{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "MJUe",
   "metadata": {},
   "source": [
    "# Home Credit Default Risk Prediction\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "vblA",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import joblib\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.preprocessing import MinMaxScaler, OneHotEncoder, OrdinalEncoder\n",
    "\n",
    "from lightgbm import LGBMClassifier\n",
    "\n",
    "from src.plots import (\n",
    "    plot_target_distribution,\n",
    "    plot_credit_amounts,\n",
    "    plot_education_levels,\n",
    "    plot_occupation,\n",
    "    plot_family_status,\n",
    "    plot_income_type,\n",
    ")\n",
    "from src.utils import get_dataset, get_features_target, get_train_test_sets\n",
    "from src.preprocessing import preprocess_data_pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "bkHC",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = get_dataset()\n",
    "X, y = get_features_target(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "lEQa",
   "metadata": {},
   "source": [
    "## 1. Exploratory Data Analysis\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "PKri",
   "metadata": {},
   "source": [
    "### 1.1 Dataset Information\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Xref",
   "metadata": {},
   "source": [
    "**a. Shape of the dataset**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "eaed528c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature dataset samples: 246008\n",
      "Number of columns/features: 122\n"
     ]
    }
   ],
   "source": [
    "print(\"Feature dataset samples: {}\".format(df.shape[0]))\n",
    "print(\"Number of columns/features: {}\".format(df.shape[1]))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "BYtC",
   "metadata": {},
   "source": [
    "**b. Dataset features**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "RGSE",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['SK_ID_CURR', 'NAME_CONTRACT_TYPE', 'CODE_GENDER', 'FLAG_OWN_CAR',\n",
       "       'FLAG_OWN_REALTY', 'CNT_CHILDREN', 'AMT_INCOME_TOTAL', 'AMT_CREDIT',\n",
       "       'AMT_ANNUITY', 'AMT_GOODS_PRICE',\n",
       "       ...\n",
       "       'FLAG_DOCUMENT_18', 'FLAG_DOCUMENT_19', 'FLAG_DOCUMENT_20',\n",
       "       'FLAG_DOCUMENT_21', 'AMT_REQ_CREDIT_BUREAU_HOUR',\n",
       "       'AMT_REQ_CREDIT_BUREAU_DAY', 'AMT_REQ_CREDIT_BUREAU_WEEK',\n",
       "       'AMT_REQ_CREDIT_BUREAU_MON', 'AMT_REQ_CREDIT_BUREAU_QRT',\n",
       "       'AMT_REQ_CREDIT_BUREAU_YEAR'],\n",
       "      dtype='object', length=121)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Kclp",
   "metadata": {},
   "source": [
    "**c. Sample from dataset**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "emfo",
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SK_ID_CURR</th>\n",
       "      <td>428247</td>\n",
       "      <td>140966</td>\n",
       "      <td>407283</td>\n",
       "      <td>434300</td>\n",
       "      <td>446788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NAME_CONTRACT_TYPE</th>\n",
       "      <td>Cash loans</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>Revolving loans</td>\n",
       "      <td>Cash loans</td>\n",
       "      <td>Cash loans</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CODE_GENDER</th>\n",
       "      <td>F</td>\n",
       "      <td>M</td>\n",
       "      <td>F</td>\n",
       "      <td>F</td>\n",
       "      <td>F</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FLAG_OWN_CAR</th>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FLAG_OWN_REALTY</th>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>Y</td>\n",
       "      <td>N</td>\n",
       "      <td>N</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>AMT_REQ_CREDIT_BUREAU_DAY</th>\n",
       "      <td>nan</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_WEEK</th>\n",
       "      <td>nan</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_MON</th>\n",
       "      <td>nan</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_QRT</th>\n",
       "      <td>nan</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AMT_REQ_CREDIT_BUREAU_YEAR</th>\n",
       "      <td>nan</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>121 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                     0           1                2  \\\n",
       "SK_ID_CURR                      428247      140966           407283   \n",
       "NAME_CONTRACT_TYPE          Cash loans  Cash loans  Revolving loans   \n",
       "CODE_GENDER                          F           M                F   \n",
       "FLAG_OWN_CAR                         N           Y                N   \n",
       "FLAG_OWN_REALTY                      N           Y                Y   \n",
       "...                                ...         ...              ...   \n",
       "AMT_REQ_CREDIT_BUREAU_DAY          nan         0.0              0.0   \n",
       "AMT_REQ_CREDIT_BUREAU_WEEK         nan         0.0              0.0   \n",
       "AMT_REQ_CREDIT_BUREAU_MON          nan         0.0              0.0   \n",
       "AMT_REQ_CREDIT_BUREAU_QRT          nan         0.0              0.0   \n",
       "AMT_REQ_CREDIT_BUREAU_YEAR         nan         0.0              0.0   \n",
       "\n",
       "                                     3           4  \n",
       "SK_ID_CURR                      434300      446788  \n",
       "NAME_CONTRACT_TYPE          Cash loans  Cash loans  \n",
       "CODE_GENDER                          F           F  \n",
       "FLAG_OWN_CAR                         N           N  \n",
       "FLAG_OWN_REALTY                      N           N  \n",
       "...                                ...         ...  \n",
       "AMT_REQ_CREDIT_BUREAU_DAY          0.0         0.0  \n",
       "AMT_REQ_CREDIT_BUREAU_WEEK         0.0         0.0  \n",
       "AMT_REQ_CREDIT_BUREAU_MON          0.0         0.0  \n",
       "AMT_REQ_CREDIT_BUREAU_QRT          1.0         0.0  \n",
       "AMT_REQ_CREDIT_BUREAU_YEAR         5.0         1.0  \n",
       "\n",
       "[121 rows x 5 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample = X.head(5).T\n",
    "sample.columns = [str(col) for col in sample.columns]  # fix integer name warning\n",
    "sample = sample.astype(str)  # avoid numeric conversion issues in viewer\n",
    "sample"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Hstk",
   "metadata": {},
   "source": [
    "**d. Target variable Distribution**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "nWHF",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "target_table, target_plot = plot_target_distribution(df=df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ZHCJ",
   "metadata": {},
   "source": [
    "**e. Number of columns of each data type**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ROlb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "float64    65\n",
       "int64      40\n",
       "object     16\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.dtypes.value_counts().sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "qnkX",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ORGANIZATION_TYPE             58\n",
