', 'FLAG_OWN_REALTY', 'FLAG_MOBIL', 'FLAG_EMP_PHONE', 'FLAG_WORK_PHONE',\n",
    "    'FLAG_CONT_MOBILE', 'FLAG_PHONE', 'FLAG_EMAIL', 'REG_REGION_NOT_LIVE_REGION', 'REG_REGION_NOT_WORK_REGION',\n",
    "    'LIVE_REGION_NOT_WORK_REGION', 'REG_CITY_NOT_LIVE_CITY', 'REG_CITY_NOT_WORK_CITY', 'LIVE_CITY_NOT_WORK_CITY',\n",
    "    'FLAG_DOCUMENT_2', 'FLAG_DOCUMENT_3', 'FLAG_DOCUMENT_4', 'FLAG_DOCUMENT_5', 'FLAG_DOCUMENT_6',\n",
    "    'FLAG_DOCUMENT_7', 'FLAG_DOCUMENT_8', 'FLAG_DOCUMENT_9', 'FLAG_DOCUMENT_10', 'FLAG_DOCUMENT_11', 'FLAG_DOCUMENT_12',\n",
    "    'FLAG_DOCUMENT_13', 'FLAG_DOCUMENT_14', 'FLAG_DOCUMENT_15', 'FLAG_DOCUMENT_16', 'FLAG_DOCUMENT_17', 'FLAG_DOCUMENT_18',\n",
    "    'FLAG_DOCUMENT_19', 'FLAG_DOCUMENT_20', 'FLAG_DOCUMENT_21',\n",
    "]\n",
    "\n",
    "CONT_COLS = [\n",
    "    'AMT_INCOME_TOTAL', 'AMT_CREDIT', 'AMT_ANNUITY', 'AMT_GOODS_PRICE', 'REGION_POPULATION_RELATIVE',\n",
    "    'DAYS_BIRTH', 'DAYS_EMPLOYED', 'DAYS_REGISTRATION', 'DAYS_ID_PUBLISH', 'CNT_FAM_MEMBERS', 'EXT_SOURCE_1', 'EXT_SOURCE_2', 'EXT_SOURCE_3',\n",
    "    'APARTMENTS_AVG', 'BASEMENTAREA_AVG', 'YEARS_BEGINEXPLUATATION_AVG', 'YEARS_BUILD_AVG',\n",
    "    'COMMONAREA_AVG', 'ELEVATORS_AVG', 'ENTRANCES_AVG', 'FLOORSMAX_AVG', 'FLOORSMIN_AVG', 'LANDAREA_AVG', 'LIVINGAPARTMENTS_AVG',\n",
    "    'LIVINGAREA_AVG', 'NONLIVINGAPARTMENTS_AVG', 'NONLIVINGAREA_AVG', 'APARTMENTS_MODE',\n",
    "    'BASEMENTAREA_MODE', 'YEARS_BEGINEXPLUATATION_MODE', 'YEARS_BUILD_MODE', 'COMMONAREA_MODE',\n",
    "    'ELEVATORS_MODE', 'ENTRANCES_MODE', 'FLOORSMAX_MODE', 'FLOORSMIN_MODE', 'LANDAREA_MODE',\n",
    "    'LIVINGAPARTMENTS_MODE', 'LIVINGAREA_MODE', 'NONLIVINGAPARTMENTS_MODE', 'NONLIVINGAREA_MODE',\n",
    "    'APARTMENTS_MEDI', 'BASEMENTAREA_MEDI', 'YEARS_BEGINEXPLUATATION_MEDI', 'YEARS_BUILD_MEDI',\n",
    "    'COMMONAREA_MEDI', 'ELEVATORS_MEDI', 'ENTRANCES_MEDI', 'FLOORSMAX_MEDI', 'FLOORSMIN_MEDI', \n",
    "    'LANDAREA_MEDI', 'LIVINGAPARTMENTS_MEDI', 'LIVINGAREA_MEDI', 'NONLIVINGAPARTMENTS_MEDI',\n",
    "    'NONLIVINGAREA_MEDI', 'TOTALAREA_MODE', 'OBS_30_CNT_SOCIAL_CIRCLE', 'DEF_30_CNT_SOCIAL_CIRCLE',\n",
    "    'OBS_60_CNT_SOCIAL_CIRCLE', 'DEF_60_CNT_SOCIAL_CIRCLE', 'DAYS_LAST_PHONE_CHANGE', 'AMT_REQ_CREDIT_BUREAU_HOUR',\n",
    "    'AMT_REQ_CREDIT_BUREAU_DAY', 'AMT_REQ_CREDIT_BUREAU_WEEK', 'AMT_REQ_CREDIT_BUREAU_MON', 'AMT_REQ_CREDIT_BUREAU_QRT',\n",
    "    'AMT_REQ_CREDIT_BUREAU_YEAR', 'PREV APP COUNT', 'TOTAL PREV LOAN AMT', 'AMT_ANNUITY_PAVG', 'AMT_APPLICATION',\n",
    "    'AMT_CREDIT_PAVG', 'AMT_DOWN_PAYMENT', 'AMT_GOODS_PRICE_PAVG', 'HOUR_APPR_PROCESS_START_PAVG', 'NFLAG_LAST_APPL_IN_DAY',\n",
    "    'RATE_DOWN_PAYMENT', 'RATE_INTEREST_PRIMARY', 'RATE_INTEREST_PRIVILEGED', 'DAYS_DECISION', 'SELLERPLACE_AREA', 'CNT_PAYMENT',\n",
    "    'DAYS_FIRST_DRAWING', 'DAYS_FIRST_DUE', 'DAYS_LAST_DUE_1ST_VERSION', 'DAYS_LAST_DUE', 'DAYS_TERMINATION', 'NFLAG_INSURED_ON_APPROVAL',\n",
    "    'MONTHS_BALANCE', 'AMT_BALANCE', 'AMT_CREDIT_LIMIT_ACTUAL', 'AMT_DRAWINGS_ATM_CURRENT', 'AMT_DRAWINGS_CURRENT', 'AMT_DRAWINGS_OTHER_CURRENT',\n",
    "    'AMT_DRAWINGS_POS_CURRENT', 'AMT_INST_MIN_REGULARITY', 'AMT_PAYMENT_CURRENT', 'AMT_PAYMENT_TOTAL_CURRENT', 'AMT_RECEIVABLE_PRINCIPAL',\n",
    "    'AMT_RECIVABLE', 'AMT_TOTAL_RECEIVABLE', 'CNT_DRAWINGS_ATM_CURRENT', 'CNT_DRAWINGS_CURRENT', 'CNT_DRAWINGS_OTHER_CURRENT',\n",
    "    'CNT_DRAWINGS_POS_CURRENT', 'CNT_INSTALMENT_MATURE_CUM', 'SK_DPD', 'SK_DPD_DEF', 'DAYS_CREDIT', 'CREDIT_DAY_OVERDUE', 'DAYS_CREDIT_ENDDATE', 'DAYS_ENDDATE_FACT',\n",
    "    'AMT_CREDIT_MAX_OVERDUE', 'CNT_CREDIT_PROLONG', 'AMT_CREDIT_SUM', 'AMT_CREDIT_SUM_DEBT', 'AMT_CREDIT_SUM_LIMIT', 'AMT_CREDIT_SUM_OVERDUE',\n",
    "    'DAYS_CREDIT_UPDATE', 'AMT_ANNUITY_BAVG', 'CNT_INSTALMENT', 'CNT_INSTALMENT_FUTURE', 'SK_DPD_CAVG', 'SK_DPD_DEF_CAVG', 'NUM_INSTALMENT_VERSION',\n",
