End-to-End-Guide-to-Building-a-Credit-Scorecard-Using-Machine-Learning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: pandas in c:\\users\\bruce\\anaconda3\\lib\\site-packages (1.1.3)\n",
      "Requirement already satisfied: pytz>=2017.2 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from pandas) (2020.1)\n",
      "Requirement already satisfied: python-dateutil>=2.7.3 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from pandas) (2.8.1)\n",
      "Requirement already satisfied: numpy>=1.15.4 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from pandas) (1.23.4)\n",
      "Requirement already satisfied: six>=1.5 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from python-dateutil>=2.7.3->pandas) (1.15.0)\n",
      "Requirement already satisfied: numpy in c:\\users\\bruce\\anaconda3\\lib\\site-packages (1.23.4)\n",
      "Requirement already satisfied: matplotlib in c:\\users\\bruce\\anaconda3\\lib\\site-packages (3.3.2)\n",
      "Requirement already satisfied: certifi>=2020.06.20 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib) (2020.6.20)\n",
      "Requirement already satisfied: cycler>=0.10 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib) (0.10.0)\n",
      "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.3 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib) (2.4.7)\n",
      "Requirement already satisfied: python-dateutil>=2.1 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib) (2.8.1)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib) (1.3.0)\n",
      "Requirement already satisfied: numpy>=1.15 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib) (1.23.4)\n",
      "Requirement already satisfied: pillow>=6.2.0 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib) (7.2.0)\n",
      "Requirement already satisfied: six in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from cycler>=0.10->matplotlib) (1.15.0)\n",
      "Requirement already satisfied: seaborn in c:\\users\\bruce\\anaconda3\\lib\\site-packages (0.11.0)\n",
      "Requirement already satisfied: matplotlib>=2.2 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from seaborn) (3.3.2)\n",
      "Requirement already satisfied: numpy>=1.15 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from seaborn) (1.23.4)\n",
      "Requirement already satisfied: scipy>=1.0 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from seaborn) (1.5.2)\n",
      "Requirement already satisfied: pandas>=0.23 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from seaborn) (1.1.3)\n",
      "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.3 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn) (2.4.7)\n",
      "Requirement already satisfied: python-dateutil>=2.1 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn) (2.8.1)\n",
      "Requirement already satisfied: certifi>=2020.06.20 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn) (2020.6.20)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn) (1.3.0)\n",
      "Requirement already satisfied: cycler>=0.10 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn) (0.10.0)\n",
      "Requirement already satisfied: pillow>=6.2.0 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn) (7.2.0)\n",
      "Requirement already satisfied: pytz>=2017.2 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from pandas>=0.23->seaborn) (2020.1)\n",
      "Requirement already satisfied: six>=1.5 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from python-dateutil>=2.1->matplotlib>=2.2->seaborn) (1.15.0)\n",
      "Requirement already satisfied: toad in c:\\users\\bruce\\appdata\\roaming\\python\\python38\\site-packages (0.1.1)\n",
      "Requirement already satisfied: pandas in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from toad) (1.1.3)\n",
      "Requirement already satisfied: scikit-learn>=0.21 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from toad) (0.23.2)\n",
      "Requirement already satisfied: joblib>=0.12 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from toad) (0.17.0)\n",
      "Requirement already satisfied: Cython>=0.29.15 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from toad) (0.29.21)\n",
      "Requirement already satisfied: seaborn>=0.10.0 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from toad) (0.11.0)\n",
      "Requirement already satisfied: scipy in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from toad) (1.5.2)\n",
      "Requirement already satisfied: numpy>=1.20 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from toad) (1.23.4)\n",
      "Requirement already satisfied: setuptools in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from toad) (50.3.1.post20201107)\n",
      "Requirement already satisfied: pytz>=2017.2 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from pandas->toad) (2020.1)\n",
      "Requirement already satisfied: python-dateutil>=2.7.3 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from pandas->toad) (2.8.1)\n",
      "Requirement already satisfied: threadpoolctl>=2.0.0 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from scikit-learn>=0.21->toad) (2.1.0)\n",
      "Requirement already satisfied: matplotlib>=2.2 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from seaborn>=0.10.0->toad) (3.3.2)\n",
      "Requirement already satisfied: six>=1.5 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from python-dateutil>=2.7.3->pandas->toad) (1.15.0)\n",
      "Requirement already satisfied: pillow>=6.2.0 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn>=0.10.0->toad) (7.2.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn>=0.10.0->toad) (1.3.0)\n",
      "Requirement already satisfied: certifi>=2020.06.20 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn>=0.10.0->toad) (2020.6.20)\n",
      "Requirement already satisfied: cycler>=0.10 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn>=0.10.0->toad) (0.10.0)\n",
      "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.3 in c:\\users\\bruce\\anaconda3\\lib\\site-packages (from matplotlib>=2.2->seaborn>=0.10.0->toad) (2.4.7)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "ERROR: Could not find a version that satisfies the requirement pickle (from versions: none)\n",
      "ERROR: No matching distribution found for pickle\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting sklearn\n",
      "  Downloading sklearn-0.0.post1.tar.gz (3.6 kB)\n",
      "Building wheels for collected packages: sklearn\n",
      "  Building wheel for sklearn (setup.py): started\n",
      "  Building wheel for sklearn (setup.py): finished with status 'done'\n",
      "  Created wheel for sklearn: filename=sklearn-0.0.post1-py3-none-any.whl size=2347 sha256=24d170a9b6f75cdca6746fc2ea61eeaa430a2a115547e9fbe66414bf09384fde\n",
      "  Stored in directory: c:\\users\\bruce\\appdata\\local\\pip\\cache\\wheels\\14\\25\\f7\\1cc0956978ae479e75140219088deb7a36f60459df242b1a72\n",
      "Successfully built sklearn\n",
      "Installing collected packages: sklearn\n",
      "Successfully installed sklearn-0.0.post1\n"
     ]
    }
   ],
   "source": [
    "import pkg_resources\n",
    "import pip\n",
    "installedPackages = {pkg.key for pkg in pkg_resources.working_set}\n",
    "required = { 'pandas','numpy', 'matplotlib', 'seaborn','toad','pickle','sklearn'}\n",
    "missing = required - installedPackages\n",
    "if missing:\n",
    "    !pip install pandas\n",
    "    !pip install numpy\n",
    "    !pip install matplotlib\n",
    "    !pip install seaborn\n",
    "    !pip install toad\n",
    "    !pip install pickle\n",
    "    !pip install sklearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.metrics import roc_auc_score,roc_curve,auc,precision_recall_curve\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression \n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "\n",
    "import numpy as np\n",
    "import glob\n",
    "import math\n",
    "import seaborn as sns \n",
    "import matplotlib.pyplot as plt\n",
    "import toad\n",
    "import pickle\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Data Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def data_split(df, start, end, date_col):\n",
    "    \"\"\"\n",
    "    split the dataset into training or testing using date\n",
    "    :param data: (df) pandas dataframe, start, end\n",
    "    :return: (df) pandas dataframe\n",
    "    \"\"\"\n",
    "    data = df[(df[date_col] >= start) & (df[date_col] < end)]\n",
    "    #data = df[(df.order_date >= start) & (df.order_date < end)]\n",
    "    data = data.reset_index(drop=True)\n",
    "    #data = data.sort_values([\"date\", \"tic\"], ignore_index=True)\n",
    "    #data.index = data.date.factorize()[0]\n",
    "    return data\n",
    "\n",
    "\n",
    "def target_info(df_target_column):\n",
    "    '''\n",
    "    Print target column information -- count for each unique values in target column\n",
    "    Input: target column in dataframe\n",
    "    '''\n",
    "    proportion_true=str(round(sum(df_target_column == True) / len(df_target_column), 2) * 100)\n",
    "    \n",
    "    proportion_false=str(round(sum(df_target_column == False) / len(df_target_column), 3) * 100)\n",
    "    \n",
    "    print('There are total {} records in our data.'.format(len(df_target_column)))\n",
    "    print('Is Fraud:')\n",
    "    print('Count: {}'.format(sum(df_target_column == True)))\n",
    "    print('Proportion (Fraud): {}'.format(proportion_true\n",
    "                                  + '%'))\n",
    "    print('Not Fraud:')\n",
    "    print('Count: {}'.format(sum(df_target_column == False)))\n",
    "    print('Proportion (Not Fraud): {}'.format(proportion_false\n",
    "                                  + '%'))\n",