       "OCCUPATION_TYPE               18\n",
       "NAME_INCOME_TYPE               8\n",
       "NAME_TYPE_SUITE                7\n",
       "WALLSMATERIAL_MODE             7\n",
       "WEEKDAY_APPR_PROCESS_START     7\n",
       "NAME_FAMILY_STATUS             6\n",
       "NAME_HOUSING_TYPE              6\n",
       "NAME_EDUCATION_TYPE            5\n",
       "FONDKAPREMONT_MODE             4\n",
       "HOUSETYPE_MODE                 3\n",
       "CODE_GENDER                    3\n",
       "FLAG_OWN_CAR                   2\n",
       "NAME_CONTRACT_TYPE             2\n",
       "FLAG_OWN_REALTY                2\n",
       "EMERGENCYSTATE_MODE            2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "categorical_cols = (\n",
    "    X.select_dtypes(include=[\"object\"]).nunique().sort_values(ascending=False)\n",
    ")\n",
    "categorical_cols"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "TqIu",
   "metadata": {},
   "source": [
    "**f. Missing data**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "Vxnm",
   "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>Count</th>\n",
       "      <th>percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>COMMONAREA_AVG</th>\n",
       "      <td>172189</td>\n",
       "      <td>69.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>COMMONAREA_MEDI</th>\n",
       "      <td>172189</td>\n",
       "      <td>69.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>COMMONAREA_MODE</th>\n",
       "      <td>172189</td>\n",
       "      <td>69.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NONLIVINGAPARTMENTS_AVG</th>\n",
       "      <td>171096</td>\n",
       "      <td>69.55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NONLIVINGAPARTMENTS_MEDI</th>\n",
       "      <td>171096</td>\n",
       "      <td>69.55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FLAG_DOCUMENT_16</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FLAG_DOCUMENT_15</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FLAG_DOCUMENT_14</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FLAG_DOCUMENT_20</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FLAG_DOCUMENT_21</th>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>121 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                           Count  percentage\n",
       "COMMONAREA_AVG            172189       69.99\n",
       "COMMONAREA_MEDI           172189       69.99\n",
       "COMMONAREA_MODE           172189       69.99\n",
       "NONLIVINGAPARTMENTS_AVG   171096       69.55\n",
       "NONLIVINGAPARTMENTS_MEDI  171096       69.55\n",
       "...                          ...         ...\n",
       "FLAG_DOCUMENT_16               0        0.00\n",
       "FLAG_DOCUMENT_15               0        0.00\n",
       "FLAG_DOCUMENT_14               0        0.00\n",
       "FLAG_DOCUMENT_20               0        0.00\n",
       "FLAG_DOCUMENT_21               0        0.00\n",
       "\n",
       "[121 rows x 2 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "missing_count = X.isna().sum().sort_values(ascending=False)\n",
    "missing_percentage = (missing_count / X.shape[0] * 100).round(2)\n",
    "\n",
    "missing_data = pd.DataFrame(\n",
    "    data={\"Count\": missing_count, \"percentage\": missing_percentage}\n",
    ")\n",
    "missing_data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "DnEU",
   "metadata": {},
   "source": [
    "### 1.2 Distribution of Variables\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ulZA",
   "metadata": {},
   "source": [
    "> Want to see how these plots were created? You can find the source code for the visualizations in [plots.py](https://huggingface.co/spaces/iBrokeTheCode/Home_Credit_Default_Risk_Prediction/blob/main/src/plots.py).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ecfG",
   "metadata": {},
   "source": [
    "**a. Credit Amounts**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "Pvdt",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "credit_plot = plot_credit_amounts(df=X)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ZBYS",
   "metadata": {},
   "source": [
    "**b. Education Level of Credit Applicants**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "aLJB",
   "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>Count</th>\n",
       "      <th>Percentage</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NAME_EDUCATION_TYPE</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Secondary / secondary special</th>\n",
       "      <td>174657</td>\n",
       "      <td>71.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Higher education</th>\n",
       "      <td>59990</td>\n",
       "      <td>24.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Incomplete higher</th>\n",
       "      <td>8248</td>\n",
       "      <td>3.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lower secondary</th>\n",
       "      <td>2984</td>\n",
       "      <td>1.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Academic degree</th>\n",
       "      <td>129</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                Count  Percentage\n",
       "NAME_EDUCATION_TYPE                              \n",
       "Secondary / secondary special  174657       71.00\n",
       "Higher education                59990       24.39\n",
       "Incomplete higher                8248        3.35\n",
       "Lower secondary                  2984        1.21\n",
       "Academic degree                   129        0.05"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "education_table, education_plot = plot_education_levels(df=X)\n",
    "education_table"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "xXTn",
   "metadata": {},
   "source": [
    "**c. Ocupation of Credit Applicants**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "AjVT",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OCCUPATION_TYPE\n",
       "NaN                      77119\n",
       "Laborers                 44108\n",
       "Sales staff              25770\n",
       "Core staff               22041\n",
       "Managers                 16992\n",
       "Drivers                  14874\n",
       "High skill tech staff     8981\n",
       "Accountants               7915\n",
       "Medicine staff            6879\n",
       "Security staff            5364\n",
       "Cooking staff             4771\n",
       "Cleaning staff            3731\n",
       "Private service staff     2140\n",
       "Low-skill Laborers        1687\n",
       "Waiters/barmen staff      1094\n",
       "Secretaries               1043\n",
       "Realty agents              625\n",
       "IT staff                   443\n",
       "HR staff                   431\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "occupation_table, occupation_plot = plot_occupation(df=X)\n",
    "occupation_table"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "NCOB",
   "metadata": {},
   "source": [
    "**d. Family Status of Applicants**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "aqbW",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NAME_FAMILY_STATUS\n",