    "    'NUM_INSTALMENT_NUMBER', 'DAYS_INSTALMENT', 'DAYS_ENTRY_PAYMENT', 'AMT_INSTALMENT', 'AMT_PAYMENT'\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Convert categorical variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dtype('O')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df['NAME_CONTRACT_TYPE'].dtype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "for col in CAT_COLUMNS + BOOL_COLS:\n",
    "    if str(train_df[col].dtype) == 'category':\n",
    "        continue\n",
    "\n",
    "    train_df[col] = train_df[col].astype('category').cat.as_ordered()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "category\n",
      "category\n"
     ]
    }
   ],
   "source": [
    "print(train_df['NAME_CONTRACT_TYPE'].dtype)\n",
    "print(train_df['FLAG_OWN_CAR'].dtype)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "apply_cats(test_df_raw, train_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>SK_ID_CURR</th>\n",
       "      <th>NAME_CONTRACT_TYPE</th>\n",
       "      <th>CODE_GENDER</th>\n",
       "      <th>FLAG_OWN_CAR</th>\n",
       "      <th>FLAG_OWN_REALTY</th>\n",
       "      <th>CNT_CHILDREN</th>\n",
       "      <th>AMT_INCOME_TOTAL</th>\n",
       "      <th>AMT_CREDIT</th>\n",
       "      <th>AMT_ANNUITY</th>\n",
       "      <th>AMT_GOODS_PRICE</th>\n",
       "      <th>...</th>\n",
       "      <th>SK_DPD_CAVG</th>\n",
       "      <th>SK_DPD_DEF_CAVG</th>\n",
       "      <th>NAME_CONTRACT_STATUS_CAVG</th>\n",
       "      <th>SK_ID_PREV_IAVG</th>\n",
       "      <th>NUM_INSTALMENT_VERSION</th>\n",
       "      <th>NUM_INSTALMENT_NUMBER</th>\n",
       "      <th>DAYS_INSTALMENT</th>\n",
       "      <th>DAYS_ENTRY_PAYMENT</th>\n",
       "      <th>AMT_INSTALMENT</th>\n",
       "      <th>AMT_PAYMENT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100001</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>F</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>135000.0</td>\n",
       "      <td>568800.0</td>\n",
       "      <td>20560.5</td>\n",
       "      <td>450000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.552756</td>\n",
       "      <td>0.552756</td>\n",
       "      <td>Completed</td>\n",
       "      <td>1.576389e+06</td>\n",
       "      <td>1.142857</td>\n",
       "      <td>2.714286</td>\n",
       "      <td>-2187.714286</td>\n",
       "      <td>-2195.000000</td>\n",
       "      <td>5885.132143</td>\n",
       "      <td>5885.132143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>100005</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>M</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>99000.0</td>\n",
       "      <td>222768.0</td>\n",
       "      <td>17370.0</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Completed</td>\n",
       "      <td>2.495675e+06</td>\n",
       "      <td>1.111111</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>-586.000000</td>\n",
       "      <td>-609.555556</td>\n",
       "      <td>6240.205000</td>\n",
       "      <td>6240.205000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100013</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>M</td>\n",
       "      <td>Y</td>\n",
       "      <td>Y</td>\n",
       "      <td>0</td>\n",
       "      <td>202500.0</td>\n",
       "      <td>663264.0</td>\n",
       "      <td>69777.0</td>\n",
       "      <td>630000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>3.030320</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Completed</td>\n",
       "      <td>1.911853e+06</td>\n",
       "      <td>0.277419</td>\n",
       "      <td>43.729032</td>\n",
       "      <td>-1352.929032</td>\n",
       "      <td>-1358.109677</td>\n",
       "      <td>10897.898516</td>\n",
       "      <td>9740.235774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100028</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>F</td>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>2</td>\n",
       "      <td>315000.0</td>\n",