    "    plt.figure(figsize=(5, 5))\n",
    "    sns.barplot(x=df_target_column.value_counts().index, y=df_target_column.value_counts())\n",
    "    plt.title('Not Fraud vs Fraud Counts')\n",
    "    plt.ylabel('Count')\n",
    "    return proportion_true\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# use pandas to load the csv file\n",
    "data = pd.read_csv('UCI_Credit_Card.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(30000, 25)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check the size of the data\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>LIMIT_BAL</th>\n",
       "      <th>SEX</th>\n",
       "      <th>EDUCATION</th>\n",
       "      <th>MARRIAGE</th>\n",
       "      <th>AGE</th>\n",
       "      <th>PAY_0</th>\n",
       "      <th>PAY_2</th>\n",
       "      <th>PAY_3</th>\n",
       "      <th>PAY_4</th>\n",
       "      <th>...</th>\n",
       "      <th>BILL_AMT4</th>\n",
       "      <th>BILL_AMT5</th>\n",
       "      <th>BILL_AMT6</th>\n",
       "      <th>PAY_AMT1</th>\n",
       "      <th>PAY_AMT2</th>\n",
       "      <th>PAY_AMT3</th>\n",
       "      <th>PAY_AMT4</th>\n",
       "      <th>PAY_AMT5</th>\n",
       "      <th>PAY_AMT6</th>\n",
       "      <th>default.payment.next.month</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>20000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>24</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>689.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>120000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>26</td>\n",
       "      <td>-1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>3272.0</td>\n",
       "      <td>3455.0</td>\n",
       "      <td>3261.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>90000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>34</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>14331.0</td>\n",
       "      <td>14948.0</td>\n",
       "      <td>15549.0</td>\n",
       "      <td>1518.0</td>\n",
       "      <td>1500.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>5000.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>50000.0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>37</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>28314.0</td>\n",
       "      <td>28959.0</td>\n",
       "      <td>29547.0</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>2019.0</td>\n",
       "      <td>1200.0</td>\n",
       "      <td>1100.0</td>\n",
       "      <td>1069.0</td>\n",
       "      <td>1000.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>50000.0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>57</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>20940.0</td>\n",
       "      <td>19146.0</td>\n",
       "      <td>19131.0</td>\n",
       "      <td>2000.0</td>\n",
       "      <td>36681.0</td>\n",
       "      <td>10000.0</td>\n",
       "      <td>9000.0</td>\n",
       "      <td>689.0</td>\n",
       "      <td>679.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ID  LIMIT_BAL  SEX  EDUCATION  MARRIAGE  AGE  PAY_0  PAY_2  PAY_3  PAY_4  \\\n",
       "0   1    20000.0    2          2         1   24      2      2     -1     -1   \n",
       "1   2   120000.0    2          2         2   26     -1      2      0      0   \n",
       "2   3    90000.0    2          2         2   34      0      0      0      0   \n",
       "3   4    50000.0    2          2         1   37      0      0      0      0   \n",
       "4   5    50000.0    1          2         1   57     -1      0     -1      0   \n",
       "\n",
       "   ...  BILL_AMT4  BILL_AMT5  BILL_AMT6  PAY_AMT1  PAY_AMT2  PAY_AMT3  \\\n",
       "0  ...        0.0        0.0        0.0       0.0     689.0       0.0   \n",
       "1  ...     3272.0     3455.0     3261.0       0.0    1000.0    1000.0   \n",
       "2  ...    14331.0    14948.0    15549.0    1518.0    1500.0    1000.0   \n",
       "3  ...    28314.0    28959.0    29547.0    2000.0    2019.0    1200.0   \n",
       "4  ...    20940.0    19146.0    19131.0    2000.0   36681.0   10000.0   \n",
       "\n",
       "   PAY_AMT4  PAY_AMT5  PAY_AMT6  default.payment.next.month  \n",
       "0       0.0       0.0       0.0                           1  \n",
       "1    1000.0       0.0    2000.0                           1  \n",
       "2    1000.0    1000.0    5000.0                           0  \n",
       "3    1100.0    1069.0    1000.0                           0  \n",
       "4    9000.0     689.0     679.0                           0  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check few lines\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#use the world 'label'\n",
    "data['label']=data['default.payment.next.month']\n",
    "data=data.drop(columns=['default.payment.next.month'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are total 30000 records in our data.\n",
      "Is Fraud:\n",
      "Count: 6636\n",
      "Proportion (Fraud): 22.0%\n",
      "Not Fraud:\n",
      "Count: 23364\n",
      "Proportion (Not Fraud): 77.9%\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'22.0'"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVoAAAE/CAYAAAAHaAufAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAVZklEQVR4nO3de7TdZX3n8fdHoojlDgExwQYl7TLaqTOklEGng2WmpK0t6qAGW4lTamYsOrZaOjBeUNuMtXVEaSszjCIXFUipDqigw4DCdBYFg4uRmwxZIhATIdwvCjXwnT/2c3Tn5ORk5/LkJOe8X2vttX/7+/s9v/08Ozuf8/s9+5aqQpLUz7OmugOSNN0ZtJLUmUErSZ0ZtJLUmUErSZ0ZtJLUmUGrnVqS7yX5V1Pdjx6SVJJDp7of2noG7QzQwujeJD8zVPv9JN8Ysf03kvz+JOvntVB4fOjyf7dB16dckqOSPDNubF+a6n4BJDkmyTVJHkuyNsnVSX57O9zvtP3j1otBO3PMAt7Z+T72rqrd2+UXx69MMqvz/feyemhcu1fVb43fYHuPLclxwN8C5wFzgQOB9wMb9E1Tz6CdOf4S+OMke0+0MsmRSb6Z5JF2fWSrLwP+BfDX7Wjur0e9w3Y0uCrJf0zyA+AzSfZJ8uV2BPZQW5471Ga9o6UkH0jy2aHbb05yV5IHkrxnkvs+IskPkuwyVHttkm+35cOTrEjyaDva/9io42rt35Lk/yQ5PcmDwAeSvDjJVa1v9yf53PDjPX4qIMk5Sf5s6PbJSdYkWZ3k9ya57wAfA/60qj5VVY9U1TNVdXVVvbVt86wk722P1X1JzkuyV1t3VJJV4/b5k8e9PebLW5vHktySZGFbdz7wQuBL7fnwJ0mem+SzbdwPt+fPgZvzeE53Bu3MsQL4BvDH41ck2Rf4CnAGsB+D/8RfSbJfVb0H+N/A29vR3Ns3836fD+wL/CywlMFz7jPt9guBHwEjhXeSBcCZwJuBF7S+zp1o26r6B+AJ4FeHym8CPt+WPwF8oqr2BF4MLN+cQTW/DHwXOABYBgT4cOvbS4CDgQ+MsqMkixj82/xrYD4w2an5z7d9XzzJNm9pl1cBLwJ2Z8THuflt4EJgb+DSsbZV9WbgbuC32vPhL4AlwF6tT/sB/57Bv6sag3ZmeT/wjiSzx9V/E7ijqs6vqnVVdQHwHTb/NPT+dkTzcJKxQH8GOK2qnqqqH1XVA1X1d1X1w6p6jEFA/csR938c8OWquqaqngLe1/a/MRcAxwMk2QP4jVYD+DFwaJL9q+rxFswb84KhcT2c5A2tvrqq/qo9Zj+qqpVVdUUb61oGf7BGHdsbgM9U1c1V9QSTB/R+7XrNJNv8DvCxqvpuVT0OnAos3owpjr+vqsuq6mngfGCDqaAhP259OrSqnq6qG6rq0RHvZ0YwaGeQqroZ+DJwyrhVLwDuGle7C5izmXexf1Xt3S4fbbW1VfXk2AZJnpfkv7VT2keBa4C9h0/xJ/EC4J6h8TwBPDDJ9p8HXpdkV+B1wLeqamycJwI/B3ynneq+epL9rB4a195VNXb0e8/wRkkOSHJhku+3sX0W2H+EcW0wNjb89xg2NuaDNrG/4X3cxWCeftRT+h8MLf8QeO4kIX0+8DXgwjbt8RdJnj3i/cwIBu3McxrwVtYP0dUMTuWHvRD4flvemq94G9/23QxOfX+5nbb/SqunXT8BPG9o++cPLa9hcHo6aJA8j58e3W14x1W3MgiYX2f9aQOq6o6qOp7Baf9HgIsz9K6MEY0f24db7Z+0sf0uPx0XDAJrpLExePw35nYGofxvJtlm/L/pC4F1wL2Me4zbH7nxZzmTWW/cVfXjqvpgVS0AjgReDZywGfub9gzaGaaqVgIXAf9hqHwZ8HNJ3pRkVpI3AgsYHP3C4D/ni7ZRF/ZgMH/3cJsbPm3c+hsZnOI+u70Ac9zQuouBVyd5ZZLnAB9i08/hzzMY668weJUegCS/m2R2VT0DPNzKT2/ZkH5iD+BxBmObA5w8bv2NwJuS7NLmZIenFZYDb0myoP0BGf+4/EQNvtv0XcD7kvzbJHu2F79emeSsttkFwB8lOSTJ7sB/Bi6qqnXA/2NwhPqb7cjzvcCumzHO9Z4PSV6V5BdaYD/KYCphax/LacWgnZk+BPzk6K2qHmBwFPJuBqelfwK8uqrub5t8Ajgug3cJnLGV9/1xYDfgfuAfgK+OW/8+Bi9OPQR8kPWPQm8BTmq1NW2bVUzuAuAo4Kqh8QAsAm5J8jiD8S0enuLYQh8E/hnwCIMXF78wbv07Gcx7P8xgDvV/jK2oqssZPDZXASvb9UZV1cXAG4HfY3D0ei/wZ8AlbZOzGZzSXwPcCTwJvKO1fQT4A+BTDM5anmDTj+OwDwPvHZqLfz6DP4KPArcBVzOYNlETv/hbkvryiFaSOjNoJakzg1aSOjNoJakzg1aSOttZv01pi+2///41b968qe6GpGnmhhtuuL+qJvzgx4wL2nnz5rFixYqp7oakaSbJRj827dSBJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHU2477rYEsddvJ5U90FbYUb/tIfZdXU8YhWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWps25Bm+TgJF9PcluSW5K8s9X3TXJFkjva9T5DbU5NsjLJ7UmOGaofluSmtu6MJGn1XZNc1OrXJZnXazyStKV6HtGuA95dVS8BjgBOSrIAOAW4sqrmA1e227R1i4GXAouATybZpe3rTGApML9dFrX6icBDVXUocDrwkY7jkaQt0i1oq2pNVX2rLT8G3AbMAY4Fzm2bnQu8pi0fC1xYVU9V1Z3ASuDwJAcBe1bVtVVVwHnj2ozt62Lg6LGjXUnaUWyXOdp2Sv9PgeuAA6tqDQzCGDigbTYHuGeo2apWm9OWx9fXa1NV64BHgP26DEKStlD3oE2yO/B3wB9W1aOTbTpBrSapT9ZmfB+WJlmRZMXatWs31WVJ2qa6Bm2SZzMI2c9V1Rda+d42HUC7vq/VVwEHDzWfC6xu9bkT1Ndrk2QWsBfw4Ph+VNVZVbWwqhbOnj17WwxNkkbW810HAT4N3FZVHxtadSmwpC0vAS4Zqi9u7yQ4hMGLXte36YXHkhzR9nnCuDZj+zoOuKrN40rSDmNWx32/AngzcFOSG1vtPwF/DixPciJwN/B6gKq6Jcly4FYG71g4qaqebu3eBpwD7AZc3i4wCPLzk6xkcCS7uON4JGmLdAvaqvp7Jp5DBTh6I22WAcsmqK8AXjZB/UlaUEvSjspPhklSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZwatJHVm0EpSZ92CNsnZSe5LcvNQ7QNJvp/kxnb5jaF1pyZZmeT2JMcM1Q9LclNbd0aStPquSS5q9euSzOs1FknaGj2PaM8BFk1QP72qXt4ulwEkWQAsBl7a2nwyyS5t+zOBpcD8dhnb54nAQ1V1KHA68JFeA5GkrdEtaKvqGuDBETc/Friwqp6qqjuBlcDhSQ4C9qyqa6uqgPOA1wy1ObctXwwcPXa0K0k7kqmYo317km+3qYV9Wm0OcM/QNqtabU5bHl9fr01VrQMeAfbr2XFJ2hLbO2jPBF4MvBxYA/yXVp/oSLQmqU/WZgNJliZZkWTF2rVrN6vDkrS1tmvQVtW9VfV0VT0D/Hfg8LZqFXDw0KZzgdWtPneC+nptkswC9mIjUxVVdVZVLayqhbNnz95Ww5GkkWzXoG1zrmNeC4y9I+FSYHF7J8EhDF70ur6q1gCPJTmizb+eAFwy1GZJWz4OuKrN40rSDmVWrx0nuQA4Ctg/ySrgNOCoJC9ncIr/PeDfAVTVLUmWA7cC64CTqurptqu3MXgHw27A5e0C8Gng/CQrGRzJLu41FknaGt2CtqqOn6D86Um2XwYsm6C+AnjZBPUngddvTR8laXvwk2GS1JlBK0mdGbSS1JlBK0mdGbSS1JlBK0mdGbSS1JlBK0mdGbSS1JlBK0mdGbSS1JlBK0mdjRS0SV4xSk2StKFRj2j/asSaJGmcSb8mMck/B44EZid519CqPYFdJm4lSRq2qe+jfQ6we9tuj6H6owx+1UCStAmTBm1VXQ1cneScqrprO/VJkqaVUX9hYdckZwHzhttU1a/26JQkTSejBu3fAv8V+BTw9Ca2lSQNGTVo11XVmV17IknT1Khv7/pSkj9IclCSfccuXXsmSdPEqEe0S9r1yUO1Al60bbsjSdPPSEFbVYf07ogkTVcjBW2SEyaqV9V527Y7kjT9jDp18EtDy88Fjga+BRi0krQJo04dvGP4dpK9gPO79EiSppkt/ZrEHwLzt2VHJGm6GnWO9ksM3mUAgy+TeQmwvFenJGk6GXWO9qNDy+uAu6pqVYf+SNK0M9LUQftyme8w+AavfYB/7NkpSZpORv2FhTcA1wOvB94AXJfEr0mUpBGMOnXwHuCXquo+gCSzgf8FXNyrY5I0XYz6roNnjYVs88BmtJWkGW3UI9qvJvkacEG7/Ubgsj5dkqTpZVO/GXYocGBVnZzkdcArgQDXAp/bDv2TpJ3epk7/Pw48BlBVX6iqd1XVHzE4mv14365J0vSwqaCdV1XfHl+sqhUMftZGkrQJmwra506ybrdt2RFJmq42FbTfTPLW8cUkJwI39OmSJE0vm3rXwR8CX0zyO/w0WBcCzwFe27FfkjRtTBq0VXUvcGSSVwEva+WvVNVV3XsmSdPEqN9H+3Xg6537IknTkp/ukqTODFpJ6syglaTODFpJ6qxb0CY5O8l9SW4equ2b5Iokd7TrfYbWnZpkZZLbkxwzVD8syU1t3RlJ0uq7Jrmo1a9LMq/XWCRpa/Q8oj0HWDSudgpwZVXNB65st0myAFgMvLS1+WSSXVqbM4GlDH4Mcv7QPk8EHqqqQ4HTgY90G4kkbYVuQVtV1wAPjisfC5zbls8FXjNUv7CqnqqqO4GVwOFJDgL2rKprq6qA88a1GdvXxcDRY0e7krQj2d5ztAdW1RqAdn1Aq88B7hnablWrzWnL4+vrtamqdcAjwH7dei5JW2hHeTFsoiPRmqQ+WZsNd54sTbIiyYq1a9duYRclacts76C9t00H0K7Hfh5nFXDw0HZzgdWtPneC+nptkswC9mLDqQoAquqsqlpYVQtnz569jYYiSaPZ3kF7KbCkLS8BLhmqL27vJDiEwYte17fphceSHNHmX08Y12ZsX8cBV7V5XEnaoYz6m2GbLckFwFHA/klWAacBfw4sb1+zeDeDny+nqm5Jshy4FVgHnFRVT7ddvY3BOxh2Ay5vF4BPA+cnWcngSHZxr7FI0tboFrRVdfxGVh29ke2XAcsmqK/gp98cNlx/khbUkrQj21FeDJOkacuglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTODFpJ6syglaTOZk11B6Tp6O4P/cJUd0Fb6IXvv2mb79MjWknqzKCVpM4MWknqzKCVpM4MWknqzKCVpM4MWknqzKCVpM4MWknqzKCVpM4MWknqzKCVpM4MWknqzKCVpM6mJGiTfC/JTUluTLKi1fZNckWSO9r1PkPbn5pkZZLbkxwzVD+s7WdlkjOSZCrGI0mTmcoj2ldV1curamG7fQpwZVXNB65st0myAFgMvBRYBHwyyS6tzZnAUmB+uyzajv2XpJHsSFMHxwLntuVzgdcM1S+sqqeq6k5gJXB4koOAPavq2qoq4LyhNpK0w5iqoC3gfya5IcnSVjuwqtYAtOsDWn0OcM9Q21WtNqctj69vIMnSJCuSrFi7du02HIYkbdpU/ZTNK6pqdZIDgCuSfGeSbSead61J6hsWq84CzgJYuHDhhNtIUi9TckRbVavb9X3AF4HDgXvbdADt+r62+Srg4KHmc4HVrT53grok7VC2e9Am+Zkke4wtA78G3AxcCixpmy0BLmnLlwKLk+ya5BAGL3pd36YXHktyRHu3wQlDbSRphzEVUwcHAl9s78SaBXy+qr6a5JvA8iQnAncDrweoqluSLAduBdYBJ1XV021fbwPOAXYDLm8XSdqhbPegrarvAr84Qf0B4OiNtFkGLJugvgJ42bbuoyRtSzvS27skaVoyaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWpM4NWkjozaCWps50+aJMsSnJ7kpVJTpnq/kjSeDt10CbZBfgb4NeBBcDxSRZMba8kaX07ddAChwMrq+q7VfWPwIXAsVPcJ0laz84etHOAe4Zur2o1SdphzJrqDmylTFCrDTZKlgJL283Hk9zetVc7p/2B+6e6E73ko0umugvTybR+rnDaRLEykp/d2IqdPWhXAQcP3Z4LrB6/UVWdBZy1vTq1M0qyoqoWTnU/tOPzubL5dvapg28C85MckuQ5wGLg0inukyStZ6c+oq2qdUneDnwN2AU4u6pumeJuSdJ6duqgBaiqy4DLprof04BTKxqVz5XNlKoNXjuSJG1DO/scrSTt8AzaGc6PMGtUSc5Ocl+Sm6e6Lzsbg3YG8yPM2kznAIumuhM7I4N2ZvMjzBpZVV0DPDjV/dgZGbQzmx9hlrYDg3ZmG+kjzJK2jkE7s430EWZJW8egndn8CLO0HRi0M1hVrQPGPsJ8G7DcjzBrY5JcAFwL/HySVUlOnOo+7Sz8ZJgkdeYRrSR1ZtBKUmcGrSR1ZtBKUmcGrSR1ZtBKUmcGrSR1ZtBKUmf/H45pqOsh8F8PAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#check the fraud proportion of the data\n",
    "target_info(data['label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set an exclude list for the scorecard package Toad\n",
    "exclude_list = ['ID','label']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    30000.000000\n",
       "mean     15000.500000\n",
       "std       8660.398374\n",
       "min          1.000000\n",
       "25%       7500.750000\n",
       "50%      15000.500000\n",
       "75%      22500.250000\n",
       "max      30000.000000\n",
       "Name: ID, dtype: float64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.ID.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# use the ID column to split the train-test data\n",
    "train = data_split(data,start = 0, end=22500,date_col='ID')\n",
    "test = data_split(data,start = 22500, end=172792,date_col='ID')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    17411\n",
       "1     5088\n",
       "Name: label, dtype: int64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.label.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    5953\n",
       "1    1548\n",
       "Name: label, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.label.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Feature Filtering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "keep: 23 drop empty: 0 drop iv: 2 drop corr: 0\n"
     ]
    }
   ],
   "source": [
    "###feature filtering by missing value, IV & corrrelation：\n",
    "##If the missing value rate is greater than the threshold, delete the feature\n",
    "##If the correlation coefficient is greater than the threshold, delete the feature\n",
    "##If the IV is smaller than the threshold, delete the features\n",
    "\n",
    "train_selected, drop_lst= toad.selection.select(frame = train,\n",
    "                                                target=train['label'], \n",
    "                                                empty = 0.7, \n",
    "                                                iv = 0.02, corr = 1, \n",
    "                                                return_drop=True, \n",
    "                                                exclude=exclude_list)\n",
    "print(\"keep:\",train_selected.shape[1],\n",
    "      \"drop empty:\",len(drop_lst['empty']),\n",
    "      \"drop iv:\",len(drop_lst['iv']),\n",
    "      \"drop corr:\",len(drop_lst['corr']))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'empty': array([], dtype=float64),\n",
       " 'iv': array(['SEX', 'MARRIAGE'], dtype=object),\n",
       " 'corr': array([], dtype=object)}"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "drop_lst"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# output the iv table to a dataframe\n",
    "def output_iv_importance(train_selected,label_col):\n",
    "    feat_import_iv = toad.quality(train_selected,label_col,iv_only=True)\n",
    "    feat_import_iv=feat_import_iv['iv']\n",
    "    feat_import_iv = feat_import_iv.reset_index()\n",
    "    feat_import_iv.columns = ['name','iv']\n",
    "    return feat_import_iv\n",
    "df_iv=output_iv_importance(train_selected,'label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(22, 2)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_iv.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>iv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PAY_0</td>\n",
       "      <td>0.864707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PAY_2</td>\n",
       "      <td>0.536068</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PAY_3</td>\n",
       "      <td>0.401513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>PAY_4</td>\n",
       "      <td>0.349813</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>PAY_5</td>\n",
       "      <td>0.340014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>PAY_6</td>\n",
       "      <td>0.282692</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>PAY_AMT1</td>\n",
       "      <td>0.189334</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>LIMIT_BAL</td>\n",
       "      <td>0.170973</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>PAY_AMT2</td>\n",
       "      <td>0.168256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>PAY_AMT3</td>\n",
       "      <td>0.128827</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>PAY_AMT4</td>\n",
       "      <td>0.121719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>PAY_AMT6</td>\n",
       "      <td>0.106017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>PAY_AMT5</td>\n",
       "      <td>0.103353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>BILL_AMT4</td>\n",
       "      <td>0.047998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>BILL_AMT5</td>\n",
       "      <td>0.046490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>BILL_AMT3</td>\n",
       "      <td>0.044047</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>EDUCATION</td>\n",
       "      <td>0.036458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>ID</td>\n",
       "      <td>0.032184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>AGE</td>\n",
       "      <td>0.027512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>BILL_AMT6</td>\n",
       "      <td>0.027461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>BILL_AMT1</td>\n",
       "      <td>0.026191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>BILL_AMT2</td>\n",
       "      <td>0.022539</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         name        iv\n",
       "0       PAY_0  0.864707\n",
       "1       PAY_2  0.536068\n",
       "2       PAY_3  0.401513\n",
       "3       PAY_4  0.349813\n",
       "4       PAY_5  0.340014\n",
       "5       PAY_6  0.282692\n",
       "6    PAY_AMT1  0.189334\n",
       "7   LIMIT_BAL  0.170973\n",
       "8    PAY_AMT2  0.168256\n",
       "9    PAY_AMT3  0.128827\n",
       "10   PAY_AMT4  0.121719\n",
       "11   PAY_AMT6  0.106017\n",
       "12   PAY_AMT5  0.103353\n",
       "13  BILL_AMT4  0.047998\n",
       "14  BILL_AMT5  0.046490\n",
       "15  BILL_AMT3  0.044047\n",
       "16  EDUCATION  0.036458\n",
       "17         ID  0.032184\n",
       "18        AGE  0.027512\n",
       "19  BILL_AMT6  0.027461\n",
       "20  BILL_AMT1  0.026191\n",
       "21  BILL_AMT2  0.022539"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_iv.head(30)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Feature Binning\n",
    "Feature binning is to transform a continuous or numerical variable into a categorical feature.\n",
    "* It simplifies the logistic regression model and reduces the risk of model overfitting\n",
    "* Logistic regression is a generalized linear model, and its expressive ability is limited; Feature binning can introduce nonlinearity into the model, which can improve the expressive ability of the model and help better model fitting\n",
    "* The discretized features are very robust to abnormal data: for example, a feature is 1 if age > 30, and 0 otherwise. If the features are not discretized, an abnormal data point \"300 years old\" will impact the model fitting\n",
    "* It can treat null data as an individual class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    17411\n",
       "1     5088\n",
       "Name: label, dtype: int64"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_selected.label.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(22499, 23)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_selected.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Steps for feature binning:\n",