       "Married                 157165\n",
       "Single / not married     36315\n",
       "Civil marriage           23841\n",
       "Separated                15803\n",
       "Widow                    12883\n",
       "Unknown                      1\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "family_status_table, family_status_plot = plot_family_status(df=X)\n",
    "family_status_table"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "TXez",
   "metadata": {},
   "source": [
    "**e. Income Type of Applicants by Target Variable**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dNNg",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "income_type_plot = plot_income_type(df=df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "yCnT",
   "metadata": {},
   "source": [
    "## 2. Preprocessing\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "wlCL",
   "metadata": {},
   "source": [
    "**a. Separate Train and Test Datasets**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "kqZH",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((196806, 121), (196806,), (49202, 121), (49202,))"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train, y_train, X_test, y_test = get_train_test_sets(X, y)\n",
    "X_train.shape, y_train.shape, X_test.shape, y_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "SFPL",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('Train dataset samples: 196806',\n",
       " 'Test dataset samples: 49202',\n",
       " 'Number of columns: 122')"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_samples = \"Train dataset samples: {}\".format(X_train.shape[0])\n",
    "test_samples = \"Test dataset samples: {}\".format(X_test.shape[0])\n",
    "columns_number = \"Number of columns: {}\".format(df.shape[1])\n",
    "\n",
    "train_samples, test_samples, columns_number"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "wAgl",
   "metadata": {},
   "source": [
    "**b. Preprocess Data**\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "rEll",
   "metadata": {},
   "source": [
    "This preprocessing perform:\n",
    "\n",
    "- Correct outliers/anomalous values in numerical columns (`DAYS_EMPLOYED` column).\n",
    "- Encode string categorical features (`dtype object`).\n",
    "  - If the feature has 2 categories, Binary Encoding is applied.\n",
    "  - One Hot Encoding for more than 2 categories.\n",
    "- Impute values for all columns with missing data (using median as imputing value).\n",
    "- Feature scaling with Min-Max scaler\n",
    "\n",
    "> Want to see how the dataset was processed? You can find the code for the preprocessing steps in [preprocessing.py](https://huggingface.co/spaces/iBrokeTheCode/Home_Credit_Default_Risk_Prediction/blob/main/src/preprocessing.py).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "dGlV",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((196806, 241), (49202, 241))"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data, test_data = preprocess_data_pipeline(train_df=X_train, test_df=X_test)\n",
    "train_data.shape, test_data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "SdmI",
   "metadata": {},
   "source": [
    "## 3. Training Models\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "lgWD",
   "metadata": {},
   "source": [
    "At this points, we will work with `train_data` and `test_data` as features sets; also `y_train` and `y_test` as target sets.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "yOPj",
   "metadata": {},
   "source": [
    "### 3.1 Logistic Regression\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fwwy",
   "metadata": {},
   "source": [
    "In Logistic Regression, C is the inverse of regularization strength:\n",
    "\n",
    "- **Small C** → Stronger regularization → Simpler model, less overfitting risk, but may underfit.\n",
    "- **Large C** → Weaker regularization → Model fits training data more closely, but may overfit.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "LJZf",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 📌 Logistic Regression\n",
    "log_reg = LogisticRegression(C=0.0001)\n",
    "log_reg.fit(train_data, y_train)\n",
    "\n",
    "# Train data predicton (class 1)\n",
    "lr_train_pred = log_reg.predict_proba(train_data)[:, 1]\n",
    "\n",
    "# Test data prediction (class 1)\n",
    "lr_test_pred = log_reg.predict_proba(test_data)[:, 1]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76169c1d",
   "metadata": {},
   "source": [
    "📈 The ROC AUC scores obtained:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "urSm",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'train_score': 0.6868418961663535, 'test_score': 0.6854973003347028}"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_reg_scores = {\n",
    "    \"train_score\": roc_auc_score(y_train, lr_train_pred),\n",
    "    \"test_score\": roc_auc_score(y_test, lr_test_pred),\n",
    "}\n",
    "log_reg_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "jxvo",
   "metadata": {},
   "source": [
    "### 3.2 Random Forest Classifier\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "mWxS",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 📌 Random Forest Classifier\n",
    "rf = RandomForestClassifier(random_state=42, n_jobs=-1)\n",
    "rf.fit(train_data, y_train)\n",
    "\n",
    "rf_train_pred = rf.predict_proba(train_data)[:, 1]\n",
    "rf_test_pred = rf.predict_proba(test_data)[:, 1]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "60b87d4f",
   "metadata": {},
   "source": [
    "📈 The ROC AUC scores obtained:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "CcZR",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'train_score': 1.0, 'test_score': 0.7066811557903828}"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_scores = {\n",
    "    \"train_score\": roc_auc_score(y_train, rf_train_pred),\n",
    "    \"test_score\": roc_auc_score(y_test, rf_test_pred),\n",
    "}\n",
    "rf_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "YWSi",
   "metadata": {},
   "source": [
    "### 3.3. Randomized Search with Cross Validations\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "zlud",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 📌 RandomizedSearchCV\n",
    "param_dist = {\"n_estimators\": [50, 100, 150], \"max_depth\": [10, 20, 30]}\n",
    "\n",
    "rf_optimized = RandomForestClassifier(random_state=42, n_jobs=-1)\n",
    "rscv = RandomizedSearchCV(\n",
    "    estimator=rf_optimized,\n",
    "    param_distributions=param_dist,\n",
    "    n_iter=5,\n",
    "    scoring=\"roc_auc\",\n",
    "    cv=3,\n",
    "    random_state=42,\n",
    "    n_jobs=-1,\n",
    ")\n",
    "\n",
    "rscv.fit(train_data, y_train)\n",
    "\n",
    "rfo_train_pred = rscv.predict_proba(train_data)[:, 1]\n",
    "rfo_test_pred = rscv.predict_proba(test_data)[:, 1]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "207ee10e",
   "metadata": {},
   "source": [