       "      <td>1575000.0</td>\n",
       "      <td>49018.5</td>\n",
       "      <td>1575000.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Completed</td>\n",
       "      <td>2.115491e+06</td>\n",
       "      <td>0.460177</td>\n",
       "      <td>30.504425</td>\n",
       "      <td>-855.548673</td>\n",
       "      <td>-858.548673</td>\n",
       "      <td>4979.282257</td>\n",
       "      <td>4356.731549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100038</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>M</td>\n",
       "      <td>Y</td>\n",
       "      <td>N</td>\n",
       "      <td>1</td>\n",
       "      <td>180000.0</td>\n",
       "      <td>625500.0</td>\n",
       "      <td>32067.0</td>\n",
       "      <td>625500.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Completed</td>\n",
       "      <td>2.327930e+06</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.500000</td>\n",
       "      <td>-622.000000</td>\n",
       "      <td>-634.250000</td>\n",
       "      <td>11100.337500</td>\n",
       "      <td>11100.337500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 206 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   SK_ID_CURR NAME_CONTRACT_TYPE CODE_GENDER FLAG_OWN_CAR FLAG_OWN_REALTY  \\\n",
       "0      100001         Cash loans           F            N               Y   \n",
       "1      100005         Cash loans           M            N               Y   \n",
       "2      100013         Cash loans           M            Y               Y   \n",
       "3      100028         Cash loans           F            N               Y   \n",
       "4      100038         Cash loans           M            Y               N   \n",
       "\n",
       "  CNT_CHILDREN  AMT_INCOME_TOTAL  AMT_CREDIT  AMT_ANNUITY  AMT_GOODS_PRICE  \\\n",
       "0            0          135000.0    568800.0      20560.5         450000.0   \n",
       "1            0           99000.0    222768.0      17370.0         180000.0   \n",
       "2            0          202500.0    663264.0      69777.0         630000.0   \n",
       "3            2          315000.0   1575000.0      49018.5        1575000.0   \n",
       "4            1          180000.0    625500.0      32067.0         625500.0   \n",
       "\n",
       "       ...      SK_DPD_CAVG SK_DPD_DEF_CAVG NAME_CONTRACT_STATUS_CAVG  \\\n",
       "0      ...         0.552756        0.552756                 Completed   \n",
       "1      ...         0.000000        0.000000                 Completed   \n",
       "2      ...         3.030320        0.000000                 Completed   \n",
       "3      ...         0.000000        0.000000                 Completed   \n",
       "4      ...         0.000000        0.000000                 Completed   \n",
       "\n",
       "  SK_ID_PREV_IAVG NUM_INSTALMENT_VERSION  NUM_INSTALMENT_NUMBER  \\\n",
       "0    1.576389e+06               1.142857               2.714286   \n",
       "1    2.495675e+06               1.111111               5.000000   \n",
       "2    1.911853e+06               0.277419              43.729032   \n",
       "3    2.115491e+06               0.460177              30.504425   \n",
       "4    2.327930e+06               1.000000               6.500000   \n",
       "\n",
       "   DAYS_INSTALMENT  DAYS_ENTRY_PAYMENT  AMT_INSTALMENT   AMT_PAYMENT  \n",
       "0     -2187.714286        -2195.000000     5885.132143   5885.132143  \n",
       "1      -586.000000         -609.555556     6240.205000   6240.205000  \n",
       "2     -1352.929032        -1358.109677    10897.898516   9740.235774  \n",
       "3      -855.548673         -858.548673     4979.282257   4356.731549  \n",
       "4      -622.000000         -634.250000    11100.337500  11100.337500  \n",
       "\n",
       "[5 rows x 206 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df_raw.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "for col in CAT_COLUMNS + BOOL_COLS:\n",
