5.0. Original Data Cleansing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ec31c411",
   "metadata": {},
   "source": [
    "# <font color='CC3D3D'> Original Data Clansing\n",
    "    \n",
    "### 1. Outlier preprocessing\n",
    "- <span style=\"color:green\">**user_spec.csv, loan_result, log_data.csv** </span>에 대하여 각 열에 대해 이상치 처리\n",
    "\n",
    "### 2. MissingValue Preprocessing\n",
    "- <span style=\"color:green\">**user_spec.csv, loan_result, log_data.csv** </span>에 대하여 각 열에 대해 결측치 처리\n",
    "    \n",
    "### 3. Deployment CSV\n",
    "- 결측치 처리된 데이터에 대해 <span style=\"color:blue\">**user_spec_clansing.csv, loan_result_clansing, log_data_clansing.csv** </span>로 생성"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d5facccb",
   "metadata": {},
   "source": [
    "# Import"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "63630aa2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "\n",
    "from sklearn.impute import KNNImputer\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from pycaret.classification import *\n",
    "\n",
    "# Visualization\n",
    "import matplotlib.pylab as plt\n",
    "from matplotlib import font_manager, rc\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "\n",
    "# EDA\n",
    "import klib\n",
    "\n",
    "plt.rc(\"font\", family = 'Malgun Gothic')\n",
    "plt.rcParams[\"figure.figsize\"] = (12,8)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bfc69b4",
   "metadata": {},
   "source": [
    "# Data Load"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "3e3dbeaf",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_root = '../Data/2022빅콘테스트_데이터분석리그_데이터분석분야_퓨처스부문_데이터셋_220908'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3fbbb9e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec = pd.read_csv(data_root+'/user_spec.csv')\n",
    "loan_result = pd.read_csv(data_root+'/loan_result.csv')\n",
    "log_data = pd.read_csv(data_root+'/log_data.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5f7268e5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1394216 entries, 0 to 1394215\n",
      "Data columns (total 17 columns):\n",
      " #   Column                               Non-Null Count    Dtype  \n",
      "---  ------                               --------------    -----  \n",
      " 0   application_id                       1394216 non-null  int64  \n",
      " 1   user_id                              1394216 non-null  int64  \n",
      " 2   birth_year                           1381255 non-null  float64\n",
      " 3   gender                               1381255 non-null  float64\n",
      " 4   insert_time                          1394216 non-null  object \n",
      " 5   credit_score                         1289101 non-null  float64\n",
      " 6   yearly_income                        1394126 non-null  float64\n",
      " 7   income_type                          1394131 non-null  object \n",
      " 8   company_enter_month                  1222456 non-null  float64\n",
      " 9   employment_type                      1394131 non-null  object \n",
      " 10  houseown_type                        1394131 non-null  object \n",
      " 11  desired_amount                       1394131 non-null  float64\n",
      " 12  purpose                              1394131 non-null  object \n",
      " 13  personal_rehabilitation_yn           806755 non-null   float64\n",
      " 14  personal_rehabilitation_complete_yn  190862 non-null   float64\n",
      " 15  existing_loan_cnt                    1195660 non-null  float64\n",
      " 16  existing_loan_amt                    1080442 non-null  float64\n",
      "dtypes: float64(10), int64(2), object(5)\n",
      "memory usage: 180.8+ MB\n"
     ]
    }
   ],
   "source": [
    "user_spec.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "92055396",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 13527363 entries, 0 to 13527362\n",
      "Data columns (total 7 columns):\n",
      " #   Column                 Dtype  \n",
      "---  ------                 -----  \n",
      " 0   application_id         int64  \n",
      " 1   loanapply_insert_time  object \n",
      " 2   bank_id                int64  \n",
      " 3   product_id             int64  \n",
      " 4   loan_limit             float64\n",
      " 5   loan_rate              float64\n",
      " 6   is_applied             float64\n",
      "dtypes: float64(3), int64(3), object(1)\n",
      "memory usage: 722.4+ MB\n"
     ]
    }
   ],
   "source": [
    "loan_result.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "cc2f373c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 17843993 entries, 0 to 17843992\n",
      "Data columns (total 6 columns):\n",
      " #   Column          Dtype \n",
      "---  ------          ----- \n",
      " 0   user_id         int64 \n",
      " 1   event           object\n",
      " 2   timestamp       object\n",
      " 3   mp_os           object\n",
      " 4   mp_app_version  object\n",
      " 5   date_cd         object\n",
      "dtypes: int64(1), object(5)\n",
      "memory usage: 816.8+ MB\n"
     ]
    }
   ],
   "source": [
    "log_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "64549ccb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                               0\n",
       "user_id                                      0\n",
       "birth_year                               12961\n",
       "gender                                   12961\n",
       "insert_time                                  0\n",
       "credit_score                            105115\n",
       "yearly_income                               90\n",
       "income_type                                 85\n",
       "company_enter_month                     171760\n",
       "employment_type                             85\n",
       "houseown_type                               85\n",
       "desired_amount                              85\n",
       "purpose                                     85\n",
       "personal_rehabilitation_yn              587461\n",
       "personal_rehabilitation_complete_yn    1203354\n",
       "existing_loan_cnt                       198556\n",
       "existing_loan_amt                       313774\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "bd760c25",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                 0\n",
       "loanapply_insert_time          0\n",
       "bank_id                        0\n",
       "product_id                     0\n",
       "loan_limit                  7495\n",
       "loan_rate                   7495\n",
       "is_applied               3257239\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_result.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4f27454f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_id                0\n",
       "event                  0\n",
       "timestamp              0\n",
       "mp_os                980\n",
       "mp_app_version    660597\n",
       "date_cd                0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_data.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "013a7da7",
   "metadata": {},
   "source": [
    "# Data Preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9c82aa6e",
   "metadata": {},
   "source": [
    "## Application_id"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "6321531e",
   "metadata": {},
   "outputs": [],
   "source": [
    "app_id = loan_result.application_id.unique().tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ad33ce86",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec = user_spec[user_spec['application_id'].isin(app_id)].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "9e58903c",
   "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>application_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>birth_year</th>\n",
       "      <th>gender</th>\n",
       "      <th>insert_time</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>company_enter_month</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>954900</td>\n",
       "      <td>553686</td>\n",
       "      <td>1968.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 14:29:03</td>\n",
       "      <td>870.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>PRIVATEBUSINESS</td>\n",
       "      <td>20070201.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>27000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>137274</td>\n",
       "      <td>59516</td>\n",
       "      <td>1997.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 21:40:22</td>\n",
       "      <td>710.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>FREELANCER</td>\n",
       "      <td>20210901.0</td>\n",
       "      <td>기타</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>10000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>15000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1570936</td>\n",
       "      <td>167320</td>\n",
       "      <td>1989.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 09:40:27</td>\n",
       "      <td>820.0</td>\n",
       "      <td>62000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>20170101.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>자가</td>\n",
       "      <td>2000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.0</td>\n",
       "      <td>344000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>967833</td>\n",
       "      <td>33400</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 08:55:07</td>\n",
       "      <td>630.0</td>\n",
       "      <td>36000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>20210901.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>5000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1614045</td>\n",
       "      <td>860410</td>\n",
       "      <td>1976.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 11:29:05</td>\n",
       "      <td>590.0</td>\n",
       "      <td>45000000.0</td>\n",
       "      <td>OTHERINCOME</td>\n",
       "      <td>NaN</td>\n",
       "      <td>기타</td>\n",
       "      <td>자가</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968748</th>\n",
       "      <td>1864587</td>\n",
       "      <td>489900</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-03-22 14:55:32</td>\n",
       "      <td>590.0</td>\n",
       "      <td>25000000.0</td>\n",
       "      <td>FREELANCER</td>\n",
       "      <td>202106.0</td>\n",
       "      <td>기타</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>5000000.0</td>\n",
       "      <td>사업자금</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968749</th>\n",
       "      <td>1327066</td>\n",
       "      <td>151422</td>\n",
       "      <td>1955.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-03-22 01:19:24</td>\n",
       "      <td>980.0</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>OTHERINCOME</td>\n",
       "      <td>NaN</td>\n",
       "      <td>기타</td>\n",
       "      <td>자가</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968750</th>\n",
       "      <td>1319606</td>\n",
       "      <td>173524</td>\n",
       "      <td>1983.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-03-22 07:34:32</td>\n",
       "      <td>750.0</td>\n",
       "      <td>75000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>200908.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>자가</td>\n",
       "      <td>100000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "      <td>200000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968751</th>\n",
       "      <td>1482466</td>\n",
       "      <td>766546</td>\n",
       "      <td>1975.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-03-22 22:12:35</td>\n",
       "      <td>640.0</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>201705.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>자가</td>\n",
       "      <td>10000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.0</td>\n",
       "      <td>117000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968752</th>\n",
       "      <td>816537</td>\n",
       "      <td>3864</td>\n",
       "      <td>1977.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2022-03-22 08:55:14</td>\n",
       "      <td>NaN</td>\n",
       "      <td>35000000.0</td>\n",
       "      <td>FREELANCER</td>\n",
       "      <td>201103.0</td>\n",
       "      <td>기타</td>\n",
       "      <td>자가</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>생활비</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>968753 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        application_id  user_id  birth_year  gender          insert_time  \\\n",
       "0               954900   553686      1968.0     1.0  2022-06-07 14:29:03   \n",
       "1               137274    59516      1997.0     1.0  2022-06-07 21:40:22   \n",
       "2              1570936   167320      1989.0     1.0  2022-06-07 09:40:27   \n",
       "3               967833    33400      2000.0     1.0  2022-06-07 08:55:07   \n",
       "4              1614045   860410      1976.0     1.0  2022-06-07 11:29:05   \n",
       "...                ...      ...         ...     ...                  ...   \n",
       "968748         1864587   489900      2000.0     1.0  2022-03-22 14:55:32   \n",
       "968749         1327066   151422      1955.0     1.0  2022-03-22 01:19:24   \n",
       "968750         1319606   173524      1983.0     1.0  2022-03-22 07:34:32   \n",
       "968751         1482466   766546      1975.0     1.0  2022-03-22 22:12:35   \n",
       "968752          816537     3864      1977.0     0.0  2022-03-22 08:55:14   \n",
       "\n",
       "        credit_score  yearly_income      income_type  company_enter_month  \\\n",
       "0              870.0     30000000.0  PRIVATEBUSINESS           20070201.0   \n",
       "1              710.0     30000000.0       FREELANCER           20210901.0   \n",
       "2              820.0     62000000.0     EARNEDINCOME           20170101.0   \n",
       "3              630.0     36000000.0     EARNEDINCOME           20210901.0   \n",
       "4              590.0     45000000.0      OTHERINCOME                  NaN   \n",
       "...              ...            ...              ...                  ...   \n",
       "968748         590.0     25000000.0       FREELANCER             202106.0   \n",
       "968749         980.0     20000000.0      OTHERINCOME                  NaN   \n",
       "968750         750.0     75000000.0     EARNEDINCOME             200908.0   \n",
       "968751         640.0     50000000.0     EARNEDINCOME             201705.0   \n",
       "968752           NaN     35000000.0       FREELANCER             201103.0   \n",
       "\n",
       "       employment_type houseown_type  desired_amount purpose  \\\n",
       "0                  정규직        기타가족소유      30000000.0    대환대출   \n",
       "1                   기타        기타가족소유      10000000.0     생활비   \n",
       "2                  정규직            자가       2000000.0     생활비   \n",
       "3                  정규직        기타가족소유       5000000.0     생활비   \n",
       "4                   기타            자가      15000000.0    대환대출   \n",
       "...                ...           ...             ...     ...   \n",
       "968748              기타        기타가족소유       5000000.0    사업자금   \n",
       "968749              기타            자가      50000000.0     생활비   \n",
       "968750             정규직            자가     100000000.0    대환대출   \n",
       "968751             정규직            자가      10000000.0    대환대출   \n",
       "968752              기타            자가      20000000.0     생활비   \n",
       "\n",
       "        personal_rehabilitation_yn  personal_rehabilitation_complete_yn  \\\n",
       "0                              0.0                                  NaN   \n",
       "1                              0.0                                  NaN   \n",
       "2                              0.0                                  NaN   \n",
       "3                              0.0                                  0.0   \n",
       "4                              0.0                                  NaN   \n",
       "...                            ...                                  ...   \n",
       "968748                         NaN                                  NaN   \n",
       "968749                         NaN                                  NaN   \n",
       "968750                         NaN                                  NaN   \n",
       "968751                         NaN                                  NaN   \n",
       "968752                         NaN                                  NaN   \n",
       "\n",
       "        existing_loan_cnt  existing_loan_amt  \n",
       "0                     1.0         27000000.0  \n",
       "1                     5.0         15000000.0  \n",
       "2                     7.0        344000000.0  \n",
       "3                     1.0         16000000.0  \n",
       "4                     1.0         20000000.0  \n",
       "...                   ...                ...  \n",
       "968748                NaN                NaN  \n",
       "968749                1.0                NaN  \n",
       "968750                8.0        200000000.0  \n",
       "968751               10.0        117000000.0  \n",
       "968752                NaN                NaN  \n",
       "\n",
       "[968753 rows x 17 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6bb3a579",
   "metadata": {},
   "source": [
    "## User_id"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "e204b45e",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_id = user_spec.user_id.unique().tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "500fa58d",
   "metadata": {},
   "outputs": [],
   "source": [
    "log_data = log_data[log_data['user_id'].isin(user_id)].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "1401407a",
   "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>user_id</th>\n",
       "      <th>event</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>mp_os</th>\n",
       "      <th>mp_app_version</th>\n",
       "      <th>date_cd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>576409</td>\n",
       "      <td>StartLoanApply</td>\n",
       "      <td>2022-03-25 11:12:09</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.8.2</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>576409</td>\n",
       "      <td>ViewLoanApplyIntro</td>\n",
       "      <td>2022-03-25 11:12:09</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.8.2</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>72878</td>\n",
       "      <td>EndLoanApply</td>\n",
       "      <td>2022-03-25 11:14:44</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.8.4</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>645317</td>\n",
       "      <td>OpenApp</td>\n",
       "      <td>2022-03-25 11:15:09</td>\n",
       "      <td>iOS</td>\n",
       "      <td>3.6.1</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>645317</td>\n",
       "      <td>UseLoanManage</td>\n",
       "      <td>2022-03-25 11:15:11</td>\n",
       "      <td>iOS</td>\n",
       "      <td>3.6.1</td>\n",
       "      <td>2022-03-25</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14344017</th>\n",
       "      <td>242442</td>\n",
       "      <td>StartLoanApply</td>\n",
       "      <td>2022-05-27 16:08:21</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.13.0</td>\n",
       "      <td>2022-05-27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14344018</th>\n",
       "      <td>242442</td>\n",
       "      <td>EndLoanApply</td>\n",
       "      <td>2022-05-27 16:08:21</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.13.0</td>\n",
       "      <td>2022-05-27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14344019</th>\n",
       "      <td>593062</td>\n",
       "      <td>OpenApp</td>\n",
       "      <td>2022-05-27 16:19:10</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.13.0</td>\n",
       "      <td>2022-05-27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14344020</th>\n",
       "      <td>593062</td>\n",
       "      <td>Login</td>\n",
       "      <td>2022-05-27 16:19:15</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.13.0</td>\n",
       "      <td>2022-05-27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14344021</th>\n",
       "      <td>593062</td>\n",
       "      <td>UseLoanManage</td>\n",
       "      <td>2022-05-27 16:19:19</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.13.0</td>\n",
       "      <td>2022-05-27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>14344022 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          user_id               event            timestamp    mp_os  \\\n",
       "0          576409      StartLoanApply  2022-03-25 11:12:09  Android   \n",
       "1          576409  ViewLoanApplyIntro  2022-03-25 11:12:09  Android   \n",
       "2           72878        EndLoanApply  2022-03-25 11:14:44  Android   \n",
       "3          645317             OpenApp  2022-03-25 11:15:09      iOS   \n",
       "4          645317       UseLoanManage  2022-03-25 11:15:11      iOS   \n",
       "...           ...                 ...                  ...      ...   \n",
       "14344017   242442      StartLoanApply  2022-05-27 16:08:21  Android   \n",
       "14344018   242442        EndLoanApply  2022-05-27 16:08:21  Android   \n",
       "14344019   593062             OpenApp  2022-05-27 16:19:10  Android   \n",
       "14344020   593062               Login  2022-05-27 16:19:15  Android   \n",
       "14344021   593062       UseLoanManage  2022-05-27 16:19:19  Android   \n",
       "\n",
       "         mp_app_version     date_cd  \n",
       "0                 3.8.2  2022-03-25  \n",
       "1                 3.8.2  2022-03-25  \n",
       "2                 3.8.4  2022-03-25  \n",
       "3                 3.6.1  2022-03-25  \n",
       "4                 3.6.1  2022-03-25  \n",
       "...                 ...         ...  \n",
       "14344017         3.13.0  2022-05-27  \n",
       "14344018         3.13.0  2022-05-27  \n",
       "14344019         3.13.0  2022-05-27  \n",
       "14344020         3.13.0  2022-05-27  \n",
       "14344021         3.13.0  2022-05-27  \n",
       "\n",
       "[14344022 rows x 6 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fa632f4",
   "metadata": {},
   "source": [
    "## Data split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "17e1ee6d",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec['month'] = user_spec['insert_time'].apply(lambda x: int(x[5:7]))\n",
    "loan_result['month'] = loan_result['loanapply_insert_time'].apply(lambda x: int(x[5:7]))\n",
    "log_data['month'] = log_data['timestamp'].apply(lambda x: int(x[5:7]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "bebd9532",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec_train = user_spec[user_spec['month'] <6].reset_index(drop=True)\n",
    "loan_result_train = loan_result[loan_result['month'] <6].reset_index(drop=True)\n",
    "log_data_train = log_data[log_data['month'] <6].reset_index(drop=True)\n",
    "\n",
    "user_spec_test = user_spec[user_spec['month'] ==6].reset_index(drop=True)\n",
    "loan_result_test = loan_result[loan_result['month'] ==6].reset_index(drop=True)\n",
    "log_data_test = log_data[log_data['month'] ==6].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "d8b1d04d",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec_train = user_spec_train.drop(['month'],axis=1)\n",
    "loan_result_train = loan_result_train.drop(['month'],axis=1)\n",
    "log_data_train = log_data_train.drop(['month'],axis=1)\n",
    "\n",
    "user_spec_test = user_spec_test.drop(['month'],axis=1)\n",
    "loan_result_test = loan_result_test.drop(['month'],axis=1)\n",
    "log_data_test= log_data_test.drop(['month'],axis=1)\n",
    "\n",
    "user_spec = user_spec.drop(['month'],axis=1)\n",
    "loan_result = loan_result.drop(['month'],axis=1)\n",
    "log_data = log_data.drop(['month'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "c7983857",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(728884, 17)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>application_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>birth_year</th>\n",
       "      <th>gender</th>\n",
       "      <th>insert_time</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>company_enter_month</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1945260</td>\n",
       "      <td>836762</td>\n",
       "      <td>1979.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-04-20 01:23:09</td>\n",
       "      <td>670.0</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>EARNEDINCOME2</td>\n",
       "      <td>202110.0</td>\n",
       "      <td>기타</td>\n",
       "      <td>전월세</td>\n",
       "      <td>25000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>76000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1019382</td>\n",
       "      <td>186886</td>\n",
       "      <td>1979.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-04-20 00:38:18</td>\n",
       "      <td>730.0</td>\n",
       "      <td>95000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>201803.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>전월세</td>\n",
       "      <td>40000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>142000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1117343</td>\n",
       "      <td>594274</td>\n",
       "      <td>1993.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2022-04-20 15:01:01</td>\n",
       "      <td>670.0</td>\n",
       "      <td>35000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>201901.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>전월세</td>\n",
       "      <td>13000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>36000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1355252</td>\n",
       "      <td>180240</td>\n",
       "      <td>1995.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2022-04-20 15:21:40</td>\n",
       "      <td>930.0</td>\n",
       "      <td>31000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>202204.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>전월세</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>주택구입</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>675491</td>\n",
       "      <td>848534</td>\n",
       "      <td>1984.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-04-20 15:44:42</td>\n",
       "      <td>580.0</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>202012.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>전월세</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>71000000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   application_id  user_id  birth_year  gender          insert_time  \\\n",