    "📈 The ROC AUC scores obtained:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "tZnO",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'train_score': 0.8196620915431655, 'test_score': 0.7308385425476998}"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfo_scores = {\n",
    "    \"train_score\": roc_auc_score(y_train, rfo_train_pred),\n",
    "    \"test_score\": roc_auc_score(y_test, rfo_test_pred),\n",
    "}\n",
    "rfo_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "xvXZ",
   "metadata": {},
   "source": [
    "🥇The best results:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "CLip",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'best_params_': {'n_estimators': 100, 'max_depth': 10},\n",
       " 'best_score_': 0.7296259755147781,\n",
       " 'best_estimator_': 'RandomForestClassifier(max_depth=10, n_jobs=-1, random_state=42)'}"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "optimized_results = {\n",
    "    \"best_params_\": {\"n_estimators\": 100, \"max_depth\": 10},\n",
    "    \"best_score_\": 0.7296259755147781,\n",
    "    \"best_estimator_\": \"RandomForestClassifier(max_depth=10, n_jobs=-1, random_state=42)\",\n",
    "}\n",
    "optimized_results"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "YECM",
   "metadata": {},
   "source": [
    "### 3.4 LightGBM\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "cEAS",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[LightGBM] [Info] Number of positive: 15784, number of negative: 181022\n",
      "[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.060479 seconds.\n",
      "You can set `force_col_wise=true` to remove the overhead.\n",
      "[LightGBM] [Info] Total Bins 11594\n",
      "[LightGBM] [Info] Number of data points in the train set: 196806, number of used features: 229\n",
      "[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.500000 -> initscore=-0.000000\n",
      "[LightGBM] [Info] Start training from score -0.000000\n"
     ]
    }
   ],
   "source": [
    "# 📌 LightGBM\n",
    "import warnings\n",
    "\n",
    "warnings.filterwarnings(\"ignore\", message=\"X does not have valid feature names\")\n",
    "\n",
    "# 📌 Get numerical and categorical variables (binary and mutiple)\n",
    "num_cols = X_train.select_dtypes(include=\"number\").columns.to_list()\n",
    "cat_cols = X_train.select_dtypes(include=\"object\").columns.to_list()\n",
    "\n",
    "binary_cols = [col for col in cat_cols if X_train[col].nunique() == 2]\n",
    "multi_cols = [col for col in cat_cols if X_train[col].nunique() > 2]\n",
    "\n",
    "# 📌 [1] Create the pipelines for different data types\n",
    "numerical_pipeline = Pipeline(\n",
    "    steps=[\n",
    "        (\"imputer\", SimpleImputer(strategy=\"median\")),\n",
    "        (\"scaler\", MinMaxScaler()),\n",
    "    ]\n",
    ")\n",
    "\n",
    "binary_pipeline = Pipeline(\n",
    "    steps=[\n",
    "        (\"imputer\", SimpleImputer(strategy=\"most_frequent\")),\n",
    "        (\"ordinal\", OrdinalEncoder()),\n",
    "        (\"scaler\", MinMaxScaler()),\n",
    "    ]\n",
    ")\n",
    "\n",
    "multi_pipeline = Pipeline(\n",
    "    steps=[\n",
    "        (\"imputer\", SimpleImputer(strategy=\"most_frequent\")),\n",
    "        (\n",
    "            \"onehot\",\n",
    "            OneHotEncoder(handle_unknown=\"ignore\", sparse_output=False),\n",
    "        ),\n",
    "        (\"scaler\", MinMaxScaler()),\n",
    "    ]\n",
    ")\n",
    "\n",
    "# 📌 [2] Create the preprocessor using ColumnTransformer\n",
    "preprocessor = ColumnTransformer(\n",
    "    transformers=[\n",
    "        (\"binary\", binary_pipeline, binary_cols),\n",
    "        (\"multi\", multi_pipeline, multi_cols),\n",
    "        (\"numerical\", numerical_pipeline, num_cols),\n",
    "    ],\n",
    "    remainder=\"passthrough\",\n",
    ")\n",
    "\n",
    "# 📌 [3] Create the Final Pipeline that combines the preprocessor and the model\n",
    "lgbm = LGBMClassifier(\n",
    "    n_estimators=500,\n",
    "    learning_rate=0.05,\n",
    "    max_depth=-1,\n",
    "    random_state=42,\n",
    "    class_weight=\"balanced\",\n",
    "    n_jobs=-1,\n",
    ")\n",
    "\n",
    "lgbm_pipeline = Pipeline(steps=[(\"preprocessor\", preprocessor), (\"classifier\", lgbm)])\n",
    "\n",
    "# 📌 [4] Fit the Final Pipeline on the ORIGINAL, unprocessed data\n",
    "# The pipeline takes care of all the preprocessing internally.\n",
    "lgbm_pipeline.fit(X_train, y_train)\n",
    "\n",
    "lgbm_train_pred = lgbm_pipeline.predict_proba(X_train)[:, 1]\n",
    "lgbm_test_pred = lgbm_pipeline.predict_proba(X_test)[:, 1]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7a3ebe11",
   "metadata": {},
   "source": [
    "📈 The ROC AUC scores obtained:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "5f483924",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'train_score': 0.8523466410959462, 'test_score': 0.7514895868142193}"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgbm_scores = {\n",
    "    \"train_score\": roc_auc_score(y_train, lgbm_train_pred),\n",
    "    \"test_score\": roc_auc_score(y_test, lgbm_test_pred),\n",
    "}\n",
    "lgbm_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ffd9c44d",
   "metadata": {},
   "source": [
    "## 4. Model Performance Analysis\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e48e2ab",
   "metadata": {},
   "source": [
    "### 4.1 Logistic Regression\n",
    "\n",
    "_The Logistic Regression model shows a `train_score` of 0.687 and a `test_score` of 0.685._\n",
    "\n",
    "**Interpretation:** This model's performance is consistent across the training and testing sets, as indicated by the very small gap between the scores. This means the model is not overfitting. However, the overall scores are relatively low for a binary classification task, suggesting that the model is likely **underfitting**. It's too simple to capture the underlying patterns in the data effectively.\n",
    "\n",
    "### 4.2 Random Forest Classifier\n",
    "\n",
    "_The base Random Forest model produced a `train_score` of 1.0 and a `test_score` of 0.707._\n",
    "\n",
    "**Interpretation:** The perfect `train_score` of 1.0 is a clear and severe sign of **overfitting**. The model has essentially memorized the training data, and this does not generalize well to unseen data, as shown by the much lower `test_score`. While the test score is better than the Logistic Regression, the model is too complex and needs to be regularized or tuned to perform better.\n",
    "\n",
    "### 4.3 Randomized Search with Cross Validations (Random Forest)\n",
    "\n",
    "_The hyperparameter-tuned Random Forest model achieved a `train_score` of 0.820 and a `test_score` of 0.731._\n",
    "\n",
    "**Interpretation:** This is a much better result than the base Random Forest. The gap between the `train_score` and `test_score` is significantly smaller, indicating that the hyperparameter tuning successfully **reduced overfitting**. The `test_score` of 0.731 is also a notable improvement, showing that the model now generalizes better to unseen data. This is a well-performing and well-tuned model.\n",
    "\n",
    "### 4.4 LightGBM\n",
    "\n",