    "    if str(test_df_raw[col].dtype) == 'category':\n",
    "        continue\n",
    "\n",
    "    test_df_raw[col] = test_df_raw[col].astype('category').cat.as_ordered()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Process dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "df, train_y, nas, mapper = proc_df(\n",
    "    train_df, 'TARGET', skip_flds=[\n",
    "        'SK_ID_CURR', 'SK_ID_PREV', 'SK_ID_PREV_CCAVG', 'SK_ID_BUREAU', 'SK_ID_PREV_IAVG'], do_scale=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>NAME_CONTRACT_TYPE</th>\n",
       "      <th>CODE_GENDER</th>\n",
       "      <th>FLAG_OWN_CAR</th>\n",
       "      <th>FLAG_OWN_REALTY</th>\n",
       "      <th>CNT_CHILDREN</th>\n",
       "      <th>AMT_INCOME_TOTAL</th>\n",
       "      <th>AMT_CREDIT</th>\n",
       "      <th>AMT_ANNUITY</th>\n",
       "      <th>AMT_GOODS_PRICE</th>\n",
       "      <th>NAME_TYPE_SUITE</th>\n",
       "      <th>...</th>\n",
       "      <th>CNT_INSTALMENT_na</th>\n",
       "      <th>CNT_INSTALMENT_FUTURE_na</th>\n",
       "      <th>SK_DPD_CAVG_na</th>\n",
       "      <th>SK_DPD_DEF_CAVG_na</th>\n",
       "      <th>NUM_INSTALMENT_VERSION_na</th>\n",
       "      <th>NUM_INSTALMENT_NUMBER_na</th>\n",
       "      <th>DAYS_INSTALMENT_na</th>\n",
       "      <th>DAYS_ENTRY_PAYMENT_na</th>\n",
       "      <th>AMT_INSTALMENT_na</th>\n",
       "      <th>AMT_PAYMENT_na</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0.142129</td>\n",
       "      <td>-0.478095</td>\n",
       "      <td>-0.166143</td>\n",
       "      <td>-0.507236</td>\n",
       "      <td>7</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.367696</td>\n",
       "      <td>-0.367394</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.426792</td>\n",
       "      <td>1.725450</td>\n",
       "      <td>0.592683</td>\n",
       "      <td>1.600873</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.367696</td>\n",
       "      <td>-0.367394</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.427196</td>\n",
       "      <td>-1.152888</td>\n",
       "      <td>-1.404669</td>\n",
       "      <td>-1.092145</td>\n",
       "      <td>7</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.367696</td>\n",
       "      <td>-0.367394</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.142533</td>\n",
       "      <td>-0.711430</td>\n",
       "      <td>0.177874</td>\n",
       "      <td>-0.653463</td>\n",
       "      <td>7</td>\n",
       "      <td>...</td>\n",
       "      <td>2.719639</td>\n",
       "      <td>2.721875</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.199466</td>\n",
       "      <td>-0.213734</td>\n",
       "      <td>-0.361749</td>\n",
       "      <td>-0.068554</td>\n",
       "      <td>7</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.367696</td>\n",
       "      <td>-0.367394</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.249839</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "      <td>-0.233257</td>\n",
       "      <td>-0.233319</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 324 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   NAME_CONTRACT_TYPE  CODE_GENDER  FLAG_OWN_CAR  FLAG_OWN_REALTY  \\\n",
       "0                   1            2             1                2   \n",
       "1                   1            1             1                1   \n",
       "2                   2            2             2                2   \n",
       "3                   1            1             1                2   \n",
       "4                   1            2             1                2   \n",
       "\n",
       "   CNT_CHILDREN  AMT_INCOME_TOTAL  AMT_CREDIT  AMT_ANNUITY  AMT_GOODS_PRICE  \\\n",
       "0             1          0.142129   -0.478095    -0.166143        -0.507236   \n",