       "0         1945260   836762      1979.0     1.0  2022-04-20 01:23:09   \n",
       "1         1019382   186886      1979.0     1.0  2022-04-20 00:38:18   \n",
       "2         1117343   594274      1993.0     0.0  2022-04-20 15:01:01   \n",
       "3         1355252   180240      1995.0     0.0  2022-04-20 15:21:40   \n",
       "4          675491   848534      1984.0     1.0  2022-04-20 15:44:42   \n",
       "\n",
       "   credit_score  yearly_income    income_type  company_enter_month  \\\n",
       "0         670.0     50000000.0  EARNEDINCOME2             202110.0   \n",
       "1         730.0     95000000.0   EARNEDINCOME             201803.0   \n",
       "2         670.0     35000000.0   EARNEDINCOME             201901.0   \n",
       "3         930.0     31000000.0   EARNEDINCOME             202204.0   \n",
       "4         580.0     50000000.0   EARNEDINCOME             202012.0   \n",
       "\n",
       "  employment_type houseown_type  desired_amount purpose  \\\n",
       "0              기타           전월세      25000000.0    대환대출   \n",
       "1             정규직           전월세      40000000.0    대환대출   \n",
       "2             정규직           전월세      13000000.0    대환대출   \n",
       "3             정규직           전월세      50000000.0    주택구입   \n",
       "4             정규직           전월세      50000000.0     생활비   \n",
       "\n",
       "   personal_rehabilitation_yn  personal_rehabilitation_complete_yn  \\\n",
       "0                         0.0                                  0.0   \n",
       "1                         0.0                                  NaN   \n",
       "2                         0.0                                  NaN   \n",
       "3                         0.0                                  0.0   \n",
       "4                         NaN                                  NaN   \n",
       "\n",
       "   existing_loan_cnt  existing_loan_amt  \n",
       "0                3.0         76000000.0  \n",
       "1                4.0        142000000.0  \n",
       "2                3.0         36000000.0  \n",
       "3                NaN                NaN  \n",
       "4                5.0         71000000.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(user_spec_train.shape)\n",
    "display(user_spec_train.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "b48b4fab",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(239869, 17)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>application_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>birth_year</th>\n",
       "      <th>gender</th>\n",
       "      <th>insert_time</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>company_enter_month</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>954900</td>\n",
       "      <td>553686</td>\n",
       "      <td>1968.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 14:29:03</td>\n",
       "      <td>870.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>PRIVATEBUSINESS</td>\n",
       "      <td>20070201.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>27000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>137274</td>\n",
       "      <td>59516</td>\n",
       "      <td>1997.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 21:40:22</td>\n",
       "      <td>710.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>FREELANCER</td>\n",
       "      <td>20210901.0</td>\n",
       "      <td>기타</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>10000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>15000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1570936</td>\n",
       "      <td>167320</td>\n",
       "      <td>1989.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 09:40:27</td>\n",
       "      <td>820.0</td>\n",
       "      <td>62000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>20170101.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>자가</td>\n",
       "      <td>2000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.0</td>\n",
       "      <td>344000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>967833</td>\n",
       "      <td>33400</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 08:55:07</td>\n",
       "      <td>630.0</td>\n",
       "      <td>36000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>20210901.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>5000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1614045</td>\n",
       "      <td>860410</td>\n",
       "      <td>1976.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2022-06-07 11:29:05</td>\n",
       "      <td>590.0</td>\n",
       "      <td>45000000.0</td>\n",
       "      <td>OTHERINCOME</td>\n",
       "      <td>NaN</td>\n",
       "      <td>기타</td>\n",
       "      <td>자가</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20000000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   application_id  user_id  birth_year  gender          insert_time  \\\n",
       "0          954900   553686      1968.0     1.0  2022-06-07 14:29:03   \n",
       "1          137274    59516      1997.0     1.0  2022-06-07 21:40:22   \n",
       "2         1570936   167320      1989.0     1.0  2022-06-07 09:40:27   \n",
       "3          967833    33400      2000.0     1.0  2022-06-07 08:55:07   \n",
       "4         1614045   860410      1976.0     1.0  2022-06-07 11:29:05   \n",
       "\n",
       "   credit_score  yearly_income      income_type  company_enter_month  \\\n",
       "0         870.0     30000000.0  PRIVATEBUSINESS           20070201.0   \n",
       "1         710.0     30000000.0       FREELANCER           20210901.0   \n",
       "2         820.0     62000000.0     EARNEDINCOME           20170101.0   \n",
       "3         630.0     36000000.0     EARNEDINCOME           20210901.0   \n",
       "4         590.0     45000000.0      OTHERINCOME                  NaN   \n",
       "\n",
       "  employment_type houseown_type  desired_amount purpose  \\\n",
       "0             정규직        기타가족소유      30000000.0    대환대출   \n",
       "1              기타        기타가족소유      10000000.0     생활비   \n",
       "2             정규직            자가       2000000.0     생활비   \n",
       "3             정규직        기타가족소유       5000000.0     생활비   \n",
       "4              기타            자가      15000000.0    대환대출   \n",
       "\n",
       "   personal_rehabilitation_yn  personal_rehabilitation_complete_yn  \\\n",
       "0                         0.0                                  NaN   \n",
       "1                         0.0                                  NaN   \n",
       "2                         0.0                                  NaN   \n",
       "3                         0.0                                  0.0   \n",
       "4                         0.0                                  NaN   \n",
       "\n",
       "   existing_loan_cnt  existing_loan_amt  \n",
       "0                1.0         27000000.0  \n",
       "1                5.0         15000000.0  \n",
       "2                7.0        344000000.0  \n",
       "3                1.0         16000000.0  \n",
       "4                1.0         20000000.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(user_spec_test.shape)\n",
    "display(user_spec_test.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "69b7fdac",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10270124, 7)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>application_id</th>\n",
       "      <th>loanapply_insert_time</th>\n",
       "      <th>bank_id</th>\n",
       "      <th>product_id</th>\n",
       "      <th>loan_limit</th>\n",
       "      <th>loan_rate</th>\n",
       "      <th>is_applied</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2157865</td>\n",
       "      <td>2022-05-09 08:44:59</td>\n",
       "      <td>54</td>\n",
       "      <td>235</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>16.5</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>576643</td>\n",
       "      <td>2022-05-09 10:54:53</td>\n",
       "      <td>54</td>\n",
       "      <td>235</td>\n",
       "      <td>11000000.0</td>\n",
       "      <td>16.5</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>576643</td>\n",
       "      <td>2022-05-09 10:54:53</td>\n",
       "      <td>11</td>\n",
       "      <td>118</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2136706</td>\n",
       "      <td>2022-05-09 10:41:06</td>\n",
       "      <td>42</td>\n",
       "      <td>216</td>\n",
       "      <td>10000000.0</td>\n",
       "      <td>13.5</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2136706</td>\n",
       "      <td>2022-05-09 10:41:07</td>\n",
       "      <td>25</td>\n",
       "      <td>169</td>\n",
       "      <td>22000000.0</td>\n",
       "      <td>15.9</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   application_id loanapply_insert_time  bank_id  product_id  loan_limit  \\\n",
       "0         2157865   2022-05-09 08:44:59       54         235  20000000.0   \n",
       "1          576643   2022-05-09 10:54:53       54         235  11000000.0   \n",
       "2          576643   2022-05-09 10:54:53       11         118   3000000.0   \n",
       "3         2136706   2022-05-09 10:41:06       42         216  10000000.0   \n",
       "4         2136706   2022-05-09 10:41:07       25         169  22000000.0   \n",
       "\n",
       "   loan_rate  is_applied  \n",
       "0       16.5         1.0  \n",
       "1       16.5         0.0  \n",
       "2       20.0         0.0  \n",
       "3       13.5         0.0  \n",
       "4       15.9         0.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(loan_result_train.shape)\n",
    "display(loan_result_train.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "1ed22b22",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3257239, 7)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>application_id</th>\n",
       "      <th>loanapply_insert_time</th>\n",
       "      <th>bank_id</th>\n",
       "      <th>product_id</th>\n",
       "      <th>loan_limit</th>\n",
       "      <th>loan_rate</th>\n",
       "      <th>is_applied</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1748340</td>\n",
       "      <td>2022-06-07 13:05:41</td>\n",
       "      <td>7</td>\n",
       "      <td>191</td>\n",
       "      <td>42000000.0</td>\n",
       "      <td>13.6</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1748340</td>\n",
       "      <td>2022-06-07 13:05:41</td>\n",
       "      <td>25</td>\n",
       "      <td>169</td>\n",
       "      <td>24000000.0</td>\n",
       "      <td>17.9</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1748340</td>\n",
       "      <td>2022-06-07 13:05:41</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>24000000.0</td>\n",
       "      <td>18.5</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1748340</td>\n",
       "      <td>2022-06-07 13:05:41</td>\n",
       "      <td>4</td>\n",
       "      <td>268</td>\n",
       "      <td>29000000.0</td>\n",
       "      <td>10.8</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1748340</td>\n",
       "      <td>2022-06-07 13:05:41</td>\n",
       "      <td>11</td>\n",
       "      <td>118</td>\n",
       "      <td>5000000.0</td>\n",
       "      <td>16.4</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   application_id loanapply_insert_time  bank_id  product_id  loan_limit  \\\n",
       "0         1748340   2022-06-07 13:05:41        7         191  42000000.0   \n",
       "1         1748340   2022-06-07 13:05:41       25         169  24000000.0   \n",
       "2         1748340   2022-06-07 13:05:41        2           7  24000000.0   \n",
       "3         1748340   2022-06-07 13:05:41        4         268  29000000.0   \n",
       "4         1748340   2022-06-07 13:05:41       11         118   5000000.0   \n",
       "\n",
       "   loan_rate  is_applied  \n",
       "0       13.6         NaN  \n",
       "1       17.9         NaN  \n",
       "2       18.5         NaN  \n",
       "3       10.8         NaN  \n",
       "4       16.4         NaN  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(loan_result_test.shape)\n",
    "display(loan_result_test.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "a62ef936",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10378704, 6)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_id</th>\n",
       "      <th>event</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>mp_os</th>\n",
       "      <th>mp_app_version</th>\n",
       "      <th>date_cd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>576409</td>\n",
       "      <td>StartLoanApply</td>\n",
       "      <td>2022-03-25 11:12:09</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.8.2</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>576409</td>\n",
       "      <td>ViewLoanApplyIntro</td>\n",
       "      <td>2022-03-25 11:12:09</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.8.2</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>72878</td>\n",
       "      <td>EndLoanApply</td>\n",
       "      <td>2022-03-25 11:14:44</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.8.4</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>645317</td>\n",
       "      <td>OpenApp</td>\n",
       "      <td>2022-03-25 11:15:09</td>\n",
       "      <td>iOS</td>\n",
       "      <td>3.6.1</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>645317</td>\n",
       "      <td>UseLoanManage</td>\n",
       "      <td>2022-03-25 11:15:11</td>\n",
       "      <td>iOS</td>\n",
       "      <td>3.6.1</td>\n",
       "      <td>2022-03-25</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id               event            timestamp    mp_os mp_app_version  \\\n",
       "0   576409      StartLoanApply  2022-03-25 11:12:09  Android          3.8.2   \n",
       "1   576409  ViewLoanApplyIntro  2022-03-25 11:12:09  Android          3.8.2   \n",
       "2    72878        EndLoanApply  2022-03-25 11:14:44  Android          3.8.4   \n",
       "3   645317             OpenApp  2022-03-25 11:15:09      iOS          3.6.1   \n",
       "4   645317       UseLoanManage  2022-03-25 11:15:11      iOS          3.6.1   \n",
       "\n",
       "      date_cd  \n",
       "0  2022-03-25  \n",
       "1  2022-03-25  \n",
       "2  2022-03-25  \n",
       "3  2022-03-25  \n",
       "4  2022-03-25  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(log_data_train.shape)\n",
    "display(log_data_train.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "d3e0e59a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3965318, 6)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_id</th>\n",
       "      <th>event</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>mp_os</th>\n",
       "      <th>mp_app_version</th>\n",
       "      <th>date_cd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>17516</td>\n",
       "      <td>EndLoanApply</td>\n",
       "      <td>2022-06-26 11:06:49</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.15.0</td>\n",
       "      <td>2022-06-26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>195705</td>\n",
       "      <td>OpenApp</td>\n",
       "      <td>2022-06-26 11:29:27</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.12.1</td>\n",
       "      <td>2022-06-26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>195705</td>\n",
       "      <td>Login</td>\n",
       "      <td>2022-06-26 11:29:34</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.12.1</td>\n",
       "      <td>2022-06-26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>743379</td>\n",
       "      <td>OpenApp</td>\n",
       "      <td>2022-06-26 11:33:09</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.15.1</td>\n",
       "      <td>2022-06-26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>855546</td>\n",
       "      <td>OpenApp</td>\n",
       "      <td>2022-06-26 11:34:23</td>\n",
       "      <td>Android</td>\n",
       "      <td>3.14.0</td>\n",
       "      <td>2022-06-26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id         event            timestamp    mp_os mp_app_version  \\\n",
       "0    17516  EndLoanApply  2022-06-26 11:06:49  Android         3.15.0   \n",
       "1   195705       OpenApp  2022-06-26 11:29:27  Android         3.12.1   \n",
       "2   195705         Login  2022-06-26 11:29:34  Android         3.12.1   \n",
       "3   743379       OpenApp  2022-06-26 11:33:09  Android         3.15.1   \n",
       "4   855546       OpenApp  2022-06-26 11:34:23  Android         3.14.0   \n",
       "\n",
       "      date_cd  \n",
       "0  2022-06-26  \n",
       "1  2022-06-26  \n",
       "2  2022-06-26  \n",
       "3  2022-06-26  \n",
       "4  2022-06-26  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(log_data_test.shape)\n",
    "display(log_data_test.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8afeb9e1",
   "metadata": {},
   "source": [
    "## Outlier"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0dbc211f",
   "metadata": {},
   "source": [
    "### Credit Score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "1835c7c8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'credit_score', data = user_spec)\n",
    "plt.title('credit_score', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e04edfa9",
   "metadata": {},
   "source": [
    "- 아무것도 안하고 넘어가기"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "279489d2",
   "metadata": {},
   "source": [
    "### yearly_income"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "0148741c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'yearly_income', data = user_spec)\n",
    "plt.title('yearly_income', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e200995",
   "metadata": {},
   "source": [
    "- 아무것도 안하고 넘어가기"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "67b52980",
   "metadata": {},
   "source": [
    "### company_enter_month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "4160c068",
   "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>user_id</th>\n",
       "      <th>enter_year</th>\n",
       "      <th>enter_month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>553686</td>\n",
       "      <td>2007.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>59516</td>\n",
       "      <td>2021.0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>167320</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>33400</td>\n",
       "      <td>2021.0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>860410</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id  enter_year  enter_month\n",
       "0   553686      2007.0          2.0\n",
       "1    59516      2021.0          9.0\n",
       "2   167320      2017.0          1.0\n",
       "3    33400      2021.0          9.0\n",
       "4   860410         NaN          NaN"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## 근속연수\n",
    "def company_enter_year(x):\n",
    "    if x > 10000000:\n",
    "        return x // 10000\n",
    "    elif x > 100000:\n",
    "        return x // 100\n",
    "\n",
    "def company_enter_month(x):\n",
    "    if x > 10000000:\n",
    "        return (x % 10000) // 100\n",
    "    elif x > 100000:\n",
    "        return x % 100\n",
    "    \n",
    "user_spec['enter_year'] = user_spec['company_enter_month'].map(company_enter_year)\n",
    "user_spec['enter_month'] = user_spec['company_enter_month'].map(company_enter_month)\n",
    "user_spec_train['enter_year'] = user_spec_train['company_enter_month'].map(company_enter_year)\n",
    "user_spec_train['enter_month'] = user_spec_train['company_enter_month'].map(company_enter_month)\n",
    "user_spec[['user_id', 'enter_year', 'enter_month']].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "53113065",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'enter_year', data = user_spec)\n",
    "plt.title('enter_year', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b3c3f151",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec['enter_birth'] = pd.to_numeric(user_spec['enter_year']) - user_spec['birth_year']\n",
    "user_spec_train['enter_birth'] = pd.to_numeric(user_spec_train['enter_year']) - user_spec_train['birth_year']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "6ad09a8e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'enter_birth', data = user_spec)\n",
    "plt.title('입사할때 나이', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "74fa52db",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.loc[user_spec['enter_birth']<15,'enter_year'] = user_spec.loc[user_spec['enter_birth']<15,'birth_year']+user_spec_train['enter_birth'].median() \n",
    "user_spec.loc[user_spec['enter_year']>2022,'enter_year'] = 2021\n",
    "user_spec.loc[user_spec['enter_birth']<15,'enter_birth'] = 15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "f02fe594",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'enter_year', data = user_spec)\n",
    "plt.title('enter_year', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "c32adfc9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'enter_birth', data = user_spec)\n",
    "plt.title('입사할때 나이', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60f511a6",
   "metadata": {},
   "source": [
    "### desired_amount"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "fafde29d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'desired_amount', data = user_spec)\n",
    "plt.title('desired_amount', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4da43242",
   "metadata": {},
   "source": [
    "- 아무것도 안하고 넘어가기"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52b042a1",
   "metadata": {},
   "source": [
    "### existing_loan_cnt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "2bf48e63",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'existing_loan_cnt', data = user_spec)\n",
    "plt.title('existing_loan_cnt', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "6b4ec04b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.500601242852262"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec['existing_loan_cnt'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "7592b4c3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "229.0"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec['existing_loan_cnt'].max()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe8cd0d2",
   "metadata": {},
   "source": [
    "- 기대출수 150개 이상 -> 냅두기"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b481f28",
   "metadata": {},
   "source": [
    "### existing_loan_amt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "6c1a4290",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'existing_loan_amt', data = user_spec)\n",
    "plt.title('existing_loan_amt', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3cf02c2",
   "metadata": {},
   "source": [
    "- 아무것도 안하기로 함"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fbfb0d7a",
   "metadata": {},
   "source": [
    "## Missing_value"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72f7e7ee",
   "metadata": {},
   "source": [
    "### Data Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "8216b764",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec['month'] = user_spec['insert_time'].apply(lambda x: int(x[5:7]))\n",
    "loan_result['month'] = loan_result['loanapply_insert_time'].apply(lambda x: int(x[5:7]))\n",
    "log_data['month'] = log_data['timestamp'].apply(lambda x: int(x[5:7]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "ee770540",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec_train = user_spec[user_spec['month'] <6].reset_index(drop=True)\n",
    "loan_result_train = loan_result[loan_result['month'] <6].reset_index(drop=True)\n",
    "log_data_train = log_data[log_data['month'] <6].reset_index(drop=True)\n",
    "\n",
    "user_spec_test = user_spec[user_spec['month'] ==6].reset_index(drop=True)\n",
    "loan_result_test = loan_result[loan_result['month'] ==6].reset_index(drop=True)\n",
    "log_data_test = log_data[log_data['month'] ==6].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac7e2a50",
   "metadata": {},
   "source": [
    "### Missing Value Check"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "ce8605bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "birth_year                               8593\n",
       "gender                                   8593\n",
       "insert_time                                 0\n",
       "credit_score                            81769\n",
       "yearly_income                               1\n",
       "income_type                                 0\n",
       "company_enter_month                     92314\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn             417763\n",
       "personal_rehabilitation_complete_yn    843993\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                              92314\n",
       "enter_month                             92314\n",
       "enter_birth                             99428\n",
       "month                                       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "0c98db24",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                 0\n",
       "loanapply_insert_time          0\n",
       "bank_id                        0\n",
       "product_id                     0\n",
       "loan_limit                  7495\n",
       "loan_rate                   7495\n",
       "is_applied               3257239\n",
       "month                          0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_result.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "9a18d08c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_id                0\n",
       "event                  0\n",
       "timestamp              0\n",
       "mp_os                377\n",
       "mp_app_version    550049\n",
       "date_cd                0\n",
       "month                  0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_data.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ca0573f",
   "metadata": {},
   "source": [
    "### Birth_year & gender"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "f44d2078",
   "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>user_id</th>\n",
       "      <th>birth_year</th>\n",
       "      <th>gender</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9</td>\n",
       "      <td>1996.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11</td>\n",
       "      <td>1997.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>14</td>\n",
       "      <td>1965.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>17</td>\n",