    "_The LightGBM model produced a `train_score` of 0.852 and a `test_score` of 0.751._\n",
    "\n",
    "**Interpretation:** The LightGBM model shows the best overall performance with the highest `test_score` of 0.751. There is a small gap between the training and testing scores, which is normal for a powerful boosting model, suggesting a good balance between capturing complex patterns and generalizing well. The model is performing exceptionally and is neither severely overfitting nor underfitting.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5d48c191",
   "metadata": {},
   "source": [
    "## 5. Model Selection\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10e2b528",
   "metadata": {},
   "source": [
    "Based on a comparison of all the models, the final model selection is clear.\n",
    "\n",
    "| Model                        | Train Score (AUC ROC) | Test Score (AUC ROC) |\n",
    "| :--------------------------- | :-------------------: | :------------------: |\n",
    "| Logistic Regression          |         0.687         |        0.685         |\n",
    "| Random Forest Classifier     |         1.000         |        0.707         |\n",
    "| Randomized Search (Tuned RF) |         0.820         |        0.731         |\n",
    "| **LightGBM**                 |         0.852         |      **0.751**       |\n",
    "\n",
    "- The **Logistic Regression** model performed poorly due to underfitting.\n",
    "- The base **Random Forest** model, while better, suffered from severe overfitting.\n",
    "- The tuned **Random Forest** model was a significant improvement and a strong contender, achieving a solid `test_score`.\n",
    "- However, the **LightGBM** model ultimately demonstrated the best performance, achieving the highest **ROC AUC test score of 0.751**. This indicates that it is the most robust and accurate model for predicting loan repayment risk on unseen data.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bd405b69",
   "metadata": {},
   "source": [
    "> 🥇 Therefore, we will select the **LightGBM** model as our final choice for deployment.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4aa60dcb",
   "metadata": {},
   "source": [
    "## 6. Saving the Model\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bca2853b",
   "metadata": {},
   "source": [
    "### 6.1 Saving the Model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3246c249",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model saved successfully as lgbm_model.joblib\n"
     ]
    }
   ],
   "source": [
    "joblib.dump(lgbm_pipeline, \"model/lgbm_model.joblib\")\n",
    "print(\"Model saved successfully as lgbm_model.joblib\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a69129f2",
   "metadata": {},
   "source": [
    "### 6.2 Feature Importances\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5db0729b",
   "metadata": {},
   "source": [
    "We will select the top 10 features based on their importances, in order to use them in the prediction interface.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "12e917d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 86,  65,  21,   0,  72,  50,   0,   9,  23,   0,   6,   1,  10,\n",
       "        34,   0,  13,   0,   9,  28,   0,   0,  46,  27,  70,  10,  12,\n",
       "        47,  23,  58,  20,  27,   0,  21,   3,  30,  12,  12,   8,  11,\n",
       "        33,   9,   7,  43,  28,   3,  15,   0,  28,  17,  10,  15,   2,\n",
       "         2,  15,   2,   6,   2,  22,  35,  20,  25,  22,  21,  21,   3,\n",
       "         4,  10,   7,   8,  31,   0,  32,   1,   7,   1,  12,   4,   1,\n",
       "         2,   0,   2,  17,   0,   0,  10,   2,   8,   0,   8,   0,  20,\n",
       "         0,   6,   9,  12,  23,   0,   6,   9,   3,  23,   0,   6,  23,\n",
       "         3,  18,  28,  10,   0,   0,   4,  12,   0,   0,   4,   4,   3,\n",
       "         6,  23,  10,   5,   0,  13,  14,  11,   7,   2,   5,   1,  10,\n",
       "         2,   2,   5,  15,  19,   0, 536,  47, 410, 642, 660, 489, 375,\n",
       "       752, 630, 556, 628, 196,   0,   2,  50,   0,  45,  13,  75,  24,\n",
       "        54, 230,   0,  23,  10,  36,  24,  16, 732, 792, 833, 130,  85,\n",
       "       100,  58,  77,  42,  66,  57,  36, 116,  75,  90,  54, 106, 106,\n",
       "       129, 112,  69, 111,  26,  39,  10,  22, 104,  69, 131,  37,  99,\n",
       "        47,  67,  93,  29,  40,   7,  23,  18,  16,  67,  34,  43,  19,\n",
       "        76, 176, 121,  68,  38,  57, 536,   0,  75,   0,   8,  10,   0,\n",
       "        10,   5,   0,  16,   0,  21,  14,  14,  33,   0,  31,   5,   0,\n",
       "         1,   2,  10,  29,  69,  82, 141], dtype=int32)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loaded_pipeline = joblib.load(\"model/lgbm_model.joblib\")\n",
    "feature_importances = loaded_pipeline.named_steps[\"classifier\"].feature_importances_\n",
    "feature_importances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "ad40c56f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['binary__NAME_CONTRACT_TYPE', 'binary__FLAG_OWN_CAR',\n",
       "       'binary__FLAG_OWN_REALTY', 'binary__EMERGENCYSTATE_MODE',\n",
       "       'multi__CODE_GENDER_F', 'multi__CODE_GENDER_M',\n",
       "       'multi__CODE_GENDER_XNA', 'multi__NAME_TYPE_SUITE_Children',\n",
       "       'multi__NAME_TYPE_SUITE_Family',\n",
       "       'multi__NAME_TYPE_SUITE_Group of people',\n",
       "       'multi__NAME_TYPE_SUITE_Other_A', 'multi__NAME_TYPE_SUITE_Other_B',\n",
       "       'multi__NAME_TYPE_SUITE_Spouse, partner',\n",
       "       'multi__NAME_TYPE_SUITE_Unaccompanied',\n",
       "       'multi__NAME_INCOME_TYPE_Businessman',\n",
       "       'multi__NAME_INCOME_TYPE_Commercial associate',\n",
       "       'multi__NAME_INCOME_TYPE_Maternity leave',\n",
       "       'multi__NAME_INCOME_TYPE_Pensioner',\n",
       "       'multi__NAME_INCOME_TYPE_State servant',\n",
       "       'multi__NAME_INCOME_TYPE_Student',\n",
       "       'multi__NAME_INCOME_TYPE_Unemployed',\n",
       "       'multi__NAME_INCOME_TYPE_Working',\n",
       "       'multi__NAME_EDUCATION_TYPE_Academic degree',\n",
       "       'multi__NAME_EDUCATION_TYPE_Higher education',\n",
       "       'multi__NAME_EDUCATION_TYPE_Incomplete higher',\n",
       "       'multi__NAME_EDUCATION_TYPE_Lower secondary',\n",
       "       'multi__NAME_EDUCATION_TYPE_Secondary / secondary special',\n",
       "       'multi__NAME_FAMILY_STATUS_Civil marriage',\n",
       "       'multi__NAME_FAMILY_STATUS_Married',\n",
       "       'multi__NAME_FAMILY_STATUS_Separated',\n",
       "       'multi__NAME_FAMILY_STATUS_Single / not married',\n",
       "       'multi__NAME_FAMILY_STATUS_Unknown',\n",
       "       'multi__NAME_FAMILY_STATUS_Widow',\n",
       "       'multi__NAME_HOUSING_TYPE_Co-op apartment',\n",
       "       'multi__NAME_HOUSING_TYPE_House / apartment',\n",
       "       'multi__NAME_HOUSING_TYPE_Municipal apartment',\n",
       "       'multi__NAME_HOUSING_TYPE_Office apartment',\n",
       "       'multi__NAME_HOUSING_TYPE_Rented apartment',\n",
       "       'multi__NAME_HOUSING_TYPE_With parents',\n",