       "1             1          0.426792    1.725450     0.592683         1.600873   \n",
       "2             1         -0.427196   -1.152888    -1.404669        -1.092145   \n",
       "3             1         -0.142533   -0.711430     0.177874        -0.653463   \n",
       "4             1         -0.199466   -0.213734    -0.361749        -0.068554   \n",
       "\n",
       "   NAME_TYPE_SUITE       ...        CNT_INSTALMENT_na  \\\n",
       "0                7       ...                -0.367696   \n",
       "1                2       ...                -0.367696   \n",
       "2                7       ...                -0.367696   \n",
       "3                7       ...                 2.719639   \n",
       "4                7       ...                -0.367696   \n",
       "\n",
       "   CNT_INSTALMENT_FUTURE_na  SK_DPD_CAVG_na  SK_DPD_DEF_CAVG_na  \\\n",
       "0                 -0.367394       -0.249839           -0.249839   \n",
       "1                 -0.367394       -0.249839           -0.249839   \n",
       "2                 -0.367394       -0.249839           -0.249839   \n",
       "3                  2.721875       -0.249839           -0.249839   \n",
       "4                 -0.367394       -0.249839           -0.249839   \n",
       "\n",
       "   NUM_INSTALMENT_VERSION_na  NUM_INSTALMENT_NUMBER_na  DAYS_INSTALMENT_na  \\\n",
       "0                  -0.233257                 -0.233257           -0.233257   \n",
       "1                  -0.233257                 -0.233257           -0.233257   \n",
       "2                  -0.233257                 -0.233257           -0.233257   \n",
       "3                  -0.233257                 -0.233257           -0.233257   \n",
       "4                  -0.233257                 -0.233257           -0.233257   \n",
       "\n",
       "   DAYS_ENTRY_PAYMENT_na  AMT_INSTALMENT_na  AMT_PAYMENT_na  \n",
       "0              -0.233319          -0.233257       -0.233319  \n",
       "1              -0.233319          -0.233257       -0.233319  \n",
       "2              -0.233319          -0.233257       -0.233319  \n",
       "3              -0.233319          -0.233257       -0.233319  \n",
       "4              -0.233319          -0.233257       -0.233319  \n",
       "\n",
       "[5 rows x 324 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_df, _, nas, mapper = proc_df(\n",
    "    test_df_raw, skip_flds=[\n",
    "        'SK_ID_CURR', 'SK_ID_PREV', 'SK_ID_PREV_CCAVG', 'SK_ID_BUREAU', 'SK_ID_PREV_IAVG'], do_scale=True,\n",
    "    mapper=mapper, na_dict=nas)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create validation set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "val_idx = get_cv_idxs(len(df), val_pct=0.2, seed=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Build and train model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "md = ColumnarModelData.from_data_frame(\n",
    "    PATH, val_idx, df, train_y.astype(np.float32),\n",
    "    cat_flds=CAT_COLUMNS + BOOL_COLS, bs=128, test_df=test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "cat_sz = [(c, len(train_df[c].cat.categories)+1) for c in CAT_COLUMNS + BOOL_COLS]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('NAME_CONTRACT_TYPE', 3),\n",
       " ('CODE_GENDER', 4),\n",
       " ('CNT_CHILDREN', 16),\n",
       " ('NAME_TYPE_SUITE', 8),\n",
       " ('NAME_INCOME_TYPE', 9),\n",
       " ('NAME_EDUCATION_TYPE', 6),\n",
       " ('NAME_FAMILY_STATUS', 7),\n",
       " ('NAME_HOUSING_TYPE', 7),\n",
       " ('OWN_CAR_AGE', 63),\n",
       " ('OCCUPATION_TYPE', 19),\n",
       " ('REGION_RATING_CLIENT', 4),\n",
       " ('REGION_RATING_CLIENT_W_CITY', 4),\n",
       " ('WEEKDAY_APPR_PROCESS_START', 8),\n",
       " ('HOUR_APPR_PROCESS_START', 25),\n",
       " ('ORGANIZATION_TYPE', 59),\n",
       " ('FONDKAPREMONT_MODE', 5),\n",
       " ('HOUSETYPE_MODE', 4),\n",
       " ('WALLSMATERIAL_MODE', 8),\n",