       "      <td>1993.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>19</td>\n",
       "      <td>1987.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312508</th>\n",
       "      <td>879692</td>\n",
       "      <td>1992.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312509</th>\n",
       "      <td>879693</td>\n",
       "      <td>2001.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312510</th>\n",
       "      <td>879695</td>\n",
       "      <td>1996.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312511</th>\n",
       "      <td>879696</td>\n",
       "      <td>1997.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312512</th>\n",
       "      <td>879697</td>\n",
       "      <td>1962.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>312513 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        user_id  birth_year  gender\n",
       "0             9      1996.0     1.0\n",
       "1            11      1997.0     1.0\n",
       "2            14      1965.0     0.0\n",
       "3            17      1993.0     1.0\n",
       "4            19      1987.0     0.0\n",
       "...         ...         ...     ...\n",
       "312508   879692      1992.0     1.0\n",
       "312509   879693      2001.0     0.0\n",
       "312510   879695      1996.0     1.0\n",
       "312511   879696      1997.0     1.0\n",
       "312512   879697      1962.0     1.0\n",
       "\n",
       "[312513 rows x 3 columns]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "birth_gender = user_spec.groupby(['user_id','birth_year','gender'],as_index=False).count().reset_index(drop=True).iloc[:,:3]\n",
    "birth_gender"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "f6e66c41",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec = user_spec.drop(['birth_year','gender'],axis=1)\n",
    "user_spec = pd.merge(user_spec,birth_gender,how='left',on='user_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "b9b7833b",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>application_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>insert_time</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>company_enter_month</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "      <th>enter_year</th>\n",
       "      <th>enter_month</th>\n",
       "      <th>enter_birth</th>\n",
       "      <th>month</th>\n",
       "      <th>birth_year</th>\n",
       "      <th>gender</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>954900</td>\n",
       "      <td>553686</td>\n",
       "      <td>2022-06-07 14:29:03</td>\n",
       "      <td>870.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>PRIVATEBUSINESS</td>\n",
       "      <td>20070201.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>27000000.0</td>\n",
       "      <td>2007.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1968.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>137274</td>\n",
       "      <td>59516</td>\n",
       "      <td>2022-06-07 21:40:22</td>\n",
       "      <td>710.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>FREELANCER</td>\n",
       "      <td>20210901.0</td>\n",
       "      <td>기타</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>10000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>2021.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1997.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1570936</td>\n",
       "      <td>167320</td>\n",
       "      <td>2022-06-07 09:40:27</td>\n",
       "      <td>820.0</td>\n",
       "      <td>62000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>20170101.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>자가</td>\n",
       "      <td>2000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.0</td>\n",
       "      <td>344000000.0</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>6</td>\n",
       "      <td>1989.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>967833</td>\n",
       "      <td>33400</td>\n",
       "      <td>2022-06-07 08:55:07</td>\n",
       "      <td>630.0</td>\n",
       "      <td>36000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>20210901.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>5000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16000000.0</td>\n",
       "      <td>2021.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>6</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1614045</td>\n",
       "      <td>860410</td>\n",
       "      <td>2022-06-07 11:29:05</td>\n",
       "      <td>590.0</td>\n",
       "      <td>45000000.0</td>\n",
       "      <td>OTHERINCOME</td>\n",
       "      <td>NaN</td>\n",
       "      <td>기타</td>\n",
       "      <td>자가</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "      <td>1976.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968748</th>\n",
       "      <td>1864587</td>\n",
       "      <td>489900</td>\n",
       "      <td>2022-03-22 14:55:32</td>\n",
       "      <td>590.0</td>\n",
       "      <td>25000000.0</td>\n",
       "      <td>FREELANCER</td>\n",
       "      <td>202106.0</td>\n",
       "      <td>기타</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>5000000.0</td>\n",
       "      <td>사업자금</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2021.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>3</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968749</th>\n",
       "      <td>1327066</td>\n",
       "      <td>151422</td>\n",
       "      <td>2022-03-22 01:19:24</td>\n",
       "      <td>980.0</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>OTHERINCOME</td>\n",
       "      <td>NaN</td>\n",
       "      <td>기타</td>\n",
       "      <td>자가</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3</td>\n",
       "      <td>1955.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968750</th>\n",
       "      <td>1319606</td>\n",
       "      <td>173524</td>\n",
       "      <td>2022-03-22 07:34:32</td>\n",
       "      <td>750.0</td>\n",
       "      <td>75000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>200908.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>자가</td>\n",
       "      <td>100000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.0</td>\n",
       "      <td>200000000.0</td>\n",
       "      <td>2009.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1983.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968751</th>\n",
       "      <td>1482466</td>\n",
       "      <td>766546</td>\n",
       "      <td>2022-03-22 22:12:35</td>\n",
       "      <td>640.0</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>EARNEDINCOME</td>\n",
       "      <td>201705.0</td>\n",
       "      <td>정규직</td>\n",
       "      <td>자가</td>\n",
       "      <td>10000000.0</td>\n",
       "      <td>대환대출</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.0</td>\n",
       "      <td>117000000.0</td>\n",
       "      <td>2017.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>42.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1975.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968752</th>\n",
       "      <td>816537</td>\n",
       "      <td>3864</td>\n",
       "      <td>2022-03-22 08:55:14</td>\n",
       "      <td>NaN</td>\n",
       "      <td>35000000.0</td>\n",
       "      <td>FREELANCER</td>\n",
       "      <td>201103.0</td>\n",
       "      <td>기타</td>\n",
       "      <td>자가</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>생활비</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2011.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1977.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>968753 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        application_id  user_id          insert_time  credit_score  \\\n",
       "0               954900   553686  2022-06-07 14:29:03         870.0   \n",
       "1               137274    59516  2022-06-07 21:40:22         710.0   \n",
       "2              1570936   167320  2022-06-07 09:40:27         820.0   \n",
       "3               967833    33400  2022-06-07 08:55:07         630.0   \n",
       "4              1614045   860410  2022-06-07 11:29:05         590.0   \n",
       "...                ...      ...                  ...           ...   \n",
       "968748         1864587   489900  2022-03-22 14:55:32         590.0   \n",
       "968749         1327066   151422  2022-03-22 01:19:24         980.0   \n",
       "968750         1319606   173524  2022-03-22 07:34:32         750.0   \n",
       "968751         1482466   766546  2022-03-22 22:12:35         640.0   \n",
       "968752          816537     3864  2022-03-22 08:55:14           NaN   \n",
       "\n",
       "        yearly_income      income_type  company_enter_month employment_type  \\\n",
       "0          30000000.0  PRIVATEBUSINESS           20070201.0             정규직   \n",
       "1          30000000.0       FREELANCER           20210901.0              기타   \n",
       "2          62000000.0     EARNEDINCOME           20170101.0             정규직   \n",
       "3          36000000.0     EARNEDINCOME           20210901.0             정규직   \n",
       "4          45000000.0      OTHERINCOME                  NaN              기타   \n",
       "...               ...              ...                  ...             ...   \n",
       "968748     25000000.0       FREELANCER             202106.0              기타   \n",
       "968749     20000000.0      OTHERINCOME                  NaN              기타   \n",
       "968750     75000000.0     EARNEDINCOME             200908.0             정규직   \n",
       "968751     50000000.0     EARNEDINCOME             201705.0             정규직   \n",
       "968752     35000000.0       FREELANCER             201103.0              기타   \n",
       "\n",
       "       houseown_type  desired_amount purpose  personal_rehabilitation_yn  \\\n",
       "0             기타가족소유      30000000.0    대환대출                         0.0   \n",
       "1             기타가족소유      10000000.0     생활비                         0.0   \n",
       "2                 자가       2000000.0     생활비                         0.0   \n",
       "3             기타가족소유       5000000.0     생활비                         0.0   \n",
       "4                 자가      15000000.0    대환대출                         0.0   \n",
       "...              ...             ...     ...                         ...   \n",
       "968748        기타가족소유       5000000.0    사업자금                         NaN   \n",
       "968749            자가      50000000.0     생활비                         NaN   \n",
       "968750            자가     100000000.0    대환대출                         NaN   \n",
       "968751            자가      10000000.0    대환대출                         NaN   \n",
       "968752            자가      20000000.0     생활비                         NaN   \n",
       "\n",
       "        personal_rehabilitation_complete_yn  existing_loan_cnt  \\\n",
       "0                                       NaN                1.0   \n",
       "1                                       NaN                5.0   \n",
       "2                                       NaN                7.0   \n",
       "3                                       0.0                1.0   \n",
       "4                                       NaN                1.0   \n",
       "...                                     ...                ...   \n",
       "968748                                  NaN                NaN   \n",
       "968749                                  NaN                1.0   \n",
       "968750                                  NaN                8.0   \n",
       "968751                                  NaN               10.0   \n",
       "968752                                  NaN                NaN   \n",
       "\n",
       "        existing_loan_amt  enter_year  enter_month  enter_birth  month  \\\n",
       "0              27000000.0      2007.0          2.0         39.0      6   \n",
       "1              15000000.0      2021.0          9.0         24.0      6   \n",
       "2             344000000.0      2017.0          1.0         28.0      6   \n",
       "3              16000000.0      2021.0          9.0         21.0      6   \n",
       "4              20000000.0         NaN          NaN          NaN      6   \n",
       "...                   ...         ...          ...          ...    ...   \n",
       "968748                NaN      2021.0          6.0         21.0      3   \n",
       "968749                NaN         NaN          NaN          NaN      3   \n",
       "968750        200000000.0      2009.0          8.0         26.0      3   \n",
       "968751        117000000.0      2017.0          5.0         42.0      3   \n",
       "968752                NaN      2011.0          3.0         34.0      3   \n",
       "\n",
       "        birth_year  gender  \n",
       "0           1968.0     1.0  \n",
       "1           1997.0     1.0  \n",
       "2           1989.0     1.0  \n",
       "3           2000.0     1.0  \n",
       "4           1976.0     1.0  \n",
       "...            ...     ...  \n",
       "968748      2000.0     1.0  \n",
       "968749      1955.0     1.0  \n",
       "968750      1983.0     1.0  \n",
       "968751      1975.0     1.0  \n",
       "968752      1977.0     0.0  \n",
       "\n",
       "[968753 rows x 21 columns]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "5a2c778c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            81769\n",
       "yearly_income                               1\n",
       "income_type                                 0\n",
       "company_enter_month                     92314\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn             417763\n",
       "personal_rehabilitation_complete_yn    843993\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                              92314\n",
       "enter_month                             92314\n",
       "enter_birth                             99428\n",
       "month                                       0\n",
       "birth_year                               6891\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "b83900ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "bg_id = user_spec[user_spec.gender.isna()].application_id.unique().tolist()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10a1f5cc",
   "metadata": {},
   "source": [
    "### Birth_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "759ae90f",
   "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>application_id</th>\n",
       "      <th>loanapply_insert_time</th>\n",
       "      <th>bank_id</th>\n",
       "      <th>product_id</th>\n",
       "      <th>loan_limit</th>\n",
       "      <th>loan_rate</th>\n",
       "      <th>is_applied</th>\n",
       "      <th>month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4260</th>\n",
       "      <td>368880</td>\n",
       "      <td>2022-06-07 14:51:54</td>\n",
       "      <td>1</td>\n",
       "      <td>61</td>\n",
       "      <td>63000000.0</td>\n",
       "      <td>15.4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4261</th>\n",
       "      <td>368880</td>\n",
       "      <td>2022-06-07 14:51:53</td>\n",
       "      <td>30</td>\n",
       "      <td>121</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>14.3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4262</th>\n",
       "      <td>368880</td>\n",
       "      <td>2022-06-07 14:52:15</td>\n",
       "      <td>55</td>\n",
       "      <td>186</td>\n",
       "      <td>24000000.0</td>\n",
       "      <td>11.6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4263</th>\n",
       "      <td>368880</td>\n",
       "      <td>2022-06-07 14:51:54</td>\n",
       "      <td>32</td>\n",
       "      <td>56</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>15.8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4264</th>\n",
       "      <td>368880</td>\n",
       "      <td>2022-06-07 14:51:55</td>\n",
       "      <td>34</td>\n",
       "      <td>243</td>\n",
       "      <td>28000000.0</td>\n",
       "      <td>18.8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</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>13526083</th>\n",
       "      <td>191834</td>\n",
       "      <td>2022-06-03 14:20:53</td>\n",
       "      <td>35</td>\n",
       "      <td>29</td>\n",
       "      <td>2000000.0</td>\n",
       "      <td>19.4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13526084</th>\n",
       "      <td>191834</td>\n",
       "      <td>2022-06-03 14:20:42</td>\n",
       "      <td>59</td>\n",
       "      <td>150</td>\n",
       "      <td>22000000.0</td>\n",
       "      <td>16.4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13526085</th>\n",
       "      <td>191834</td>\n",
       "      <td>2022-06-03 14:20:43</td>\n",
       "      <td>35</td>\n",
       "      <td>168</td>\n",
       "      <td>13000000.0</td>\n",
       "      <td>19.4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13526086</th>\n",
       "      <td>191834</td>\n",
       "      <td>2022-06-03 14:20:44</td>\n",
       "      <td>58</td>\n",
       "      <td>175</td>\n",
       "      <td>16000000.0</td>\n",
       "      <td>13.8</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13526087</th>\n",
       "      <td>191834</td>\n",
       "      <td>2022-06-03 14:20:43</td>\n",
       "      <td>6</td>\n",
       "      <td>36</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>17.7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>108227 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          application_id loanapply_insert_time  bank_id  product_id  \\\n",
       "4260              368880   2022-06-07 14:51:54        1          61   \n",
       "4261              368880   2022-06-07 14:51:53       30         121   \n",
       "4262              368880   2022-06-07 14:52:15       55         186   \n",
       "4263              368880   2022-06-07 14:51:54       32          56   \n",
       "4264              368880   2022-06-07 14:51:55       34         243   \n",
       "...                  ...                   ...      ...         ...   \n",
       "13526083          191834   2022-06-03 14:20:53       35          29   \n",
       "13526084          191834   2022-06-03 14:20:42       59         150   \n",
       "13526085          191834   2022-06-03 14:20:43       35         168   \n",
       "13526086          191834   2022-06-03 14:20:44       58         175   \n",
       "13526087          191834   2022-06-03 14:20:43        6          36   \n",
       "\n",
       "          loan_limit  loan_rate  is_applied  month  \n",
       "4260      63000000.0       15.4         NaN      6  \n",
       "4261      20000000.0       14.3         NaN      6  \n",
       "4262      24000000.0       11.6         NaN      6  \n",
       "4263      15000000.0       15.8         NaN      6  \n",
       "4264      28000000.0       18.8         NaN      6  \n",
       "...              ...        ...         ...    ...  \n",
       "13526083   2000000.0       19.4         NaN      6  \n",
       "13526084  22000000.0       16.4         NaN      6  \n",
       "13526085  13000000.0       19.4         NaN      6  \n",
       "13526086  16000000.0       13.8         NaN      6  \n",
       "13526087  15000000.0       17.7         NaN      6  \n",
       "\n",
       "[108227 rows x 8 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_result[loan_result['application_id'].isin(bg_id)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "33fd961c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def age(x):\n",
    "    if x > 2004:\n",
    "        return 10\n",
    "    elif x > 1994:\n",
    "        return 20\n",
    "    elif x > 1984:\n",
    "        return 30\n",
    "    elif x > 1974:\n",
    "        return 40\n",
    "    elif x > 1964:\n",
    "        return 50\n",
    "    elif x > 1954:\n",
    "        return 60\n",
    "    elif x > 1944:\n",
    "        return 70\n",
    "    elif x > 1934:\n",
    "        return 80\n",
    "    elif x > 1924:\n",
    "        return 90\n",
    "    else:\n",
    "        return np.nan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "646317b5",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec_train['age'] = user_spec_train.birth_year.apply(lambda x : age(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "66f160db",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec['age'] = user_spec.birth_year.apply(lambda x : age(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "7c8f2149",
   "metadata": {},
   "outputs": [],
   "source": [
    "lab_df = user_spec_train[['age','credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "951cafc6",
   "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>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>20.0</th>\n",
       "      <td>657.417245</td>\n",
       "      <td>2.895319e+07</td>\n",
       "      <td>2.141487e+07</td>\n",
       "      <td>3.501217</td>\n",
       "      <td>3.260615e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30.0</th>\n",
       "      <td>674.987638</td>\n",
       "      <td>3.995539e+07</td>\n",
       "      <td>3.532712e+07</td>\n",
       "      <td>4.329993</td>\n",
       "      <td>6.769406e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40.0</th>\n",
       "      <td>674.919001</td>\n",
       "      <td>4.938365e+07</td>\n",
       "      <td>3.901918e+07</td>\n",
       "      <td>4.852480</td>\n",
       "      <td>9.400945e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50.0</th>\n",
       "      <td>670.245368</td>\n",
       "      <td>5.958612e+07</td>\n",
       "      <td>3.925237e+07</td>\n",
       "      <td>4.956305</td>\n",
       "      <td>9.173035e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60.0</th>\n",
       "      <td>674.318777</td>\n",
       "      <td>4.708209e+07</td>\n",
       "      <td>3.618451e+07</td>\n",
       "      <td>4.803957</td>\n",
       "      <td>7.125581e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70.0</th>\n",
       "      <td>689.375000</td>\n",
       "      <td>3.824627e+07</td>\n",
       "      <td>2.527701e+07</td>\n",
       "      <td>5.191629</td>\n",
       "      <td>5.918268e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80.0</th>\n",
       "      <td>710.346715</td>\n",
       "      <td>3.342415e+07</td>\n",
       "      <td>2.220177e+07</td>\n",
       "      <td>4.294340</td>\n",
       "      <td>4.427854e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90.0</th>\n",
       "      <td>728.148148</td>\n",
       "      <td>2.213333e+07</td>\n",
       "      <td>8.833333e+06</td>\n",
       "      <td>3.526316</td>\n",
       "      <td>7.882353e+06</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      credit_score  yearly_income  desired_amount  existing_loan_cnt  \\\n",
       "age                                                                    \n",
       "20.0    657.417245   2.895319e+07    2.141487e+07           3.501217   \n",
       "30.0    674.987638   3.995539e+07    3.532712e+07           4.329993   \n",
       "40.0    674.919001   4.938365e+07    3.901918e+07           4.852480   \n",
       "50.0    670.245368   5.958612e+07    3.925237e+07           4.956305   \n",
       "60.0    674.318777   4.708209e+07    3.618451e+07           4.803957   \n",
       "70.0    689.375000   3.824627e+07    2.527701e+07           5.191629   \n",
       "80.0    710.346715   3.342415e+07    2.220177e+07           4.294340   \n",
       "90.0    728.148148   2.213333e+07    8.833333e+06           3.526316   \n",
       "\n",
       "      existing_loan_amt  \n",
       "age                      \n",
       "20.0       3.260615e+07  \n",
       "30.0       6.769406e+07  \n",
       "40.0       9.400945e+07  \n",
       "50.0       9.173035e+07  \n",
       "60.0       7.125581e+07  \n",
       "70.0       5.918268e+07  \n",
       "80.0       4.427854e+07  \n",
       "90.0       7.882353e+06  "
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lab_df1 = lab_df.groupby('age')[['credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']].mean()\n",
    "lab_df1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "fe4c8824",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_id = user_spec[user_spec.birth_year.isna()].user_id.values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "efad1fe5",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_list = ['credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "376d3d0b",
   "metadata": {},
   "outputs": [],
   "source": [
    "age_list=[20,30,40,50,60,70,80,90]\n",
    "def age_feature(x,feature):\n",
    "    diff = []\n",
    "    for i in feature:\n",
    "        diff.append(abs(i-x))\n",
    "    idx = diff.index(np.min(diff))\n",
    "    return age_list[idx]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "37dcb775",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████████████████████████████████████████| 6891/6891 [00:48<00:00, 142.71it/s]\n"
     ]
    }
   ],
   "source": [
    "for idd in tqdm(user_id):\n",
    "    user_birth_list = []\n",
    "    for f in feature_list:\n",
    "        x = np.mean(user_spec[user_spec.user_id == idd][f'{f}'].fillna(0).tolist())\n",
    "        feature = lab_df1[f'{f}'].values.tolist()\n",
    "        user_birth_list.append(age_feature(x,feature))\n",
    "    age = int(np.mean(user_birth_list))\n",
    "    missing_value = 2022-age\n",
    "    user_spec.loc[user_spec.user_id == idd,'birth_year'] = missing_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "fbfc7878",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.drop('age',axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "57312b9c",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            81769\n",
       "yearly_income                               1\n",
       "income_type                                 0\n",
       "company_enter_month                     92314\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn             417763\n",
       "personal_rehabilitation_complete_yn    843993\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                              92314\n",
       "enter_month                             92314\n",
       "enter_birth                             99428\n",
       "month                                       0\n",
       "birth_year                                  0\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "92b56c73",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1982.0    38223\n",
       "1981.0    35571\n",
       "1992.0    35543\n",
       "1983.0    34806\n",