       "       'multi__OCCUPATION_TYPE_Accountants',\n",
       "       'multi__OCCUPATION_TYPE_Cleaning staff',\n",
       "       'multi__OCCUPATION_TYPE_Cooking staff',\n",
       "       'multi__OCCUPATION_TYPE_Core staff',\n",
       "       'multi__OCCUPATION_TYPE_Drivers',\n",
       "       'multi__OCCUPATION_TYPE_HR staff',\n",
       "       'multi__OCCUPATION_TYPE_High skill tech staff',\n",
       "       'multi__OCCUPATION_TYPE_IT staff',\n",
       "       'multi__OCCUPATION_TYPE_Laborers',\n",
       "       'multi__OCCUPATION_TYPE_Low-skill Laborers',\n",
       "       'multi__OCCUPATION_TYPE_Managers',\n",
       "       'multi__OCCUPATION_TYPE_Medicine staff',\n",
       "       'multi__OCCUPATION_TYPE_Private service staff',\n",
       "       'multi__OCCUPATION_TYPE_Realty agents',\n",
       "       'multi__OCCUPATION_TYPE_Sales staff',\n",
       "       'multi__OCCUPATION_TYPE_Secretaries',\n",
       "       'multi__OCCUPATION_TYPE_Security staff',\n",
       "       'multi__OCCUPATION_TYPE_Waiters/barmen staff',\n",
       "       'multi__WEEKDAY_APPR_PROCESS_START_FRIDAY',\n",
       "       'multi__WEEKDAY_APPR_PROCESS_START_MONDAY',\n",
       "       'multi__WEEKDAY_APPR_PROCESS_START_SATURDAY',\n",
       "       'multi__WEEKDAY_APPR_PROCESS_START_SUNDAY',\n",
       "       'multi__WEEKDAY_APPR_PROCESS_START_THURSDAY',\n",
       "       'multi__WEEKDAY_APPR_PROCESS_START_TUESDAY',\n",
       "       'multi__WEEKDAY_APPR_PROCESS_START_WEDNESDAY',\n",
       "       'multi__ORGANIZATION_TYPE_Advertising',\n",
       "       'multi__ORGANIZATION_TYPE_Agriculture',\n",
       "       'multi__ORGANIZATION_TYPE_Bank',\n",
       "       'multi__ORGANIZATION_TYPE_Business Entity Type 1',\n",
       "       'multi__ORGANIZATION_TYPE_Business Entity Type 2',\n",
       "       'multi__ORGANIZATION_TYPE_Business Entity Type 3',\n",
       "       'multi__ORGANIZATION_TYPE_Cleaning',\n",
       "       'multi__ORGANIZATION_TYPE_Construction',\n",
       "       'multi__ORGANIZATION_TYPE_Culture',\n",
       "       'multi__ORGANIZATION_TYPE_Electricity',\n",
       "       'multi__ORGANIZATION_TYPE_Emergency',\n",
       "       'multi__ORGANIZATION_TYPE_Government',\n",
       "       'multi__ORGANIZATION_TYPE_Hotel',\n",
       "       'multi__ORGANIZATION_TYPE_Housing',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 1',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 10',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 11',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 12',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 13',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 2',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 3',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 4',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 5',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 6',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 7',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 8',\n",
       "       'multi__ORGANIZATION_TYPE_Industry: type 9',\n",
       "       'multi__ORGANIZATION_TYPE_Insurance',\n",
       "       'multi__ORGANIZATION_TYPE_Kindergarten',\n",
       "       'multi__ORGANIZATION_TYPE_Legal Services',\n",
       "       'multi__ORGANIZATION_TYPE_Medicine',\n",
       "       'multi__ORGANIZATION_TYPE_Military',\n",
       "       'multi__ORGANIZATION_TYPE_Mobile',\n",
       "       'multi__ORGANIZATION_TYPE_Other',\n",
       "       'multi__ORGANIZATION_TYPE_Police',\n",
       "       'multi__ORGANIZATION_TYPE_Postal',\n",
       "       'multi__ORGANIZATION_TYPE_Realtor',\n",
       "       'multi__ORGANIZATION_TYPE_Religion',\n",
       "       'multi__ORGANIZATION_TYPE_Restaurant',\n",
       "       'multi__ORGANIZATION_TYPE_School',\n",
       "       'multi__ORGANIZATION_TYPE_Security',\n",
       "       'multi__ORGANIZATION_TYPE_Security Ministries',\n",
       "       'multi__ORGANIZATION_TYPE_Self-employed',\n",
       "       'multi__ORGANIZATION_TYPE_Services',\n",
       "       'multi__ORGANIZATION_TYPE_Telecom',\n",
       "       'multi__ORGANIZATION_TYPE_Trade: type 1',\n",
       "       'multi__ORGANIZATION_TYPE_Trade: type 2',\n",
       "       'multi__ORGANIZATION_TYPE_Trade: type 3',\n",
       "       'multi__ORGANIZATION_TYPE_Trade: type 4',\n",
       "       'multi__ORGANIZATION_TYPE_Trade: type 5',\n",
       "       'multi__ORGANIZATION_TYPE_Trade: type 6',\n",
       "       'multi__ORGANIZATION_TYPE_Trade: type 7',\n",
       "       'multi__ORGANIZATION_TYPE_Transport: type 1',\n",
       "       'multi__ORGANIZATION_TYPE_Transport: type 2',\n",
       "       'multi__ORGANIZATION_TYPE_Transport: type 3',\n",
       "       'multi__ORGANIZATION_TYPE_Transport: type 4',\n",
       "       'multi__ORGANIZATION_TYPE_University',\n",
       "       'multi__ORGANIZATION_TYPE_XNA',\n",
       "       'multi__FONDKAPREMONT_MODE_not specified',\n",
       "       'multi__FONDKAPREMONT_MODE_org spec account',\n",
       "       'multi__FONDKAPREMONT_MODE_reg oper account',\n",
       "       'multi__FONDKAPREMONT_MODE_reg oper spec account',\n",
       "       'multi__HOUSETYPE_MODE_block of flats',\n",
       "       'multi__HOUSETYPE_MODE_specific housing',\n",
       "       'multi__HOUSETYPE_MODE_terraced house',\n",
       "       'multi__WALLSMATERIAL_MODE_Block',\n",
       "       'multi__WALLSMATERIAL_MODE_Mixed',\n",
       "       'multi__WALLSMATERIAL_MODE_Monolithic',\n",
       "       'multi__WALLSMATERIAL_MODE_Others',\n",
       "       'multi__WALLSMATERIAL_MODE_Panel',\n",
       "       'multi__WALLSMATERIAL_MODE_Stone, brick',\n",
       "       'multi__WALLSMATERIAL_MODE_Wooden', 'numerical__SK_ID_CURR',\n",
       "       'numerical__CNT_CHILDREN', 'numerical__AMT_INCOME_TOTAL',\n",
       "       'numerical__AMT_CREDIT', 'numerical__AMT_ANNUITY',\n",
       "       'numerical__AMT_GOODS_PRICE',\n",
       "       'numerical__REGION_POPULATION_RELATIVE', 'numerical__DAYS_BIRTH',\n",
       "       'numerical__DAYS_EMPLOYED', 'numerical__DAYS_REGISTRATION',\n",
       "       'numerical__DAYS_ID_PUBLISH', 'numerical__OWN_CAR_AGE',\n",
       "       'numerical__FLAG_MOBIL', 'numerical__FLAG_EMP_PHONE',\n",
       "       'numerical__FLAG_WORK_PHONE', 'numerical__FLAG_CONT_MOBILE',\n",
       "       'numerical__FLAG_PHONE', 'numerical__FLAG_EMAIL',\n",
       "       'numerical__CNT_FAM_MEMBERS', 'numerical__REGION_RATING_CLIENT',\n",
       "       'numerical__REGION_RATING_CLIENT_W_CITY',\n",
       "       'numerical__HOUR_APPR_PROCESS_START',\n",
       "       'numerical__REG_REGION_NOT_LIVE_REGION',\n",
       "       'numerical__REG_REGION_NOT_WORK_REGION',\n",
       "       'numerical__LIVE_REGION_NOT_WORK_REGION',\n",
       "       'numerical__REG_CITY_NOT_LIVE_CITY',\n",
       "       'numerical__REG_CITY_NOT_WORK_CITY',\n",
       "       'numerical__LIVE_CITY_NOT_WORK_CITY', 'numerical__EXT_SOURCE_1',\n",
       "       'numerical__EXT_SOURCE_2', 'numerical__EXT_SOURCE_3',\n",
       "       'numerical__APARTMENTS_AVG', 'numerical__BASEMENTAREA_AVG',\n",
       "       'numerical__YEARS_BEGINEXPLUATATION_AVG',\n",