       " ('EMERGENCYSTATE_MODE', 3),\n",
       " ('NAME_CONTRACT_TYPE_BMODE', 5),\n",
       " ('WEEKDAY_APPR_PROCESS_START_BMODE', 8),\n",
       " ('FLAG_LAST_APPL_PER_CONTRACT', 3),\n",
       " ('NAME_CASH_LOAN_PURPOSE', 26),\n",
       " ('NAME_CONTRACT_STATUS', 5),\n",
       " ('NAME_PAYMENT_TYPE', 5),\n",
       " ('CODE_REJECT_REASON', 10),\n",
       " ('NAME_TYPE_SUITE_BMODE', 9),\n",
       " ('NAME_CLIENT_TYPE', 5),\n",
       " ('NAME_GOODS_CATEGORY', 27),\n",
       " ('NAME_PORTFOLIO', 6),\n",
       " ('NAME_PRODUCT_TYPE', 4),\n",
       " ('CHANNEL_TYPE', 9),\n",
       " ('NAME_SELLER_INDUSTRY', 12),\n",
       " ('NAME_YIELD_GROUP', 6),\n",
       " ('PRODUCT_COMBINATION', 19),\n",
       " ('NAME_CONTRACT_STATUS_CCAVG', 5),\n",
       " ('NAME_CONTRACT_STATUS_CAVG', 8),\n",
       " ('FLAG_OWN_CAR', 3),\n",
       " ('FLAG_OWN_REALTY', 3),\n",
       " ('FLAG_MOBIL', 3),\n",
       " ('FLAG_EMP_PHONE', 3),\n",
       " ('FLAG_WORK_PHONE', 3),\n",
       " ('FLAG_CONT_MOBILE', 3),\n",
       " ('FLAG_PHONE', 3),\n",
       " ('FLAG_EMAIL', 3),\n",
       " ('REG_REGION_NOT_LIVE_REGION', 3),\n",
       " ('REG_REGION_NOT_WORK_REGION', 3),\n",
       " ('LIVE_REGION_NOT_WORK_REGION', 3),\n",
       " ('REG_CITY_NOT_LIVE_CITY', 3),\n",
       " ('REG_CITY_NOT_WORK_CITY', 3),\n",
       " ('LIVE_CITY_NOT_WORK_CITY', 3),\n",
       " ('FLAG_DOCUMENT_2', 3),\n",
       " ('FLAG_DOCUMENT_3', 3),\n",
       " ('FLAG_DOCUMENT_4', 3),\n",
       " ('FLAG_DOCUMENT_5', 3),\n",
       " ('FLAG_DOCUMENT_6', 3),\n",
       " ('FLAG_DOCUMENT_7', 3),\n",
       " ('FLAG_DOCUMENT_8', 3),\n",
       " ('FLAG_DOCUMENT_9', 3),\n",
       " ('FLAG_DOCUMENT_10', 3),\n",
       " ('FLAG_DOCUMENT_11', 3),\n",
       " ('FLAG_DOCUMENT_12', 3),\n",
       " ('FLAG_DOCUMENT_13', 3),\n",
       " ('FLAG_DOCUMENT_14', 3),\n",
       " ('FLAG_DOCUMENT_15', 3),\n",
       " ('FLAG_DOCUMENT_16', 3),\n",
       " ('FLAG_DOCUMENT_17', 3),\n",
       " ('FLAG_DOCUMENT_18', 3),\n",
       " ('FLAG_DOCUMENT_19', 3),\n",
       " ('FLAG_DOCUMENT_20', 3),\n",
       " ('FLAG_DOCUMENT_21', 3)]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_sz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "emb_szs = [(c, min(50, (c+1)//2)) for _,c in cat_sz]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(3, 2),\n",
       " (4, 2),\n",
       " (16, 8),\n",
       " (8, 4),\n",
       " (9, 5),\n",
       " (6, 3),\n",
       " (7, 4),\n",
       " (7, 4),\n",
       " (63, 32),\n",
       " (19, 10),\n",
       " (4, 2),\n",
       " (4, 2),\n",
       " (8, 4),\n",
       " (25, 13),\n",
       " (59, 30),\n",
       " (5, 3),\n",
       " (4, 2),\n",
       " (8, 4),\n",
       " (3, 2),\n",
       " (5, 3),\n",
       " (8, 4),\n",
       " (3, 2),\n",
       " (26, 13),\n",
       " (5, 3),\n",
       " (5, 3),\n",
       " (10, 5),\n",
       " (9, 5),\n",
       " (5, 3),\n",
       " (27, 14),\n",
       " (6, 3),\n",
       " (4, 2),\n",
       " (9, 5),\n",
       " (12, 6),\n",
       " (6, 3),\n",
       " (19, 10),\n",
       " (5, 3),\n",
       " (8, 4),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2),\n",
       " (3, 2)]"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "emb_szs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "m = md.get_learner(\n",
    "    emb_szs, len(df.columns)-len(CAT_COLUMNS + BOOL_COLS),\n",
    "    0.04, 1, [1000, 500], [0.001, 0.01], y_range=[0, 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "58f482f63d8f404f900bdca353e698b0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=1), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss                                  \n",
      "    0      0.081166   0.080469  \n",
      "\n"
     ]
    }
   ],
   "source": [
    "m.lr_find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "m.sched.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr = 2e-5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fastai_roc_auc_score(preds, targs):\n",