       "1991.0    34454\n",
       "1980.0    34017\n",
       "1993.0    34007\n",
       "1984.0    32231\n",
       "1979.0    31680\n",
       "1990.0    31679\n",
       "1985.0    31678\n",
       "1994.0    31574\n",
       "1995.0    31401\n",
       "1986.0    30244\n",
       "1987.0    30238\n",
       "1989.0    30180\n",
       "1988.0    30073\n",
       "1996.0    26684\n",
       "1978.0    25075\n",
       "1977.0    24654\n",
       "1976.0    23508\n",
       "1997.0    23398\n",
       "1974.0    22630\n",
       "1975.0    22360\n",
       "1973.0    21086\n",
       "1972.0    19142\n",
       "1971.0    18580\n",
       "1998.0    17368\n",
       "1970.0    16871\n",
       "1969.0    15538\n",
       "1999.0    13921\n",
       "1968.0    13742\n",
       "1967.0    11006\n",
       "2000.0    10447\n",
       "1966.0     9723\n",
       "1965.0     8488\n",
       "1964.0     7759\n",
       "1962.0     6673\n",
       "1963.0     6406\n",
       "2001.0     5954\n",
       "1960.0     5918\n",
       "1961.0     5455\n",
       "1959.0     4291\n",
       "1958.0     3732\n",
       "1957.0     2939\n",
       "2002.0     2391\n",
       "1956.0     2362\n",
       "1955.0     2163\n",
       "1954.0     2042\n",
       "1952.0     1479\n",
       "1953.0     1292\n",
       "1946.0      853\n",
       "1950.0      848\n",
       "1951.0      805\n",
       "2003.0      668\n",
       "1948.0      631\n",
       "1949.0      531\n",
       "1947.0      430\n",
       "1945.0      256\n",
       "1944.0      242\n",
       "1942.0      176\n",
       "1943.0      173\n",
       "1941.0       93\n",
       "1940.0       74\n",
       "1938.0       62\n",
       "1939.0       54\n",
       "1936.0       48\n",
       "1935.0       39\n",
       "1937.0       38\n",
       "1934.0       36\n",
       "1932.0       11\n",
       "1933.0        7\n",
       "1930.0        1\n",
       "1931.0        1\n",
       "Name: birth_year, dtype: int64"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.birth_year.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "1cd32879",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/MAAAKUCAYAAABIemj1AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAA9hAAAPYQGoP6dpAABCL0lEQVR4nO3de5TXdb3v8ddPLnKRkIDNprJASTRhvAa1QxRS8YZRmMZNt7pteyrxZEZHqPba5t2im5p5AU3bpFsM2CrtXUqQZVIiB3SxwwsQ3YRzHMy46Fx+5w+PkxOgw8yPmfnK47FWa83v+/n+fvP+slrOPOf7/f2+pXK5XA4AAABQGHu19QAAAADArhHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAUACPPfZYBg8enH//939v0v5nnXVWxo4dm1deeaUi3/++++7L4MGD84tf/KIirwcAtIyYB4C3oD/84Q95/vnnU1NTs8vP3bp1626YCACopI5tPQAAUHkLFixIXV1dunfv3uTn1NfX5/TTT8/gwYNz1VVX7cbpAICWcmYeAN6CunXrlh49euzSc+rr6/PUU0+lXC7vpqkAgEoR8wCwB6uvr2/rEdoVf8gAoCjEPAAUzNq1azN16tQMHz48hx56aM4444w89NBDjfaZMGFCRo8e3Wjb//pf/yuDBw/Otm3bctlll+Woo47KcccdlyQZPXp0DjnkkCTJD3/4wwwePDiDBw/OY489tt33f/TRRzNhwoQcdthhGT58eL7whS/kxRdfbPL8tbW1GTFiRE444YQdrr/yyisZPnx4zj///IZt5XI599xzT8aNG5eqqqq8//3vzyc/+cmsWLFiu+evWLEin//85zNmzJgceuihOe644/Iv//Iv2834+g/1u/vuuzNy5MgcdNBB+d3vftfkYwGAtuI98wBQIGvXrs3Xvva1jBw5Mp/97GdTXV2df//3f8+nP/3pXH311Rk3btybvsbXv/71PPvss/nc5z6XP//5z0mSadOm5S9/+UtmzJiRI444Ih//+MeTJPvvv3+j5/7sZz/Lfffdl9NPPz1jxozJQw89lHnz5uX//J//k9tuu61Jx9CxY8d85CMfya233pqVK1dm6NChjdaXLFmSTZs2NcyQJDNmzMj8+fNz6qmn5uMf/3g2bdqU++67L5MmTcr3vve9HH744Q37Tpo0KQceeGDGjh2b3r17Z8WKFbn77ruzdu3a3HHHHdvNs3jx4ixatCjnnntuNm/enK5duzbpOACgLYl5ACiQ2bNn5+tf/3rGjBnTsG3ChAn5yEc+kiuuuCInnHBCunXr9oavsXr16syePTt77fXXC/ROPPHE1NbWZsaMGXnPe96Tj33sYzt87g9+8IPcd999GThwYJJkypQpmTBhQh555JGsX78+++23X5OOY/z48bn11lvzwAMPbBfz8+fPT58+fTJq1Kgkybx587JgwYLceuut+eAHP9iw32u337v22mszZ86chu3XXXddTjzxxEb/Pp06dcrdd9+d3/zmNxk8eHCj7/fAAw9k3rx56dOnT5NmB4D2wGX2AFAgH/rQhxqFfJLsu++++eQnP5k///nP+fnPf/6mrzFp0qRGIb8rJk2a1BDySdKhQ4d85CMfSZL85je/afLr7L///jniiCOycOHCRu9Tf/HFF/PTn/40H/nIR9KpU6ckyR133JEPfvCDGTRoUDZu3Njwv23btmX48OFZvnx5o9vpvT7kX/OhD30oyatXNvytk046ScgDUDjOzANAgRx11FE73D5kyJAkyZo1a970NV7btzl29Nx3vOMdSZK//OUvu/Ra48ePz4wZM/LrX/8673//+5MkCxcuzCuvvNJwif2WLVuyatWqlMvljBgxYqevtWnTpkaXx69evTq/+tWv8txzz2XdunV59tlnk6ThbQVvdkwA0N6JeQAokLe//e073P7apfU1NTVv+hq9e/du9vd/29vett22vffeO8mufzL+SSedlCuuuCL3339/Q8zPnz8/Rx11VMPZ/5deeinlcjnHHXdczjzzzJ2+Vq9evZIkL7zwQj73uc/lF7/4Rd75znfmwAMPzLvf/e70798/99xzzw6f66w8AEUk5gGgQF555ZUdbv/DH/6QpGmhXiqVKjpTc3Xv3j0nnXRS/vM//zNf+tKX8sc//jHLli3LNddc02if5NXL+UeOHPmmr3nllVfmiSeeyPe+970MHz68Yfsjjzyy05hvL/8eALArvGceAApkZ+9LX7JkSZLksMMOa/Zrvxa1rXnv+fHjx6e6ujq/+MUvMn/+/PTo0aPRe9732WefvPOd78yyZcuybdu2N329Rx55JB/4wAcahXySPP300xWfHQDakpgHgAL54Q9/mKeeeqrRtmeffTb33ntvDj/88Bx00EHNfu0OHTqkR48e+dOf/tTSMZvsyCOPzMCBA7Nw4cL8x3/8R0499dR06dKl0T7jxo3Lxo0b853vfGe75//ud7/L0qVLGx536NAhf/rTnxp9qN7//b//t8m3zQOAonCZPQAUyOGHH54JEybk9NNPz3vf+96sX78+99xzT7p27Zorrriixa9/1FFHZfHixbn22mvzd3/3dxkxYkQGDRpUgcl37mMf+1huvPHGbN26NTNnztxu/ZOf/GR++tOf5qabbsqTTz6ZY489NnvttVdWrFiRhQsX5gtf+EKGDRuWJDnllFNyxx135J/+6Z8yevTobNiwIffee2/+4R/+IQsWLNitxwEArUnMA0CBjB07NmeeeWZuuumm3HPPPenZs2c+/OEP58ILL8y73vWuFr/+l770pbz88suZM2dO9tlnn0b3dd9dPvrRj+Yb3/hG3ve+9+WQQw7Zbr1Lly753ve+l5tvvjkLFy7MNddck06dOuWAAw7IRRddlPHjxzfse8kll6Rr1665//77c8011+Q973lPvvCFL6Rfv35iHoC3lFL59dehAQC0sqeffjqnnnpqvvzlL2fSpEltPQ4AFIIz8wBAxdTV1eWFF154w306deqUfffdt+Hx97///XTt2jWnnXbabp4OAN46xDwAUDF//OMf8+EPf/gN9xk2bFjuvPPOJMnzzz+fefPmZcKECenRo0drjAgAbwkuswcAKmbbtm2NPl1+R/bdd9+sWbMmL774Yr7//e+nvr4+9957b3r27NlKUwJA8Yl5AKDVfepTn8qjjz6a4cOH54tf/GJFPrwPAPYkYh4AAAAKZq+2HgAAAADYNT4A7w088cQTKZfL6dSpU1uPAgAAwB6gpqYmpVIphx9++BvuJ+bfQLlcjnchAAAA0Fqa2qBi/g28dkZ+6NChbTwJAAAAe4KVK1c2aT/vmQcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABRMx7YeAGBP89JLL2Xt2rVtPQbsVlu2bEmSdOvWrY0ngd1rwIAB6dGjR1uPAeyBxDxAK3rppZdy6qmn5qWXXmrrUQCogB49euT+++8X9ECrc5k9AAAAFIwz8wCt6LUzOC6z563s2WefzWWXXZYk+fKXv5wDDjigjSeC3cdl9kBbEfMAraxHjx4ZOnRoW48BreKAAw7w/3cA2A1cZg8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBtIuYf/TRR/OJT3wixx9/fI477rjceeedDWu/+93vcs4552TUqFE5/vjjM3/+/EbPvf/++3PSSSdl5MiRmTJlStavX9+wtm3btnzpS1/KqFGjMnLkyFx77bUpl8utdlwAAACwO7SLmH/ooYdy5ZVX5sc//nFmz56dW265JUuWLEldXV3+x//4Hxk7dmwWLVqU73znO7niiiuyatWqJMkTTzyRmTNn5rbbbsuSJUvyoQ99KBdddFHD61599dWpr6/Pj3/84zzwwAN57LHHctddd7XVYQIAAEBFtIuY/+IXv5j9998/SbLffvvlpJNOyi9/+cs8+uij6dChQz72sY8lSQYNGpTTTjstP/zhD5Mkd911V84+++y84x3vSJL80z/9U37/+9/nv//7v7N58+bMmzcvn//859OxY8f06NEj//zP/5y5c+e2zUECAABAhbSLmP9bL7zwQnr06JHly5fniCOOaLRWVVXV6Mz869c7duyY973vfVm1alWeeuqpvOtd78q+++7bsH7ooYfm6aefTl1dXascBwAAAOwOHdt6gL+1YsWK/PSnP83UqVNzyy23pF+/fo3We/funU2bNiVJNm7cmD59+jRaf/vb355Nmzalc+fO6d2793ZrtbW1eemllxpF/hspl8vZsmVLs48HAPY027Zta/S1n6MA0HTlcjmlUulN92tXMf/AAw/kyiuvzNVXX5399tsvdXV1231gXV1dXcOB7Wi9vr4+pVIptbW1O1xL0qR/mNfU1NQ0XAkAALy5tWvXNvp6r73a5YWAANBude7c+U33aRcxX1dXl6985St57LHHctttt+Wggw5KkvTs2bPhLPxrXnjhhfTt27dhvbq6uuE986+t9+nTJz169Eh1dfV2z917773To0ePJs/WqVOnDBo0qJlHBgB7ntf+eJ4kAwYMyMEHH9yG0wBAsTzzzDNN2q9dxPyVV16Z9evXZ+7cuenWrVvD9iFDhuTWW29ttO8TTzyRww47LElyyCGH5IknnsghhxySJHnllVfy1FNP5fLLL0+XLl2yZs2avPjii+nZs2fDcw899NBdOkNQKpUazQQAvLEuXbo0+trPUQBouqZeSd7m1729/PLLmTNnTq666qrtftiPGjUqGzZsaLi3/MqVK/PQQw/l4x//eJLkzDPPzKxZs/KnP/0pdXV1ufHGGzN8+PDst99+6du3b44++ujMnDkztbW1eeGFF/Kd73wnZ599dqsfIwAAAFRSm5+ZX79+ferr6/OJT3yi0faBAwfmtttuy0033ZQvfelLufrqq9OnT5987Wtfy9///d8nSY4//visW7cup59+esrlcoYNG5Yrr7yy4TWuuOKKzJgxIyNGjEi3bt1y7rnn5rjjjmvV4wMAAIBKK5X/9lPiaLBy5cokydChQ9t4EgAojpUrVzZcCXfHHXf4OQoAu6CpHdrml9kDAAAAu0bMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgOrb1APCaLVu25Nlnn23rMQBoodf/t9x/1wHeGg444IB069atrcfgdcQ87cazzz6bs88+u63HAKCCLrvssrYeAYAKuOOOOzJ06NC2HoPXcZk9AAAAFIwz87RL1ft/IDXd9m3rMQBoplJdTZKk3KFTG08CQHN12rIpvZ77ZVuPwU6Iedqlmm77pqZH37YeAwAAoF1qN5fZl8vlzJs3L2eeeWaj7fPnz8/YsWMzYsSInHHGGVm9enXD2l/+8pd8+ctfzgknnJAPfvCD+fKXv5yampqG9erq6kydOjXHHntsRo0alVmzZrXa8QAAAMDu0i5ifsmSJTnttNNyww035MUXX2zYvnjx4tx44425+eab88gjj2TixImZOnVqyuVykuSLX/xiOnbsmIULF2bRokV5/vnnc9tttzU8f9q0aTnwwAOzaNGi/OAHP8hdd92Vhx9+uNWPDwAAACqpXcT81q1bc8kll+Tyyy9vtH3BggWZMmVK+vfvnyQZN25cunfvnqVLl2bbtm358Y9/nIsvvjgdOnRIly5dcskll+See+5JkqxZsyZPPvlkLrjggpRKpfTr1y9TpkzJ3LlzW/34AAAAoJLaRcyPGTMmxxxzzHbba2pqUltb22hbr169snbt2tTW1qauri719fWN1n7/+9/nlVdeyfLly1NVVZWOHf/6sQBVVVVZtWrV7jsQAAAAaAXt+gPwTjzxxMycOTMjR47MwIEDs2jRojz++OM58sgjs88++2TEiBG57rrrMn369JTL5XzrW99KqVRKdXV1Nm7cmN69ezd6vd69e2fTpk27NEO5XM6WLVsqeFTszLZt29p6BAAAYAe2bdumi1pJuVxOqVR60/3adcyffPLJ2bRpUy666KJs3rw5I0aMyPDhw9OtW7ckyXXXXZdrr702p5xySrp3756zzjord999d7p165ba2tqG99a/pq6urkn/KK9XU1PjbH4rWbt2bVuPAAAA7MDatWuz117t4sLuPULnzp3fdJ92HfNJMnHixEycOLHh8fjx4zNw4MAkr15Wf9VVVzWsPf300+nTp0969OiRnj17ZsWKFY1eq7q6On369Nml79+pU6cMGjSoBUdAU73+LRMAAED7MWDAgBx88MFtPcYe4ZlnnmnSfu0+5l/vueeey7p16zJs2LAdri9YsCCjR49OkgwZMiQ33HBD6uvrG/6CtGzZshx++OG79D1LpVLDlQDsXl26dGnrEQAAgB3o0qWLLmolTb2avF1fJ/Hae9+TZMOGDZkxY0YuvPDChuhbs2ZNwwfkLVmyJPPnz88FF1yQ5NUPu+vbt29uueWW1NfXZ/369ZkzZ04mT57cNgcDAAAAFdKuz8xXV1fn/PPPT11dXbp27ZrJkydn0qRJDesPP/xwZs+enU6dOuXd7353vvvd7+ad73xnklf/mnH99ddn+vTpmT17dnr27Jlp06ZlyJAhbXU4AAAAUBHtKuaHDx+eH/3oRw2P999//zz00EM73f+8887Leeedt9P1/fbbL3feeWdFZwQAAIC21q4vswcAAAC2J+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMB3begDYkY5bNrX1CAAAsEfzO3n7JuZpN7Zs2dLw9duf+2UbTgIAALze639Xp31wmT0AAAAUjDPztBvdunVr+PqF/T+Q2m77tt0wAACwh+u4ZVPDFbOv/12d9kHM0y7Vdts3NT36tvUYAAAA7ZLL7AEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAF025ivlwuZ968eTnzzDMbbZ8/f37Gjh2bESNG5Iwzzsjq1asb1mpqanL55ZfnmGOOyejRo/P5z38+f/7znxvWq6urM3Xq1Bx77LEZNWpUZs2a1WrHAwAAALtLu4j5JUuW5LTTTssNN9yQF198sWH74sWLc+ONN+bmm2/OI488kokTJ2bq1Kkpl8tJkptvvjmrV6/Ogw8+mP/6r/9Kx44dc+WVVzY8f9q0aTnwwAOzaNGi/OAHP8hdd92Vhx9+uNWPDwAAACqpXcT81q1bc8kll+Tyyy9vtH3BggWZMmVK+vfvnyQZN25cunfvnqVLlyZJVq1alRNOOCHdu3dPx44dM3bs2Dz55JNJkjVr1uTJJ5/MBRdckFKplH79+mXKlCmZO3du6x4cAAAAVFjHth4gScaMGZMkeeyxxxptr6mpSW1tbaNtvXr1ytq1azN8+PCMGTMmd955Z0466aR07do1c+bMydixY5Mky5cvT1VVVTp2/OshVlVV5c4779yl2crlcrZs2dKcw2IXbdu2ra1HAAAAdmDbtm26qJWUy+WUSqU33a9dxPzOnHjiiZk5c2ZGjhyZgQMHZtGiRXn88cdz5JFHJklOOeWUPPDAAzn66KPToUOHHHTQQfnqV7+aJNm4cWN69+7d6PV69+6dTZs27dIMNTU1WbVqVUWOhze2du3ath4BAADYgbVr12avvdrFhd17hM6dO7/pPu065k8++eRs2rQpF110UTZv3pwRI0Zk+PDh6datW5Lk6quvbrjsvmPHjvnqV7+aiy++ODfccENqa2sb3lv/mrq6uib9heP1OnXqlEGDBlXsmNi5+vr6th4BAADYgQEDBuTggw9u6zH2CM8880yT9mvXMZ8kEydOzMSJExsejx8/PgMHDszWrVvzb//2b/n5z3+effbZJ0ly6aWX5gMf+EDWrl2bnj17ZsWKFY1eq7q6On369Nml718qlRr+eMDu1aVLl7YeAQAA2IEuXbroolbS1BPQhbpO4rnnnsu6desybNiw1NXVpa6uLh06dGhY32uvvbLXXnulpqYmQ4YMyYoVKxqd7V22bFkOP/zwthgdAAAAKqZdx3x1dXU2btyYJNmwYUNmzJiRCy+8MF26dMk+++yTo48+OjNnzmy4pP6GG27I3/3d32X//fdPVVVV+vbtm1tuuSX19fVZv3595syZk8mTJ7fxUQEAAEDLtOvL7Kurq3P++eenrq4uXbt2zeTJkzNp0qSG9WuvvTbXXnttTjjhhJTL5RxyyCH5zne+03C2/vrrr8/06dMze/bs9OzZM9OmTcuQIUPa6nAAAACgItpVzA8fPjw/+tGPGh7vv//+eeihh3a6/7777psrr7xyp+v77bffLt+KDgAAANq7dn2ZPQAAALA9MQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMB3begDYkU5bNrX1CAC0QKmuJklS7tCpjScBoLn8Tt6+iXnapV7P/bKtRwAAAGi3XGYPAAAABePMPO3GAQcckDvuuKOtxwCghZ599tlcdtllSZIvf/nLOeCAA9p4IgBayn/L2x8xT7vRrVu3DB06tK3HAKCCDjjgAP9tB4DdwGX2AAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYDo294mvvPJKOnfuXLFByuVy5s+fnzlz5uTuu+9u2D5//vzceuutqa6uzjve8Y5cfvnlOfDAA/Piiy/mox/96Hav8Yc//CFz587NkCFDUl1dnX/5l3/JihUrUiqVMmXKlJx77rkVmxkAAADaQrNivlwu57DDDsujjz6anj17tniIJUuW5Lrrrsu2bdvSoUOHhu2LFy/OjTfemNtvvz39+/fPvHnzMnXq1CxcuDA9e/bMww8/3Oh1HnjggXz/+9/PkCFDkiTTpk3LoYcemm9+85vZsGFDJkyYkAEDBmT06NEtnhkAAADaSrMusy+VSjnkkEPy+9//viJDbN26NZdcckkuv/zyRtsXLFiQKVOmpH///kmScePGpXv37lm6dOl2r1FXV5dvfetb+exnP5skWbNmTZ588slccMEFKZVK6devX6ZMmZK5c+dWZGYAAABoK82+zP6f//mf85WvfCUnnXRS3ve+96Vr166N1g855JAmv9aYMWOSJI899lij7TU1NamtrW20rVevXlm7dm2GDx/eaPuDDz6Yfv365f3vf3+SZPny5amqqkrHjn89xKqqqtx5551NngsAAADao2bH/Gc+85kkyRNPPLHdWqlUyqpVq5o/1f934oknZubMmRk5cmQGDhyYRYsW5fHHH8+RRx653b6zZ8/Opz/96YbHGzduTO/evRvt07t372zatGmXZiiXy9myZUuz5geAPdG2bdsafe3nKAA0XblcTqlUetP9mh3z//3f/93cpzbZySefnE2bNuWiiy7K5s2bM2LEiAwfPjzdunVrtN9TTz2VF198MaNGjWrYVltbm3K53Gi/urq6Jv2jvF5NTU1F/jABAHuKtWvXNvp6r73cPAcAdkVTPmy+2THfWiZOnJiJEyc2PB4/fnwGDhzYaJ+5c+fm1FNPbfTLQs+ePbNixYpG+1VXV6dPnz679P07deqUQYMGNWNyANgz1dfXN3w9YMCAHHzwwW04DQAUyzPPPNOk/VoU8w8++GCWL1+el156abu1q666qiUvvUPPPfdc1q1bl2HDhjVsq6urywMPPJBZs2Y12nfIkCG54YYbUl9f3xD5y5Yty+GHH75L37NUKm13JQAAsHNdunRp9LWfowDQdE29mrzZ171dc801+dd//dds2rQp999/f7p27Zrf/e53WbRoUXr06NHcl22kuro6GzduTJJs2LAhM2bMyIUXXtjol4SVK1emXC5v94F7VVVV6du3b2655ZbU19dn/fr1mTNnTiZPnlyR2QAAAKCtNPvM/P3335877rgjBx10UH71q1/l4osvzj777JObb745mzdvrshw1dXVOf/881NXV5euXbtm8uTJmTRpUqN9/vf//t87vHyvVCrl+uuvz/Tp0zN79uz07Nkz06ZNa7gHPQAAABRVs2N+y5Ytede73pUkefe7353169fn4IMPzsSJE/PhD3+44X7vu2L48OH50Y9+1PB4//33z0MPPfSGzzn77LNz9tln73Btv/32cys6AAAA3nKafZn9e9/73qxcuTJJMnTo0DzyyCNJkt/97neVmQwAAADYoWafmT/vvPPys5/9LB/84Afz8Y9/POPHj8+SJUvy9NNPZ+zYsZWcEQAAAHidZsf88ccfn+OPPz5J8p73vCc//OEP89BDD2XChAk56aSTKjYgAAAA0FiL7zNfLpezYcOG7LfffvnHf/zHCowEAAAAvJFmv2f+L3/5Sy699NJUVVVl1KhRDdtvvfXWPP744xUZDgAAANhes2P+2muvzaZNm/LDH/4w9fX1DdsHDBiQ66+/viLDAQAAANtrdswvWrQo//qv/5pBgwalVCo1bB82bFhWrVpVkeEAAACA7TU75jdv3pyuXbtut33Lli3Ztm1bi4YCAAAAdq7ZMX/MMcdk1qxZjbbV19fnW9/6Vo444ogWDwYAAADsWLM/zX7GjBk555xz8utf/zrlcjmf//zns3z58rz88su57bbbKjkjAAAA8DrNjvk+ffpk7ty5efDBB/Pe9743nTp1yvnnn5+xY8fu8PJ7AAAAoDKaHfM33HBDxo8fn3HjxmXcuHEVHAkAAAB4I81+z/yyZcty/PHH57zzzsuDDz6YV155pZJzAQAAADvR7Ji/7bbb8tOf/jTHHnts7rjjjhx99NG57LLL8tRTT1VyPgAAAOBvNPsy+yTp3bt3pkyZkilTpmT9+vX5j//4j1x66aVJkvHjx+e0005Lr169KjIoAAAA8Kpmn5n/W1u2bMnLL7+cl19+ObW1tfn5z3+eD3/4w7nmmmtSU1NTqW8DAAAAe7wWnZl/+umns3DhwixcuDAvvPBCTj755Fx33XWpqqpKkvzxj3/M1VdfnS9+8Yu55pprKjIwAAAA7OmaHfOnnHJK1q1blw996EOZOnVqPvzhD6dz586N9unfv3+uuOKKHH300WIeAAAAKqTZMf/ae+L79Omz031+9atfpaqqKl/5ylea+20AAACAv9Hs98yfe+65bxjySXLWWWdl7733zqmnntrcbwMAAAD8jYp9AN6OlMvl3fnyAAAAsEfarTFfKpV258sDAADAHmm3xjwAAABQeWIeAAAACsZ75gEAAKBgdmvMf+9739udLw8AAAB7pGbfZ76+vj7z58/Pr3/967zwwgvbnYW/6aabMmzYsBYPCAAAADTW7Ji/4oorct999+WEE07IQQcdlA4dOlRyLgAAAGAnmh3zDzzwQL773e86+w4AAACtrEXvmT/kkEMqNQcAAADQRM2O+dNPPz0LFiyo5CwAAABAEzT5MvvZs2c3etyzZ8/cfPPNefTRRzNkyJB06tSp0fo555xTmQkBAACARpoc8w8//PB22971rneluro6P/vZzxptL5VKYh4AAAB2kybH/J133rk75wAAAACaqNnvmb/00kt3uP3FF1/MV7/61WYPBAAAALyxZsf8vHnzdri9rq7OWXwAAADYjXbpPvPr16/PFVdckSQpl8u54IILGq2Xy+WsXr06RxxxROUmBAAAABrZpZh/17velaqqqtTX12fx4sUZMmRIo/VSqZTjjz8+J598ckWHBAAAAP5ql2K+VCrlU5/6VJLkvvvuy2c+85ndMhQAAACwc81+z/xtt92WV155pZKzAAAAAE3Q7JgfO3ZsampqKjkLAAAA0ATNjvnhw4dnxYoVlZwFAAAAaIJdes/865100km56qqr8oEPfCDve9/70q1bt0brJ5xwQouHAwAAALbX7Ji/8cYbkyQ/+clP8pOf/KTRWqlUEvMAAACwmzQ75h9++OFKzgEAAAA0UbPfMw8AAAC0jSafmf/617+ez372sw2PL7300jfc/6qrrmr+VAAAAMBONTnm6+rqGj3+2w+8AwAAAFpHk2P+kksuafT4S1/6UpLkN7/5TdatW5cePXpk0KBB6du3b2UnBAAAABpp9gfgPf/88/nc5z6X5cuXp1+/ftmyZUteeumlfPzjH8+ll16azp07V3JOAAAA4P9r9gfgffnLX06XLl2yePHiPPTQQ3n00Udz7733Zvny5fnqV79ayRkBAACA12l2zC9dujSXXXZZevfu3bDtoIMOylVXXZX777+/IsMBAAAA22t2zHfv3n2HH4K33377ZevWrS0aCgAAANi5Zsf8hAkTctddd223/cEHH8xxxx3XoqEAAACAnWvyB+DNnj270eOuXbvm3/7t37JixYocccQR6dy5c55++uk89NBDOeussyo+KAAAAPCqJsf8ww8/vN22/v37Z+vWrfn5z3/esG3w4MFZunRpZaYDAAAAttPkmL/zzjt35xwAAABAEzX7PfMAAABA2xDzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCaTcxXy6XM2/evJx55pmNts+fPz9jx47NiBEjcsYZZ2T16tWN1jdu3JiLL744xx57bEaMGJFrr722Ya26ujpTp07Nsccem1GjRmXWrFmtciwAAACwO3Vs6wGSZMmSJbnuuuuybdu2dOjQoWH74sWLc+ONN+b2229P//79M2/evEydOjULFy5MqVTKyy+/nHPOOScf/ehHc91116VDhw7505/+1PD8adOm5dBDD803v/nNbNiwIRMmTMiAAQMyevTotjhMAAAAqIh2cWZ+69atueSSS3L55Zc32r5gwYJMmTIl/fv3T5KMGzcu3bt3z9KlS5Mk99xzT/r165fzzjuv4Y8Af//3f58kWbNmTZ588slccMEFKZVK6devX6ZMmZK5c+e24pEBAABA5bWLmB8zZkyOOeaY7bbX1NSktra20bZevXpl7dq1SZL//M//zMc+9rEdvuby5ctTVVWVjh3/evFBVVVVVq1aVbnBAQAAoA20i8vsd+bEE0/MzJkzM3LkyAwcODCLFi3K448/niOPPDJJsnr16rz88suZMGFCnn/++QwaNCiXXnppBg4cmI0bN6Z3796NXq93797ZtGnTLs1QLpezZcuWSh0SALzlbdu2rdHXfo4CQNOVy+WUSqU33a9dx/zJJ5+cTZs25aKLLsrmzZszYsSIDB8+PN26dUuSbN68Of/1X/+Vb33rW+nVq1dmzZqVCy64IPfff39qa2tTLpcbvV5dXV2T/lFer6amxtl8ANgFr11B99rXe+3VLi4EBIDC6Ny585vu065jPkkmTpyYiRMnNjweP358Bg4cmOTVS+7PO++89O3bN0ly/vnn57bbbstzzz2Xnj17ZsWKFY1eq7q6On369Nml79+pU6cMGjSohUcBAHuO+vr6hq8HDBiQgw8+uA2nAYBieeaZZ5q0X7uP+dd77rnnsm7dugwbNixJcsABB2Tz5s0N66VSKaVSKXvvvXeGDBmSG264IfX19Q1nBJYtW5bDDz98l75nqVRquBIAAHhzXbp0afS1n6MA0HRNvZq8XV/3Vl1dnY0bNyZJNmzYkBkzZuTCCy9s+CXhE5/4RL797W+nuro6SXLbbbflPe95T97znvekqqoqffv2zS233JL6+vqsX78+c+bMyeTJk9vseAAAAKAS2vWZ+erq6px//vmpq6tL165dM3ny5EyaNKlh/aSTTsratWtz2mmnpVOnThkyZEi+/e1vN/wl4/rrr8/06dMze/bs9OzZM9OmTcuQIUPa6nAAAACgIkrlv/2UOBqsXLkySTJ06NA2ngQAimPlypU5++yzkyR33HGHn6MAsAua2qHt+jJ7AAAAYHtiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGDEPAAAABSMmAcAAICCEfMAAABQMGIeAAAACkbMAwAAQMGIeQAAACgYMQ8AAAAFI+YBAACgYDq29QAAe5qXXnopa9eubesxYLd59tlnd/g1vBUNGDAgPXr0aOsxgD2QmAdoRS+99FJOPfXUvPTSS209CrSKyy67rK1HgN2qR48euf/++wU90OpcZg8AAAAF48w8QCt67QyOy+x5q9uyZUuSpFu3bm08CexeLrMH2oqYB2hlPXr0yNChQ9t6DAAACkzMAwAVd8QRRzR8vWzZsjacBADemrxnHgCoqNeH/I4eAwAtJ+YBAACgYMQ8AFAxOzsL7+w8AFSWmAcAKuLNgl3QA0DliHkAAAAoGDEPAAAABSPmAYCKeLNb0LlFHQBUjpgHACpmZ8Eu5AGgssQ8AAAAFIyYBwAqxq3pAKB1iHkAoCLcmg4AWo+YBwAAgIIR8wAAAFAwYh4AqAi3pgOA1iPmAYCKcWs6AGgdYh4AAAAKRswDAABAwYh5AKBi3GceAFqHmAcAKsJ95gGg9Yh5AAAAKBgxDwAAAAUj5gGAinCfeQBoPe0m5svlcubNm5czzzyz0fb58+dn7NixGTFiRM4444ysXr26Ye1HP/pRhgwZktGjRzf878EHH2xYr66uztSpU3Psscdm1KhRmTVrVqsdDwDsidxnHgBaR8e2HiBJlixZkuuuuy7btm1Lhw4dGrYvXrw4N954Y26//fb0798/8+bNy9SpU7Nw4cKUSqUkyWGHHZa77rprh687bdq0HHroofnmN7+ZDRs2ZMKECRkwYEBGjx7dKscFAAAAu0O7ODO/devWXHLJJbn88ssbbV+wYEGmTJmS/v37J0nGjRuX7t27Z+nSpQ379OjRY4evuWbNmjz55JO54IILUiqV0q9fv0yZMiVz587dfQcCAAAAraBdxPyYMWNyzDHHbLe9pqYmtbW1jbb16tUra9eubXj8tre9bYevuXz58lRVVaVjx79efFBVVZVVq1ZVZmgAYDvuMw8AraNdXGa/MyeeeGJmzpyZkSNHZuDAgVm0aFEef/zxHHnkkQ37/OQnP8mxxx6bt7/97fnoRz+ayZMnp1QqZePGjendu3ej1+vdu3c2bdq0SzOUy+Vs2bKlEocDAG9pI0aMeMP1I444Io888kgrTQMAxVQulxveVv5G2nXMn3zyydm0aVMuuuiibN68OSNGjMjw4cPTrVu3JK+e0T/xxBOTJL/5zW9y8cUXp1wu56yzzkptbW3K5XKj16urq2vSP8rr1dTUOJsPABXiZyoAvLnOnTu/6T7tOuaTZOLEiZk4cWLD4/Hjx2fgwIFJ0ijMBw8enE9/+tO56667ctZZZ6Vnz55ZsWJFo9eqrq5Onz59dun7d+rUKYMGDWrBEQAArzn44IPbegQAaNeeeeaZJu3X7mP+9Z577rmsW7cuw4YN2+F6bW1tOnXqlCQZMmRIbrjhhtTX12evvV79aIBly5bl8MMP36XvWSqVGq4EAAB2btmyZW/43ni3pwOAN9fUq8nbxQfg7Ux1dXU2btyYJNmwYUNmzJiRCy+8MF26dEmS/OpXv2p4P/u6dety44035rTTTkvy6ofd9e3bN7fcckvq6+uzfv36zJkzJ5MnT26bgwGAPYD7zANA62jXZ+arq6tz/vnnp66uLl27ds3kyZMzadKkhvVf/vKXueiii9K5c+fss88+OeecczJ+/Pgkr/414/rrr8/06dMze/bs9OzZM9OmTcuQIUPa6nAAAACgIkrlv/2UOBqsXLkySTJ06NA2ngQAisFl9gDQMk3t0HZ9mT0AUBxvdi9595oHgMoR8wAAAFAwYh4AAAAKRswDABXxZu+J9555AKgcMQ8AVIxb0wFA6xDzAAAAUDBiHgAAAApGzAMAFbOz28+5LR0AVJaYBwAqwn3mAaD1iHkAAAAoGDEPAAAABSPmAYCKcJ95AGg9Yh4AqBj3mQeA1iHmAQAAoGDEPAAAABSMmAcAKsZ95gGgdYh5AKAi3GceAFqPmAcAAICCEfMAAABQMGIeAKgI95kHgNYj5gGAinGfeQBoHWIeAAAACkbMAwAV49Z0ANA6xDwAUBFuTQcArUfMAwAAQMGIeQAAACgYMQ8AVIRb0wFA6xHzAEDFuDUdALQOMQ8AAAAFI+YBAACgYMQ8AFAx7jMPAK1DzAMAFeE+8wDQesQ8AAAAFIyYBwAAgIIR8wBARbjPPAC0HjEPAFSM+8wDQOsQ8wAAAFAwYh4AAAAKRswDABXjPvMA0DrEPABQEe4zDwCtR8wDAABAwYh5AAAAKBgxDwBUhPvMA0DrEfMAQMW4zzwAtA4xDwAAAAUj5gGAinFrOgBoHWIeAKiI8ePHt2gdAGg6MQ8AVMSaNWtatA4ANJ2YBwAqYuDAgS1aBwCaTswDABUxd+7cFq0DAE0n5gGAinFrOgBoHWIeAAAACkbMAwAAQMGIeQCgYtxnHgBah5gHACpi3LhxLVoHAJpOzAMAFfHb3/62ResAQNOJeQCgIt797ne3aB0AaDoxDwBUxLx581q0DgA0nZgHACrGfeYBoHWIeQAAACgYMQ8AAAAFI+YBgIpxn3kAaB1iHgCoiDvvvLNF6wBA04l5AKAivv71r7doHQBoOjEPAFTEZz/72RatAwBNJ+YBgIqYMmVKi9YBgKZrNzFfLpczb968nHnmmY22z58/P2PHjs2IESNyxhlnZPXq1Tt8/oMPPpjBgwdn48aNDduqq6szderUHHvssRk1alRmzZq1W48BAPZ07jMPAK2jY1sPkCRLlizJddddl23btqVDhw4N2xcvXpwbb7wxt99+e/r375958+Zl6tSpWbhwYUqlUsN+dXV1+e53v7vd606bNi2HHnpovvnNb2bDhg2ZMGFCBgwYkNGjR7fKcQEAAMDu0C7OzG/dujWXXHJJLr/88kbbFyxYkClTpqR///5JknHjxqV79+5ZunRpo/3mzJmTI488stG2NWvW5Mknn8wFF1yQUqmUfv36ZcqUKZk7d+7uPRgA2IO5NR0AtI52cWZ+zJgxSZLHHnus0faamprU1tY22tarV6+sXbs2w4cPT5I8//zzuf3223Pvvffm+9//fsN+y5cvT1VVVTp2/OshVlVV7fJtccrlcrZs2bJLzwGAPdGDDz74huv33ntvTj755FaaBgCKqVwuN7oSfWfaRczvzIknnpiZM2dm5MiRGThwYBYtWpTHH3+84Sx8uVzO9OnT85nPfCb77rtvo+du3LgxvXv3brStd+/e2bRp0y7NUFNTk1WrVrXkMABgj3DllVe+6frAgQNbaRoAKK7OnTu/6T7tOuZPPvnkbNq0KRdddFE2b96cESNGZPjw4enWrVuS5I477ki3bt0ybty47Z5bW1ubcrncaFtdXV2T/sLxep06dcqgQYOafQwAsKeYPn36Gwb99OnTc/DBB7fiRABQPM8880yT9mvXMZ8kEydOzMSJExsejx8/PgMHDsyvfvWr3Hnnnbn33nt3+LyePXtmxYoVjbZVV1enT58+u/T9S6VSwx8PAICdO/30098w5k8//fRWnAYAiqmpJ6DbxQfgNdVzzz2XdevWZdiwYfn+97+fF154Iccff3yOOuqoHHXUUUlevTR/7ty5GTJkSFasWJH6+vqG5y9btiyHH354W40PAG95bk0HAK2jXZ+Zr66uTm1tbfr27ZsNGzZkxowZufDCC9OlS5d84xvf2G7/wYMH50c/+lH69u2bcrmcvn375pZbbsn555+f3//+95kzZ06+/e1vt/6BAAAAQAW1+5g///zzU1dXl65du2by5MmZNGlSk55bKpVy/fXXZ/r06Zk9e3Z69uyZadOmZciQIbt5agAAANi9SuW//ZQ4GqxcuTJJMnTo0DaeBACK4Y3uJ+9SewB4c03t0EK9Zx4AaL+WLl3aonUAoOnEPABQERdccEGL1gGAphPzAEBF3HTTTS1aBwCaTswDABUxbNiwFq0DAE0n5gGAinGfeQBoHWIeAAAACkbMAwAAQMGIeQCgYnZ2n/k3uv88ALDrxDwAUBFLlixp0ToA0HRiHgCoiP/5P/9ni9YBgKYT8wBARXzjG99o0ToA0HRiHgCoiJEjR7ZoHQBoOjEPAFSM+8wDQOsQ8wAAAFAwYh4AqBi3pgOA1iHmAYCKWL16dYvWAYCmE/MAQEV84hOfaNE6ANB0Yh4AqIgf/OAHLVoHAJpOzAMAFXHggQe2aB0AaDoxDwBUjFvTAUDrEPMAAABQMGIeAAAACkbMAwAV4z7zANA6xDwAUBHLly9v0ToA0HRiHgCoiHPPPbdF6wBA04l5AKAiZs2a1aJ1AKDpxDwAUBGHHXZYi9YBgKYT8wBAxbjPPAC0DjEPAAAABSPmAQAAoGDEPABQMe4zDwCtQ8wDABXxm9/8pkXrAEDTiXkAoCImTJjQonUAoOnEPABQEXPmzGnROgDQdGIeAKiIwYMHt2gdAGg6MQ8AVIz7zANA6xDzAAAAUDBiHgCoGLemA4DWIeYBgIp45plnWrQOADSdmAcAKuKMM85o0ToA0HRiHgCoiHvuuadF6wBA04l5AKAiBg0a1KJ1AKDpxDwAUDFuTQcArUPMAwAV9Y53vOMNHwMALSfmAYCK+sMf/vCGjwGAlhPzAEDFTJs2bZe2AwDNI+YBgIp4+eWX85Of/GSHaz/5yU/y8ssvt/JEAPDWJeYBgIq4+OKLW7QOADSdmAcAKmLmzJktWgcAmk7MAwAVsffee+e4447b4dqYMWOy9957t/JEAPDWJeYBgIq59tprd7j9qquuauVJAOCtTcwDABV11llnveFjAKDlxDwAUFHf+9733vAxANByYh4AqJgzzjhjl7YDAM0j5gGAiti8eXOeeeaZHa4988wz2bx5cytPBABvXWIeAKiIM888s0XrAEDTiXkAoCLuvvvuFq0DAE0n5gGAiujevXsGDRq0w7X3vve96d69eytPBABvXWIeAKiYe+65Z4fbnZUHgMoS8wBARR199NFv+BgAaDkxDwBUzG9/+9v84he/aLTt0UcfzW9/+9s2mggA3prEPABQEeVyOddcc81Ot5fL5TaYCgDemsQ8AFARa9asyaOPPpq6urpG2+vq6vLoo49mzZo1bTQZALz1iHkAoCIGDhyYD37wg+nQoUOj7R06dMg//MM/ZODAgW00GQC89Yh5AKAiSqVSvvCFL+x0e6lUaoOpAOCtqd3EfLlczrx583LmmWc22j5//vyMHTs2I0aMyBlnnJHVq1c3rC1evDgf+chHMnr06IwePTo33XRTo+dWV1dn6tSpOfbYYzNq1KjMmjWrVY4FAPZU7373u3POOec0hHupVMo555yT/fbbr40nA4C3lo5tPUCSLFmyJNddd122bdvW6NK8xYsX58Ybb8ztt9+e/v37Z968eZk6dWoWLlyYUqmU3r1754477si+++6b559/PmeeeWYOPPDAjB49Okkybdq0HHroofnmN7+ZDRs2ZMKECRkwYEDDOgBQeRMmTMisWbNSLpdTKpXyiU98oq1HAoC3nHZxZn7r1q255JJLcvnllzfavmDBgkyZMiX9+/dPkowbNy7du3fP0qVLkyRDhgzJvvvumyTp169fhg4dmueffz7Jqx/C8+STT+aCCy5IqVRKv379MmXKlMydO7f1DgwA9kBz5sxJfX19kqS+vj4/+MEP2ngiAHjraRcxP2bMmBxzzDHbba+pqUltbW2jbb169cratWsbbauvr8/Pf/7zrFmzJieccEKSZPny5amqqkrHjn+9+KCqqiqrVq2q/AEAAElevc/87NmzG22bPXu2+8wDQIW1i8vsd+bEE0/MzJkzM3LkyAwcODCLFi3K448/niOPPLJhn/Hjx2f16tXp1atXrrrqqvTu3TtJsnHjxoavX9O7d+9s2rRpl2Yol8vZsmVLi48FAN7qyuVyrrzyyp1u/9rXvuZD8ADgTbz2NrU3065j/uSTT86mTZty0UUXZfPmzRkxYkSGDx+ebt26Newzd+7c1NXVZfny5bn00ktz0UUX5ZRTTkltbW3K5XKj16urq9vlXyJqamqczQeAJvjjH//Y8Fa416urq8vSpUuzaNGihrfOAQA717lz5zfdp13HfJJMnDgxEydObHg8fvz47e5T26FDhxx55JH51Kc+lbvuuiunnHJKevbsmRUrVjTar7q6On369Nml79+pU6cMGjSo+QcAAHuIgw46KMOGDcvjjz+eurq6hu0dOnTIUUcdlVGjRjkzDwBv4plnnmnSfu0+5l/vueeey7p16zJs2LAdrnfu3DldunRJ8uqH491www2pr6/PXnu9+tEAy5Yty+GHH75L37NUKjW6EgAA2Lnp06dn/PjxjbaVSqVMnz493bt3b6OpAKA4mvqH73bxAXg7U11dnY0bNyZJNmzYkBkzZuTCCy9sCPa77rorf/nLX5Ikv//973PTTTflox/9aJJXP+yub9++ueWWW1JfX5/169dnzpw5mTx5ctscDADsAdxnHgBaR7s+M19dXZ3zzz8/dXV16dq1ayZPnpxJkyY1rK9evTonnnhiOnfunLe97W0599xzc9pppyV59ZeH66+/PtOnT8/s2bPTs2fPTJs2LUOGDGmrwwGAPcI555yTBQsWZMOGDenbt2/+8R//sa1HAoC3nFL5bz8ljgYrV65MkgwdOrSNJwGAYlm8eHGuvfbaTJs2bYe3nwUAdqypHdquz8wDAMV0zDHHiHgA2I3a9XvmAQAAgO2JeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwYh5AAAAKBgxDwAAAAUj5gEAAKBgxDwAAAAUjJgHAACAghHzAAAAUDBiHgAAAApGzAMAAEDBiHkAAAAoGDEPAAAABSPmAQAAoGA6tvUA7VlNTU3K5XJWrlzZ1qMAAACwB3jllVdSKpXedD8x/waa8g8IAAAAlVIqlZrUoqVyuVxuhXkAAACACvGeeQAAACgYMQ8AAAAFI+YBAACgYMQ8AAAAFIyYBwAAgIIR8wAAAFAwYh4AAAAKRswDAABAwfw/ciGWanhix+cAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'birth_year', data = user_spec)\n",
    "plt.title('birth_year', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "abb56460",
   "metadata": {},
   "source": [
    "### Credit_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "792fb1a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# credit_score가 nan인 User_id\n",
    "\n",
    "user_id = user_spec[user_spec.credit_score.isna()].user_id"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "f50db913",
   "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>user_id</th>\n",
       "      <th>credit_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>14</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>19</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>48</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>74</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58674</th>\n",
       "      <td>879634</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58675</th>\n",
       "      <td>879685</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58676</th>\n",
       "      <td>879688</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58677</th>\n",
       "      <td>879695</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58678</th>\n",
       "      <td>879697</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>58679 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       user_id  credit_score\n",
       "0           14             0\n",
       "1           19             0\n",
       "2           41             1\n",
       "3           48             1\n",
       "4           74             0\n",
       "...        ...           ...\n",
       "58674   879634             0\n",
       "58675   879685             0\n",
       "58676   879688             0\n",
       "58677   879695             0\n",
       "58678   879697             0\n",
       "\n",
       "[58679 rows x 2 columns]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# credit_score가 nan값인 user의 credit score unique 갯수\n",
    "\n",
    "cs_df = user_spec[user_spec['user_id'].isin(user_id)].groupby('user_id')['credit_score'].nunique().reset_index() ; cs_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "843f36a1",
   "metadata": {},
   "source": [
    "- 1이면 자기 자신으로 채움  \n",
    "- 2이상이면 보간\n",
    "- 0은 어떻게"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "ddd4ff4d",
   "metadata": {},
   "outputs": [],
   "source": [
    "cs_df_id = cs_df.user_id.to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "19c47760",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████████████████████████████████████| 58679/58679 [01:14<00:00, 782.72it/s]\n"
     ]
    }
   ],
   "source": [
    "for idd in tqdm(cs_df_id):\n",
    "    if cs_df[cs_df.user_id == idd].credit_score.values[0] == 1:\n",
    "        miss_list = user_spec[user_spec.user_id == idd]['credit_score'].unique().tolist()\n",
    "        miss = [x for x in miss_list if pd.isnull(x) == False][0]\n",
    "        user_spec[user_spec.user_id == idd] = user_spec[user_spec.user_id == idd].fillna({'credit_score':miss})\n",
    "    elif cs_df[cs_df.user_id == idd].credit_score.values[0] > 1:\n",
    "        user_spec.loc[user_spec.user_id == idd, 'credit_score'] = user_spec[user_spec.user_id == idd].sort_values(['credit_score']).loc[user_spec.user_id == idd, 'credit_score'].fillna(method='ffill')\n",
    "        user_spec.loc[user_spec.user_id == idd, 'credit_score'] = user_spec[user_spec.user_id == idd].sort_values(['credit_score']).loc[user_spec.user_id == idd, 'credit_score'].fillna(method='bfill')\n",
    "    else:\n",
    "        pass\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "f3bd8c60",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            69483\n",