    "#### Step 1. Initialization: c = toad.transform.Combiner() \n",
    "#### Step 2. Training binning: \n",
    "c.fit(dataframe, \n",
    "      y = 'target', \n",
    "      method = 'chi', \n",
    "      min_samples = 0.05, \n",
    "      n_bins = None, \n",
    "      empty_separate = False)\n",
    "* y: target column\n",
    "* method: binning method, supports chi (chi-square binning), dt (decision tree binning), kmean, quantile, step (equal step size binning)\n",
    "* min_samples: Each box contains the least number of samples, which can be a number or a proportion \n",
    "* n_bins: the number of bins; If it is not possible to divide so many boxes the maximum number of bins will be divided.\n",
    "* empty_separate: Whether to separate empty boxes separately \n",
    "\n",
    "#### Step 3. check binning nodes: c.export() \n",
    "#### Step 4. Manually adjust binning: c.load(dict) \n",
    "#### Step 5. Apply binning results: c.transform(dataframe, labels=False)\n",
    "* labels: Whether to convert the binning results into box labels. If False, output 0, 1, 2… (discrete variables are sorted according to the proportion), and if True output (-inf, 0], (0,10], (10, inf)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.6334667126337687\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "start = time.time()\n",
    "combiner = toad.transform.Combiner()\n",
    "# use the filtered features for training\n",
    "# Use the stable chi-square binning, \n",
    "# specifying that each bin has at least 5% data to ensure stability\n",
    "# empty values will be automatically assigned to the best bin\n",
    "combiner.fit(X=train_selected,\n",
    "             y=train_selected['label'],\n",
    "             method='chi',\n",
    "             min_samples = 0.05,\n",
    "             exclude=exclude_list)\n",
    "end = time.time()\n",
    "print((end-start)/60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<toad.transform.Combiner at 0x21423d74460>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "combiner"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save 1: pickle combiner\n",
    "filename = 'CreditScore_save1_combiner.pkl'\n",
    "pickle.dump(combiner, open(filename, 'wb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'CreditScore_save1_combiner.pkl'"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filename"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load the model from disk\n",
    "#combiner = pickle.load(open(filename, 'rb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "#output binning\n",
    "bins = combiner.export()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(bins)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'LIMIT_BAL': [50000.0,\n",
       "  60000.0,\n",
       "  80000.0,\n",
       "  110000.0,\n",
       "  150000.0,\n",
       "  230000.0,\n",
       "  370000.0],\n",
       " 'EDUCATION': [2],\n",
       " 'AGE': [26, 29, 34, 46],\n",
       " 'PAY_0': [-1, 0, 1, 2],\n",
       " 'PAY_2': [-1, 2],\n",
       " 'PAY_3': [-1, 0, 2],\n",
       " 'PAY_4': [-1, 0, 1],\n",
       " 'PAY_5': [-1, 0, 2],\n",
       " 'PAY_6': [-1, 0, 2],\n",
       " 'BILL_AMT1': [2501.0],\n",
       " 'BILL_AMT2': [2400.0, 8287.0, 35213.0],\n",
       " 'BILL_AMT3': [2400.0, 7458.0, 81890.0],\n",
       " 'BILL_AMT4': [2501.0],\n",
       " 'BILL_AMT5': [2395.0, 8110.0, 35288.0],\n",
       " 'BILL_AMT6': [19195.0, 50442.0],\n",
       " 'PAY_AMT1': [6.0, 4902.0],\n",
       " 'PAY_AMT2': [92.0, 1603.0, 4520.0, 15002.0],\n",
       " 'PAY_AMT3': [1.0, 2921.0],\n",
       " 'PAY_AMT4': [1.0, 1506.0, 3946.0],\n",
       " 'PAY_AMT5': [1.0, 1880.0, 10000.0],\n",
       " 'PAY_AMT6': [28.0, 480.0, 926.0, 1078.0, 1501.0, 4060.0, 9567.0]}"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bins"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "#apply binning\n",
    "train_selected_bin = combiner.transform(train_selected)\n",
    "test_bin = combiner.transform(test[train_selected_bin.columns])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>LIMIT_BAL</th>\n",
       "      <th>EDUCATION</th>\n",
       "      <th>AGE</th>\n",
       "      <th>PAY_0</th>\n",
       "      <th>PAY_2</th>\n",
       "      <th>PAY_3</th>\n",
       "      <th>PAY_4</th>\n",
       "      <th>PAY_5</th>\n",
       "      <th>PAY_6</th>\n",
       "      <th>...</th>\n",
       "      <th>BILL_AMT4</th>\n",
       "      <th>BILL_AMT5</th>\n",
       "      <th>BILL_AMT6</th>\n",
       "      <th>PAY_AMT1</th>\n",
       "      <th>PAY_AMT2</th>\n",
       "      <th>PAY_AMT3</th>\n",
       "      <th>PAY_AMT4</th>\n",
       "      <th>PAY_AMT5</th>\n",
       "      <th>PAY_AMT6</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ID  LIMIT_BAL  EDUCATION  AGE  PAY_0  PAY_2  PAY_3  PAY_4  PAY_5  PAY_6  \\\n",
       "0   1          0          1    0      4      2      1      1      0      0   \n",
       "1   2          4          1    1      1      2      2      2      2      3   \n",
       "2   3          3          1    3      2      1      2      2      2      2   \n",
       "3   4          1          1    3      2      1      2      2      2      2   \n",
       "4   5          1          1    4      1      1      1      2      2      2   \n",
       "\n",
       "   ...  BILL_AMT4  BILL_AMT5  BILL_AMT6  PAY_AMT1  PAY_AMT2  PAY_AMT3  \\\n",
       "0  ...          0          0          0         0         1         0   \n",
       "1  ...          1          1          0         0         1         1   \n",
       "2  ...          1          2          0         1         1         1   \n",
       "3  ...          1          2          1         1         2         1   \n",
       "4  ...          1          2          0         1         4         2   \n",
       "\n",
       "   PAY_AMT4  PAY_AMT5  PAY_AMT6  label  \n",
       "0         0         0         0      1  \n",
       "1         1         0         5      1  \n",
       "2         1         1         6      0  \n",
       "3         1         1         3      0  \n",
       "4         3         1         2      0  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_selected_bin.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>LIMIT_BAL</th>\n",
       "      <th>EDUCATION</th>\n",
       "      <th>AGE</th>\n",
       "      <th>PAY_0</th>\n",
       "      <th>PAY_2</th>\n",
       "      <th>PAY_3</th>\n",
       "      <th>PAY_4</th>\n",
       "      <th>PAY_5</th>\n",
       "      <th>PAY_6</th>\n",
       "      <th>...</th>\n",
       "      <th>BILL_AMT4</th>\n",
       "      <th>BILL_AMT5</th>\n",
       "      <th>BILL_AMT6</th>\n",
       "      <th>PAY_AMT1</th>\n",
       "      <th>PAY_AMT2</th>\n",
       "      <th>PAY_AMT3</th>\n",
       "      <th>PAY_AMT4</th>\n",
       "      <th>PAY_AMT5</th>\n",
       "      <th>PAY_AMT6</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>22500</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>22501</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>22502</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22503</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>22504</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      ID  LIMIT_BAL  EDUCATION  AGE  PAY_0  PAY_2  PAY_3  PAY_4  PAY_5  PAY_6  \\\n",
       "0  22500          7          1    3      2      1      2      1      2      2   \n",
       "1  22501          5          1    2      1      1      1      1      1      1   \n",
       "2  22502          6          0    2      2      1      2      2      2      2   \n",
       "3  22503          1          1    3      2      1      3      3      3      2   \n",
       "4  22504          6          1    2      2      1      2      2      2      2   \n",
       "\n",
       "   ...  BILL_AMT4  BILL_AMT5  BILL_AMT6  PAY_AMT1  PAY_AMT2  PAY_AMT3  \\\n",
       "0  ...          1          3          2         2         2         2   \n",
       "1  ...          1          0          0         0         1         2   \n",
       "2  ...          1          3          1         1         2         1   \n",
       "3  ...          1          2          1         1         2         0   \n",
       "4  ...          1          3          2         2         3         2   \n",
       "\n",
       "   PAY_AMT4  PAY_AMT5  PAY_AMT6  label  \n",
       "0         3         2         6      0  \n",
       "1         2         2         5      0  \n",
       "2         2         2         5      0  \n",
       "3         0         1         2      1  \n",
       "4         3         2         6      1  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_bin.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_list = [feat for feat in train_selected_bin.columns if feat not in exclude_list]\n",
    "len(features_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='PAY_AMT1', ylabel='prop'>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Fine tune bins\n",
    "from toad.plot import  bin_plot,badrate_plot\n",
    "bin_plot(train_selected_bin,x='PAY_AMT1',target='label')\n",
    "bin_plot(test_bin,x='PAY_AMT1',target='label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "#setting rules\n",
    "#rule = {'PAY_AMT1':[['0', 'nan'],['1'], ['2'], ['3']]}\n",
    "\n",
    "#Adjust binning\n",
    "#c.set_rules(rule)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    22499.000000\n",
       "mean         1.988711\n",
       "std          1.038531\n",
       "min          0.000000\n",
       "25%          1.000000\n",
       "50%          2.000000\n",
       "75%          2.000000\n",
       "max          4.000000\n",