       "       'numerical__YEARS_BUILD_AVG', 'numerical__COMMONAREA_AVG',\n",
       "       'numerical__ELEVATORS_AVG', 'numerical__ENTRANCES_AVG',\n",
       "       'numerical__FLOORSMAX_AVG', 'numerical__FLOORSMIN_AVG',\n",
       "       'numerical__LANDAREA_AVG', 'numerical__LIVINGAPARTMENTS_AVG',\n",
       "       'numerical__LIVINGAREA_AVG', 'numerical__NONLIVINGAPARTMENTS_AVG',\n",
       "       'numerical__NONLIVINGAREA_AVG', 'numerical__APARTMENTS_MODE',\n",
       "       'numerical__BASEMENTAREA_MODE',\n",
       "       'numerical__YEARS_BEGINEXPLUATATION_MODE',\n",
       "       'numerical__YEARS_BUILD_MODE', 'numerical__COMMONAREA_MODE',\n",
       "       'numerical__ELEVATORS_MODE', 'numerical__ENTRANCES_MODE',\n",
       "       'numerical__FLOORSMAX_MODE', 'numerical__FLOORSMIN_MODE',\n",
       "       'numerical__LANDAREA_MODE', 'numerical__LIVINGAPARTMENTS_MODE',\n",
       "       'numerical__LIVINGAREA_MODE',\n",
       "       'numerical__NONLIVINGAPARTMENTS_MODE',\n",
       "       'numerical__NONLIVINGAREA_MODE', 'numerical__APARTMENTS_MEDI',\n",
       "       'numerical__BASEMENTAREA_MEDI',\n",
       "       'numerical__YEARS_BEGINEXPLUATATION_MEDI',\n",
       "       'numerical__YEARS_BUILD_MEDI', 'numerical__COMMONAREA_MEDI',\n",
       "       'numerical__ELEVATORS_MEDI', 'numerical__ENTRANCES_MEDI',\n",
       "       'numerical__FLOORSMAX_MEDI', 'numerical__FLOORSMIN_MEDI',\n",
       "       'numerical__LANDAREA_MEDI', 'numerical__LIVINGAPARTMENTS_MEDI',\n",
       "       'numerical__LIVINGAREA_MEDI',\n",
       "       'numerical__NONLIVINGAPARTMENTS_MEDI',\n",
       "       'numerical__NONLIVINGAREA_MEDI', 'numerical__TOTALAREA_MODE',\n",
       "       'numerical__OBS_30_CNT_SOCIAL_CIRCLE',\n",
       "       'numerical__DEF_30_CNT_SOCIAL_CIRCLE',\n",
       "       'numerical__OBS_60_CNT_SOCIAL_CIRCLE',\n",
       "       'numerical__DEF_60_CNT_SOCIAL_CIRCLE',\n",
       "       'numerical__DAYS_LAST_PHONE_CHANGE', 'numerical__FLAG_DOCUMENT_2',\n",
       "       'numerical__FLAG_DOCUMENT_3', 'numerical__FLAG_DOCUMENT_4',\n",
       "       'numerical__FLAG_DOCUMENT_5', 'numerical__FLAG_DOCUMENT_6',\n",
       "       'numerical__FLAG_DOCUMENT_7', 'numerical__FLAG_DOCUMENT_8',\n",
       "       'numerical__FLAG_DOCUMENT_9', 'numerical__FLAG_DOCUMENT_10',\n",
       "       'numerical__FLAG_DOCUMENT_11', 'numerical__FLAG_DOCUMENT_12',\n",
       "       'numerical__FLAG_DOCUMENT_13', 'numerical__FLAG_DOCUMENT_14',\n",
       "       'numerical__FLAG_DOCUMENT_15', 'numerical__FLAG_DOCUMENT_16',\n",
       "       'numerical__FLAG_DOCUMENT_17', 'numerical__FLAG_DOCUMENT_18',\n",
       "       'numerical__FLAG_DOCUMENT_19', 'numerical__FLAG_DOCUMENT_20',\n",
       "       'numerical__FLAG_DOCUMENT_21',\n",
       "       'numerical__AMT_REQ_CREDIT_BUREAU_HOUR',\n",
       "       'numerical__AMT_REQ_CREDIT_BUREAU_DAY',\n",
       "       'numerical__AMT_REQ_CREDIT_BUREAU_WEEK',\n",
       "       'numerical__AMT_REQ_CREDIT_BUREAU_MON',\n",
       "       'numerical__AMT_REQ_CREDIT_BUREAU_QRT',\n",
       "       'numerical__AMT_REQ_CREDIT_BUREAU_YEAR'], dtype=object)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get then names of the final features after preprocessing\n",
    "preprocessor = loaded_pipeline.named_steps[\"preprocessor\"]\n",
    "final_features_names = preprocessor.get_feature_names_out()\n",
    "final_features_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "336a580f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a DataFrame to store the feature names and their corresponding importances\n",
    "feature_importances_df = pd.DataFrame(\n",
    "    {\"feature\": final_features_names, \"importance\": feature_importances}\n",
    ")\n",
    "\n",
    "sorted_feature_importance = feature_importances_df.sort_values(\n",
    "    by=\"importance\", ascending=False\n",
    ").reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e86a01f7",
   "metadata": {},
   "source": [
    "**Top 10 most important features**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "c7bc9e30",
   "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>feature</th>\n",
       "      <th>importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>numerical__EXT_SOURCE_3</td>\n",
       "      <td>833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>numerical__EXT_SOURCE_2</td>\n",
       "      <td>792</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>numerical__DAYS_BIRTH</td>\n",
       "      <td>752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>numerical__EXT_SOURCE_1</td>\n",
       "      <td>732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>numerical__AMT_ANNUITY</td>\n",
       "      <td>660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>numerical__AMT_CREDIT</td>\n",
       "      <td>642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>numerical__DAYS_EMPLOYED</td>\n",
       "      <td>630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>numerical__DAYS_ID_PUBLISH</td>\n",
       "      <td>628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>numerical__DAYS_REGISTRATION</td>\n",
       "      <td>556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>numerical__SK_ID_CURR</td>\n",
       "      <td>536</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        feature  importance\n",
       "0       numerical__EXT_SOURCE_3         833\n",
       "1       numerical__EXT_SOURCE_2         792\n",
       "2         numerical__DAYS_BIRTH         752\n",
       "3       numerical__EXT_SOURCE_1         732\n",
       "4        numerical__AMT_ANNUITY         660\n",
       "5         numerical__AMT_CREDIT         642\n",
       "6      numerical__DAYS_EMPLOYED         630\n",
       "7    numerical__DAYS_ID_PUBLISH         628\n",
       "8  numerical__DAYS_REGISTRATION         556\n",
       "9         numerical__SK_ID_CURR         536"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted_feature_importance.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "37d77ecb",
   "metadata": {},
   "source": [
    "**Calculate default values for remaining features**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "0c5f45cb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'SK_ID_CURR': 277659.5,\n",
       " 'CNT_CHILDREN': 0.0,\n",
       " 'AMT_INCOME_TOTAL': 147150.0,\n",
       " 'AMT_CREDIT': 512997.75,\n",
       " 'AMT_ANNUITY': 24885.0,\n",
       " 'AMT_GOODS_PRICE': 450000.0,\n",
       " 'REGION_POPULATION_RELATIVE': 0.01885,\n",
       " 'DAYS_BIRTH': -15743.5,\n",
       " 'DAYS_EMPLOYED': -1219.0,\n",
       " 'DAYS_REGISTRATION': -4492.0,\n",
       " 'DAYS_ID_PUBLISH': -3254.0,\n",
       " 'OWN_CAR_AGE': 9.0,\n",
       " 'FLAG_MOBIL': 1.0,\n",
       " 'FLAG_EMP_PHONE': 1.0,\n",
       " 'FLAG_WORK_PHONE': 0.0,\n",
       " 'FLAG_CONT_MOBILE': 1.0,\n",
       " 'FLAG_PHONE': 0.0,\n",
       " 'FLAG_EMAIL': 0.0,\n",
       " 'CNT_FAM_MEMBERS': 2.0,\n",
       " 'REGION_RATING_CLIENT': 2.0,\n",
       " 'REGION_RATING_CLIENT_W_CITY': 2.0,\n",
       " 'HOUR_APPR_PROCESS_START': 12.0,\n",
       " 'REG_REGION_NOT_LIVE_REGION': 0.0,\n",