       "yearly_income                               1\n",
       "income_type                                 0\n",
       "company_enter_month                     92314\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn             417763\n",
       "personal_rehabilitation_complete_yn    843993\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                              92314\n",
       "enter_month                             92314\n",
       "enter_birth                             99428\n",
       "month                                       0\n",
       "birth_year                                  0\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a07cafb0",
   "metadata": {},
   "source": [
    "- 나머지 값들은 knn Imputer로 처리"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf2772e5",
   "metadata": {},
   "source": [
    "### yearly_income"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "bfae9e80",
   "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>application_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>insert_time</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>company_enter_month</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "      <th>enter_year</th>\n",
       "      <th>enter_month</th>\n",
       "      <th>enter_birth</th>\n",
       "      <th>month</th>\n",
       "      <th>birth_year</th>\n",
       "      <th>gender</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>415357</th>\n",
       "      <td>341149</td>\n",
       "      <td>670502</td>\n",
       "      <td>2022-06-30 21:32:18</td>\n",
       "      <td>930.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>OTHERINCOME</td>\n",
       "      <td>NaN</td>\n",
       "      <td>기타</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>사업자금</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "      <td>1981.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        application_id  user_id          insert_time  credit_score  \\\n",
       "415357          341149   670502  2022-06-30 21:32:18         930.0   \n",
       "\n",
       "        yearly_income  income_type  company_enter_month employment_type  \\\n",
       "415357            NaN  OTHERINCOME                  NaN              기타   \n",
       "\n",
       "       houseown_type  desired_amount purpose  personal_rehabilitation_yn  \\\n",
       "415357        기타가족소유       6000000.0    사업자금                         0.0   \n",
       "\n",
       "        personal_rehabilitation_complete_yn  existing_loan_cnt  \\\n",
       "415357                                  0.0                1.0   \n",
       "\n",
       "        existing_loan_amt  enter_year  enter_month  enter_birth  month  \\\n",
       "415357          3000000.0         NaN          NaN          NaN      6   \n",
       "\n",
       "        birth_year  gender  \n",
       "415357      1981.0     1.0  "
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec[user_spec.yearly_income.isna()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "27389ef4",
   "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>application_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>insert_time</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>company_enter_month</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "      <th>enter_year</th>\n",
       "      <th>enter_month</th>\n",
       "      <th>enter_birth</th>\n",
       "      <th>month</th>\n",
       "      <th>birth_year</th>\n",
       "      <th>gender</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>393331</th>\n",
       "      <td>1746224</td>\n",
       "      <td>670502</td>\n",
       "      <td>2022-06-30 22:41:11</td>\n",
       "      <td>930.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>OTHERINCOME</td>\n",
       "      <td>NaN</td>\n",
       "      <td>기타</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>사업자금</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "      <td>1981.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>415357</th>\n",
       "      <td>341149</td>\n",
       "      <td>670502</td>\n",
       "      <td>2022-06-30 21:32:18</td>\n",
       "      <td>930.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>OTHERINCOME</td>\n",
       "      <td>NaN</td>\n",
       "      <td>기타</td>\n",
       "      <td>기타가족소유</td>\n",
       "      <td>6000000.0</td>\n",
       "      <td>사업자금</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "      <td>1981.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        application_id  user_id          insert_time  credit_score  \\\n",
       "393331         1746224   670502  2022-06-30 22:41:11         930.0   \n",
       "415357          341149   670502  2022-06-30 21:32:18         930.0   \n",
       "\n",
       "        yearly_income  income_type  company_enter_month employment_type  \\\n",
       "393331            0.0  OTHERINCOME                  NaN              기타   \n",
       "415357            NaN  OTHERINCOME                  NaN              기타   \n",
       "\n",
       "       houseown_type  desired_amount purpose  personal_rehabilitation_yn  \\\n",
       "393331        기타가족소유       6000000.0    사업자금                         0.0   \n",
       "415357        기타가족소유       6000000.0    사업자금                         0.0   \n",
       "\n",
       "        personal_rehabilitation_complete_yn  existing_loan_cnt  \\\n",
       "393331                                  0.0                1.0   \n",
       "415357                                  0.0                1.0   \n",
       "\n",
       "        existing_loan_amt  enter_year  enter_month  enter_birth  month  \\\n",
       "393331          3000000.0         NaN          NaN          NaN      6   \n",
       "415357          3000000.0         NaN          NaN          NaN      6   \n",
       "\n",
       "        birth_year  gender  \n",
       "393331      1981.0     1.0  \n",
       "415357      1981.0     1.0  "
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec[user_spec.user_id == 670502]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "718b109f",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.yearly_income.fillna(0,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "c8471fd8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            69483\n",
       "yearly_income                               0\n",
       "income_type                                 0\n",
       "company_enter_month                     92314\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn             417763\n",
       "personal_rehabilitation_complete_yn    843993\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                              92314\n",
       "enter_month                             92314\n",
       "enter_birth                             99428\n",
       "month                                       0\n",
       "birth_year                                  0\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90b6df2a",
   "metadata": {},
   "source": [
    "### Company_enter_month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "9ad83603",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.drop(['company_enter_month'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11d5277f",
   "metadata": {},
   "source": [
    "- enter year 5년 단위로 중앙\n",
    "- enter month 5년 단위로 중앙"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "989583ad",
   "metadata": {},
   "source": [
    "#### enter_year"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "117839a7",
   "metadata": {},
   "source": [
    "- enter_year를 들어간 나이로 바꿈\n",
    "- 그 이후에 birth_year와 같은 방식으로 처리\n",
    "- 해당 나이를 birth_year에 더하기\n",
    "- 이후 2022년 넘는 것을 처리"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "4937844f",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.enter_birth = user_spec.enter_year - user_spec.birth_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "0fcb71c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "def enter_age(x):\n",
    "    if x > 90:\n",
    "        return 90\n",
    "    elif x > 85:\n",
    "        return 85\n",
    "    elif x > 80:\n",
    "        return 80\n",
    "    elif x > 75:\n",
    "        return 75\n",
    "    elif x > 70:\n",
    "        return 70\n",
    "    elif x > 65:\n",
    "        return 65\n",
    "    elif x > 60:\n",
    "        return 60\n",
    "    elif x > 55:\n",
    "        return 55\n",
    "    elif x > 50:\n",
    "        return 50\n",
    "    elif x > 45:\n",
    "        return 45\n",
    "    elif x > 40:\n",
    "        return 40\n",
    "    elif x > 35:\n",
    "        return 35\n",
    "    elif x > 30:\n",
    "        return 30\n",
    "    elif x > 25:\n",
    "        return 25\n",
    "    elif x > 20:\n",
    "        return 20\n",
    "    elif x > 15:\n",
    "        return 15\n",
    "    elif x > 10:\n",
    "        return 10\n",
    "    else:\n",
    "        return np.nan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "472be88a",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec_train['enter_age'] = user_spec_train.enter_birth.apply(lambda x : enter_age(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "9c12d5ef",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec['enter_age'] = user_spec.enter_birth.apply(lambda x : enter_age(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "4ada4694",
   "metadata": {},
   "outputs": [],
   "source": [
    "lab_df = user_spec_train[['enter_age','credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "bbf79835",
   "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>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>enter_age</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <td>700.396476</td>\n",
       "      <td>1.802162e+08</td>\n",
       "      <td>6.724324e+07</td>\n",
       "      <td>4.083333</td>\n",
       "      <td>8.150000e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15.0</th>\n",
       "      <td>674.235415</td>\n",
       "      <td>4.832597e+07</td>\n",
       "      <td>3.166094e+07</td>\n",
       "      <td>4.580959</td>\n",
       "      <td>8.776464e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20.0</th>\n",
       "      <td>666.544187</td>\n",
       "      <td>4.365129e+07</td>\n",
       "      <td>3.221169e+07</td>\n",
       "      <td>4.359892</td>\n",
       "      <td>7.146314e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25.0</th>\n",
       "      <td>675.730827</td>\n",
       "      <td>4.552838e+07</td>\n",
       "      <td>3.709679e+07</td>\n",
       "      <td>4.589030</td>\n",
       "      <td>8.173685e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30.0</th>\n",
       "      <td>678.303698</td>\n",
       "      <td>4.571831e+07</td>\n",
       "      <td>3.999636e+07</td>\n",
       "      <td>4.567256</td>\n",
       "      <td>8.529000e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35.0</th>\n",
       "      <td>670.528771</td>\n",
       "      <td>4.902173e+07</td>\n",
       "      <td>3.694254e+07</td>\n",
       "      <td>4.556442</td>\n",
       "      <td>8.482858e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40.0</th>\n",
       "      <td>664.746224</td>\n",
       "      <td>4.876196e+07</td>\n",
       "      <td>3.354689e+07</td>\n",
       "      <td>4.640788</td>\n",
       "      <td>7.907058e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45.0</th>\n",
       "      <td>662.074221</td>\n",
       "      <td>4.821837e+07</td>\n",
       "      <td>3.496127e+07</td>\n",
       "      <td>4.538454</td>\n",
       "      <td>7.658089e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50.0</th>\n",
       "      <td>661.574123</td>\n",
       "      <td>4.904413e+07</td>\n",
       "      <td>3.939689e+07</td>\n",
       "      <td>4.544116</td>\n",
       "      <td>6.920438e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55.0</th>\n",
       "      <td>664.279192</td>\n",
       "      <td>4.430083e+07</td>\n",
       "      <td>3.320476e+07</td>\n",
       "      <td>4.534173</td>\n",
       "      <td>6.577873e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60.0</th>\n",
       "      <td>664.647709</td>\n",
       "      <td>3.660148e+07</td>\n",
       "      <td>2.906150e+07</td>\n",
       "      <td>4.749876</td>\n",
       "      <td>5.854262e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65.0</th>\n",
       "      <td>660.913955</td>\n",
       "      <td>3.630423e+07</td>\n",
       "      <td>2.132912e+07</td>\n",
       "      <td>5.708101</td>\n",
       "      <td>5.564390e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70.0</th>\n",
       "      <td>704.337568</td>\n",
       "      <td>2.710543e+07</td>\n",
       "      <td>3.567132e+07</td>\n",
       "      <td>4.561243</td>\n",
       "      <td>1.115082e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75.0</th>\n",
       "      <td>664.230769</td>\n",
       "      <td>2.450394e+07</td>\n",
       "      <td>2.577165e+07</td>\n",
       "      <td>4.240741</td>\n",
       "      <td>4.685567e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80.0</th>\n",
       "      <td>703.333333</td>\n",
       "      <td>5.842857e+07</td>\n",
       "      <td>2.280952e+07</td>\n",
       "      <td>5.333333</td>\n",
       "      <td>1.136364e+07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           credit_score  yearly_income  desired_amount  existing_loan_cnt  \\\n",
       "enter_age                                                                   \n",
       "10.0         700.396476   1.802162e+08    6.724324e+07           4.083333   \n",
       "15.0         674.235415   4.832597e+07    3.166094e+07           4.580959   \n",
       "20.0         666.544187   4.365129e+07    3.221169e+07           4.359892   \n",
       "25.0         675.730827   4.552838e+07    3.709679e+07           4.589030   \n",
       "30.0         678.303698   4.571831e+07    3.999636e+07           4.567256   \n",
       "35.0         670.528771   4.902173e+07    3.694254e+07           4.556442   \n",
       "40.0         664.746224   4.876196e+07    3.354689e+07           4.640788   \n",
       "45.0         662.074221   4.821837e+07    3.496127e+07           4.538454   \n",
       "50.0         661.574123   4.904413e+07    3.939689e+07           4.544116   \n",
       "55.0         664.279192   4.430083e+07    3.320476e+07           4.534173   \n",
       "60.0         664.647709   3.660148e+07    2.906150e+07           4.749876   \n",
       "65.0         660.913955   3.630423e+07    2.132912e+07           5.708101   \n",
       "70.0         704.337568   2.710543e+07    3.567132e+07           4.561243   \n",
       "75.0         664.230769   2.450394e+07    2.577165e+07           4.240741   \n",
       "80.0         703.333333   5.842857e+07    2.280952e+07           5.333333   \n",
       "\n",
       "           existing_loan_amt  \n",
       "enter_age                     \n",
       "10.0            8.150000e+07  \n",
       "15.0            8.776464e+07  \n",
       "20.0            7.146314e+07  \n",
       "25.0            8.173685e+07  \n",
       "30.0            8.529000e+07  \n",
       "35.0            8.482858e+07  \n",
       "40.0            7.907058e+07  \n",
       "45.0            7.658089e+07  \n",
       "50.0            6.920438e+07  \n",
       "55.0            6.577873e+07  \n",
       "60.0            5.854262e+07  \n",
       "65.0            5.564390e+07  \n",
       "70.0            1.115082e+08  \n",
       "75.0            4.685567e+07  \n",
       "80.0            1.136364e+07  "
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lab_df1 = lab_df.groupby('enter_age')[['credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']].mean()\n",
    "lab_df1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "8fc299d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_id = user_spec[user_spec.enter_year.isna()].user_id.values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "6e251990",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_list = ['credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "866ce7b1",
   "metadata": {},
   "outputs": [],
   "source": [
    "age_list=[10,15,20,25,30,35,40,45,50,55,60,65,70,75,80]\n",
    "def age_feature(x,feature):\n",
    "    diff = []\n",
    "    for i in feature:\n",
    "        diff.append(abs(i-x))\n",
    "    idx = diff.index(np.min(diff))\n",
    "    return age_list[idx]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "2176aec4",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████████████████████████████████████| 92314/92314 [11:55<00:00, 129.01it/s]\n"
     ]
    }
   ],
   "source": [
    "for idd in tqdm(user_id):\n",
    "    user_enter_list = []\n",
    "    for f in feature_list:\n",
    "        x = np.mean(user_spec[user_spec.user_id == idd][f'{f}'].fillna(0).tolist())\n",
    "        feature = lab_df1[f'{f}'].values.tolist()\n",
    "        user_enter_list.append(age_feature(x,feature))\n",
    "    age = int(np.mean(user_enter_list))\n",
    "    missing_value = np.unique(user_spec[user_spec.user_id == idd]['birth_year'].values)[0]+age\n",
    "    if missing_value > 2022:\n",
    "        missing_value = 2021\n",
    "    else:\n",
    "        pass\n",
    "    user_spec.loc[user_spec.user_id == idd,'enter_year'] = missing_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "1883d687",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.drop('enter_age',axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "756799cc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            69483\n",
       "yearly_income                               0\n",
       "income_type                                 0\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn             417763\n",
       "personal_rehabilitation_complete_yn    843993\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                                  0\n",
       "enter_month                             92314\n",
       "enter_birth                             92314\n",
       "month                                       0\n",
       "birth_year                                  0\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "2d5a6910",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'enter_year', data = user_spec)\n",
    "plt.title('enter_year', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "ccc6ba3b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2021.0    351490\n",
       "2022.0    133616\n",
       "2020.0    112109\n",
       "2019.0     78735\n",
       "2018.0     55281\n",
       "2017.0     40388\n",
       "2016.0     31603\n",
       "2015.0     23761\n",
       "2014.0     18043\n",
       "2013.0     16864\n",
       "2012.0     14645\n",
       "2011.0     13130\n",
       "2010.0     11521\n",
       "2009.0      7986\n",
       "2008.0      6876\n",
       "2007.0      6472\n",
       "2006.0      6142\n",
       "2005.0      5161\n",
       "2004.0      4694\n",
       "2003.0      4089\n",
       "2002.0      3805\n",
       "2000.0      3342\n",
       "2001.0      2849\n",
       "1999.0      2640\n",
       "1996.0      2047\n",
       "1997.0      1835\n",
       "1995.0      1830\n",
       "1998.0      1202\n",
       "1994.0      1161\n",
       "1993.0      1088\n",
       "1992.0      1035\n",
       "1991.0       818\n",
       "1990.0       743\n",
       "1989.0       468\n",
       "1988.0       418\n",
       "1986.0       278\n",
       "1987.0       255\n",
       "1985.0        90\n",
       "1984.0        76\n",
       "1983.0        56\n",
       "1982.0        40\n",
       "1980.0        24\n",
       "1981.0        16\n",
       "1979.0        12\n",
       "1977.0         3\n",
       "1970.0         3\n",
       "1974.0         3\n",
       "1976.0         2\n",
       "1975.0         2\n",
       "1942.0         2\n",
       "1969.0         2\n",
       "1972.0         1\n",
       "1978.0         1\n",
       "Name: enter_year, dtype: int64"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.enter_year.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "aac15d16",
   "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>application_id</th>\n",
       "      <th>user_id</th>\n",
       "      <th>insert_time</th>\n",
       "      <th>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>income_type</th>\n",
       "      <th>employment_type</th>\n",
       "      <th>houseown_type</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>purpose</th>\n",
       "      <th>personal_rehabilitation_yn</th>\n",
       "      <th>personal_rehabilitation_complete_yn</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "      <th>enter_year</th>\n",
       "      <th>enter_month</th>\n",
       "      <th>enter_birth</th>\n",
       "      <th>month</th>\n",
       "      <th>birth_year</th>\n",
       "      <th>gender</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [application_id, user_id, insert_time, credit_score, yearly_income, income_type, employment_type, houseown_type, desired_amount, purpose, personal_rehabilitation_yn, personal_rehabilitation_complete_yn, existing_loan_cnt, existing_loan_amt, enter_year, enter_month, enter_birth, month, birth_year, gender]\n",
       "Index: []"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec[user_spec.enter_year > 2022]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cad4c43f",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "dff7c39c",
   "metadata": {},
   "source": [
    "#### enter_month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "7509a880",
   "metadata": {},
   "outputs": [],
   "source": [
    "lab_df = user_spec_train[['enter_month','credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "7625c8d4",
   "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>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>enter_month</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1.0</th>\n",
       "      <td>670.242402</td>\n",
       "      <td>4.816826e+07</td>\n",
       "      <td>3.534250e+07</td>\n",
       "      <td>4.573935</td>\n",
       "      <td>7.942155e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.0</th>\n",
       "      <td>665.445263</td>\n",
       "      <td>4.330703e+07</td>\n",
       "      <td>3.744789e+07</td>\n",
       "      <td>4.444444</td>\n",
       "      <td>7.388758e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3.0</th>\n",
       "      <td>672.394614</td>\n",
       "      <td>4.701587e+07</td>\n",
       "      <td>3.479107e+07</td>\n",
       "      <td>4.619108</td>\n",
       "      <td>7.722391e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4.0</th>\n",
       "      <td>672.088236</td>\n",
       "      <td>4.601787e+07</td>\n",
       "      <td>3.545405e+07</td>\n",
       "      <td>4.581807</td>\n",
       "      <td>7.936082e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5.0</th>\n",
       "      <td>672.781322</td>\n",
       "      <td>4.720423e+07</td>\n",
       "      <td>3.641927e+07</td>\n",
       "      <td>4.474324</td>\n",
       "      <td>8.180455e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6.0</th>\n",
       "      <td>671.210821</td>\n",
       "      <td>4.749210e+07</td>\n",
       "      <td>3.936609e+07</td>\n",
       "      <td>4.648449</td>\n",
       "      <td>8.343572e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7.0</th>\n",
       "      <td>675.944930</td>\n",
       "      <td>4.950934e+07</td>\n",
       "      <td>3.798585e+07</td>\n",
       "      <td>4.662807</td>\n",
       "      <td>8.924416e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8.0</th>\n",
       "      <td>672.881996</td>\n",
       "      <td>4.693330e+07</td>\n",
       "      <td>3.608808e+07</td>\n",
       "      <td>4.632850</td>\n",
       "      <td>8.358506e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9.0</th>\n",
       "      <td>671.301971</td>\n",
       "      <td>4.607789e+07</td>\n",
       "      <td>3.787175e+07</td>\n",
       "      <td>4.547576</td>\n",
       "      <td>7.950747e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10.0</th>\n",
       "      <td>669.986702</td>\n",
       "      <td>4.475739e+07</td>\n",
       "      <td>3.593673e+07</td>\n",
       "      <td>4.483876</td>\n",
       "      <td>7.759008e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11.0</th>\n",
       "      <td>668.302230</td>\n",
       "      <td>4.494149e+07</td>\n",
       "      <td>3.660967e+07</td>\n",
       "      <td>4.490621</td>\n",
       "      <td>7.659418e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12.0</th>\n",
       "      <td>670.196491</td>\n",
       "      <td>4.562845e+07</td>\n",
       "      <td>3.461414e+07</td>\n",
       "      <td>4.450024</td>\n",
       "      <td>7.768864e+07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             credit_score  yearly_income  desired_amount  existing_loan_cnt  \\\n",
       "enter_month                                                                   \n",
       "1.0            670.242402   4.816826e+07    3.534250e+07           4.573935   \n",
       "2.0            665.445263   4.330703e+07    3.744789e+07           4.444444   \n",
       "3.0            672.394614   4.701587e+07    3.479107e+07           4.619108   \n",
       "4.0            672.088236   4.601787e+07    3.545405e+07           4.581807   \n",
       "5.0            672.781322   4.720423e+07    3.641927e+07           4.474324   \n",
       "6.0            671.210821   4.749210e+07    3.936609e+07           4.648449   \n",
       "7.0            675.944930   4.950934e+07    3.798585e+07           4.662807   \n",
       "8.0            672.881996   4.693330e+07    3.608808e+07           4.632850   \n",
       "9.0            671.301971   4.607789e+07    3.787175e+07           4.547576   \n",
       "10.0           669.986702   4.475739e+07    3.593673e+07           4.483876   \n",
       "11.0           668.302230   4.494149e+07    3.660967e+07           4.490621   \n",
       "12.0           670.196491   4.562845e+07    3.461414e+07           4.450024   \n",
       "\n",
       "             existing_loan_amt  \n",
       "enter_month                     \n",
       "1.0               7.942155e+07  \n",
       "2.0               7.388758e+07  \n",
       "3.0               7.722391e+07  \n",
       "4.0               7.936082e+07  \n",
       "5.0               8.180455e+07  \n",
       "6.0               8.343572e+07  \n",
       "7.0               8.924416e+07  \n",
       "8.0               8.358506e+07  \n",
       "9.0               7.950747e+07  \n",
       "10.0              7.759008e+07  \n",
       "11.0              7.659418e+07  \n",