    "    return roc_auc_score(targs, preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9da50bf6aea1468db5108a639ed96ec3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=3), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss   fastai_roc_auc_score           \n",
      "    0      0.06756    0.068763   0.733859  \n",
      "    1      0.067249   0.068299   0.744699                       \n",
      "    2      0.067494   0.06788    0.750594                       \n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[array([0.06788]), 0.7505936721970905]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m.fit(lr, 3, metrics=[fastai_roc_auc_score])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "24a3fbcc4c8b47ec9daebff496dcd2eb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=5), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss   fastai_roc_auc_score           \n",
      "    0      0.067121   0.067746   0.753427  \n",
      "    1      0.067049   0.067676   0.755136                       \n",
      "    2      0.067066   0.067684   0.755508                       \n",
      "    3      0.065886   0.067597   0.756942                       \n",
      "    4      0.067783   0.067599   0.757334                       \n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[array([0.0676]), 0.7573341633351035]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m.fit(lr, 5, metrics=[fastai_roc_auc_score], cycle_len=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "60d84c3cba7544d7b7b4c2066c95791a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=8), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss   fastai_roc_auc_score           \n",
      "    0      0.064621   0.06757    0.757407  \n",
      "    1      0.06695    0.067553   0.758008                       \n",
      "    2      0.065544   0.06753    0.758794                       \n",
      "    3      0.06537    0.067514   0.759036                       \n",
      "    4      0.067812   0.06763    0.758022                       \n",
      "    5      0.064845   0.067524   0.758856                       \n",
      "    6      0.065109   0.067486   0.759571                       \n",
      "    7      0.065478   0.067509   0.759701                       \n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[array([0.06751]), 0.7597008453585836]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m.fit(lr, 2, metrics=[fastai_roc_auc_score], cycle_len=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "m.save('nn_retrain_v1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "lr = 1e-6"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c995039cbe004af3b4a19475494268f3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=7), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss   fastai_roc_auc_score           \n",
      "    0      0.064187   0.067511   0.759739  \n",
      "    1      0.062675   0.067525   0.759751                       \n",
      "    2      0.067521   0.067506   0.759836                       \n",
      "    3      0.063839   0.067537   0.759824                       \n",
      "    4      0.063248   0.067531   0.759976                       \n",
      "    5      0.062112   0.067497   0.759945                       \n",
      "    6      0.065107   0.067494   0.759948                       \n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[array([0.06749]), 0.7599478769124758]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m.fit(lr, 3, metrics=[fastai_roc_auc_score], cycle_len=1, cycle_mult=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "m.save('nn_retrain_v2')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train on whole dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "341f71383b704b28a3a5d24b77508101",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=3), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss                                  \n",