       " 'REG_REGION_NOT_WORK_REGION': 0.0,\n",
       " 'LIVE_REGION_NOT_WORK_REGION': 0.0,\n",
       " 'REG_CITY_NOT_LIVE_CITY': 0.0,\n",
       " 'REG_CITY_NOT_WORK_CITY': 0.0,\n",
       " 'LIVE_CITY_NOT_WORK_CITY': 0.0,\n",
       " 'EXT_SOURCE_1': 0.5068839442599388,\n",
       " 'EXT_SOURCE_2': 0.5662837032261614,\n",
       " 'EXT_SOURCE_3': 0.5370699579791587,\n",
       " 'APARTMENTS_AVG': 0.0876,\n",
       " 'BASEMENTAREA_AVG': 0.0764,\n",
       " 'YEARS_BEGINEXPLUATATION_AVG': 0.9816,\n",
       " 'YEARS_BUILD_AVG': 0.7552,\n",
       " 'COMMONAREA_AVG': 0.0211,\n",
       " 'ELEVATORS_AVG': 0.0,\n",
       " 'ENTRANCES_AVG': 0.1379,\n",
       " 'FLOORSMAX_AVG': 0.1667,\n",
       " 'FLOORSMIN_AVG': 0.2083,\n",
       " 'LANDAREA_AVG': 0.0483,\n",
       " 'LIVINGAPARTMENTS_AVG': 0.0756,\n",
       " 'LIVINGAREA_AVG': 0.0746,\n",
       " 'NONLIVINGAPARTMENTS_AVG': 0.0,\n",
       " 'NONLIVINGAREA_AVG': 0.0035,\n",
       " 'APARTMENTS_MODE': 0.084,\n",
       " 'BASEMENTAREA_MODE': 0.0748,\n",
       " 'YEARS_BEGINEXPLUATATION_MODE': 0.9816,\n",
       " 'YEARS_BUILD_MODE': 0.7648,\n",
       " 'COMMONAREA_MODE': 0.0191,\n",
       " 'ELEVATORS_MODE': 0.0,\n",
       " 'ENTRANCES_MODE': 0.1379,\n",
       " 'FLOORSMAX_MODE': 0.1667,\n",
       " 'FLOORSMIN_MODE': 0.2083,\n",
       " 'LANDAREA_MODE': 0.0459,\n",
       " 'LIVINGAPARTMENTS_MODE': 0.0771,\n",
       " 'LIVINGAREA_MODE': 0.0731,\n",
       " 'NONLIVINGAPARTMENTS_MODE': 0.0,\n",
       " 'NONLIVINGAREA_MODE': 0.0011,\n",
       " 'APARTMENTS_MEDI': 0.0864,\n",
       " 'BASEMENTAREA_MEDI': 0.0761,\n",
       " 'YEARS_BEGINEXPLUATATION_MEDI': 0.9816,\n",
       " 'YEARS_BUILD_MEDI': 0.7585,\n",
       " 'COMMONAREA_MEDI': 0.0209,\n",
       " 'ELEVATORS_MEDI': 0.0,\n",
       " 'ENTRANCES_MEDI': 0.1379,\n",
       " 'FLOORSMAX_MEDI': 0.1667,\n",
       " 'FLOORSMIN_MEDI': 0.2083,\n",
       " 'LANDAREA_MEDI': 0.0488,\n",
       " 'LIVINGAPARTMENTS_MEDI': 0.0765,\n",
       " 'LIVINGAREA_MEDI': 0.0749,\n",
       " 'NONLIVINGAPARTMENTS_MEDI': 0.0,\n",
       " 'NONLIVINGAREA_MEDI': 0.003,\n",
       " 'TOTALAREA_MODE': 0.0687,\n",
       " 'OBS_30_CNT_SOCIAL_CIRCLE': 0.0,\n",
       " 'DEF_30_CNT_SOCIAL_CIRCLE': 0.0,\n",
       " 'OBS_60_CNT_SOCIAL_CIRCLE': 0.0,\n",
       " 'DEF_60_CNT_SOCIAL_CIRCLE': 0.0,\n",
       " 'DAYS_LAST_PHONE_CHANGE': -755.0,\n",
       " 'FLAG_DOCUMENT_2': 0.0,\n",
       " 'FLAG_DOCUMENT_3': 1.0,\n",
       " 'FLAG_DOCUMENT_4': 0.0,\n",
       " 'FLAG_DOCUMENT_5': 0.0,\n",
       " 'FLAG_DOCUMENT_6': 0.0,\n",
       " 'FLAG_DOCUMENT_7': 0.0,\n",
       " 'FLAG_DOCUMENT_8': 0.0,\n",
       " 'FLAG_DOCUMENT_9': 0.0,\n",
       " 'FLAG_DOCUMENT_10': 0.0,\n",
       " 'FLAG_DOCUMENT_11': 0.0,\n",
       " 'FLAG_DOCUMENT_12': 0.0,\n",
       " 'FLAG_DOCUMENT_13': 0.0,\n",
       " 'FLAG_DOCUMENT_14': 0.0,\n",
       " 'FLAG_DOCUMENT_15': 0.0,\n",
       " 'FLAG_DOCUMENT_16': 0.0,\n",
       " 'FLAG_DOCUMENT_17': 0.0,\n",
       " 'FLAG_DOCUMENT_18': 0.0,\n",
       " 'FLAG_DOCUMENT_19': 0.0,\n",
       " 'FLAG_DOCUMENT_20': 0.0,\n",
       " 'FLAG_DOCUMENT_21': 0.0,\n",
       " 'AMT_REQ_CREDIT_BUREAU_HOUR': 0.0,\n",
       " 'AMT_REQ_CREDIT_BUREAU_DAY': 0.0,\n",
       " 'AMT_REQ_CREDIT_BUREAU_WEEK': 0.0,\n",
       " 'AMT_REQ_CREDIT_BUREAU_MON': 0.0,\n",
       " 'AMT_REQ_CREDIT_BUREAU_QRT': 0.0,\n",
       " 'AMT_REQ_CREDIT_BUREAU_YEAR': 1.0,\n",
       " 'NAME_CONTRACT_TYPE': 'Cash loans',\n",
       " 'CODE_GENDER': 'F',\n",
       " 'FLAG_OWN_CAR': 'N',\n",
       " 'FLAG_OWN_REALTY': 'Y',\n",
       " 'NAME_TYPE_SUITE': 'Unaccompanied',\n",
       " 'NAME_INCOME_TYPE': 'Working',\n",
       " 'NAME_EDUCATION_TYPE': 'Secondary / secondary special',\n",
       " 'NAME_FAMILY_STATUS': 'Married',\n",
       " 'NAME_HOUSING_TYPE': 'House / apartment',\n",
       " 'OCCUPATION_TYPE': 'Laborers',\n",
       " 'WEEKDAY_APPR_PROCESS_START': 'TUESDAY',\n",
       " 'ORGANIZATION_TYPE': 'Business Entity Type 3',\n",
       " 'FONDKAPREMONT_MODE': 'reg oper account',\n",
       " 'HOUSETYPE_MODE': 'block of flats',\n",
       " 'WALLSMATERIAL_MODE': 'Panel',\n",
       " 'EMERGENCYSTATE_MODE': 'No'}"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_features = X_train.columns.to_list()\n",
    "ui_features = sorted_feature_importance[\"feature\"].head(10).tolist()\n",
    "\n",
    "default_values = {}\n",
    "\n",
    "num_default = X_train.select_dtypes(include=[\"number\"]).median().to_dict()\n",
    "default_values.update(num_default)\n",
    "\n",
    "cat_defaults = X_train.select_dtypes(include=[\"object\"]).mode().iloc[0].to_dict()\n",
    "default_values.update(cat_defaults)\n",
    "\n",
    "default_values\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c744b94",
   "metadata": {},
   "source": [
    "**Calculate the minimum and maximum values for each feature in the dataset**\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "5ddefb61",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'EXT_SOURCE_3': (np.float64(0.0005272652387098),\n",
       "  np.float64(0.8960095494948396)),\n",
       " 'EXT_SOURCE_2': (np.float64(5.002108762101576e-06),\n",
       "  np.float64(0.8549996664047012)),\n",
       " 'DAYS_BIRTH': (np.int64(-25229), np.int64(-7673)),\n",
       " 'EXT_SOURCE_1': (np.float64(0.0145681324124455),\n",
       "  np.float64(0.962692770561306)),\n",
       " 'AMT_ANNUITY': (np.float64(1980.0), np.float64(258025.5)),\n",
       " 'AMT_CREDIT': (np.float64(45000.0), np.float64(4050000.0)),\n",
       " 'DAYS_EMPLOYED': (np.int64(-17583), np.int64(365243)),\n",
       " 'DAYS_ID_PUBLISH': (np.int64(-7197), np.int64(0)),\n",
       " 'DAYS_REGISTRATION': (np.float64(-24672.0), np.float64(0.0)),\n",
       " 'SK_ID_CURR': (np.int64(100003), np.int64(456253))}"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "min_max_values = {\n",
    "    \"EXT_SOURCE_3\": (X_train[\"EXT_SOURCE_3\"].min(), X_train[\"EXT_SOURCE_3\"].max()),\n",
    "    \"EXT_SOURCE_2\": (X_train[\"EXT_SOURCE_2\"].min(), X_train[\"EXT_SOURCE_2\"].max()),\n",
    "    \"DAYS_BIRTH\": (X_train[\"DAYS_BIRTH\"].min(), X_train[\"DAYS_BIRTH\"].max()),\n",
    "    \"EXT_SOURCE_1\": (X_train[\"EXT_SOURCE_1\"].min(), X_train[\"EXT_SOURCE_1\"].max()),\n",
    "    \"AMT_ANNUITY\": (X_train[\"AMT_ANNUITY\"].min(), X_train[\"AMT_ANNUITY\"].max()),\n",
    "    \"AMT_CREDIT\": (X_train[\"AMT_CREDIT\"].min(), X_train[\"AMT_CREDIT\"].max()),\n",
    "    \"DAYS_EMPLOYED\": (X_train[\"DAYS_EMPLOYED\"].min(), X_train[\"DAYS_EMPLOYED\"].max()),\n",
    "    \"DAYS_ID_PUBLISH\": (\n",
    "        X_train[\"DAYS_ID_PUBLISH\"].min(),\n",
    "        X_train[\"DAYS_ID_PUBLISH\"].max(),\n",
    "    ),\n",
    "    \"DAYS_REGISTRATION\": (\n",
    "        X_train[\"DAYS_REGISTRATION\"].min(),\n",
    "        X_train[\"DAYS_REGISTRATION\"].max(),\n",
    "    ),\n",
    "    \"SK_ID_CURR\": (X_train[\"SK_ID_CURR\"].min(), X_train[\"SK_ID_CURR\"].max()),\n",
    "}\n",
    "min_max_values"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Home_Credit_Default_Risk_Prediction",
   "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": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}