       "Name: PAY_0, dtype: float64"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_selected_bin.PAY_0.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Transform to WOE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "##转化成WOE映射\n",
    "t=toad.transform.WOETransformer()\n",
    "#transform training set\n",
    "train_woe = t.fit_transform(X=train_selected_bin,\n",
    "                            y=train_selected_bin['label'], \n",
    "                            exclude=exclude_list)\n",
    "#transform testing set\n",
    "test_woe = t.transform(test_bin)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>LIMIT_BAL</th>\n",
       "      <th>EDUCATION</th>\n",
       "      <th>AGE</th>\n",
       "      <th>PAY_0</th>\n",
       "      <th>PAY_2</th>\n",
       "      <th>PAY_3</th>\n",
       "      <th>PAY_4</th>\n",
       "      <th>PAY_5</th>\n",
       "      <th>PAY_6</th>\n",
       "      <th>...</th>\n",
       "      <th>BILL_AMT4</th>\n",
       "      <th>BILL_AMT5</th>\n",
       "      <th>BILL_AMT6</th>\n",
       "      <th>PAY_AMT1</th>\n",
       "      <th>PAY_AMT2</th>\n",
       "      <th>PAY_AMT3</th>\n",
       "      <th>PAY_AMT4</th>\n",
       "      <th>PAY_AMT5</th>\n",
       "      <th>PAY_AMT6</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.666522</td>\n",
       "      <td>0.089338</td>\n",
       "      <td>0.260065</td>\n",
       "      <td>2.019269</td>\n",
       "      <td>1.459809</td>\n",
       "      <td>-0.417749</td>\n",
       "      <td>-0.383479</td>\n",
       "      <td>-0.146325</td>\n",
       "      <td>-0.117947</td>\n",
       "      <td>...</td>\n",
       "      <td>0.028510</td>\n",
       "      <td>-0.006521</td>\n",
       "      <td>-0.034304</td>\n",
       "      <td>0.662394</td>\n",
       "      <td>0.124160</td>\n",
       "      <td>0.488183</td>\n",
       "      <td>0.454327</td>\n",
       "      <td>0.370272</td>\n",
       "      <td>0.343887</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>-0.034630</td>\n",
       "      <td>0.089338</td>\n",
       "      <td>-0.100163</td>\n",
       "      <td>-0.308169</td>\n",
       "      <td>1.459809</td>\n",
       "      <td>-0.296556</td>\n",
       "      <td>-0.230510</td>\n",
       "      <td>-0.203587</td>\n",
       "      <td>1.282840</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.009858</td>\n",
       "      <td>-0.239312</td>\n",
       "      <td>-0.034304</td>\n",
       "      <td>0.662394</td>\n",
       "      <td>0.124160</td>\n",
       "      <td>0.016060</td>\n",
       "      <td>0.056501</td>\n",
       "      <td>0.370272</td>\n",
       "      <td>-0.082401</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.108342</td>\n",
       "      <td>0.089338</td>\n",
       "      <td>-0.028816</td>\n",
       "      <td>-0.677822</td>\n",
       "      <td>-0.408118</td>\n",
       "      <td>-0.296556</td>\n",
       "      <td>-0.230510</td>\n",
       "      <td>-0.203587</td>\n",
       "      <td>-0.201216</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.009858</td>\n",
       "      <td>0.140562</td>\n",
       "      <td>-0.034304</td>\n",
       "      <td>-0.009085</td>\n",
       "      <td>0.124160</td>\n",
       "      <td>0.016060</td>\n",
       "      <td>0.056501</td>\n",
       "      <td>0.084926</td>\n",
       "      <td>-0.391360</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.200392</td>\n",
       "      <td>0.089338</td>\n",
       "      <td>-0.028816</td>\n",
       "      <td>-0.677822</td>\n",
       "      <td>-0.408118</td>\n",
       "      <td>-0.296556</td>\n",
       "      <td>-0.230510</td>\n",
       "      <td>-0.203587</td>\n",
       "      <td>-0.201216</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.009858</td>\n",
       "      <td>0.140562</td>\n",
       "      <td>0.162615</td>\n",
       "      <td>-0.009085</td>\n",
       "      <td>-0.076846</td>\n",
       "      <td>0.016060</td>\n",
       "      <td>0.056501</td>\n",
       "      <td>0.084926</td>\n",
       "      <td>-0.119586</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.200392</td>\n",
       "      <td>0.089338</td>\n",
       "      <td>0.156679</td>\n",
       "      <td>-0.308169</td>\n",
       "      <td>-0.408118</td>\n",
       "      <td>-0.417749</td>\n",
       "      <td>-0.230510</td>\n",
       "      <td>-0.203587</td>\n",
       "      <td>-0.201216</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.009858</td>\n",
       "      <td>0.140562</td>\n",
       "      <td>-0.034304</td>\n",
       "      <td>-0.009085</td>\n",
       "      <td>-1.078550</td>\n",
       "      <td>-0.372758</td>\n",
       "      <td>-0.457199</td>\n",
       "      <td>0.084926</td>\n",
       "      <td>0.257636</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ID  LIMIT_BAL  EDUCATION       AGE     PAY_0     PAY_2     PAY_3     PAY_4  \\\n",
       "0   1   0.666522   0.089338  0.260065  2.019269  1.459809 -0.417749 -0.383479   \n",
       "1   2  -0.034630   0.089338 -0.100163 -0.308169  1.459809 -0.296556 -0.230510   \n",
       "2   3   0.108342   0.089338 -0.028816 -0.677822 -0.408118 -0.296556 -0.230510   \n",
       "3   4   0.200392   0.089338 -0.028816 -0.677822 -0.408118 -0.296556 -0.230510   \n",
       "4   5   0.200392   0.089338  0.156679 -0.308169 -0.408118 -0.417749 -0.230510   \n",
       "\n",
       "      PAY_5     PAY_6  ...  BILL_AMT4  BILL_AMT5  BILL_AMT6  PAY_AMT1  \\\n",
       "0 -0.146325 -0.117947  ...   0.028510  -0.006521  -0.034304  0.662394   \n",
       "1 -0.203587  1.282840  ...  -0.009858  -0.239312  -0.034304  0.662394   \n",
       "2 -0.203587 -0.201216  ...  -0.009858   0.140562  -0.034304 -0.009085   \n",
       "3 -0.203587 -0.201216  ...  -0.009858   0.140562   0.162615 -0.009085   \n",
       "4 -0.203587 -0.201216  ...  -0.009858   0.140562  -0.034304 -0.009085   \n",
       "\n",
       "   PAY_AMT2  PAY_AMT3  PAY_AMT4  PAY_AMT5  PAY_AMT6  label  \n",
       "0  0.124160  0.488183  0.454327  0.370272  0.343887      1  \n",
       "1  0.124160  0.016060  0.056501  0.370272 -0.082401      1  \n",
       "2  0.124160  0.016060  0.056501  0.084926 -0.391360      0  \n",
       "3 -0.076846  0.016060  0.056501  0.084926 -0.119586      0  \n",
       "4 -1.078550 -0.372758 -0.457199  0.084926  0.257636      0  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_woe.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_data_woe = pd.concat([train_woe,test_woe])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "#save 2: pickle woe_transform\n",
    "filename = 'CreditScore_save2_woe_transform.pkl'\n",
    "pickle.dump(t, open(filename, 'wb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "features_use = [feat for feat in final_data_woe.columns if feat not in exclude_list]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(features_use)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate PSI\n",
    "* PSI (Population Stability Index) reflects the stability of the distribution. We often use it to screen features and evaluate model stability. The industry level is to drop features with a PSI greater than 0.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "#get the feature name\n",
    "features_list = [feat for feat in train_woe.columns if feat not in exclude_list]\n",
    "#calculate PSI using toad\n",
    "psi_df = toad.metrics.PSI(train_woe[features_list], test_woe[features_list]).sort_values(0)\n",
    "#put into a dataframe\n",
    "psi_df = psi_df.reset_index()\n",
    "psi_df = psi_df.rename(columns = {'index' : 'feature',0:'psi'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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>psi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>EDUCATION</td>\n",
       "      <td>0.000258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BILL_AMT4</td>\n",
       "      <td>0.000330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BILL_AMT6</td>\n",
       "      <td>0.001591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BILL_AMT5</td>\n",
       "      <td>0.002520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>BILL_AMT1</td>\n",
       "      <td>0.002634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>PAY_AMT5</td>\n",
       "      <td>0.003066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>PAY_AMT1</td>\n",
       "      <td>0.003402</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>PAY_AMT4</td>\n",
       "      <td>0.004556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>BILL_AMT2</td>\n",
       "      <td>0.005398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>PAY_AMT6</td>\n",
       "      <td>0.007230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>BILL_AMT3</td>\n",
       "      <td>0.007589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>PAY_AMT2</td>\n",
       "      <td>0.010605</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>PAY_4</td>\n",
       "      <td>0.011666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>PAY_3</td>\n",
       "      <td>0.011933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>PAY_2</td>\n",
       "      <td>0.012077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>PAY_5</td>\n",
       "      <td>0.014513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>PAY_AMT3</td>\n",
       "      <td>0.016165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>LIMIT_BAL</td>\n",
       "      <td>0.018541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>PAY_6</td>\n",
       "      <td>0.019574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>PAY_0</td>\n",
       "      <td>0.022814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>AGE</td>\n",
       "      <td>0.035298</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      feature       psi\n",