       "12.0              7.768864e+07  "
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lab_df1 = lab_df.groupby('enter_month')[['credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']].mean()\n",
    "lab_df1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "55200a2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_id = user_spec[user_spec.enter_month.isna()].user_id.values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "4618c305",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_list = ['credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "60022186",
   "metadata": {},
   "outputs": [],
   "source": [
    "age_list=[1,2,3,4,5,6,7,8,9,10,11,12]\n",
    "def age_feature(x,feature):\n",
    "    diff = []\n",
    "    for i in feature:\n",
    "        diff.append(abs(i-x))\n",
    "    idx = diff.index(np.min(diff))\n",
    "    return age_list[idx]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "c07b6e90",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████████████████████████████████████| 92314/92314 [11:05<00:00, 138.79it/s]\n"
     ]
    }
   ],
   "source": [
    "for idd in tqdm(user_id):\n",
    "    user_enter_list = []\n",
    "    for f in feature_list:\n",
    "        x = np.mean(user_spec[user_spec.user_id == idd][f'{f}'].fillna(0).tolist())\n",
    "        feature = lab_df1[f'{f}'].values.tolist()\n",
    "        user_enter_list.append(age_feature(x,feature))\n",
    "    missing_value = int(np.mean(user_enter_list))\n",
    "    user_spec.loc[user_spec.user_id == idd,'enter_month'] = missing_value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "2496ccf8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            69483\n",
       "yearly_income                               0\n",
       "income_type                                 0\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn             417763\n",
       "personal_rehabilitation_complete_yn    843993\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                                  0\n",
       "enter_month                                 0\n",
       "enter_birth                             92314\n",
       "month                                       0\n",
       "birth_year                                  0\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "9724c24c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.countplot(y = 'enter_month', data = user_spec)\n",
    "plt.title('enter_month', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "48b6804f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.0     121746\n",
       "1.0     117008\n",
       "4.0     111222\n",
       "3.0     106628\n",
       "2.0      76264\n",
       "6.0      73739\n",
       "12.0     65864\n",
       "11.0     63799\n",
       "7.0      63305\n",
       "10.0     61320\n",
       "9.0      54810\n",
       "8.0      53048\n",
       "Name: enter_month, dtype: int64"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.enter_month.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23b7790a",
   "metadata": {},
   "source": [
    "#### enter_birth"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "119d55a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.enter_birth = user_spec.enter_year - user_spec.birth_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "24466aaf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'enter_birth', data = user_spec)\n",
    "plt.title('enter_birth', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "95848152",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            69483\n",
       "yearly_income                               0\n",
       "income_type                                 0\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn             417763\n",
       "personal_rehabilitation_complete_yn    843993\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                                  0\n",
       "enter_month                                 0\n",
       "enter_birth                                 0\n",
       "month                                       0\n",
       "birth_year                                  0\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "8d4fdc8a",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.loc[user_spec['enter_birth']<15,'enter_year'] = user_spec.loc[user_spec['enter_birth']<15,'birth_year']+user_spec_train['enter_birth'].median() \n",
    "user_spec.loc[user_spec['enter_year']>2022,'enter_year'] = 2021\n",
    "user_spec.loc[user_spec['enter_birth']<15,'enter_birth'] = 15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "00427580",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'enter_birth', data = user_spec)\n",
    "plt.title('enter_birth', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2690b92c",
   "metadata": {},
   "source": [
    "### Personal_rehabilitation_yn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "6b0e1be1",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.personal_rehabilitation_yn = user_spec.personal_rehabilitation_yn.fillna(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "348cffd1",
   "metadata": {},
   "source": [
    "### Personal_rehabilitation_complete_yn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "13f37475",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.personal_rehabilitation_complete_yn = user_spec.personal_rehabilitation_complete_yn.fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "1770bb38",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            69483\n",
       "yearly_income                               0\n",
       "income_type                                 0\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn                  0\n",
       "personal_rehabilitation_complete_yn         0\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                                  0\n",
       "enter_month                                 0\n",
       "enter_birth                                 0\n",
       "month                                       0\n",
       "birth_year                                  0\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed40bb60",
   "metadata": {},
   "source": [
    "### existing_loan_cnt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "8cd1aa2c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA+wAAAKUCAYAAACe37KQAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAA9hAAAPYQGoP6dpAABDOklEQVR4nO3deXxU9b0//veQBFlEKYoUqVKUghugqFgRRBQFxfoFl1ariCtuLVbFre7WKmqlili1boj7DlX03t5KtVLjClgUrGKjKFeBCiKrJGF+f/SSnzERZzIhc5g8n49HHw/P533mzGuwD8kr58w5qXQ6nQ4AAAAgUZrkOwAAAABQk8IOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAOvB448/HrvuumuUlpbW2zH//e9/R9++feOSSy6pt2PWl5tvvjm6du0aH330Ub6jAEDBUNgBYD2YP39+LFu2LBYuXFin169cubLG2ooVK2Lx4sUxb968XOORB7X9OwWAdSnOdwAAKESnn356HH744dGuXbusX3v77bfHrbfeGjNmzKi2vvXWW8ff/va32HjjjespJQ3ltNNOixUrVsS9996b7ygAbECcYQeA9SCVStWprEdEvP/++996NrZNmzbRtGnTXKKRB++8806sWbMm3zEA2MAo7AAAAJBACjsA/J8///nPceSRR8bOO+8cPXv2jGHDhsXf//73qvm//vWv6NatW5xwwgk1XnvDDTfE9ttvX3UZ+2OPPRZdu3aNV199tdp+r7/+eowYMSL22muv6N69ewwcODAefPDBiPj/b9z29NNPR0RE165do2vXrnHBBRdERERFRUW17bX23XffOOqoo2LFihXx29/+Nvr27RvdunWLoUOHxosvvljrZ33vvfdi5MiRsccee0T37t3j0EMPjeeff3693Dzuueeei+HDh0evXr1ip512iv322y+uuuqq+Pzzz2vs+8EHH8Sll14agwcPjl122SX22WefOOecc+Kzzz6rtt+rr74aXbt2jcceeyzefvvtOOGEE6Jnz56x2267xa9+9as63zsgIuLvf/97nHLKKfHjH/84dtppp+jfv39ce+21VfOv/xlNnjw5hgwZEt27d48+ffrE9ddfHxUVFVX7Dhs2LLp27Rrz58+P1157rerf6ZNPPlnnfAA0Hr7DDgARMW7cuLj55ptj//33j7POOiu++uqrmDx5cpx44olx4403xqBBg2KbbbaJESNGxLhx42Lq1KnRp0+fiIiYO3dujB8/Po4++ujYeeedv/U9nnnmmRg1alTsvvvuceqpp8bq1atj1qxZ8dZbb8XPf/7zGDBgQHTo0CEeffTRmD59elxzzTUR8Z/vrn+XysrKOOmkk6JZs2Zx4oknxr///e944IEH4rTTTotHH300dtppp6p933jjjTjppJOiRYsWceSRR8YWW2wRs2bNijPPPDN22GGH3P4gv+GSSy6JRx99NHbdddc49dRTo0WLFvH222/HI488En/5y1/iwQcfjC233LJq/zPOOCNSqVQccMAB0b59+5gzZ048/PDD8c4778QzzzwTxcXVf3SZPXt2XH/99XHIIYdE//79o7S0NJ577rmYO3duPP7449GkSXbnJm666ab4wx/+ED/60Y9i+PDhsckmm8ScOXOq/eJmrYcffjieffbZOPTQQ+MnP/lJPP3003HnnXdGeXl5/PrXv46IiJNPPjmGDh0ao0ePjjZt2sSIESMiImLXXXfN9o8SgMYoDQCN3GuvvZbu2rVr+qmnnqq2Xl5enh46dGi6X79+6TVr1qTT6XT6q6++Sg8cODB9yCGHpCsrK9PpdDp9yimnpPfZZ5/0smXLql776KOPprt06ZJ+5ZVXqtaGDh2a7t+/f7q8vLza+6xevbra9jnnnJPu0qVLjZzl5eXpLl26pM8///xq6/3790936dIlffnll1dbf+GFF2rsX15enu7fv396r732Sn/22WfV9p86dWp6u+22S3fp0iX94Ycf1vpn9W3Gjh1b43VPPfVUukuXLunf/OY3NfZ//fXX0127dk2feOKJ1db/+7//u+rPda3bb7893aVLl/Tzzz9ftfbKK6+ku3Tpkt5hhx3S06ZNq7b/yJEja/zZZ+L5559Pd+nSJX3WWWfV+He0YsWKGp+1f//+6c8//7xqfeXKlel99tkn3b1792r7p9PpdN++fdPHHHNMVnkAwCXxADR69957b/zoRz+KvfbaKxYuXFj1v8WLF0ffvn3j008/jbKysoiIaNq0aVxxxRXx7rvvxqRJk+Kll16Kv/71r3H55ZdHy5Ytv/O9Vq9eHZWVldXWSkpKcv4MG220UZxzzjnV1vr16xebb755vPPOO1VrU6dOjXnz5sVJJ51U46Z4e+21V+yzzz45Z1nr3nvvjbZt28bZZ59dY7bbbrvFQQcdFC+99FLMnz+/av2AAw6ocVa8d+/eERHx4Ycf1jjOoEGDYpdddqm2dthhh0VEVPvcmbjzzjujdevWcdVVV9U4k9+8efMa+59yyinRpk2bqu1mzZrF4MGDY9WqVfHBBx9k9d4AUBuXxAPQ6E2bNi0+//zzqkvca7No0aLYZpttIiJijz32iKFDh8ZNN90UG220URx88MHRr1+/73yfI444Ii6//PI4/PDD47TTTosDDjigRjGsq44dO9b6uLf27dvHv//976rttd+x32uvvWo9TpcuXWLKlCk551m1alXMnj07Bg4cGC1atKh1n169esXkyZNj1qxZ1X55MHfu3Hj11Vdjzpw58dFHH1X9smTJkiU1jvH1S/3XWnuJ/dKlSzPOu3r16njrrbfioIMO+ta8mbx3+/bts35vAPg2CjsAjd6SJUtil112idNPP/1b9+ncuXO17bPPPjv69esXa9asibvvvjuj9znqqKNik002ibFjx8ZZZ50VW2yxRZx44okxbNiwKCoqyukzfNuz2Zs3b17tcWJrb/S2tlh+U339AmHJkiWRTqe/9X0ioqqkL1u2LCL+U/J//etfx+TJk6Nt27bRtWvX6NChQ2yzzTZx11131XqM2j53s2bNIiKyeoza4sWLo6KiItq2bZvxa2p777Vn4tPpdMbHAYBvo7AD0Oi1bNkyysvLY++99874NQ888EBsuumm8dVXX8X9998f559/fkavGzx4cBx00EExderUuOOOO+Kaa66J9957L66++uq6xs/K2kK+ZMmSWgvn4sWL6+V91n49YMGCBd+6z9rZpptuGhERt956azz33HMxZsyYGDx4cNV+ZWVl31rY68vakv/1y/MBIN98hx2ARq9r167x/vvvr7Ncft1bb70Vd9xxR4waNSpGjBgR9957b/zjH//I+P1SqVT07ds3JkyYEH379o0nn3wyli9fXm0ekd0Z4kytvax/+vTpNWbpdLrW9brYeOONo1OnTvHGG2/EqlWrat3njTfeiOLi4ujevXtE/Of79Z07d65W1iMi3n///XrJtC6bbrppdOjQId5888346quv6v34qVRqvfz7BKCwKewANHpDhgyJr776Kq677roalzIvWrQoXnjhhartVatWxXnnnRfdu3ePww47LE488cTo0KFDXHTRRVFeXr7O95k3b16NtbZt29Yoc61bt46IqPHs8fpwwAEHRElJSdx8881Vl6Kv9dBDD8Xs2bPr7b2OOeaYmD9/fvz+97+vMZsxY0Y888wzMWTIkKrPW1RUFIsWLapW8FesWBHjxo2rt0zrctRRR8Wnn34aN910U41ZrlcetG7der38+wSgsLkkHoBGb+jQofHnP/85nn766fj4449j0KBB0axZs3j33Xdj8uTJccQRR1TdPf13v/tdfPzxx3HjjTdGKpWKpk2bxq9//es49dRT47bbbotf/vKX3/o+xx9/fHTp0iV22223aNasWcyaNSsmTpwY/+///b9o1apV1X677757TJgwIS644IIYNGhQbLzxxnHIIYfUy2dt165djBo1Kq655poYOnRoHH744bHJJpvE66+/Hq+//nocccQR8eijj9bLex199NHxxhtvxPjx4+Ptt9+OAQMGRPPmzas+90477RQXXnhh1f6DBw+Oq6++Oo499tg4+OCD48svv4xJkybFzjvvHP/85z/rJdO6nHDCCfHmm2/GXXfdFTNnzox99903iouLY9asWTF9+vT4r//6rzofe/fdd4/77rsvLr744thuu+1iu+22i912260e0wNQiBR2ABq9Jk2axLhx42L8+PExceLEGDNmTDRp0iQ6duwYw4YNi+OPPz4iIkpLS+P++++PY445Jrbffvuq1/fv3z/22WefuP322+OAAw6Irl271vo+P/3pT2PSpEnx97//PYqLi6Njx45x6aWXxk9/+tNq++2///4xYsSIePLJJ+O6666LU045pV4/73HHHRdt27aNu+66K8aNGxctW7aMPn36xKOPPhoPP/xwRPznMXG5SqVSMWbMmOjdu3c8/vjjcdNNN8WaNWvihz/8YfziF7+I4cOHV3ufY489NtLpdDz00EPxu9/9Ltq1axdHH3107LfffvGnP/0p5zzfpaioKG655ZZ4+OGH44knnogbb7wxSkpKYtttt40RI0bkdOxf/OIXsWDBgviv//qv+Mtf/hI33HBDPaUGoJCl0m5jCgD8nwsuuCAmTZoUM2bMqJfSDgDUnTPsAEBERFRWVsbLL78cXbt2jY022igqKytj0aJF63xNSUlJ1XfQk2rhwoXrnBcVFUWbNm0aKA0AZE5hB4BGZvbs2dUu6V/r97//fcyfPz9OPfXUiIj49NNPY7/99lvnsXr16hX33XffeslZX/r06bPOeYcOHWLKlCkNlAYAMueSeABoZPr06RMdOnSIvn37Rrt27WLx4sUxZcqUmD59euy3334xbty4aNKkSaxatSpee+21dR6rdevWVY9lS6q//e1v65w3a9YsevXq1UBpACBzCjsANDITJ06MP/3pT/H+++/H4sWLo3nz5tGlS5cYMmRIHHbYYdGkiae+AkASKOwAAACQQH6FDgAAAAnUqG86N3369Ein01FSUpLvKAAAADQC5eXlkUqlYpdddvnOfRt1YU+n0+EbAQAAADSUbDpooy7sa8+sd+vWLc9JAAAAaAxmzpyZ8b6+ww4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAFk74YQTomfPnnHCCSfkOwoAFCyFHQDISllZWcyYMSMiImbMmBFlZWX5DQQABUphBwCycuyxx65zGwCoHwo7AJCx8ePHx/Lly6utLV++PMaPH5+fQABQwBR2ACAjFRUVMXbs2FpnY8eOjYqKigZOBACFTWEHADLyxz/+Mac5AJAdhR0AyMiIESNymgMA2VHYAYCMFBcXx8iRI2udnXXWWVFcXNzAiQCgsCnsAEDGjjvuuGjZsmW1tZYtW8awYcPylAgACpfCDgBkZcKECevcBgDqh8IOAGSlU6dOsfPOO0dExM477xydOnXKbyAAKFC+bAYAZO3uu+/OdwQAKHjOsAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmksAMAAEACKewAAACQQAo7AAAAJJDCDgAAAAmUiMJeWloaRx55ZOy///4xYMCAuO+++6pmn3zySRx//PHRv3//2H///WPSpEnVXvvMM8/EgQceGHvvvXcMGzYsPv7444aODwAAAPWuON8BIiKef/75uPrqq2ObbbaJjz/+OI4++ujo2LFj7LXXXnHaaafF8ccfH4ceemjMmTMnfv7zn0eXLl1i++23j+nTp8eYMWPi/vvvjy233DJuu+22OPPMM+PJJ5/M90cCAACAnCTiDPvFF18c22yzTUREbLXVVnHggQfGK6+8EqWlpVFUVBSHHnpoRER07tw5DjnkkHjqqaciIuL++++P4cOHx5ZbbhkRESeddFLMmzcv3n333fx8EAAAAKgniTjD/k2LFi2KbbbZJmbMmBE9e/asNuvevXs89thjERExffr0OO6446pmxcXFscMOO8Ts2bNju+22y+i90ul0rFixot6yAwAAwLdJp9ORSqUy2jdxhf0f//hHvPDCCzFy5Mi44447ol27dtXmm222WXzxxRcREbFw4cLYfPPNq83btGlTNc9EeXl5zJ49O9fYAAAAkJGmTZtmtF+iCvvkyZPj6quvjtGjR8dWW20VlZWVkU6nq+1TWVlZ9duI2uZr1qzJ+LcVERElJSXRuXPn3MMDAADAd5gzZ07G+yaisFdWVsZvfvObePXVV+Ouu+6qupx90003rXG2fNGiRdG2bduq+eLFi6u+w752/s2z7uuSSqWiRYsWuX8IAAAA+A7ZnGBOxE3nrr766vj444/jiSeeqPbd85122immTZtWbd/p06fHzjvvHBERO+64Y0yfPr1qtnr16njnnXeiR48eDZIbAAAA1pe8F/avvvoqHnroobjmmmtqnOnu379/LFiwoOrZ6zNnzoznn38+jjjiiIiI+NnPfhZ33313fPbZZ1FZWRl/+MMfYo899oitttqqwT8HAAAA1Ke8XxL/8ccfx5o1a+LII4+stt6pU6e466674rbbbotLLrkkRo8eHZtvvnnccMMN8f3vfz8iIvbff//46KOP4vDDD490Oh29evWKq6++Oh8fAwAAAOpVKv3Nu7Y1IjNnzoyIiG7duuU5CQAAAI1BNj0075fEAwAAADUp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBAxfkOAABseHr27Fn1z9OmTctjEgAoXM6wAwBZeeihh9a5DQDUD4UdAMjK9ddfv85tAKB+KOwAQMYGDx6c1ToAUHcKOwCQkaVLl8ann35a6+zTTz+NpUuXNnAiAChsCjsAkJEhQ4bkNAcAsqOwAwAZmThxYk5zACA7CjsAkJFWrVpF+/bta51tueWW0apVqwZOBACFTWEHADI2efLkWtefeeaZBk4CAIVPYQcAsnLuueeucxsAqB8KOwCQlaOOOmqd2wBA/SjOdwAAYMMzbdq0fEcAgILnDDsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBACjsAAAAkkMIOAAAACaSwAwAAQAIp7AAAAJBAxfkOAABseHr27Fn1z9OmTctjEgAoXM6wAwBZueuuu9a5DQDUD4UdAMjKLbfcss5tAKB+KOwAQMYGDBiQ1ToAUHcKOwCQkSVLlsSiRYtqnS1atCiWLFnSwIkAoLAp7ABARg4++OCc5gBAdhR2ACAjEydOzGkOAGRHYQcAMvLII4/kNAcAsqOwAwAZGTFiRE5zACA7CjsAkJHi4uIYOXJkrbOzzjoriouLGzgRABQ2hR0AyNhxxx0XLVu2rLbWsmXLGDZsWJ4SAUDhUtgBgKxMmDBhndsAQP1Q2AGArHTq1Cl23nnniIjYeeedo1OnTvkNBAAFypfNAICs3X333fmOAAAFzxl2AAAASCCFHQAAABJIYQcAAIAESsx32NPpdEyaNCkeeuiheOSRR6rWd9lll9h4442jpKQkIiJ22mmnGDt2bNV8/Pjx8cADD8SqVauie/fucdVVV8X3vve9Bs8PAAAA9SkRhf1vf/tbXH/99bFq1aooKiqqMX/wwQdjq622qrH+7LPPxqRJk+Kxxx6LVq1axZVXXhmXXnpp3HzzzQ0RGwAAANabRFwSv3Llyhg1alRcddVVtc432WSTWtcnTJgQZ5xxRrRu3TqKiorizDPPjClTpsQXX3yxHtMCAADA+peIwj5w4MDo169frbNUKhWtWrWqsV5RURFvv/129OzZs2qtTZs20aFDh3jvvffWW1YAAABoCIm4JH5dUqlUDBgwIEpKSmLXXXeNM888M9q1axeLFy+OysrKaNOmTbX927Rpk9UZ9nQ6HStWrKjn1AAAAFBTOp2OVCqV0b6JL+yvv/56NGnSJJYuXRo33nhjnHbaafHEE09EZWVlRNT8sGvWrMn4w0dElJeXx+zZs+s9NwAAANSmadOmGe2X+MLepMl/rtpv1apVXHTRRbHrrrvGxx9/HJtvvnmk0+lYsmRJtG7dumr/RYsWxeabb57x8UtKSqJz5871HRsAAABqmDNnTsb7Jr6wf92aNWtizZo1UVJSEi1atIhOnTrF9OnTo3///hERsWDBgvj8889ju+22y/iYqVQqWrRosb4iAwAAQJVsrghPxE3nvs3cuXOjrKwsIiJWr14dv/3tb6Nbt27Rvn37iIj42c9+FuPGjYsvv/wyVq9eHWPGjIkjjjgimjdvns/YAAAAkLNEn2H/4osv4pxzzolVq1ZF06ZNY88994yxY8dWzY899tiYP39+DBw4MIqKimK//faLUaNG5TExAAAA1I9UOp1O5ztEvsycOTMiIrp165bnJAAAADQG2fTQRF8SDwAAAI2Vwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEB1Luzjxo2rdf3LL7+MBx54oM6BAAAAgBwK+y233FL7AZs0ieuuu67OgQAAAICI4mx2njdvXtxzzz0REZFOp+Oqq66qNl+zZk3MnDkzfvSjH9VfQgAAAGiEsirsW2yxRaxcuTLWrFkTqVQqli9fXm2eSqWiT58+ceSRR9ZrSAAAAGhsUul0Ol2XF2633Xbx7rvv1neeBjVz5syIiOjWrVuekwAAANAYZNND6/wd9g29rAMAAECSZXVJ/DfNmDEjSktL4/PPP68xu/jii3M5NAAAADRqdS7sd999d1x//fXRs2fP6NChQxQVFdVnLgAAAGjU6lzYJ0yYEL///e9j0KBB9ZkHAAAAiBy+w7506dLYe++96zMLAAAA8H/qXNgHDRoUL730Un1mAQAAAP5PnS+J33333WPMmDHxj3/8I3bccccoLq5+qAMOOCDncAAAANBY1fk57Pvuu++3HzSViueff77OoRqK57ADAADQkLLpoXU+wz5lypS6vhQAAAD4DnX+DvvMmTPjX//6V431f/7zn7WuAwAAAJmrc2E/++yzY9myZTXWFy9eHBdddFFOoQAAAKCxq3Nh/+yzz6JLly411nv06BHvvvtuTqEAAACgsatzYd9iiy3ivffeq7H+ySefRNOmTXMKBQAAAI1dnQv7T37yk7j00kvj008/rVpbsmRJ/Pa3v40BAwbUSzgAAABorOpc2M8444zYeuutY9CgQfHTn/40jjnmmNhvv/1i+fLlcf7559dnRgAgYV588cUYPHhwvPjii/mOAgAFq87PYV/rH//4R0ybNi1Wr14dXbt2jb59+0aTJnX+PUCD8hx2AMjeypUrY+jQobFgwYLYYost4qmnnormzZvnOxYAbBAa5Dnsa3Xv3j26d+9e6+zYY4+NCRMm5PoWAECC3HPPPbFw4cKIiFi4cGGMHz8+TjvttDynAoDCs15Phb/++uvr8/AAQAObO3du3HPPPbH2Ar10Oh333HNPzJ07N8/JAKDwbBjXrgMAeZdOp+Paa6/91vUcv2UHAHyDwg4AZKSsrCxKS0ujsrKy2nplZWWUlpZGWVlZnpIBQGFS2AGAjHTq1Cn23HPPKCoqqrZeVFQUvXv3jk6dOuUpGQAUpvVa2F0aBwCFI5VK1fro1rXrqVQqD6kAoHCt18K+++67r8/DAwANbOutt44jjjii2toRRxwRW221VZ4SAUDhyvmxbvPnz4/PP/+8xtn0HXfcMe67775cDw8AJMyUKVNqbJ977rl5SgMAhavOhX3WrFkxatSoKCsrq1HWU6lUzJ49O+dwAECyPP3007FgwYJqa/Pnz4+nn346fvKTn+QpFQAUplS6jl80P+yww+KHP/xh/OpXv4oOHTpEkyYb3v3rZs6cGRER3bp1y3MSAEi+ysrK2HPPPaOioqLGrLi4OEpLS2vckA4AqC6bHlrnlv3BBx/EZZddFltttdUGWdYBgOw88cQTtZb1iIiKiop44oknGjgRABS2Ojftzp07x6JFi+ozCwCQYIcddlgUF9f+bbri4uI47LDDGjgRABS2Ohf2yy+/PEaPHh0ffPBBfeYBABKqqKgoLrnkklpnl112mcvhAaCe1fk77LvsskuUl5dHZWVlNG3atMZf0tOmTauXgOuT77ADQPYOPPDAmD9/ftX297///Xj22WfzmAgANhzZ9NA63yX+9ttvr+tLAYAN2N133x2DBw+u2r7rrrvymAYACledC3uvXr3qMwcAsIFo37597L///vE///M/sf/++0f79u3zHQkAClKdL4mPiPjss8/i7bffjmXLltWYDRkyJJdcDcIl8QAAADSkBrkkftKkSXHRRRfF5ptvHgsXLoxOnTrFvHnzoqSkJDp06LBBFHYAAABIqjrfJf4Pf/hDjB49Ol544YVo3bp1PPnkk/Hyyy9H37594+STT67PjAAAANDo1Lmwz58/PwYMGBARET/4wQ9i3rx50bx58zjvvPNizJgx9RYQAAAAGqM6F/Ytt9wyPvzww4iI6Nq1a7z11lsR8Z9ntH7++ef1Eg4AAAAaqzp/h/3ggw+OO+64I2644YYYPHhwjBo1Kj755JN4+eWXY/fdd6/PjAAAANDo5HSX+K+bPHlyTJ48Odq2bRsjR46MzTbbrD4Ou165SzwAAAANqUHuEv9NgwcPjsGDB9fX4QAAAKBRq/N32CMiSktLY9iwYfHjH/84evbsGcOHD4833nijvrIBAABAo1Xnwv7ss8/G6aefHrvttlvccMMNccstt8Suu+4ap59+erz66qv1mREASJi+fftGz549o2/fvvmOAgAFq86XxN92221xxRVXxCGHHFK1tueee0anTp3ixhtvjIceeqheAgIAyTJ16tRYvnx5REQsX748pk6dGn369MlzKgAoPHU+w/7hhx/GvvvuW2O9f//+MWvWrJxCAQDJNXLkyHVuAwD1o86FvVWrVrF48eIa64sXL46mTZvmFAoASKZf/OIXWa0DAHVX58Ler1+/uPbaa2P16tVVa+Xl5XH99dfH3nvvXS/hAIDkWLVqVbz88su1zl5++eVYtWpVAycCgMJW5++wjxo1Ko499tg44IADYtddd40mTZrEm2++GSUlJXHffffVZ0YAIAFOO+2075zfc889DZQGAApfKp1Op+v64tWrV8ezzz4bM2fOjIqKithhhx3ikEMOiebNm9dnxvUmmwfWA0Bjt2rVqujdu/e3zl9++eVo1qxZAyYCgA1PNj20zmfYIyKaNm0aQ4YMiSFDhuRyGABgA9CsWbPo3bt3rZfF9+3bV1kHgHqWcWH/85//nNWBDzjggKzDAADJNm7cuOjZs2eN9ZtuuikPaQCgsGVc2EePHp3xQVOplMIOAAVq7Nix1R7lNnbs2DymAYDClXFhnzJlyvrMAQBsIPr06RMtW7aM5cuXR8uWLaNPnz75jgQABSmn77B/l/322y+ef/759fkWAEAevPTSS/mOAAAFr87PYc/EvHnz1ufhAQAAoGCt18KeSqXW5+EBAACgYK3Xwg4AAADUjcIOAAAACbReC3s6nV6fhwcAAICCtV4L+9ChQ9fn4QEAAKBg1fmxbv/7v//7rbOioqLYZJNN4pprrqnr4QEAAKBRq3Nh33fffavuAp9Op6vdEX7tdufOnePUU0+NwYMH554UAAAAGpE6F/Zbbrklbrnllrjggguic+fOUV5eHm+++Wbce++9cdFFF0VRUVH89a9/jUsuuSRSqVQcdNBB9ZkbAMijE044IWbMmBE777xz3H333fmOAwAFqc6F/aabboprrrkmdtxxx6q1gw46KL73ve/FzTffHHfccUfsuOOO0a5du7jrrrsUdgAoEGVlZTFjxoyIiJgxY0aUlZVFp06d8hsKAApQnW86V1ZWFh07dqyx3qNHj3jjjTeqtgcMGBBz5syp69sAAAlz7LHHrnMbAKgfdS7sHTt2jOeff77GemlpabRs2bJqu6KiIkpKSur6NgBAgowfPz6WL19ebW358uUxfvz4/AQCgAJW50vif/nLX8YFF1wQc+bMiV69ekWzZs1i2rRpceedd8ZJJ51Utd+LL74YnTt3rpewAED+VFRUxNixY2udjR07No455pgoLq7zjxYAwDfU+W/VgQMHRtu2beO2226LiRMnxrJly+KHP/xhXHjhhXHooYdW7de0adM4++yz6yUsAJA/f/zjH79zfvrppzdQGgAofKl0Op3Od4h8mTlzZkREdOvWLc9JACD5KioqolevXt86f+2115xhB4DvkE0Pzflv1fnz58fnn38e3+z9X797PACw4SsuLo6RI0fWeln8WWedpawDQD2r89+ss2bNilGjRkVZWVmNsp5KpWL27Nk5hwMAkmXfffettbD369cvD2kAoLDVubBfcsklsf3228ftt98eHTp0iCZN6nzDeQBgA5BOp+Paa6+tdXbttdfGuHHjIpVKNXAqAChcdS7sH3zwQdxzzz2xySab1GceACChysrKorS0tNZZaWlplJWVxTbbbNPAqQCgcNX5tHjnzp1j0aJF9ZkFAEiwrbfeOqc5AJCdOhf2yy+/PEaPHh0ffPBBfeYBABLqySefzGkOAGSnzo9122WXXaK8vDwqKyujadOmUVRUVG0+bdq0egm4PnmsGwBkrrKyMvbcc8+oqKioMSsuLo7S0tIaPw8AANU1yGPdbr/99rq+FADYABUVFcUll1wSl112WY3ZZZddpqwDQD2r8xn2QuAMOwBkb999940vvviiart169YxZcqU/AUCgA3IejnD/u6778Z2221Xtf3666+vc//dd98900MDABuQr5f12rYBgPqRcWE/66yz4rnnnqvaHjFixLfum0qlNojvsAMA2fnFL37xrevjxo1r4DQAUNgyLux/+tOfqm1Pnz693sMAAMm1atWqePnll2udvfzyy7Fq1apo1qxZA6cCgMKV8WPdSkpKqm1PnDix1v2WLVsWf/3rX3MKBQAkz2mnnZbTHADITp2fw37hhRfWul5cXBxnnXVWnQMBAMl066235jQHALKT1WPd5s+fH88++2xERKTT6bjnnnuqzdesWRPTpk2LLbfcsv4SAgCJ0KxZs+jdu3etl8X37dvX5fAAUM+yKuybbrppvPjii1FZWRmpVKrGI1yaNGkS7dq1ixtvvDHrIOl0OiZNmhQPPfRQPPLII1Xrs2bNissvvzwWLlwYzZs3j4suuij22muvqvn48ePjgQceiFWrVkX37t3jqquuiu9973tZvz8A8N3GjRsXPXv2rLF+00035SENABS2rAp7s2bNYvz48RERsd1228V9991XLyH+9re/xfXXXx+rVq2KoqKiqvVly5bFqaeeGqNHj47evXvHa6+9Fqeffno899xz0bZt23j22Wdj0qRJ8dhjj0WrVq3iyiuvjEsvvTRuvvnmeskFANQ0duzYGDlyZLVtAKD+1fk77M8//3y9hVi5cmWMGjUqrrrqqmrrkydPjm7dukXv3r0jIqJXr16x2267VV2WP2HChDjjjDOidevWUVRUFGeeeWZMmTLF82ABYD3q06dPtGzZMiIiWrZsGX369MlzIgAoTHUu7K+88krMmDGjavvFF1+MQw89NEaOHJl1YR44cGD069evxvqMGTNqXHbXo0ePePfdd6OioiLefvvtavM2bdpEhw4d4r333svq/QGA7Lz00ksxbdq0eOmll/IdBQAKVlaXxH/dmDFj4sEHH4yIiEWLFsXZZ58dJ598cnzyySdx9dVXx3XXXZdzuAULFsQee+xRbW2zzTaLt956KxYvXhyVlZXRpk2bavM2bdpk9QuDdDodK1asyDkrAAAAfJd0Oh2pVCqjfetc2JcvXx7t27ePiIiHHnooevfuHaeeemosW7YsDjjggLoetprKyspa11KpVNXsmx92zZo1GX/4iIjy8vKYPXt27mEBAAAgA02bNs1ovzoX9u222y5efvnl6N69ezz44INVd4ctLy+PlStX1vWw1Wy66aaxePHiamuLFi2Ktm3bxiabbBLpdDqWLFkSrVu3rjbffPPNM36PkpKS6Ny5c73kBQAAgHWZM2dOxvvWubCfeeaZcdppp0V5eXkMHDgwdtttt4iImDp1amyzzTZ1PWw1O+64Y0yfPj2OP/74qrXp06fHQQcdFC1atIhOnTrF9OnTo3///hHxn0voP//889huu+0yfo9UKhUtWrSol7wA0Fjsueee8dVXX8VGG20UpaWl+Y4DABuMbK4Ir/NN5/bcc8+YMmVKPPHEEzFmzJiq9R122KHadi4OOeSQKC0trfpB4MUXX4x//etfMWjQoIiI+NnPfhbjxo2LL7/8MlavXh1jxoyJI444Ipo3b14v7w8A1DRlypT46quvIiLiq6++iilTpuQ5EQAUpjqfYY/4zw3evnnTt2233TanQF/3/e9/P8aMGRNXXHFFLFmyJDp27Bi33npr1RnxY489NubPnx8DBw6MoqKi2G+//WLUqFH19v4AQE3f/Lt21KhRMW3atDylAYDClUqn0+lMdhw1alT87ne/q9oeNmzYOk/lT5gwIfd069nMmTMjIqJbt255TgIAG4YTTzwxpk+fXmN9l112ibvuuisPiQBgw5JND834DPs3n4e+9nvjAEDjsHLlylrLesR/7jGzcuVKX0sDgHqU8Rn2QuQMOwBk7qijjop//vOf3zrv2rVrPPTQQw2YCAA2PNn00DrfdG7kyJGxaNGiGutffvllnH/++XU9LACQUHfffXdOcwAgO3Uu7K1bt47BgwfHU089VbX2l7/8JQ466KBaizwAsGFr1qxZbLzxxrXONt5442jWrFkDJwKAwlbnu8RfeeWVceSRR8ZvfvObePrpp2PTTTeNmTNnxuWXXx4DBgyoz4wAQAKUlZXFsmXLap0tW7YsysrKYptttmngVABQuOp8hj3iP89cv/DCC+O9996LF154Ia6++mplHQAKVKdOnWLPPfesdda7d+/o1KlTAycCgMJW5zPs8+bNizFjxsQrr7wS5557brRp0ybOO++86NevX5xzzjmxySab1GdOACDPUqlUnH/++XHYYYdFZWVl1XpxcXGcf/7563zcKwCQvTqfYT/wwAOjefPmMXny5BgyZEjsvffeMWnSpFi2bFkMGjSoPjMCAAmx9dZbx/HHH19VzlOpVBx//PGx1VZb5TkZABSeOj/W7Y033ogf/ehH8eKLL8b//u//xqmnnlo1++///u8YOHBgvYVcXzzWDQCyt3Llyhg6dGgsWLAgtthii3jqqac8fx0AMtQgj3XbaKONYtCgQXHffffFTTfdVLV+7rnnxkcffVTXwwIACde8efO48MILo3379nHhhRcq6wCwntS5sI8ePTrOOeeceOyxx+LrJ+lPPPHE+NOf/lQv4QCAZOrXr19Mnjw5+vXrl+8oAFCw6lzYZ82aFYMHD46IqHaTmY4dO8a8efNyTwYAAACNWJ0L+/e+971ai/msWbOiZcuWOYUCAACAxq7Ohf2YY46Jiy++OD755JOIiKioqIjXXnstLrrooqoz7wAAAEDd1Pk57CeccEKsWrUqDjnkkEin09GjR49o0qRJHHnkkXHOOefUZ0YAAABodOr8WLe1vvrqq3jvvfeipKQkOnbsuEHdKdZj3QAAAGhI2fTQOp9hX2ujjTZSeAGgkdl1110jnU5HKpWKN998M99xAKAg1fk77ABA4/T4449XPdI1nU7H448/nudEAFCYFHYAICtXX331OrcBgPqhsAMAGRsyZEhW6wBA3SnsAEBGli1bFnPnzq11Nnfu3Fi2bFkDJwKAwqawAwAZOeyww3KaAwDZUdgBgIw88cQTOc0BgOwo7ABARjbeeOPYeuuta5117NgxNt544wZOBACFTWEHADI2ceLEWtefeuqphg0CAI2Awg4AAAAJpLADABm78cYbs1oHAOpOYQcAMlJeXh4TJkyodTZhwoQoLy9v4EQAUNgUdgAgI9ddd11OcwAgOwo7AJCR8847L6c5AJAdhR0AyEhRUVE0bdq01lnTpk2jqKiogRMBQGFT2AGAjEydOjVWr15d62z16tUxderUBk4EAIVNYQcAMrLXXntFcXFxrbPi4uLYa6+9GjgRABQ2hR0AyMhHH30UFRUVtc4qKirio48+auBEAFDYFHYAICOdOnWKPffcs9ZZ7969o1OnTg2cCAAKm8IOAGQklUrF+eefH6lUqtp6kyZNal0HAHKjsAMAGdt6663jxBNPrLZ24oknxlZbbZWnRABQuBR2ACArRx11VDRp8p8fIZo0aRJHHnlknhMBQGFS2AGArDz00EORTqcjIiKdTsfDDz+c50QAUJgUdgAgY3Pnzo177rmnWmG/5557Yu7cuXlOBgCFR2EHADKSTqfj2muv/db1tSUeAKgfCjsAkJGysrIoLS2NysrKauuVlZVRWloaZWVleUoGAIVJYQcAMuI57ADQsBR2ACAjqVQq2rVrV+usXbt2nsMOAPVMYQcAMlJeXh4TJ06sdfbUU09FeXl5wwYCgAKnsAMAGbnuuutymgMA2VHYAYCMnHfeeTnNAYDsKOwAQEZKSkri2GOPrXV2/PHHR0lJSQMnAoDCprADABn71a9+VaOYl5SUxC9/+cs8JQKAwqWwAwBZOeqoo9a5DQDUD4UdAMjKhAkT1rkNANQPhR0AyNjgwYOzWgcA6k5hBwAysnTp0vj0009rnX366aexdOnSBk4EAIVNYQcAMjJkyJCc5gBAdhR2ACAjjz32WE5zACA7CjsAkJFbb701pzkAkB2FHQDIyHnnnZfTHADIjsIOAGQklUrlNAcAsqOwAwAZ+eMf/5jTHADIjsIOAGTk5JNPzmkOAGRHYQcAMvLxxx/nNAcAsqOwAwAZ6dSpU7Rt27bWWdu2baNTp04NnAgACpvCDgBkpLy8PBYuXFjrbOHChVFeXt7AiQCgsCnsAEBGLrnkkpzmAEB2FHYAICO/+c1vcpoDANlR2AGAjDRt2jSGDBlS6+ywww6Lpk2bNmwgAChwCjsAkLPKysp8RwCAgqOwAwAZWb16dUycOLHW2cSJE2P16tUNGwgACpzCDgBkxE3nAKBhKewAQEYUdgBoWAo7AJARd4kHgIalsAMAGVHYAaBhKewAQEY81g0AGpbCDgBk7Ljjjqt1fdiwYQ0bBAAaAYUdAMhIOp2Oa6+9ttbZtddeG+l0uoETAUBhU9gBgIyUlZVFaWlprbPS0tIoKytr4EQAUNgUdgAgIz/4wQ9ymgMA2VHYAYCMXH/99TnNAYDsKOwAQEbOO++8nOYAQHYUdgAgI6lUKqc5AJAdhR0AyMgf//jHnOYAQHYUdgAgIyNGjMhpDgBkR2EHADJSXFwcI0eOrHV21llnRXFxcQMnAoDCprADABm78847a12//fbbGzgJABQ+hR0AyMiiRYtixYoVtc5WrFgRixYtauBEAFDYFHYAICMHHnhgTnMAIDsKOwCQkYkTJ+Y0BwCyo7ADABm56qqrcpoDANlR2AGAjIwZMyanOQCQHYUdAMjIqlWrcpoDANlR2AGAjBx88ME5zQGA7CjsAEBGHnjggZzmAEB2FHYAICNHH310TnMAIDsKOwCQkWeeeSanOQCQHYUdAMhI8+bNc5oDANlR2AGAjFxyySU5zQGA7CjsAEBGLr300pzmAEB2FHYAICM33nhjTnMAIDsKOwCQkTPOOCOnOQCQHYUdAMjIMccck9McAMiOwg4AZOSRRx7JaQ4AZEdhBwAy4rFuANCwFHYAICNTp07NaQ4AZEdhBwAy0qNHj5zmAEB2FHYAICOHHnpoTnMAIDsKOwCQkYkTJ+Y0BwCyo7ADABlZunRpTnMAIDsKOwCQkYMPPjinOQCQHYUdAMjI3XffndMcAMiOwg4AZOTEE0/MaQ4AZEdhBwAy8txzz+U0BwCyo7ADABnZfPPNc5oDANlJfGG/8sorY9ddd41999236n/z5s2LiIhZs2bFT3/60+jfv38cdNBB8fe//z3PaQGgcE2dOjWnOQCQneJ8B8jE8OHDY+TIkdXWli1bFqeeemqMHj06evfuHa+99lqcfvrp8dxzz0Xbtm3zlBQAClePHj1ymgMA2Un8GfaIiE022aTG2uTJk6Nbt27Ru3fviIjo1atX7LbbbvHss882dDwAaBQOPfTQnOYAQHY2iMLeqlWrGmszZsyInj17Vlvr0aNHvPvuuw0VCwAalYkTJ+Y0BwCys0FcEj9mzJi4+eabY6uttopTTjkl+vTpEwsWLIg99tij2n6bbbZZvPXWW1kdO51Ox4oVK+ozLgAUpE8++eQ75x07dmygNACwYUqn05FKpTLaN/GF/eKLL45LL700KisrY+rUqfGrX/0qxo8fH5WVlTX2rayszPiDr1VeXh6zZ8+ur7gAULBOPvnkdc6PPvrouOOOOxooDQBsuJo2bZrRfokv7E2a/Oeq/aKioujXr18MHjw4/vKXv8Smm24aixcvrrbvokWLsr7hXElJSXTu3Lne8gJAobr33ntj+PDh65xvu+22DZgIADY8c+bMyXjfxBf2b6qsrIySkpLYcccdY/r06XH88cdXzaZPnx4HHXRQVsdLpVLRokWL+o4JAAXnpZde+s55t27dGigNAGyYsrkqPPE3nXvppZdizZo1EfGf57v++c9/joEDB8YhhxwSpaWlUVpaGhERL774YvzrX/+KQYMG5TMuABSsESNG5DQHALKT+DPs48ePj/POOy+aN28e7du3j3HjxlVdwj5mzJi44oorYsmSJdGxY8e49dZbnS0HgPXkhRde+M75gAEDGiYMADQCqXQ6nc53iHyZOXNmRITL9wAgA998nGptpk2b1gBJAGDDlU0PTfwl8QBAMvz617/OaQ4AZEdhBwAy8vvf/z6nOQCQHYUdAMjIww8/nNMcAMiOwg4AZOS4447LaQ4AZEdhBwAyMnHixJzmAEB2FHYAICPz5s3LaQ4AZEdhBwAy8vOf/zynOQCQHYUdAMjI1VdfndMcAMiOwg4AZMRz2AGgYSnsAEBGnnzyyZzmAEB2FHYAICMrV67MaQ4AZEdhBwAycvTRR+c0BwCyo7ADABm55pprcpoDANlR2AGAjFx44YU5zQGA7CjsAEBGJk6cmNMcAMiOwg4AZGThwoU5zQGA7CjsAEBGTj755JzmAEB2FHYAICO/+93vcpoDANlR2AGAjFx22WU5zQGA7CjsAEBGnnnmmZzmAEB2FHYAICNffPFFTnMAIDsKOwCQkaFDh+Y0BwCyo7ADABm58MILc5oDANlR2AGAjFxzzTU5zQGA7CjsAEBGbrvttpzmAEB2FHYAICPNmjXLaQ4AZEdhBwAyctxxx+U0BwCyo7ADABk599xzc5oDANlR2AGAjFx//fU5zQGA7CjsAEBGdt1115zmAEB2FHYAICNvvvlmTnMAIDsKOwCQkWOPPTanOQCQHYUdAMhI06ZNc5oDANlR2AGAjNx55505zQGA7CjsAEBG2rVrl9McAMiOwg4AZGT+/Pk5zQGA7CjsAEBGrrjiipzmAEB2FHYAICO33XZbTnMAIDsKOwCQkU8//TSnOQCQHYUdAAAAEkhhBwAAgARS2AGAjFx44YU5zQGA7CjsAEBGSkpKcpoDANlR2AGAjFx55ZU5zQGA7CjsAAAAkEAKOwAAACSQwg4AZGT77bfPaQ4AZEdhBwAy0qTJun9s+K45AJAdf7MCABl55513cpoDANlR2AEAACCBFHYAAABIIIUdAMhIz549c5oDANlR2AGAjEybNi2nOQCQHYUdAAAAEkhhBwAAgARS2AEAACCBFHYAAABIIIUdAAAAEkhhBwAAgARS2AEAACCBFHYAAABIIIUdAAAAEkhhBwAAgARS2AEAACCBFHYAAABIIIUdAAAAEkhhBwAAgARS2AEAACCBFHYAAABIIIUdAAAAEkhhBwAAgARS2AEAACCBFHYAAABIIIUdAAAAEkhhBwAAgARS2AEAACCBFHYAAABIIIUdAAAAEkhhBwAAgARS2AEAACCBFHYAAABIIIUdAAAAEkhhBwAAgARS2AEAACCBFHYAAABIoOJ8BwAoNEuXLo0PP/ww3zEgL2bOnJnvCFCvfvjDH0arVq3yHQNopBR2gHq0dOnSOPjgg2Pp0qX5jgJ5MXz48HxHgHrVqlWreOaZZ5R2IC9cEg8AAAAJ5Aw7QD1aeybGJfEUstrOot977715SALrn0vigXxS2AHqWatWraJbt275jgENyv/nAaD+uSQeAMjKN8+mT5s2LU9JAKCwKewAQJ25FB4A1h+FHQAAABJIYQcAAIAEUtgBAAAggRR2AAAASCCFHQAAABJIYQcAAIAEUtgBAAAggRR2AAAASKDifAeg8VixYkV88MEH+Y4BQI6+/t9y/10HKAzbbrtttGjRIt8x+AaFnQbzwQcfxPDhw/MdA4B6dOWVV+Y7AgD14N57741u3brlOwbf4JJ4AAAASCBn2MmLxdv8OMpbtM53DADqKFVZHhER6aKSPCcBoK5KVnwR3/vXK/mOwToo7ORFeYvWUd6qbb5jAAAAJJbCTl4Ur/gi3xEAAKBR8zN58insNJgVK1ZU/XMbl94AAEBifP1ndZLDTecAAAAggZxhp8F8/bmOi7b5cVS46RwAAORN8Yovqq589Qz2ZFLYyYsKN50DAABYJ4WdvChxgwuADZrHugFs+PxMnnwKO3nheY8AAADrVhA3nVu1alVccskl0b9//9h7773juuuui3Q6ne9YAAAAUGcFcYZ99OjRsWbNmvif//mfWLlyZRx33HFx//33x7Bhw/Idja/Zdttt49577813DFivVqxYEeecc45HowAUiBYtWsQNN9zghlwUvG233TbfEajFBl/Yly9fHhMnTowXXnghiouLo1WrVnHKKafEH/7wB4U9YVq0aBHdunXLdwxYr5YuXRpFRUX5jgFAPSkqKooddtghWrVqle8oQCO0wRf2d955J37wgx9E69atq9Z69OgR77//flRWVn7nD87pdNqZMKDeFBUVxWOPPRYfffRRvqPAerVy5cqIiGjevHmek8D61bFjxygqKvLzIlBv0ul0pFKpjPbd4Av7woULY7PNNqu21qZNm6ioqIilS5dWK/K1KS8vj9mzZ6/HhEBj1KRJQdwiBL5Vy5Yt8x0BGsTHH3+c7whAAWratGlG+23whb2ioqLGDebWrFkTEZHRby1KSkqic+fO6yUbAAAAfN2cOXMy3neDL+ytW7eOxYsXV1tbtGhRbLTRRhl91yiVSrmJCAAAAA0i08vhIwrgsW477LBDlJWVxZIlS6rWpk+fHj169HBJKgAAABusDb7Rtm3bNvr27RtjxoyJioqKWLRoUdx6660xfPjwfEcDAACAOtvgC3tExG9/+9tYsGBB9OnTJw4//PD42c9+FgMGDMh3LAAAAKizDf477BH/uSv8rbfemu8YAAAAUG8K4gw7AAAAFBqFHQAAABJIYQcAAIAEUtgBAAAggRR2AAAASCCFHQAAABJIYQcAAIAEUtgBAAAggRR2AAAASCCFHQAAABJIYQcAAIAEUtgBAAAggRR2AAAASCCFHQAAABJIYQcAAIAEUtgBAAAggRR2AAAASCCFHQAAABKoON8B8qm8vDzS6XTMnDkz31EAAABoBFavXh2pVCqjfRt1Yc/0DwkAAADqQyqVyriLptLpdHo95wEAAACy5DvsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAinsAAAAkEAKOwAAACSQwg4AAAAJpLADAABAAv1//4uBH9qTSSMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'existing_loan_cnt', data = user_spec)\n",