      "    0      0.071503   0.340369  \n",
      "    1      0.063977   0.380292                                  \n",
      "    2      0.065456   0.323353                                  \n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a501ee9d642d452a90145f8e1eed9fe5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=5), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss                                  \n",
      "    0      0.066998   0.3113    \n",
      "    1      0.067516   0.311775                                  \n",
      "    2      0.066976   0.316702                                  \n",
      "    3      0.066012   0.307013                                  \n",
      "    4      0.064431   0.301919                                  \n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "774eb4ada80d4ea2b73d5f76868b1856",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=8), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss                                  \n",
      "    0      0.068765   0.33349   \n",
      "    1      0.067686   0.333234                                  \n",
      "    2      0.065454   0.306597                                  \n",
      "    3      0.066401   0.321467                                  \n",
      "    4      0.067011   0.334695                                  \n",
      "    5      0.067325   0.308287                                  \n",
      "    6      0.061197   0.354947                                  \n",
      "    7      0.065977   0.323473                                  \n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d0c635f905304a60a24a44a4f97c6276",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=7), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss                                  \n",
      "    0      0.066235   0.319138  \n",
      "    1      0.064263   0.320816                                  \n",
      "    2      0.065062   0.324307                                  \n",
      "    3      0.067223   0.315286                                  \n",
      "    4      0.067398   0.317028                                  \n",
      "    5      0.064751   0.320859                                  \n",
      "    6      0.064      0.318914                                  \n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[array([0.31891])]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr = 2e-5\n",
    "md_full = ColumnarModelData.from_data_frame(\n",
    "    PATH, np.array([0]), df, train_y.astype(np.float32),\n",
    "    cat_flds=CAT_COLUMNS + BOOL_COLS, bs=128, test_df=test_df)\n",
    "m_full = md_full.get_learner(\n",
    "    emb_szs, len(df.columns)-len(CAT_COLUMNS + BOOL_COLS),\n",
    "    0.04, 1, [1000, 500], [0.001, 0.01], y_range=[0, 1])\n",
    "m_full.fit(lr, 3)\n",
    "m_full.fit(lr, 5, cycle_len=1)\n",
    "m_full.fit(lr, 2, cycle_len=4)\n",
    "lr = 1e-6\n",
    "m_full.fit(lr, 3, cycle_len=1, cycle_mult=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds = m_full.predict(is_test=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(48744, 1)"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.DataFrame(\n",
    "    {'SK_ID_CURR': list(test_df['SK_ID_CURR']),\n",
    "     'TARGET': [p[0] for p in preds]}\n",
    ").to_csv(f'{PATH}/sub2.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='data/home-credit-default-risk/sub2.csv' target='_blank'>data/home-credit-default-risk/sub2.csv</a><br>"
      ],
      "text/plain": [
       "/Users/lex/code/study-notes/moocs/fast.ai/deep-learning-2018/data/home-credit-default-risk/sub2.csv"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FileLink(f'{PATH}/sub2.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"./images/home-credit-default-risk-2.png\">"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}