       "0   EDUCATION  0.000258\n",
       "1   BILL_AMT4  0.000330\n",
       "2   BILL_AMT6  0.001591\n",
       "3   BILL_AMT5  0.002520\n",
       "4   BILL_AMT1  0.002634\n",
       "5    PAY_AMT5  0.003066\n",
       "6    PAY_AMT1  0.003402\n",
       "7    PAY_AMT4  0.004556\n",
       "8   BILL_AMT2  0.005398\n",
       "9    PAY_AMT6  0.007230\n",
       "10  BILL_AMT3  0.007589\n",
       "11   PAY_AMT2  0.010605\n",
       "12      PAY_4  0.011666\n",
       "13      PAY_3  0.011933\n",
       "14      PAY_2  0.012077\n",
       "15      PAY_5  0.014513\n",
       "16   PAY_AMT3  0.016165\n",
       "17  LIMIT_BAL  0.018541\n",
       "18      PAY_6  0.019574\n",
       "19      PAY_0  0.022814\n",
       "20        AGE  0.035298"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "psi_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# features less than 0.25\n",
    "psi025 = list(psi_df[psi_df.psi<0.25].feature)\n",
    "# features geater than 0.25\n",
    "psi_remove = list(psi_df[psi_df.psi>=0.25].feature)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "psi_remove"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# keep exclude list\n",
    "for i in exclude_list:\n",
    "    if i in psi025:\n",
    "        pass\n",
    "    else:\n",
    "       psi025.append(i) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove features that are geater than 0.25\n",
    "train_selected_woe_psi = train_woe[psi025]\n",
    "off_woe_psi = test_woe[psi025]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "# output our final data table\n",
    "final_data_woe = pd.concat([train_selected_woe_psi,off_woe_psi])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(30000, 23)\n"
     ]
    }
   ],
   "source": [
    "print(final_data_woe.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "#save 3: final data table with transformed woe\n",
    "final_data_woe.to_csv('CreditScore_save3_final_data_woe.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 5. Output IV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_use = [feat for feat in final_data_woe.columns if feat not in exclude_list]\n",
    "len(features_use)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_iv=output_iv_importance(final_data_woe[features_use+['label']],'label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>iv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PAY_0</td>\n",
       "      <td>0.873565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PAY_2</td>\n",
       "      <td>0.546857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PAY_3</td>\n",
       "      <td>0.412420</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>PAY_4</td>\n",
       "      <td>0.359289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>PAY_5</td>\n",
       "      <td>0.333734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>PAY_6</td>\n",
       "      <td>0.285164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>LIMIT_BAL</td>\n",
       "      <td>0.181621</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>PAY_AMT1</td>\n",
       "      <td>0.173425</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>PAY_AMT2</td>\n",
       "      <td>0.162703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>PAY_AMT3</td>\n",
       "      <td>0.114630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>PAY_AMT4</td>\n",
       "      <td>0.104007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>PAY_AMT6</td>\n",
       "      <td>0.101453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>PAY_AMT5</td>\n",
       "      <td>0.087383</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>AGE</td>\n",
       "      <td>0.021193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>EDUCATION</td>\n",
       "      <td>0.015987</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>BILL_AMT5</td>\n",
       "      <td>0.012865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>BILL_AMT2</td>\n",
       "      <td>0.008361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>BILL_AMT6</td>\n",
       "      <td>0.007382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>BILL_AMT3</td>\n",
       "      <td>0.006877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>BILL_AMT1</td>\n",
       "      <td>0.002872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>BILL_AMT4</td>\n",
       "      <td>0.000979</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         name        iv\n",
       "0       PAY_0  0.873565\n",
       "1       PAY_2  0.546857\n",
       "2       PAY_3  0.412420\n",
       "3       PAY_4  0.359289\n",
       "4       PAY_5  0.333734\n",
       "5       PAY_6  0.285164\n",
       "6   LIMIT_BAL  0.181621\n",
       "7    PAY_AMT1  0.173425\n",
       "8    PAY_AMT2  0.162703\n",
       "9    PAY_AMT3  0.114630\n",
       "10   PAY_AMT4  0.104007\n",
       "11   PAY_AMT6  0.101453\n",
       "12   PAY_AMT5  0.087383\n",
       "13        AGE  0.021193\n",
       "14  EDUCATION  0.015987\n",
       "15  BILL_AMT5  0.012865\n",
       "16  BILL_AMT2  0.008361\n",
       "17  BILL_AMT6  0.007382\n",
       "18  BILL_AMT3  0.006877\n",
       "19  BILL_AMT1  0.002872\n",
       "20  BILL_AMT4  0.000979"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_iv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "#save 4: information value after woe transformation\n",
    "df_iv.to_csv('CreditScore_save4_IV.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 6. Model Tuning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "def check_train_test_auc(x_train,y_train,x_test,y_test):\n",
    "    # 用逻辑回归建模\n",
    "    from sklearn.linear_model import LogisticRegression\n",
    "    lr = LogisticRegression(random_state=42,C= 0.1, penalty='l2', solver='newton-cg')\n",
    "\n",
    "\n",
    "    lr = LogisticRegression(class_weight='balanced')\n",
    "    lr.fit(x_train, y_train)\n",
    "\n",
    "    # 预测训练和隔月的OOT\n",
    "    pred_train = lr.predict_proba(x_train)[:,1]\n",
    "    from toad.metrics import KS, AUC\n",
    "\n",
    "    print('train KS',KS(pred_train, y_train))\n",
    "    print('train AUC',AUC(pred_train, y_train))\n",
    "    \n",
    "    pred_OOT =lr.predict_proba(x_test)[:,1]\n",
    "    print('Test KS',KS(pred_OOT, y_test))\n",
    "    print('Test AUC',AUC(pred_OOT, y_test))\n",
    "    \n",
    "    from sklearn.metrics import confusion_matrix, accuracy_score, roc_auc_score, plot_roc_curve, classification_report\n",
    "\n",
    "    fig, ax = plt.subplots(figsize=(12, 8))\n",
    "    plot_roc_curve(lr, x_test, y_test, color='blue', ax=ax)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train KS 0.40988185783295383\n",
      "train AUC 0.7703247664492447\n",
      "Test KS 0.4264763906414198\n",
      "Test AUC 0.7834675891381715\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#train & test\n",
    "check_train_test_auc(x_train = train_woe[features_use],y_train=train_woe['label'],\n",
    "                     x_test =test_woe[features_use] ,y_test = test_woe['label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_evaluation_scores(label, predictions):\n",
    "    from sklearn.metrics import classification_report, confusion_matrix, accuracy_score\n",
    "    from sklearn.metrics import balanced_accuracy_score\n",
    "    tp, fn, fp, tn = confusion_matrix(label,predictions,labels=[1,0]).reshape(-1)\n",
    "    print('True Positive：',tp)\n",
    "    print('True Negative：',tn)\n",
    "    print('False Positive：',fp)\n",
    "    print('False Negative：',fn)\n",
    "    accuracy = (tp+tn)/(tp+fn+fp+tn)\n",
    "    print('accuracy: ',accuracy)\n",
    "    recall = tp/(tp+fn)\n",
    "    print('（recall）: ',recall)\n",
    "    precision = tp/(tp+fp)\n",
    "    print('（precision）: ',precision)\n",
    "    #f1 score = 2*(P*R)/(P+R)\n",
    "    f1 = 2*precision*recall/(precision+recall)\n",
    "    print('F1 score: ',f1)\n",
    "    \n",
    "    print(classification_report(label, predictions))\n",
    "    \n",
    "    print('balanced_accuracy_score: ',balanced_accuracy_score(label,predictions))\n",
    "    return precision, recall\n",
    "\n",
    "def evaluate_result(df_train,df_test,features_name):\n",
    "    from sklearn.ensemble import AdaBoostClassifier, GradientBoostingClassifier, RandomForestClassifier, ExtraTreesClassifier\n",
    "    import seaborn as sns\n",
    "    import matplotlib.pyplot as plt\n",
    "    start = time.time()\n",
    "    x_train = df_train[features_name]\n",
    "    y_train = df_train['label']\n",
    "\n",
    "    x_test  = df_test[features_name]\n",
    "    y_test  = df_test['label']\n",
    "\n",
    "    model = GradientBoostingClassifier(n_estimators=250,random_state=0)\n",
    "    model.fit(x_train,y_train)\n",
    "    predictions = model.predict(x_test)\n",
    "    get_evaluation_scores(label = y_test, predictions=predictions)\n",
    "    feat_importances = pd.Series(model.feature_importances_, index=features_name)\n",
    "    feat_importances=pd.DataFrame(feat_importances).reset_index()\n",
    "    feat_importances.columns=['feature_name','feature_importance']\n",
    "    feat_importances=feat_importances.sort_values(['feature_importance'],ascending=False)\n",
    "    import matplotlib.pyplot as plt\n",
    "    plt.figure(figsize=(15,15))\n",
    "\n",
    "    sns_plot1=sns.barplot(feat_importances.feature_importance,feat_importances.feature_name,estimator=sum)\n",
    "    plt.title(\"Features Importance\",size=18)\n",
    "    plt.ylabel('', size = 15)\n",
    "    plt.tick_params(labelsize=18)\n",
    "    return feat_importances,model,x_train,y_train,x_test,y_test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train a GBDT and check the feauture importance table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True Positive： 542\n",