    "plt.title('existing_loan_cnt', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ea4cbf8e",
   "metadata": {},
   "source": [
    "### existing_loan_amt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "ca478aa6",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(y = 'existing_loan_amt', data = user_spec)\n",
    "plt.title('existing_loan_amt', size = 15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3ec57ea",
   "metadata": {},
   "source": [
    "- credit_score, existing_loan_cnt,amt -> knn_imputer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "faea31af",
   "metadata": {},
   "source": [
    "### Knn Imputer V2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "cadc7029",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                              0\n",
       "user_id                                     0\n",
       "insert_time                                 0\n",
       "credit_score                            69483\n",
       "yearly_income                               0\n",
       "income_type                                 0\n",
       "employment_type                             0\n",
       "houseown_type                               0\n",
       "desired_amount                              0\n",
       "purpose                                     0\n",
       "personal_rehabilitation_yn                  0\n",
       "personal_rehabilitation_complete_yn         0\n",
       "existing_loan_cnt                      146290\n",
       "existing_loan_amt                      225046\n",
       "enter_year                                  0\n",
       "enter_month                                 0\n",
       "enter_birth                                 0\n",
       "month                                       0\n",
       "birth_year                                  0\n",
       "gender                                   6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "id": "720e2b2e",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec_train = user_spec[user_spec['month'] <6].reset_index(drop=True)\n",
    "loan_result_train = loan_result[loan_result['month'] <6].reset_index(drop=True)\n",
    "log_data_train = log_data[log_data['month'] <6].reset_index(drop=True)\n",
    "\n",
    "user_spec_test = user_spec[user_spec['month'] ==6].reset_index(drop=True)\n",
    "loan_result_test = loan_result[loan_result['month'] ==6].reset_index(drop=True)\n",
    "log_data_test = log_data[log_data['month'] ==6].reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "id": "de10d098",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(968753, 20)"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "e7140a86",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(728884, 20)"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "099299ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "knn_col = ['credit_score','yearly_income','desired_amount','existing_loan_cnt','existing_loan_amt']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "3e2adc35",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler().fit(user_spec_train[knn_col])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "25a4c9bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_knn = scaler.transform(user_spec_train[knn_col])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "09a59eaf",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_knn = scaler.transform(user_spec[knn_col])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "6a76efb3",
   "metadata": {},
   "outputs": [],
   "source": [
    "imputer = KNNImputer(n_neighbors = 5).fit(train_knn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "7429e72a",
   "metadata": {},
   "outputs": [],
   "source": [
    "knn_test = imputer.transform(test_knn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "bc53ad6f",
   "metadata": {},
   "outputs": [],
   "source": [
    "knn_test_scaler = scaler.inverse_transform(knn_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "8e794da4",
   "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>credit_score</th>\n",
       "      <th>yearly_income</th>\n",
       "      <th>desired_amount</th>\n",
       "      <th>existing_loan_cnt</th>\n",
       "      <th>existing_loan_amt</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>870.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>27000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>710.0</td>\n",
       "      <td>30000000.0</td>\n",
       "      <td>10000000.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>15000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>820.0</td>\n",
       "      <td>62000000.0</td>\n",
       "      <td>2000000.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>344000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>630.0</td>\n",
       "      <td>36000000.0</td>\n",
       "      <td>5000000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>16000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>590.0</td>\n",
       "      <td>45000000.0</td>\n",
       "      <td>15000000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968748</th>\n",
       "      <td>590.0</td>\n",
       "      <td>25000000.0</td>\n",
       "      <td>5000000.0</td>\n",
       "      <td>11.2</td>\n",
       "      <td>68800000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968749</th>\n",
       "      <td>980.0</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>65600000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968750</th>\n",
       "      <td>750.0</td>\n",
       "      <td>75000000.0</td>\n",
       "      <td>100000000.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>200000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968751</th>\n",
       "      <td>640.0</td>\n",
       "      <td>50000000.0</td>\n",
       "      <td>10000000.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>117000000.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>968752</th>\n",
       "      <td>656.0</td>\n",
       "      <td>35000000.0</td>\n",
       "      <td>20000000.0</td>\n",
       "      <td>6.6</td>\n",
       "      <td>36200000.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>968753 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        credit_score  yearly_income  desired_amount  existing_loan_cnt  \\\n",
       "0              870.0     30000000.0      30000000.0                1.0   \n",
       "1              710.0     30000000.0      10000000.0                5.0   \n",
       "2              820.0     62000000.0       2000000.0                7.0   \n",
       "3              630.0     36000000.0       5000000.0                1.0   \n",
       "4              590.0     45000000.0      15000000.0                1.0   \n",
       "...              ...            ...             ...                ...   \n",
       "968748         590.0     25000000.0       5000000.0               11.2   \n",
       "968749         980.0     20000000.0      50000000.0                1.0   \n",
       "968750         750.0     75000000.0     100000000.0                8.0   \n",
       "968751         640.0     50000000.0      10000000.0               10.0   \n",
       "968752         656.0     35000000.0      20000000.0                6.6   \n",
       "\n",
       "        existing_loan_amt  \n",
       "0              27000000.0  \n",
       "1              15000000.0  \n",
       "2             344000000.0  \n",
       "3              16000000.0  \n",
       "4              20000000.0  \n",
       "...                   ...  \n",
       "968748         68800000.0  \n",
       "968749         65600000.0  \n",
       "968750        200000000.0  \n",
       "968751        117000000.0  \n",
       "968752         36200000.0  \n",
       "\n",
       "[968753 rows x 5 columns]"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn_df = pd.DataFrame(knn_test_scaler,columns = knn_col);knn_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "16512375",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.credit_score = knn_df.credit_score\n",
    "user_spec.existing_loan_cnt = knn_df.existing_loan_cnt\n",
    "user_spec.existing_loan_amt = knn_df.existing_loan_amt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "b56a56f8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                            0\n",
       "user_id                                   0\n",
       "insert_time                               0\n",
       "credit_score                              0\n",
       "yearly_income                             0\n",
       "income_type                               0\n",
       "employment_type                           0\n",
       "houseown_type                             0\n",
       "desired_amount                            0\n",
       "purpose                                   0\n",
       "personal_rehabilitation_yn                0\n",
       "personal_rehabilitation_complete_yn       0\n",
       "existing_loan_cnt                         0\n",
       "existing_loan_amt                         0\n",
       "enter_year                                0\n",
       "enter_month                               0\n",
       "enter_birth                               0\n",
       "month                                     0\n",
       "birth_year                                0\n",
       "gender                                 6891\n",
       "dtype: int64"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "a4ee615d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(968753, 20)"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7a14b3b",
   "metadata": {},
   "source": [
    "### Gender"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a45957ec",
   "metadata": {},
   "source": [
    "- Feature\n",
    "    - 다 갖다 씀\n",
    "    - Model : Pycaret Model에서 성능 높은 모델로 채우자"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb5cbbec",
   "metadata": {},
   "source": [
    "#### Pycaret"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "d36dfae5",
   "metadata": {},
   "outputs": [],
   "source": [
    "gender_df = user_spec[user_spec.gender.notna()]\n",
    "gender_train = gender_df[gender_df.month < 6]\n",
    "gender_test = user_spec[user_spec.gender.isna()].drop('gender',axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "a5526389",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_90011_row44_col1 {\n",
       "  background-color: lightgreen;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_90011\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_90011_level0_col0\" class=\"col_heading level0 col0\" >Description</th>\n",
       "      <th id=\"T_90011_level0_col1\" class=\"col_heading level0 col1\" >Value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row0\" class=\"row_heading level0 row0\" >0</th>\n",
       "      <td id=\"T_90011_row0_col0\" class=\"data row0 col0\" >session_id</td>\n",
       "      <td id=\"T_90011_row0_col1\" class=\"data row0 col1\" >123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row1\" class=\"row_heading level0 row1\" >1</th>\n",
       "      <td id=\"T_90011_row1_col0\" class=\"data row1 col0\" >Target</td>\n",
       "      <td id=\"T_90011_row1_col1\" class=\"data row1 col1\" >gender</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row2\" class=\"row_heading level0 row2\" >2</th>\n",
       "      <td id=\"T_90011_row2_col0\" class=\"data row2 col0\" >Target Type</td>\n",
       "      <td id=\"T_90011_row2_col1\" class=\"data row2 col1\" >Binary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row3\" class=\"row_heading level0 row3\" >3</th>\n",
       "      <td id=\"T_90011_row3_col0\" class=\"data row3 col0\" >Label Encoded</td>\n",
       "      <td id=\"T_90011_row3_col1\" class=\"data row3 col1\" >0.0: 0, 1.0: 1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row4\" class=\"row_heading level0 row4\" >4</th>\n",
       "      <td id=\"T_90011_row4_col0\" class=\"data row4 col0\" >Original Data</td>\n",
       "      <td id=\"T_90011_row4_col1\" class=\"data row4 col1\" >(724051, 20)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row5\" class=\"row_heading level0 row5\" >5</th>\n",
       "      <td id=\"T_90011_row5_col0\" class=\"data row5 col0\" >Missing Values</td>\n",
       "      <td id=\"T_90011_row5_col1\" class=\"data row5 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row6\" class=\"row_heading level0 row6\" >6</th>\n",
       "      <td id=\"T_90011_row6_col0\" class=\"data row6 col0\" >Numeric Features</td>\n",
       "      <td id=\"T_90011_row6_col1\" class=\"data row6 col1\" >11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row7\" class=\"row_heading level0 row7\" >7</th>\n",
       "      <td id=\"T_90011_row7_col0\" class=\"data row7 col0\" >Categorical Features</td>\n",
       "      <td id=\"T_90011_row7_col1\" class=\"data row7 col1\" >7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row8\" class=\"row_heading level0 row8\" >8</th>\n",
       "      <td id=\"T_90011_row8_col0\" class=\"data row8 col0\" >Ordinal Features</td>\n",
       "      <td id=\"T_90011_row8_col1\" class=\"data row8 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row9\" class=\"row_heading level0 row9\" >9</th>\n",
       "      <td id=\"T_90011_row9_col0\" class=\"data row9 col0\" >High Cardinality Features</td>\n",
       "      <td id=\"T_90011_row9_col1\" class=\"data row9 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row10\" class=\"row_heading level0 row10\" >10</th>\n",
       "      <td id=\"T_90011_row10_col0\" class=\"data row10 col0\" >High Cardinality Method</td>\n",
       "      <td id=\"T_90011_row10_col1\" class=\"data row10 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row11\" class=\"row_heading level0 row11\" >11</th>\n",
       "      <td id=\"T_90011_row11_col0\" class=\"data row11 col0\" >Transformed Train Set</td>\n",
       "      <td id=\"T_90011_row11_col1\" class=\"data row11 col1\" >(506835, 79)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row12\" class=\"row_heading level0 row12\" >12</th>\n",
       "      <td id=\"T_90011_row12_col0\" class=\"data row12 col0\" >Transformed Test Set</td>\n",
       "      <td id=\"T_90011_row12_col1\" class=\"data row12 col1\" >(217216, 79)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row13\" class=\"row_heading level0 row13\" >13</th>\n",
       "      <td id=\"T_90011_row13_col0\" class=\"data row13 col0\" >Shuffle Train-Test</td>\n",
       "      <td id=\"T_90011_row13_col1\" class=\"data row13 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row14\" class=\"row_heading level0 row14\" >14</th>\n",
       "      <td id=\"T_90011_row14_col0\" class=\"data row14 col0\" >Stratify Train-Test</td>\n",
       "      <td id=\"T_90011_row14_col1\" class=\"data row14 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row15\" class=\"row_heading level0 row15\" >15</th>\n",
       "      <td id=\"T_90011_row15_col0\" class=\"data row15 col0\" >Fold Generator</td>\n",
       "      <td id=\"T_90011_row15_col1\" class=\"data row15 col1\" >StratifiedKFold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row16\" class=\"row_heading level0 row16\" >16</th>\n",
       "      <td id=\"T_90011_row16_col0\" class=\"data row16 col0\" >Fold Number</td>\n",
       "      <td id=\"T_90011_row16_col1\" class=\"data row16 col1\" >10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row17\" class=\"row_heading level0 row17\" >17</th>\n",
       "      <td id=\"T_90011_row17_col0\" class=\"data row17 col0\" >CPU Jobs</td>\n",
       "      <td id=\"T_90011_row17_col1\" class=\"data row17 col1\" >-1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row18\" class=\"row_heading level0 row18\" >18</th>\n",
       "      <td id=\"T_90011_row18_col0\" class=\"data row18 col0\" >Use GPU</td>\n",
       "      <td id=\"T_90011_row18_col1\" class=\"data row18 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row19\" class=\"row_heading level0 row19\" >19</th>\n",
       "      <td id=\"T_90011_row19_col0\" class=\"data row19 col0\" >Log Experiment</td>\n",
       "      <td id=\"T_90011_row19_col1\" class=\"data row19 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row20\" class=\"row_heading level0 row20\" >20</th>\n",
       "      <td id=\"T_90011_row20_col0\" class=\"data row20 col0\" >Experiment Name</td>\n",
       "      <td id=\"T_90011_row20_col1\" class=\"data row20 col1\" >clf-default-name</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row21\" class=\"row_heading level0 row21\" >21</th>\n",
       "      <td id=\"T_90011_row21_col0\" class=\"data row21 col0\" >USI</td>\n",
       "      <td id=\"T_90011_row21_col1\" class=\"data row21 col1\" >d6a7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row22\" class=\"row_heading level0 row22\" >22</th>\n",
       "      <td id=\"T_90011_row22_col0\" class=\"data row22 col0\" >Imputation Type</td>\n",
       "      <td id=\"T_90011_row22_col1\" class=\"data row22 col1\" >simple</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row23\" class=\"row_heading level0 row23\" >23</th>\n",
       "      <td id=\"T_90011_row23_col0\" class=\"data row23 col0\" >Iterative Imputation Iteration</td>\n",
       "      <td id=\"T_90011_row23_col1\" class=\"data row23 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row24\" class=\"row_heading level0 row24\" >24</th>\n",
       "      <td id=\"T_90011_row24_col0\" class=\"data row24 col0\" >Numeric Imputer</td>\n",
       "      <td id=\"T_90011_row24_col1\" class=\"data row24 col1\" >mean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row25\" class=\"row_heading level0 row25\" >25</th>\n",
       "      <td id=\"T_90011_row25_col0\" class=\"data row25 col0\" >Iterative Imputation Numeric Model</td>\n",
       "      <td id=\"T_90011_row25_col1\" class=\"data row25 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row26\" class=\"row_heading level0 row26\" >26</th>\n",
       "      <td id=\"T_90011_row26_col0\" class=\"data row26 col0\" >Categorical Imputer</td>\n",
       "      <td id=\"T_90011_row26_col1\" class=\"data row26 col1\" >constant</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row27\" class=\"row_heading level0 row27\" >27</th>\n",
       "      <td id=\"T_90011_row27_col0\" class=\"data row27 col0\" >Iterative Imputation Categorical Model</td>\n",
       "      <td id=\"T_90011_row27_col1\" class=\"data row27 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row28\" class=\"row_heading level0 row28\" >28</th>\n",
       "      <td id=\"T_90011_row28_col0\" class=\"data row28 col0\" >Unknown Categoricals Handling</td>\n",
       "      <td id=\"T_90011_row28_col1\" class=\"data row28 col1\" >least_frequent</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row29\" class=\"row_heading level0 row29\" >29</th>\n",
       "      <td id=\"T_90011_row29_col0\" class=\"data row29 col0\" >Normalize</td>\n",
       "      <td id=\"T_90011_row29_col1\" class=\"data row29 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row30\" class=\"row_heading level0 row30\" >30</th>\n",
       "      <td id=\"T_90011_row30_col0\" class=\"data row30 col0\" >Normalize Method</td>\n",
       "      <td id=\"T_90011_row30_col1\" class=\"data row30 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row31\" class=\"row_heading level0 row31\" >31</th>\n",
       "      <td id=\"T_90011_row31_col0\" class=\"data row31 col0\" >Transformation</td>\n",
       "      <td id=\"T_90011_row31_col1\" class=\"data row31 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row32\" class=\"row_heading level0 row32\" >32</th>\n",
       "      <td id=\"T_90011_row32_col0\" class=\"data row32 col0\" >Transformation Method</td>\n",
       "      <td id=\"T_90011_row32_col1\" class=\"data row32 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row33\" class=\"row_heading level0 row33\" >33</th>\n",
       "      <td id=\"T_90011_row33_col0\" class=\"data row33 col0\" >PCA</td>\n",
       "      <td id=\"T_90011_row33_col1\" class=\"data row33 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row34\" class=\"row_heading level0 row34\" >34</th>\n",
       "      <td id=\"T_90011_row34_col0\" class=\"data row34 col0\" >PCA Method</td>\n",
       "      <td id=\"T_90011_row34_col1\" class=\"data row34 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row35\" class=\"row_heading level0 row35\" >35</th>\n",
       "      <td id=\"T_90011_row35_col0\" class=\"data row35 col0\" >PCA Components</td>\n",
       "      <td id=\"T_90011_row35_col1\" class=\"data row35 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row36\" class=\"row_heading level0 row36\" >36</th>\n",
       "      <td id=\"T_90011_row36_col0\" class=\"data row36 col0\" >Ignore Low Variance</td>\n",
       "      <td id=\"T_90011_row36_col1\" class=\"data row36 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row37\" class=\"row_heading level0 row37\" >37</th>\n",
       "      <td id=\"T_90011_row37_col0\" class=\"data row37 col0\" >Combine Rare Levels</td>\n",
       "      <td id=\"T_90011_row37_col1\" class=\"data row37 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row38\" class=\"row_heading level0 row38\" >38</th>\n",
       "      <td id=\"T_90011_row38_col0\" class=\"data row38 col0\" >Rare Level Threshold</td>\n",
       "      <td id=\"T_90011_row38_col1\" class=\"data row38 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row39\" class=\"row_heading level0 row39\" >39</th>\n",
       "      <td id=\"T_90011_row39_col0\" class=\"data row39 col0\" >Numeric Binning</td>\n",
       "      <td id=\"T_90011_row39_col1\" class=\"data row39 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row40\" class=\"row_heading level0 row40\" >40</th>\n",
       "      <td id=\"T_90011_row40_col0\" class=\"data row40 col0\" >Remove Outliers</td>\n",
       "      <td id=\"T_90011_row40_col1\" class=\"data row40 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row41\" class=\"row_heading level0 row41\" >41</th>\n",
       "      <td id=\"T_90011_row41_col0\" class=\"data row41 col0\" >Outliers Threshold</td>\n",
       "      <td id=\"T_90011_row41_col1\" class=\"data row41 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row42\" class=\"row_heading level0 row42\" >42</th>\n",
       "      <td id=\"T_90011_row42_col0\" class=\"data row42 col0\" >Remove Multicollinearity</td>\n",
       "      <td id=\"T_90011_row42_col1\" class=\"data row42 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row43\" class=\"row_heading level0 row43\" >43</th>\n",
       "      <td id=\"T_90011_row43_col0\" class=\"data row43 col0\" >Multicollinearity Threshold</td>\n",
       "      <td id=\"T_90011_row43_col1\" class=\"data row43 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row44\" class=\"row_heading level0 row44\" >44</th>\n",
       "      <td id=\"T_90011_row44_col0\" class=\"data row44 col0\" >Remove Perfect Collinearity</td>\n",
       "      <td id=\"T_90011_row44_col1\" class=\"data row44 col1\" >True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row45\" class=\"row_heading level0 row45\" >45</th>\n",
       "      <td id=\"T_90011_row45_col0\" class=\"data row45 col0\" >Clustering</td>\n",
       "      <td id=\"T_90011_row45_col1\" class=\"data row45 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row46\" class=\"row_heading level0 row46\" >46</th>\n",
       "      <td id=\"T_90011_row46_col0\" class=\"data row46 col0\" >Clustering Iteration</td>\n",
       "      <td id=\"T_90011_row46_col1\" class=\"data row46 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row47\" class=\"row_heading level0 row47\" >47</th>\n",
       "      <td id=\"T_90011_row47_col0\" class=\"data row47 col0\" >Polynomial Features</td>\n",
       "      <td id=\"T_90011_row47_col1\" class=\"data row47 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row48\" class=\"row_heading level0 row48\" >48</th>\n",
       "      <td id=\"T_90011_row48_col0\" class=\"data row48 col0\" >Polynomial Degree</td>\n",
       "      <td id=\"T_90011_row48_col1\" class=\"data row48 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row49\" class=\"row_heading level0 row49\" >49</th>\n",
       "      <td id=\"T_90011_row49_col0\" class=\"data row49 col0\" >Trignometry Features</td>\n",
       "      <td id=\"T_90011_row49_col1\" class=\"data row49 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row50\" class=\"row_heading level0 row50\" >50</th>\n",
       "      <td id=\"T_90011_row50_col0\" class=\"data row50 col0\" >Polynomial Threshold</td>\n",
       "      <td id=\"T_90011_row50_col1\" class=\"data row50 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row51\" class=\"row_heading level0 row51\" >51</th>\n",
       "      <td id=\"T_90011_row51_col0\" class=\"data row51 col0\" >Group Features</td>\n",
       "      <td id=\"T_90011_row51_col1\" class=\"data row51 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row52\" class=\"row_heading level0 row52\" >52</th>\n",
       "      <td id=\"T_90011_row52_col0\" class=\"data row52 col0\" >Feature Selection</td>\n",
       "      <td id=\"T_90011_row52_col1\" class=\"data row52 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row53\" class=\"row_heading level0 row53\" >53</th>\n",
       "      <td id=\"T_90011_row53_col0\" class=\"data row53 col0\" >Feature Selection Method</td>\n",
       "      <td id=\"T_90011_row53_col1\" class=\"data row53 col1\" >classic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row54\" class=\"row_heading level0 row54\" >54</th>\n",
       "      <td id=\"T_90011_row54_col0\" class=\"data row54 col0\" >Features Selection Threshold</td>\n",
       "      <td id=\"T_90011_row54_col1\" class=\"data row54 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row55\" class=\"row_heading level0 row55\" >55</th>\n",
       "      <td id=\"T_90011_row55_col0\" class=\"data row55 col0\" >Feature Interaction</td>\n",
       "      <td id=\"T_90011_row55_col1\" class=\"data row55 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row56\" class=\"row_heading level0 row56\" >56</th>\n",
       "      <td id=\"T_90011_row56_col0\" class=\"data row56 col0\" >Feature Ratio</td>\n",
       "      <td id=\"T_90011_row56_col1\" class=\"data row56 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row57\" class=\"row_heading level0 row57\" >57</th>\n",
       "      <td id=\"T_90011_row57_col0\" class=\"data row57 col0\" >Interaction Threshold</td>\n",
       "      <td id=\"T_90011_row57_col1\" class=\"data row57 col1\" >None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row58\" class=\"row_heading level0 row58\" >58</th>\n",
       "      <td id=\"T_90011_row58_col0\" class=\"data row58 col0\" >Fix Imbalance</td>\n",
       "      <td id=\"T_90011_row58_col1\" class=\"data row58 col1\" >False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_90011_level0_row59\" class=\"row_heading level0 row59\" >59</th>\n",
       "      <td id=\"T_90011_row59_col0\" class=\"data row59 col0\" >Fix Imbalance Method</td>\n",
       "      <td id=\"T_90011_row59_col1\" class=\"data row59 col1\" >SMOTE</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1bf6ce79430>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "exp_clf = setup(data = gender_train, target = 'gender', session_id=123)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "6754bad5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "#T_65c70 th {\n",
       "  text-align: left;\n",
       "}\n",
       "#T_65c70_row0_col0, #T_65c70_row0_col3, #T_65c70_row0_col4, #T_65c70_row1_col0, #T_65c70_row1_col1, #T_65c70_row1_col2, #T_65c70_row1_col3, #T_65c70_row1_col5, #T_65c70_row1_col6, #T_65c70_row1_col7, #T_65c70_row2_col0, #T_65c70_row2_col1, #T_65c70_row2_col2, #T_65c70_row2_col3, #T_65c70_row2_col4, #T_65c70_row2_col5, #T_65c70_row2_col6, #T_65c70_row2_col7, #T_65c70_row3_col0, #T_65c70_row3_col1, #T_65c70_row3_col2, #T_65c70_row3_col3, #T_65c70_row3_col4, #T_65c70_row3_col5, #T_65c70_row3_col6, #T_65c70_row3_col7, #T_65c70_row4_col0, #T_65c70_row4_col1, #T_65c70_row4_col2, #T_65c70_row4_col3, #T_65c70_row4_col4, #T_65c70_row4_col5, #T_65c70_row4_col6, #T_65c70_row4_col7, #T_65c70_row5_col0, #T_65c70_row5_col1, #T_65c70_row5_col2, #T_65c70_row5_col3, #T_65c70_row5_col4, #T_65c70_row5_col5, #T_65c70_row5_col6, #T_65c70_row5_col7, #T_65c70_row6_col0, #T_65c70_row6_col1, #T_65c70_row6_col2, #T_65c70_row6_col3, #T_65c70_row6_col4, #T_65c70_row6_col5, #T_65c70_row6_col6, #T_65c70_row6_col7, #T_65c70_row7_col0, #T_65c70_row7_col1, #T_65c70_row7_col2, #T_65c70_row7_col3, #T_65c70_row7_col4, #T_65c70_row7_col5, #T_65c70_row7_col6, #T_65c70_row7_col7, #T_65c70_row8_col0, #T_65c70_row8_col1, #T_65c70_row8_col2, #T_65c70_row8_col4, #T_65c70_row8_col5, #T_65c70_row8_col6, #T_65c70_row8_col7, #T_65c70_row9_col0, #T_65c70_row9_col1, #T_65c70_row9_col2, #T_65c70_row9_col4, #T_65c70_row9_col5, #T_65c70_row9_col6, #T_65c70_row9_col7, #T_65c70_row10_col0, #T_65c70_row10_col1, #T_65c70_row10_col2, #T_65c70_row10_col3, #T_65c70_row10_col4, #T_65c70_row10_col5, #T_65c70_row10_col6, #T_65c70_row10_col7, #T_65c70_row11_col0, #T_65c70_row11_col1, #T_65c70_row11_col2, #T_65c70_row11_col3, #T_65c70_row11_col4, #T_65c70_row11_col5, #T_65c70_row11_col6, #T_65c70_row11_col7, #T_65c70_row12_col0, #T_65c70_row12_col1, #T_65c70_row12_col2, #T_65c70_row12_col3, #T_65c70_row12_col4, #T_65c70_row12_col5, #T_65c70_row12_col6, #T_65c70_row12_col7, #T_65c70_row13_col0, #T_65c70_row13_col1, #T_65c70_row13_col2, #T_65c70_row13_col3, #T_65c70_row13_col4, #T_65c70_row13_col5, #T_65c70_row13_col6, #T_65c70_row13_col7, #T_65c70_row14_col0, #T_65c70_row14_col1, #T_65c70_row14_col2, #T_65c70_row14_col3, #T_65c70_row14_col4, #T_65c70_row14_col5, #T_65c70_row14_col6, #T_65c70_row14_col7, #T_65c70_row15_col0, #T_65c70_row15_col1, #T_65c70_row15_col2, #T_65c70_row15_col3, #T_65c70_row15_col4, #T_65c70_row15_col5, #T_65c70_row15_col6, #T_65c70_row15_col7 {\n",
       "  text-align: left;\n",
       "}\n",
       "#T_65c70_row0_col1, #T_65c70_row0_col2, #T_65c70_row0_col5, #T_65c70_row0_col6, #T_65c70_row0_col7, #T_65c70_row1_col4, #T_65c70_row8_col3, #T_65c70_row9_col3 {\n",
       "  text-align: left;\n",
       "  background-color: yellow;\n",
       "}\n",
       "#T_65c70_row0_col8, #T_65c70_row1_col8, #T_65c70_row2_col8, #T_65c70_row3_col8, #T_65c70_row4_col8, #T_65c70_row5_col8, #T_65c70_row6_col8, #T_65c70_row7_col8, #T_65c70_row8_col8, #T_65c70_row10_col8, #T_65c70_row11_col8, #T_65c70_row12_col8, #T_65c70_row13_col8, #T_65c70_row14_col8, #T_65c70_row15_col8 {\n",
       "  text-align: left;\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "#T_65c70_row9_col8 {\n",
       "  text-align: left;\n",
       "  background-color: yellow;\n",
       "  background-color: lightgrey;\n",
       "}\n",
       "</style>\n",
       "<table id=\"T_65c70\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_65c70_level0_col0\" class=\"col_heading level0 col0\" >Model</th>\n",
       "      <th id=\"T_65c70_level0_col1\" class=\"col_heading level0 col1\" >Accuracy</th>\n",
       "      <th id=\"T_65c70_level0_col2\" class=\"col_heading level0 col2\" >AUC</th>\n",
       "      <th id=\"T_65c70_level0_col3\" class=\"col_heading level0 col3\" >Recall</th>\n",
       "      <th id=\"T_65c70_level0_col4\" class=\"col_heading level0 col4\" >Prec.</th>\n",
       "      <th id=\"T_65c70_level0_col5\" class=\"col_heading level0 col5\" >F1</th>\n",
       "      <th id=\"T_65c70_level0_col6\" class=\"col_heading level0 col6\" >Kappa</th>\n",
       "      <th id=\"T_65c70_level0_col7\" class=\"col_heading level0 col7\" >MCC</th>\n",
       "      <th id=\"T_65c70_level0_col8\" class=\"col_heading level0 col8\" >TT (Sec)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row0\" class=\"row_heading level0 row0\" >rf</th>\n",
       "      <td id=\"T_65c70_row0_col0\" class=\"data row0 col0\" >Random Forest Classifier</td>\n",
       "      <td id=\"T_65c70_row0_col1\" class=\"data row0 col1\" >0.8361</td>\n",
       "      <td id=\"T_65c70_row0_col2\" class=\"data row0 col2\" >0.9030</td>\n",
       "      <td id=\"T_65c70_row0_col3\" class=\"data row0 col3\" >0.9449</td>\n",
       "      <td id=\"T_65c70_row0_col4\" class=\"data row0 col4\" >0.8347</td>\n",
       "      <td id=\"T_65c70_row0_col5\" class=\"data row0 col5\" >0.8864</td>\n",
       "      <td id=\"T_65c70_row0_col6\" class=\"data row0 col6\" >0.5960</td>\n",
       "      <td id=\"T_65c70_row0_col7\" class=\"data row0 col7\" >0.6110</td>\n",
       "      <td id=\"T_65c70_row0_col8\" class=\"data row0 col8\" >19.6180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row1\" class=\"row_heading level0 row1\" >dt</th>\n",
       "      <td id=\"T_65c70_row1_col0\" class=\"data row1 col0\" >Decision Tree Classifier</td>\n",
       "      <td id=\"T_65c70_row1_col1\" class=\"data row1 col1\" >0.7918</td>\n",
       "      <td id=\"T_65c70_row1_col2\" class=\"data row1 col2\" >0.7636</td>\n",
       "      <td id=\"T_65c70_row1_col3\" class=\"data row1 col3\" >0.8431</td>\n",
       "      <td id=\"T_65c70_row1_col4\" class=\"data row1 col4\" >0.8483</td>\n",
       "      <td id=\"T_65c70_row1_col5\" class=\"data row1 col5\" >0.8457</td>\n",
       "      <td id=\"T_65c70_row1_col6\" class=\"data row1 col6\" >0.5255</td>\n",
       "      <td id=\"T_65c70_row1_col7\" class=\"data row1 col7\" >0.5255</td>\n",
       "      <td id=\"T_65c70_row1_col8\" class=\"data row1 col8\" >2.0730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row2\" class=\"row_heading level0 row2\" >et</th>\n",
       "      <td id=\"T_65c70_row2_col0\" class=\"data row2 col0\" >Extra Trees Classifier</td>\n",
       "      <td id=\"T_65c70_row2_col1\" class=\"data row2 col1\" >0.7525</td>\n",
       "      <td id=\"T_65c70_row2_col2\" class=\"data row2 col2\" >0.8068</td>\n",
       "      <td id=\"T_65c70_row2_col3\" class=\"data row2 col3\" >0.9485</td>\n",
       "      <td id=\"T_65c70_row2_col4\" class=\"data row2 col4\" >0.7512</td>\n",
       "      <td id=\"T_65c70_row2_col5\" class=\"data row2 col5\" >0.8384</td>\n",
       "      <td id=\"T_65c70_row2_col6\" class=\"data row2 col6\" >0.3390</td>\n",
       "      <td id=\"T_65c70_row2_col7\" class=\"data row2 col7\" >0.3852</td>\n",
       "      <td id=\"T_65c70_row2_col8\" class=\"data row2 col8\" >29.8000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row3\" class=\"row_heading level0 row3\" >catboost</th>\n",
       "      <td id=\"T_65c70_row3_col0\" class=\"data row3 col0\" >CatBoost Classifier</td>\n",
       "      <td id=\"T_65c70_row3_col1\" class=\"data row3 col1\" >0.7568</td>\n",
       "      <td id=\"T_65c70_row3_col2\" class=\"data row3 col2\" >0.7949</td>\n",
       "      <td id=\"T_65c70_row3_col3\" class=\"data row3 col3\" >0.8976</td>\n",
       "      <td id=\"T_65c70_row3_col4\" class=\"data row3 col4\" >0.7776</td>\n",
       "      <td id=\"T_65c70_row3_col5\" class=\"data row3 col5\" >0.8333</td>\n",
       "      <td id=\"T_65c70_row3_col6\" class=\"data row3 col6\" >0.3927</td>\n",
       "      <td id=\"T_65c70_row3_col7\" class=\"data row3 col7\" >0.4068</td>\n",
       "      <td id=\"T_65c70_row3_col8\" class=\"data row3 col8\" >22.7730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row4\" class=\"row_heading level0 row4\" >xgboost</th>\n",
       "      <td id=\"T_65c70_row4_col0\" class=\"data row4 col0\" >Extreme Gradient Boosting</td>\n",
       "      <td id=\"T_65c70_row4_col1\" class=\"data row4 col1\" >0.7480</td>\n",
       "      <td id=\"T_65c70_row4_col2\" class=\"data row4 col2\" >0.7810</td>\n",
       "      <td id=\"T_65c70_row4_col3\" class=\"data row4 col3\" >0.8923</td>\n",
       "      <td id=\"T_65c70_row4_col4\" class=\"data row4 col4\" >0.7713</td>\n",
       "      <td id=\"T_65c70_row4_col5\" class=\"data row4 col5\" >0.8274</td>\n",
       "      <td id=\"T_65c70_row4_col6\" class=\"data row4 col6\" >0.3697</td>\n",
       "      <td id=\"T_65c70_row4_col7\" class=\"data row4 col7\" >0.3835</td>\n",
       "      <td id=\"T_65c70_row4_col8\" class=\"data row4 col8\" >22.6420</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row5\" class=\"row_heading level0 row5\" >lightgbm</th>\n",
       "      <td id=\"T_65c70_row5_col0\" class=\"data row5 col0\" >Light Gradient Boosting Machine</td>\n",
       "      <td id=\"T_65c70_row5_col1\" class=\"data row5 col1\" >0.7380</td>\n",
       "      <td id=\"T_65c70_row5_col2\" class=\"data row5 col2\" >0.7650</td>\n",
       "      <td id=\"T_65c70_row5_col3\" class=\"data row5 col3\" >0.8947</td>\n",
       "      <td id=\"T_65c70_row5_col4\" class=\"data row5 col4\" >0.7605</td>\n",
       "      <td id=\"T_65c70_row5_col5\" class=\"data row5 col5\" >0.8222</td>\n",
       "      <td id=\"T_65c70_row5_col6\" class=\"data row5 col6\" >0.3369</td>\n",
       "      <td id=\"T_65c70_row5_col7\" class=\"data row5 col7\" >0.3535</td>\n",
       "      <td id=\"T_65c70_row5_col8\" class=\"data row5 col8\" >1.3120</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row6\" class=\"row_heading level0 row6\" >gbc</th>\n",
       "      <td id=\"T_65c70_row6_col0\" class=\"data row6 col0\" >Gradient Boosting Classifier</td>\n",
       "      <td id=\"T_65c70_row6_col1\" class=\"data row6 col1\" >0.7258</td>\n",
       "      <td id=\"T_65c70_row6_col2\" class=\"data row6 col2\" >0.7462</td>\n",
       "      <td id=\"T_65c70_row6_col3\" class=\"data row6 col3\" >0.8951</td>\n",
       "      <td id=\"T_65c70_row6_col4\" class=\"data row6 col4\" >0.7489</td>\n",
       "      <td id=\"T_65c70_row6_col5\" class=\"data row6 col5\" >0.8155</td>\n",
       "      <td id=\"T_65c70_row6_col6\" class=\"data row6 col6\" >0.2980</td>\n",
       "      <td id=\"T_65c70_row6_col7\" class=\"data row6 col7\" >0.3167</td>\n",
       "      <td id=\"T_65c70_row6_col8\" class=\"data row6 col8\" >33.3530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row7\" class=\"row_heading level0 row7\" >ada</th>\n",
       "      <td id=\"T_65c70_row7_col0\" class=\"data row7 col0\" >Ada Boost Classifier</td>\n",
       "      <td id=\"T_65c70_row7_col1\" class=\"data row7 col1\" >0.7179</td>\n",
       "      <td id=\"T_65c70_row7_col2\" class=\"data row7 col2\" >0.7275</td>\n",
       "      <td id=\"T_65c70_row7_col3\" class=\"data row7 col3\" >0.8803</td>\n",
       "      <td id=\"T_65c70_row7_col4\" class=\"data row7 col4\" >0.7478</td>\n",
       "      <td id=\"T_65c70_row7_col5\" class=\"data row7 col5\" >0.8086</td>\n",
       "      <td id=\"T_65c70_row7_col6\" class=\"data row7 col6\" >0.2858</td>\n",
       "      <td id=\"T_65c70_row7_col7\" class=\"data row7 col7\" >0.3000</td>\n",
       "      <td id=\"T_65c70_row7_col8\" class=\"data row7 col8\" >6.8810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row8\" class=\"row_heading level0 row8\" >lr</th>\n",
       "      <td id=\"T_65c70_row8_col0\" class=\"data row8 col0\" >Logistic Regression</td>\n",
       "      <td id=\"T_65c70_row8_col1\" class=\"data row8 col1\" >0.6769</td>\n",
       "      <td id=\"T_65c70_row8_col2\" class=\"data row8 col2\" >0.5879</td>\n",
       "      <td id=\"T_65c70_row8_col3\" class=\"data row8 col3\" >1.0000</td>\n",
       "      <td id=\"T_65c70_row8_col4\" class=\"data row8 col4\" >0.6769</td>\n",
       "      <td id=\"T_65c70_row8_col5\" class=\"data row8 col5\" >0.8074</td>\n",
       "      <td id=\"T_65c70_row8_col6\" class=\"data row8 col6\" >0.0000</td>\n",
       "      <td id=\"T_65c70_row8_col7\" class=\"data row8 col7\" >0.0000</td>\n",
       "      <td id=\"T_65c70_row8_col8\" class=\"data row8 col8\" >2.9810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row9\" class=\"row_heading level0 row9\" >dummy</th>\n",
       "      <td id=\"T_65c70_row9_col0\" class=\"data row9 col0\" >Dummy Classifier</td>\n",
       "      <td id=\"T_65c70_row9_col1\" class=\"data row9 col1\" >0.6769</td>\n",
       "      <td id=\"T_65c70_row9_col2\" class=\"data row9 col2\" >0.5000</td>\n",
       "      <td id=\"T_65c70_row9_col3\" class=\"data row9 col3\" >1.0000</td>\n",
       "      <td id=\"T_65c70_row9_col4\" class=\"data row9 col4\" >0.6769</td>\n",
       "      <td id=\"T_65c70_row9_col5\" class=\"data row9 col5\" >0.8074</td>\n",
       "      <td id=\"T_65c70_row9_col6\" class=\"data row9 col6\" >0.0000</td>\n",
       "      <td id=\"T_65c70_row9_col7\" class=\"data row9 col7\" >0.0000</td>\n",
       "      <td id=\"T_65c70_row9_col8\" class=\"data row9 col8\" >0.0980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row10\" class=\"row_heading level0 row10\" >ridge</th>\n",
       "      <td id=\"T_65c70_row10_col0\" class=\"data row10 col0\" >Ridge Classifier</td>\n",
       "      <td id=\"T_65c70_row10_col1\" class=\"data row10 col1\" >0.6788</td>\n",
       "      <td id=\"T_65c70_row10_col2\" class=\"data row10 col2\" >0.0000</td>\n",
       "      <td id=\"T_65c70_row10_col3\" class=\"data row10 col3\" >0.9880</td>\n",
       "      <td id=\"T_65c70_row10_col4\" class=\"data row10 col4\" >0.6813</td>\n",
       "      <td id=\"T_65c70_row10_col5\" class=\"data row10 col5\" >0.8064</td>\n",
       "      <td id=\"T_65c70_row10_col6\" class=\"data row10 col6\" >0.0246</td>\n",
       "      <td id=\"T_65c70_row10_col7\" class=\"data row10 col7\" >0.0445</td>\n",
       "      <td id=\"T_65c70_row10_col8\" class=\"data row10 col8\" >0.9890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row11\" class=\"row_heading level0 row11\" >nb</th>\n",
       "      <td id=\"T_65c70_row11_col0\" class=\"data row11 col0\" >Naive Bayes</td>\n",
       "      <td id=\"T_65c70_row11_col1\" class=\"data row11 col1\" >0.6751</td>\n",
       "      <td id=\"T_65c70_row11_col2\" class=\"data row11 col2\" >0.6198</td>\n",
       "      <td id=\"T_65c70_row11_col3\" class=\"data row11 col3\" >0.9940</td>\n",
       "      <td id=\"T_65c70_row11_col4\" class=\"data row11 col4\" >0.6772</td>\n",
       "      <td id=\"T_65c70_row11_col5\" class=\"data row11 col5\" >0.8055</td>\n",
       "      <td id=\"T_65c70_row11_col6\" class=\"data row11 col6\" >0.0014</td>\n",
       "      <td id=\"T_65c70_row11_col7\" class=\"data row11 col7\" >0.0060</td>\n",
       "      <td id=\"T_65c70_row11_col8\" class=\"data row11 col8\" >0.2610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row12\" class=\"row_heading level0 row12\" >lda</th>\n",
       "      <td id=\"T_65c70_row12_col0\" class=\"data row12 col0\" >Linear Discriminant Analysis</td>\n",
       "      <td id=\"T_65c70_row12_col1\" class=\"data row12 col1\" >0.6834</td>\n",
       "      <td id=\"T_65c70_row12_col2\" class=\"data row12 col2\" >0.6223</td>\n",
       "      <td id=\"T_65c70_row12_col3\" class=\"data row12 col3\" >0.9631</td>\n",
       "      <td id=\"T_65c70_row12_col4\" class=\"data row12 col4\" >0.6910</td>\n",
       "      <td id=\"T_65c70_row12_col5\" class=\"data row12 col5\" >0.8046</td>\n",
       "      <td id=\"T_65c70_row12_col6\" class=\"data row12 col6\" >0.0771</td>\n",
       "      <td id=\"T_65c70_row12_col7\" class=\"data row12 col7\" >0.1226</td>\n",
       "      <td id=\"T_65c70_row12_col8\" class=\"data row12 col8\" >4.0460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row13\" class=\"row_heading level0 row13\" >knn</th>\n",
       "      <td id=\"T_65c70_row13_col0\" class=\"data row13 col0\" >K Neighbors Classifier</td>\n",
       "      <td id=\"T_65c70_row13_col1\" class=\"data row13 col1\" >0.7151</td>\n",
       "      <td id=\"T_65c70_row13_col2\" class=\"data row13 col2\" >0.7204</td>\n",
       "      <td id=\"T_65c70_row13_col3\" class=\"data row13 col3\" >0.8347</td>\n",
       "      <td id=\"T_65c70_row13_col4\" class=\"data row13 col4\" >0.7656</td>\n",
       "      <td id=\"T_65c70_row13_col5\" class=\"data row13 col5\" >0.7987</td>\n",
       "      <td id=\"T_65c70_row13_col6\" class=\"data row13 col6\" >0.3148</td>\n",
       "      <td id=\"T_65c70_row13_col7\" class=\"data row13 col7\" >0.3183</td>\n",
       "      <td id=\"T_65c70_row13_col8\" class=\"data row13 col8\" >5.2350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row14\" class=\"row_heading level0 row14\" >svm</th>\n",
       "      <td id=\"T_65c70_row14_col0\" class=\"data row14 col0\" >SVM - Linear Kernel</td>\n",
       "      <td id=\"T_65c70_row14_col1\" class=\"data row14 col1\" >0.6058</td>\n",
       "      <td id=\"T_65c70_row14_col2\" class=\"data row14 col2\" >0.0000</td>\n",
       "      <td id=\"T_65c70_row14_col3\" class=\"data row14 col3\" >0.7673</td>\n",
       "      <td id=\"T_65c70_row14_col4\" class=\"data row14 col4\" >0.6949</td>\n",
       "      <td id=\"T_65c70_row14_col5\" class=\"data row14 col5\" >0.6945</td>\n",
       "      <td id=\"T_65c70_row14_col6\" class=\"data row14 col6\" >0.0370</td>\n",
       "      <td id=\"T_65c70_row14_col7\" class=\"data row14 col7\" >0.0452</td>\n",
       "      <td id=\"T_65c70_row14_col8\" class=\"data row14 col8\" >2.2970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_65c70_level0_row15\" class=\"row_heading level0 row15\" >qda</th>\n",
       "      <td id=\"T_65c70_row15_col0\" class=\"data row15 col0\" >Quadratic Discriminant Analysis</td>\n",
       "      <td id=\"T_65c70_row15_col1\" class=\"data row15 col1\" >0.5996</td>\n",
       "      <td id=\"T_65c70_row15_col2\" class=\"data row15 col2\" >0.6014</td>\n",
       "      <td id=\"T_65c70_row15_col3\" class=\"data row15 col3\" >0.6566</td>\n",
       "      <td id=\"T_65c70_row15_col4\" class=\"data row15 col4\" >0.7264</td>\n",
       "      <td id=\"T_65c70_row15_col5\" class=\"data row15 col5\" >0.6878</td>\n",
       "      <td id=\"T_65c70_row15_col6\" class=\"data row15 col6\" >0.1301</td>\n",
       "      <td id=\"T_65c70_row15_col7\" class=\"data row15 col7\" >0.1325</td>\n",
       "      <td id=\"T_65c70_row15_col8\" class=\"data row15 col8\" >3.2610</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x1bf6ce605e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "best_model = compare_models(sort = \"F1\",n_select=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "e13aa392",
   "metadata": {},
   "outputs": [],
   "source": [
    "prediction = predict_model(best_model, data = gender_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "id": "83d5dbbc",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.loc[user_spec.gender.isna(),'gender']= prediction.Label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "e7adaddb",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.gender = pd.to_numeric(user_spec.gender)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "a983a8d9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 0.])"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.gender.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "484e52c4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                         0\n",
       "user_id                                0\n",
       "insert_time                            0\n",
       "credit_score                           0\n",
       "yearly_income                          0\n",
       "income_type                            0\n",
       "employment_type                        0\n",
       "houseown_type                          0\n",
       "desired_amount                         0\n",
       "purpose                                0\n",
       "personal_rehabilitation_yn             0\n",
       "personal_rehabilitation_complete_yn    0\n",
       "existing_loan_cnt                      0\n",
       "existing_loan_amt                      0\n",
       "enter_year                             0\n",
       "enter_month                            0\n",
       "enter_birth                            0\n",
       "month                                  0\n",
       "birth_year                             0\n",
       "gender                                 0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_spec.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03505a1b",
   "metadata": {},
   "source": [
    "#### Deployment CSV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "id": "54eb06b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_spec.to_csv('../Data/user_spec_clansing.csv',index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c974a98",
   "metadata": {},
   "source": [
    "### loan_limit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "id": "21787b8c",
   "metadata": {},
   "outputs": [],
   "source": [
    "bank_id = loan_result[loan_result.loan_limit.isna()].bank_id.unique().tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "id": "240f2d05",
   "metadata": {},
   "outputs": [],
   "source": [
    "ll_df = loan_result.groupby('bank_id')['loan_limit'].mean().reset_index()\n",
    "ll_df.loan_limit = pd.to_numeric(ll_df.loan_limit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "id": "2714cf65",
   "metadata": {},
   "outputs": [],
   "source": [
    "for idd in bank_id:\n",
    "     loan_result.loc[(loan_result.bank_id == idd) & (loan_result['loan_limit'].isna()),'loan_limit'] = int(ll_df[ll_df.bank_id == idd]['loan_limit'].values[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "id": "48d0f340",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                 0\n",
       "loanapply_insert_time          0\n",
       "bank_id                        0\n",
       "product_id                     0\n",
       "loan_limit                     0\n",
       "loan_rate                   7495\n",
       "is_applied               3257239\n",
       "dtype: int64"
      ]
     },
     "execution_count": 218,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_result.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27c07a21",
   "metadata": {},
   "source": [
    "### loan_rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "id": "c1a39326",
   "metadata": {},
   "outputs": [],
   "source": [
    "bank_id = loan_result[loan_result.loan_rate.isna()].bank_id.unique().tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "id": "4c5d4314",
   "metadata": {},
   "outputs": [],
   "source": [
    "ll_df = loan_result.groupby('bank_id')['loan_rate'].mean().reset_index()\n",
    "ll_df.loan_rate = pd.to_numeric(ll_df.loan_rate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "id": "41cb38b8",
   "metadata": {},
   "outputs": [],
   "source": [
    "for idd in bank_id:\n",
    "     loan_result.loc[(loan_result.bank_id == idd) & (loan_result['loan_rate'].isna()),'loan_rate'] = int(ll_df[ll_df.bank_id == idd]['loan_rate'].values[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "id": "88fde6d0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "application_id                 0\n",
       "loanapply_insert_time          0\n",
       "bank_id                        0\n",
       "product_id                     0\n",
       "loan_limit                     0\n",
       "loan_rate                      0\n",
       "is_applied               3257239\n",
       "dtype: int64"
      ]
     },
     "execution_count": 222,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loan_result.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a25e4174",
   "metadata": {},
   "source": [
    "#### Deployment CSV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "id": "d4f42812",
   "metadata": {},
   "outputs": [],
   "source": [
    "loan_result.to_csv('../Data/loan_result_clansing.csv',index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b1d26b0f",
   "metadata": {},
   "source": [
    "### Mp_Os & Os_Version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "id": "71211223",
   "metadata": {},
   "outputs": [],
   "source": [
    "log_data.drop(['mp_os','mp_app_version'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "id": "5da8d119",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_id      0\n",
       "event        0\n",
       "timestamp    0\n",
       "date_cd      0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_data.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b04063cc",
   "metadata": {},
   "source": [
    "#### Deployment CSV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "id": "62c54a2a",
   "metadata": {},
   "outputs": [],
   "source": [
    "log_data.to_csv('../Data/log_data_clansing.csv',index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}