      "True Negative： 5725\n",
      "False Positive： 228\n",
      "False Negative： 1006\n",
      "accuracy:  0.8354886015197973\n",
      "（recall）:  0.35012919896640826\n",
      "（precision）:  0.7038961038961039\n",
      "F1 score:  0.46764452113891286\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.85      0.96      0.90      5953\n",
      "           1       0.70      0.35      0.47      1548\n",
      "\n",
      "    accuracy                           0.84      7501\n",
      "   macro avg       0.78      0.66      0.69      7501\n",
      "weighted avg       0.82      0.84      0.81      7501\n",
      "\n",
      "balanced_accuracy_score:  0.6559145910840776\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Bruce\\anaconda3\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variables as keyword args: x, y. From version 0.12, the only valid positional argument will be `data`, and passing other arguments without an explicit keyword will result in an error or misinterpretation.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fet_importance_GBDT_reason,model,x_train,y_train,x_test,y_test = evaluate_result(df_train=train_woe,\n",
    "                df_test=test_woe,\n",
    "                features_name=features_use)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "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_name</th>\n",
       "      <th>feature_importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>PAY_0</td>\n",
       "      <td>0.641879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>PAY_2</td>\n",
       "      <td>0.080198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>PAY_3</td>\n",
       "      <td>0.044386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>LIMIT_BAL</td>\n",
       "      <td>0.030078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>PAY_AMT2</td>\n",
       "      <td>0.028890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>PAY_5</td>\n",
       "      <td>0.027666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>PAY_4</td>\n",
       "      <td>0.023903</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>PAY_6</td>\n",
       "      <td>0.019000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>PAY_AMT1</td>\n",
       "      <td>0.018523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>PAY_AMT6</td>\n",
       "      <td>0.017613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>PAY_AMT3</td>\n",
       "      <td>0.011601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>PAY_AMT4</td>\n",
       "      <td>0.011289</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>AGE</td>\n",
       "      <td>0.008551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>BILL_AMT1</td>\n",
       "      <td>0.006361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>PAY_AMT5</td>\n",
       "      <td>0.006068</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BILL_AMT6</td>\n",
       "      <td>0.005579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>BILL_AMT2</td>\n",
       "      <td>0.004624</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BILL_AMT5</td>\n",
       "      <td>0.004110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>BILL_AMT3</td>\n",
       "      <td>0.004073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>EDUCATION</td>\n",
       "      <td>0.003597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BILL_AMT4</td>\n",
       "      <td>0.002012</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   feature_name  feature_importance\n",
       "19        PAY_0            0.641879\n",
       "14        PAY_2            0.080198\n",
       "13        PAY_3            0.044386\n",
       "17    LIMIT_BAL            0.030078\n",
       "11     PAY_AMT2            0.028890\n",
       "15        PAY_5            0.027666\n",
       "12        PAY_4            0.023903\n",
       "18        PAY_6            0.019000\n",
       "6      PAY_AMT1            0.018523\n",
       "9      PAY_AMT6            0.017613\n",
       "16     PAY_AMT3            0.011601\n",
       "7      PAY_AMT4            0.011289\n",
       "20          AGE            0.008551\n",
       "4     BILL_AMT1            0.006361\n",
       "5      PAY_AMT5            0.006068\n",
       "2     BILL_AMT6            0.005579\n",
       "8     BILL_AMT2            0.004624\n",
       "3     BILL_AMT5            0.004110\n",
       "10    BILL_AMT3            0.004073\n",
       "0     EDUCATION            0.003597\n",
       "1     BILL_AMT4            0.002012"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fet_importance_GBDT_reason"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_roc_pre_recall_curve(labels, probs):\n",
    "    from sklearn.metrics import precision_recall_curve\n",
    "    # Get ROC curve FPR and TPR from true labels vs score values\n",
    "    fpr, tpr, _ = roc_curve(labels, probs)\n",
    "\n",
    "    # Calculate ROC Area Under the Curve (AUC) from FPR and TPR data points\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "\n",
    "    # Calculate precision and recall from true labels vs score values\n",
    "    precision, recall, _ = precision_recall_curve(labels, probs)\n",
    "\n",
    "    plt.figure(figsize=(8, 3))\n",
    "\n",
    "    plt.subplot(1,2,1)\n",
    "    lw = 2\n",
    "    plt.plot(fpr, tpr, color='darkorange', lw=lw, label='ROC curve (area = %0.4f)' % roc_auc)\n",
    "    plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n",
    "    plt.xlim([0.0, 1.0])\n",
    "    plt.ylim([0.0, 1.05])\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.title('ROC Curve')\n",
    "    plt.legend(loc=\"lower right\")\n",
    "    plt.grid(True)\n",
    "\n",
    "    plt.subplot(1,2,2)\n",
    "    plt.step(recall, precision, color='orange', where='post')\n",
    "    # plt.fill_between(recall, precision, step='post', alpha=0.5, color='orange')\n",
    "    plt.xlabel('Recall')\n",
    "    plt.ylabel('Precision')\n",
    "    plt.ylim([0.0, 1.05])\n",
    "    plt.xlim([0.0, 1.0])\n",
    "    plt.title('Precision Recall Curve')\n",
    "    plt.grid(True)\n",
    "\n",
    "    left  = 0.125  # the left side of the subplots of the figure\n",
    "    right = 0.9    # the right side of the subplots of the figure\n",
    "    bottom = 0.1   # the bottom of the subplots of the figure \n",
    "    top = 0.9      # the top of the subplots of the figure\n",
    "    wspace = 0.5   # the amount of width reserved for blank space between subplots\n",
    "    hspace = 0.2   # the amount of height reserved for white space between subplots\n",
    "    plt.subplots_adjust(left, bottom, right, top, wspace, hspace)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAADqCAYAAAClduxYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAABebklEQVR4nO3dd3QUVR/G8e9ueq+bUKUqJdTQewfBUEMRUKogIKIoKL2FJiAI+KJU6b0jSCdSlar0bkhAkmx6z7Z5/4gsRggJJJvZJPdzDudkZndmnl1272+n3auQJElCEARBEIR8Syl3AEEQBEEQTEsUe0EQBEHI50SxFwRBEIR8ThR7QRAEQcjnRLEXBEEQhHxOFHtBEARByOcs5Q4gZK5cuXK88847KJVKFAoFycnJODo6MmXKFCpXrgxAUlISixcv5vjx41hbWwPQvHlzhg4diq2trXFdu3btYvPmzaSkpKDVaqlRowajR4/G2dn5pdt+3ecLgpDm8ePHtGrVinfeecc4T5Ik+vTpQ9euXXNkGwsXLqREiRJ06tQpw+d07NiRdevW5ch3dvHixWzYsAFvb28g7fUkJCTQqlUrxowZg0KhyPY2/m3MmDG8/fbbDBw4kHLlynHu3Dnc3d1feN6JEydYtWoVcXFx6HQ63n77bb7++msKFy6co3nyNEkwe++8844UGRmZbt6KFSuk7t27S5IkSVqtVurevbs0c+ZMKSkpSZIkSUpKSpICAgKkXr16SVqtVpIkSfrhhx+knj17Smq1WpIkSdJoNNKUKVOknj17vnS7r/t8QRCeCwkJkapVq5ZuXmhoqFSzZk3p1q1bMqXKnkWLFklTp05NNy8mJkZq3LixdPLkyRzf3tdffy2tWLFCkqSXt4OSJEl79+6V2rZtKwUFBUmSJEkGg0H68ccfpZYtW0qpqak5nimvEnv2eZBOp+Pp06e4uLgAcPDgQQwGA2PHjjU+x87OjvHjx9OpUyeOHDlCkyZNWLp0Kbt27cLT0xMAKysrvvrqK44cOYJGozEeEYC0IwWZPX/p0qVER0czadIkIO1X/7PpDz/8EBcXFx4+fEiPHj1YsmQJp06dwtraGr1eT9OmTVm9ejVeXl7MmDGDu3fvotVqqVevHl999RWWluKjKeQ/3t7elChRgqCgIG7evMn27duNR+rWrVvHtm3b2LRpEwaDAVdXVyZOnEiZMmVITExk+vTpXL58GQsLC1q2bMnIkSMZO3ascc930aJFHDlyBCsrK9zc3Jg1axZeXl7p9oj/97//sX//fiwsLChVqhQTJ05EpVLx4YcfUq1aNS5fvszTp0+pV68eAQEBKJWZn+mNiIggJSXF2B49ePCAGTNmEBMTg16v58MPPzQeydi+fTs//fQTSqUSNzc3vvnmG7y9vZk5cyZ//vkniYmJSJLE9OnTqVGjRpbe0wULFhAQEECJEiUAUCgUDB48mMKFC6PRaPj55585dOgQS5cuBWDnzp3G6TFjxhATE0NISAgNGjRg+/btHDp0CJVKBUC3bt0YPnw49erVY968eVy4cAG9Xk/FihWZMGECjo6Or/0ZkItoUfOIvn37AhAdHY2NjQ3NmjVj1qxZAFy5coWaNWu+sIxCoaBevXpcunSJ4sWLY2trS8mSJdM9x87Ojg4dOryw7MOHD1/r+S/j7OzMgQMHADhy5AjHjx/n3Xff5fTp0xQrVowyZcowduxYfHx8mD17Nnq9njFjxvDTTz8xaNCgLG1DEPKSK1euEBwcTNWqVTl37hz379/n+PHjODo6cv78eXbv3s2GDRuws7Pj9OnTDB8+nF9++YVFixaRmprKgQMH0Ov1DBgwgPPnzxvX+/TpU9asWcO5c+ewtrZm1apVXL16lZYtWxqfs2PHDk6dOsX27duxt7dn8eLFjBkzhpUrVwIQHBzMunXrSEpKom3btpw/f566deu+8BoOHDjApUuXSE5OJjY2looVKzJ16lSqVKmCTqdjxIgRzJkzBx8fH+Lj4+nRowdly5bF1taWefPmsWvXLgoXLszq1av54Ycf6Ny5M+Hh4WzZsgWlUsmyZctYvnx5lop9dHQ0T548wdfXN918hUKR5XYqJSWF/fv3AxAbG8vevXsZOHAgDx48ICIigkaNGrFkyRIsLCzYuXMnCoWC+fPnM2/ePKZMmZKlbZgDUezziDVr1uDu7s6NGzcYPHgwderUwcPDw/i4Tqd76XIajQYLCwuUSiUGgyHL23vd57/Mv3+AdO3alV27dvHuu++yc+dOunfvDkBgYCDXrl1j+/btQNoXTxDyi5SUFDp27AiAXq/Hzc2NuXPnGs8llytXzrh3GBgYyKNHj3j//feNy8fFxRETE8PZs2cZO3YsFhYWWFhYsH79eiDtmhpIO2JQvnx5OnfuTOPGjWncuDH16tVLl+XkyZN06dIFe3t7APr06cOPP/6IRqMBoFmzZiiVShwdHSlRogSxsbEvfU3t2rVj0qRJaDQaAgICuH//Ps2bNwcgKCiI4OBgxo0bl+49uHnzJqmpqTRs2ND42vv162d8jouLC5s3byYkJITff/8dBweHLL2/z448ZKet+vePim7dujF16lQGDhzIjh078Pf3R6lUEhgYSHx8PGfPngVAq9Wma3/zAlHs8xgfHx/Gjh3LmDFjqFChAsWKFcPX15cVK1ZgMBjSHXYzGAxcuHCBoUOHUrZsWXQ6HUFBQen21lNTUxk+fDjTp083XnQDZOn5CoUC6V9DK2i12nRZnzUqAG3btmX27Nk8ePCACxcuMHv2bGPGhQsXUqZMGSCtccvpi3wEQS62trbs2bMnw8f//R0xGAx07NiR0aNHG6fDw8NxcXHB0tIy3ffi6dOn6S68VSqVrF+/nmvXrnHu3DlmzpxJo0aN+Oqrr9Kt/9/rMBgM6XYS/r2+/363X8ba2pqJEyfi7+/PnDlzmDBhAnq9Hicnp3SvOSIiAicnJ7Zs2ZJu+ykpKTx58oSQkBBmzJhB//79adGiBaVLl2bv3r2v3PYzLi4ulCxZkj///JP69eune+yzzz5j6NChr9VO1axZE51Ox9WrV/n555/ZsmULkPZejRs3jiZNmgCQmJhIampqljKaC3HrXR7k5+dHlSpVjIfx27Rpg52dHTNnzjTuGaekpBAQEICDgwOtWrXC2tqaQYMGMX78eCIiIoC0vf6ZM2eSnJycrtADWXq+m5sbN27cMF6Re+LEiQwz29jY8N577zFmzBhat26NnZ0dAA0bNmT16tVIkoRGo2Ho0KHGvRZBKEgaNmzI/v37CQ8PB2DTpk3G03f16tVj165dGAwGNBoNI0aM4MKFC8Zlb9++jZ+fH2XKlOHjjz+mX79+XLt2Ld36GzVqxI4dO0hKSgJg3bp11KpVK921Oq/L2tqayZMns3HjRm7evEmpUqXS/cB5+vQpfn5+XL9+nTp16nDu3Dnj69u8eTNz587lzJkzNGvWjF69elGpUiWOHj2KXq/Pcobhw4czY8YMHj16BKQdQVmyZAm3b9+mdOnSuLu7c+/ePVJTU9FqtRw6dOiV6+vWrRsBAQGUK1fOeBSiYcOGbNiwAY1Gg8FgYOLEicyfP/9N3jLZiD37PGrixIl06NCBU6dO0ahRI1atWsWSJUvo0qULSqUSvV5P8+bNWbVqFVZWVgAMGTIEOzs7Bg4cCKTtpdeuXZslS5a8dBuZPf/Z9lu3bo23tze1a9d+5d5At27dWL9+fbrzXOPHj2fGjBm0b98erVZL/fr1+eijj3LiLRKEPKVhw4YMGjSIAQMGoFAocHR05Pvvv0ehUBgLWseOHdHr9bRr147WrVtz/PhxAMqXL0/btm3x9/fH3t4eW1tbJkyYkG79Xbt25enTp3Tr1g2DwUCJEiWYN29etnPXrFmT9u3bM23aNDZt2sSSJUuYMWMGK1asQKfT8dlnnxkPlY8ePdr4/VapVMycOZOEhAS+/PJL2rdvj06no0GDBhw+fDjLh+bbt2+PJEl88cUX6HQ6UlNT8fHxYc2aNVhbW9OgQQNq1apF27ZtUalU1KlThzt37mS4vk6dOjF//vx0xXzYsGF88803dO7cGb1eT4UKFRgzZkw23rXcp5AyO1YjCIIgCEKeJg7jC4IgCEI+J4q9IAiCIORzotgLgiAIQj4nir0gCIIg5HOi2AuCIAhCPieKvSAIgiDkc3n6Pvvo6EQMBvO4c9DDw5HIyAS5YwDmlQVEnldRKhW4uWWta1Ahe0R78XLmlAVEnoxkt63I08XeYJDM5ssLiCyvIPIIchPtRcbMKQuIPKYgDuMLgiAIQj5n0mKfkJCAn58fjx8/fuGxW7du0aVLF9q0acP48eMzHLVNEISCQbQXgmA6Jiv2f/75Jz179iQoKOilj48ePZpJkyZx6NAhJEli69atpooiCIKZE+2FIJiWyYr91q1bmTx5Ml5eXi889uTJE1JSUqhWrRoAXbp04eDBg6aKIgjmR5JQpEZhGX1d7iRmIVfaC0PWR1IThPzGZBfozZgxI8PHwsPDUalUxmmVSkVYWJipogiCPLQJKJPDsIy+jmXsHazCziBZOWETvBdJYYlC0oFzCSgbJHdS2eVGe+FxtipxJceQWqT3G2UUhLxMlqvxDQYDCoXCOC1JUrrprPLwcMzJWNmmUjnJHcHInLJAPs+TGgs310HYRVBagz4lbToDGp0Flx4Xol6Z8JzLkI/lVHsB4MwTMJPPojl9J8wpC4g8piBLsS9UqBBqtdo4HRER8dLDd5mJjEwwm1siVCon1Op4uWMA5pUF8ngeXRKWMTdRpMZgGXMLlJag16DQxmGlvoB1aGDmq3D1QaFLJKVUVxIM7vSaZs3pS0ls3NiFpk1L4JG9l5Pv5VR7IaEkKSmVJDP4LJrTd8KcsoDIkxGlUpGtHVxZin3RokWxsbHh0qVL1KhRgz179tC4cWM5oggC6JKxiriIdch+rJ8cRqFLAkmPQp+KUhOT5dVoVXXQFG6K3qkUKG3QOxZH51kL/tkLjYpKpnfvXVy6FIqnpz1ubrYmekH5i2gvBCH7crXYDxo0iBEjRlC5cmXmzZvHhAkTSEhIwMfHhz59+uRmFKEAUiY+QZkcCk9uYhcTjWXUn1hFXMQi/q9Ml9V4N0Bh0CNZOaJzrQBKaySFEsnGDU2hxujdq7xy+adP4+nefSd37kRSvLgz27b5U7q0W069tHxJtBeCkHMUkiSZx3HwNyAO47+cOWWBXM5j0GIVfg5lUiiWUX+g0CVjE7QTpSY600U13o2QLO1Jeac/Og9fJAtrJAtbsLAz7p2/iQcPounefQchIXGUK+fB1q1dKFw47Rxgdg/NCVkn7SlNkldXkspOlDuKWX1HzSkLiDwZyZOH8QUhp1k/PojNg03YPtqV+ZMdCqNx8UHrWROde2UMDm+hc68Mipy/E1WnM9Cz505CQuKoUaMwGzd2ws3NLse3I2ROgQGHv+YiWTiQXOoLueMIQq4SxV7Is6xCT+Lwxwysws+98JjBxoPUt9qDJKFT1UbvWBytdyNQWqBSORGbS7/ULS2VzJ/fimXLrvDDD+1wcLDKle0KGXO8P0UUe6HAEcVeyFskAxYxN7G79SN299eme0hTuBmaIi1IebsvkrWLTAHThIUl4O2ddsitYcO3aNjwLVnzCM9p3JvIHUEQcp0o9oL5kySsws5gd2cZNo92v/BwfO25pJbsimRrHjexbd58g6++OsqqVe1p2bK03HGE/zBYucsdQRBynSj2gnnSJWF/fT52d1eDpEeZGpnuYUmhJKnSFyRVHQtK8zk0vmTJRaZMOQnA1avhotibIduwXcQbVpjV50YQTE0Ue8GsKBMfY3t3FQ7X5qWbb7B2Q1OsNSll+6D1qg9KC5kSvpwkScyYcZpFiy4AMGNGUwYN8pU5lfBvUXVO4368OADWEUfQeLWTOZEg5B5R7AXZKTSxWIWewvb+Wmwepx/gJLVoa5IrfYHWq162bn8zJb3ewOjRR1m//joWFgoWLWpDt24V5Y4l/Idk5Wr82+XP99G4NiCu2gYkcVhfKABEsRdkZf/HDByufmOclpTW6NwqkVKmJylleoOV+d+D/uWXR9i48Qa2thasXNmeVq3EoXtzlVjqKxz+mgOAdcwZbEK3k1J8sMypBMH0RLEXZOFwcTz2NxcbpyUUaIs0J67Bj0h23jIme309e1bi+PEgli3zo27donLHEV7BIvlhummDTWGZkghC7hLFXsgVCm089tfmYxF9HZsnh9I9llTpCxJ9p8iS603pdAYsLdM64alTpyjnzw/E1lZ8ncydUhuTbtr26WY0Xu3lCSMIuSjnuwwThH8oUiKxfnwQtrfBc1NR7K9/+0Khj+z8R54r9I8fx9G8+Tr2779nnCcKfd4QV3kFWqfqxmmrqFMyphGE3CNaKCFHWUb+gc1f27H5awsWyWEvPK7xbkRijQB0HtXN9oK7V7l7N5Lu3Xfw998JLFp0nrZty6JU5r3XUVBJVu7E1P0VhTYKz8CSKHUxuP7ejITy36JzEXdPCPmXKPZCjlBoYnE+8T7WYWdefNClFPHvDCWl3MA8fW/zlSuh9Oy5k6ioFGrXLsKGDZ1Eoc+znt+6aRV3CadrA4hu+Id8cQTBxESxF7LHoMM6ZD8uv36YbnZCzVmklOqOZKdCpXIixQxGjcqOkyeD6dt3D4mJWlq2LMWKFX7Y2+fdHy4FnWSVvjtly/9cuCcI+Y0o9sIbUaREYnfvJxyuTEs3P7lsXxLqLgBl/vloHThwn8GD96PR6OnatQILF7bGysq8OvURskdS5J/PqyC8jPiEC6/FOuQX7G4twTr013TzU0p0Ianq1+hdK8iUzHQKF3bEykpJ375VCAhoKg7d5xMGS1eUuhgAUgt1lzeMIJiYKPZCltn8tQ3nUwON05LSCq2qLvH1FmFwLiNjMtOqXr0QgYF9eOstZxR58KJC4eWi653FOuIwTrc+xyZsF/GVfpQ7kiCYjCj2wqsZdDif7IdN8F7jLK1nDeIar8HgmD+HbZUkialTT1K5shf+/mlHKkqUkHfIXCHnGWyLkVJsAE63PkdhSMYueAnJbw2TO5YgmIQo9sLLGXTYPtiAw5WpKFMijLMTq40nqcrXMgYzLZ3OwBdfHGHz5hvY21vSuHEJVCp7uWMJucDxzhhR7IV8SxR74QU2DzbheP4rlNpY4zytWxViW+0xmzHjTSElRcfgwfs5ePAB9vaWrFrVQRR6QRDyBVHsBSPLiMvY3l2F3f21AEiWDmi8GxLXZA1Y5u+iFxeXSp8+ezh79jGurjZs3NiZmjWLyB1LyAU6+zJYJj0AwONEcZS6WPTW3sRV34bOuZq84QQhh4hiL2D9+CC2d1elG15W492Q2Ja7wcJavmC5RK1O4v33d3LtWjiFCjmwdas/5ct7yh1LyCXRDa6gOuIMgFKXdjTLQhOG2++NifHdi9ajqYzpBCFniGJfUEkGbO+sxCZ4H9ahgekeim22GU2xtnmyO9s3oVYn8uhRLKVKubJtmz9vvSUuxhPSKAypIEkF5rsg5F+i2BdAyviHeOyqlm6e3rEE8XUWoPWun+8P2f9XxYoqtm7tQrFiznh5OcgdR5BBqldHbML3vDDf5Y9uGS6jbhkDCjGWmJA3mPSTum/fPtq1a0fr1q3ZsGHDC4/fuHEDf39/OnTowMcff0xcXJwp4xR41o9247nONV2hN1i5EFf/B6La/4a2aMsCU+gvXvybrVtvGqd9fQuLQi8zOduLuKrrULeKM/6L8/kh02Us468CYBu8FLughWDQ5VgeQchpJiv2YWFhLFiwgI0bN7J79262bNnC/fv30z1nxowZjBgxgr1791KqVClWrlxpqjgFmy4J951VcPm1DwrJAKSdk49+9wiRPUNILdsbrApOoTt06D5du25nxIhD/P77E7njCJhfe6FR+WX6HMfbX+B0/WOc7ozG8d5EVMfcQZ9sskyCkB2ZFvvExESmTp1K3759iYmJYdKkSSQmJma64rNnz1K3bl1cXV2xt7enTZs2HDx4MN1zDAaDcV3JycnY2tq+4csQMmLzcDOqjYWwSAgCQOdakahOl4htcwCdVx15w8lg9+47tG+/iaQkHd27V6RGjcJyRxIwv/ZCsnJJ28tvHo66eWi6vf7YqpsAsIq9iO3TTemWUx33xvVCm7Tz/IJgRjIt9tOnT8fZ2ZnIyEhsbGxISEhg0qRJma44PDwclUplnPby8iIsLP345mPGjGHChAk0bNiQs2fP8v7777/BSxBeRqGJw+VwB5xPDzbOi6+7kOgOv6F3flvGZPL56ac/+fjj/Wi1BoYOrcHCha2xtBTnXM2B2bYXFrZgkf7UlsbrvXTTKYXfJ7nYR8Zpq5hzuP1Wz/TZBOE1ZHqB3q1bt5g1axa//vordnZ2zJs3Dz+/zA9xGQyGdP2IS5KUbjolJYXx48ezevVqqlSpwk8//cTXX3/NsmXLshzew8Mxy8/NDSqVk9wRIPQC/Pwtnne2pJ//SRROtm7IlVDO90aSJKZPP8mkSYEAzJ7dgq++aiD6uTcjea69aHECjjUDwLbZJtAmwLYVxoctE26ickoGW68MV2EW7cU/zCkLiDymkGmxVyrT7/no9foX5r1MoUKFuHjxonFarVbj5fX8g3/37l1sbGyoUqUKAD169GDhwoVZDg4QGZmAwWAeh8tUKifUMo/Z7nT6Y2wfpj+smFKyC/ENl0O8JcTLk0/u9yYsLJHvvvsNpVLB3Lkt+OKLBrL/Xz2jVCrM7kerHPJce6GsAa3+uUDw2Wfpn+ln9+yz0zvt4RZqUFinu31P7u/Ev5lTFhB5MpLdtiLTql2rVi3mzp1LSkoKp06d4tNPP6V27dqZrrh+/fqcO3eOqKgokpOTOXz4MI0bNzY+XqJECUJDQ3n48CEAx44do3Llym/8QgoyZUIIqrXO6Qp9Qs2ZRPQIIr7xalBayRfODHh7O7BxY2eWL3+PDz+sIncc4SXyU3sRXftEumnVMRWqoy44X/aXKZEggEKSXn0liVarZdmyZQQGBqLX62nUqBGffPIJ1taZ96y2b98+li5dilarpWvXrgwaNIhBgwYxYsQIKleuzK+//sq3336LJEl4eHgQEBBA8eLFsxxe7NmDIjUKzy0ljdM6p9JYDryFOio117NkRI73JilJy+nTIbRuXdos8mRE7Nk/l5/aC+Pe/UskvB2AYxFforSF0duXlv1efXP6PoDIk5HsthWZFvujR4/SsmXLdPN2795Np06d3nijOaXAFnuDHpvgPTie+xSl9vk242t/S0q5j1B5OZvFh/OZ3P6yxMam0Lv3bi5c+Jtly96jY8dysuZ5FVHsc49c7YUy6SEeZ6pl+HhEk7+QrOUbYMqcvg8g8mQku21Fhufsjx8/jk6nY86cOUiSxLPfBDqdjsWLF5tFsS9wJAnbOytwOv/lCw/F11tMytt9ZQhlXsLCEujRYyc3b0ZQpIgjFSuqMl9IEEzIYF+aiCYPcb3UEcuEay887vlrKXSOFYmucxqUolNTwTQy/GTdunWL3377jcjISNauXft8AUtL+vXrlxvZhH9RJj7BY0eFdPNSi79HcvmP0RZqIvruBoKCYujWbQePHsVStqwbW7f6U6xYxodTBSG3SNaeRNc7AwYtCn08nkVKYNjmiVIbCfxz9f4xd5KL9iOh4iKZ0wr5UYbF/pNPPuGTTz5hw4YN9O7dOzczCf9QaOOxebDxn8FqThrna1W1iW22JV+PLf+6btxQ06PHTsLDE6lWzZuNGzvj6Vkwuv4V8hClFZLSHYDIpn8B6c/v2z1Zjc7Rh5TiH4FkKPAX1wo5J9NjRt26dePIkSPGnqv0ej3BwcGMHDnS5OEKLMmAw8Xx2N/63wsPxTVcQWrp7jKEMl96vYFBg34mPDyRRo2Ks2ZNRxwd8//QvEL+oG4Vh0ITieevpQBwujMapzujX3iepLAgudhHGOxKoHOqgta98QvPEYSMZFrsR44cSUhICGq1mooVK/Lnn39m6dY74Q0YdFg/OYzLifQ9g6UWf4/kdwagLdJC9it3zZGFhZKlS99j+fIrzJ3bAhsbcd5TyFskaw/UreJeeRW/QtJjH7L0hfkRTYORrFxNmE7ID7LUg97hw4eZMmUK/fv3x2AwMGXKlFyIVrC87Jy83rEkUZ0uiUN5Gbh3L4q33047JFq5sheLFrWROZEgZI+61fOR/DyPeaEwpJBYeix6h7dxvjbgpct4Br6VtmzLWHHtjpChTHcTvby8sLS0pGTJkty9e5e3336beJl6YsuvbO+tS1fo9XaFiW26kaguV0Whz8Dy5Zdp2HA1mzZdlzuKIJhERItw1K3iSCozltRCXdMNxqNu9jd62/R9DKiOuqDQRsmUVjB3me7Z29vbs2/fPsqXL8/WrVspXbo0SUlJuZGtQLB+tBunc58Yp2ObbUZTvJ2MicybJEnMmXOOb7/9DYCYGPPpPEgQco2lI1GNbgCgTHmCx6m0nQXPwJKoW0SJW/iEF2S6Zz9p0iRu3bpFgwYNUCqVfPDBBwwcODA3suV7ysQnuPzaxzit7vlUFPpXMBgkxow5zrffpvVzv3Bha4YOrSF3LEGQlcG2aFqB/4fHr6VBnwL6pLQr+gWBLOzZlyxZkq+++gqA7777DoD79++bNFRB4HB5CvbX5xunI3oEgZWDfIHMnEaj59NPD7Jr1x1sbCxYuvQ92rUrK3csQTAPSkui6p7G/beGKHUxqI4/H0QoutZRdK7iouqCLsM9+8ePH/Pll18SEBBAcnIyAImJicyePVv0npdNNg83pyv0cQ1+QLJxlzGR+Rs9+ii7dt3B0dGazZu7iEIvCP+hd6pCYqlRL8x3u9ASy9jLMiQSzEmGxX7cuHG4ubmhVqtZunQpV69e5b333uPUqVOsWLEio8WETFiFnsL59GDjdGTXu6SWEZ0WZWbYsJq88447u3d3o0GDrA9+IggFSVLZSc8v4msZg96mCABu55uiOuKMIjVc5oSCXDI8jB8aGsratWtJSUmhS5cubN26lX79+jFgwAAsLcXFH29CmfQU18PvGacjuv8lesF7hcRELQ4OaXcjlCvnwa+/9sHCQvQzIAhZolAS1fg2DnfGYh+c1kGX58my6W7vEwqODFtOe/u0rkZtbW2JjY1lzpw5DB48WBT6N2R/dQ4e25+Pvhbd9qgo9K/w8GE0TZqs4aef/jTOE4VeEF5fYrlZaffg/0OhjX3Fs4X8Kkutp7u7Ow0bNjR1lnzL+Xh3HP6YbpyO6nAenUpcMJORa9fC8fPbTHBwHNu330KnE1cUC0K2KBSkFO4BgGXcFZnDCHLIsNgr/tUTk9ibf3NOpz/G5vFB43REj0foXcvLmMi8nT0bQqdOW4mISKZp0xJs3eqPpaXYoxeE7NI7pB1ZdL3cAbugxTKnEXJbhlX8zp07+Pr6ApCSkmL8W5IkFAoFly+LqztfSZuI86n+6Qq9+oNoUFrIGMq8HTz4gEGDfiY1VU+nTuX4/vt3sbYW75cg5ISkUqNwuD8NAMd747GO+IXYmgdkTiXklgyL/ZEjR3IzR/6hT8H+6hwcrs0zzjLYqojsdl/0W/0Ke/bcYciQA+j1Ev36VWXWrGbiHL0g5DB1y1isIw7i8kcPrKNPpw280y0WEG1TfpdhsS9atGhu5sgftAmoNhVJNyu53GASas8VhT4TlSt74eZmR58+lfn66/rpTiMJec/58+eJjY1FkiTjvNatW8uYSABAoUCjaktysYHYPV6ZNu9wA6h9Vt5cgsmJk/E5JSUG958bGCc13g2Ja74FycpJxlDm7dkpIYDSpd04daovHh52MqcSsmvChAmcPHmSEiVKGOcpFApR7M1IQoUFJJSbi+qYO8Rexy5oIcklP5M7lmBCotjnAIvom7C2LhaAwdaT2GZb0KlqyR3LrOn1Br766hhly7ob+7cXhT5/OHfuHAcOHMDR0VHuKMKrKC3ROVbEMuEmjvcmonWth86lljgKmU+Jk6LZZPPXDtz31TVOx7TaJwp9JlJTdQwatJ91667xzTdnCA1NkDuSkIMKFy4sCn0eEV3vN7BI+5HtdqElqqMu8K9TL0L+kWmxV6vVDB48mDZt2hAREcHAgQMJDxddLgJYxNzC+VR/43R0u0D0bj4yJjJ/CQkaevXazc8/38PZ2YbNm/0pVEgUhvzE19eXkSNHsnfvXg4fPmz8J5ipzk/QeDQ3ToqCnz9lehh/6tSptGzZkvXr1+Pi4kL58uWZMGECy5Yty418Zss6aBcuJ/s+n/FpHDrRMdUrRUYm06vXTq5cCUOlsmfLFn8qVVLJHUvIYVeupHXasm3bNuM8cc7ejFm7Eeu7G4UuDs8TxYC0gh9V9zR6pyoyhxNySqbF/smTJ3Tv3p2NGzdiZWXF6NGjad++fZZWvm/fPn744Qd0Oh19+/ald+/0A748fPiQyZMnExsbi0qlYv78+bi4uLzZK8lFDpcmYX/jOwD0doWIbbUPd2snIF7WXObsyZN4unffwb17Ubz1lgvbtvlTqpSr3LEEE1i3bh0AOp0OSZKwsrLK0nL5tb3IKyRLZyKaBuMZ+BYA7r81JK7yalILdZE5mZATMj2Mr1AoMBied1eakJCQbjojYWFhLFiwgI0bN7J79262bNnC/fv3jY9LksTQoUMZNGgQe/fupUKFCnniaIHdze+NhV7rVoWoLlfRu5Z79UICer2BhAQNFSp4sn9/D1Ho87HIyEg++ugjqlWrRpUqVejTpw9hYWGvXCa/thd5jWTlirplLKlenQBwvtYP25AVYEiVN5iQbZkW+9atWzNq1Cji4+PZvHkzffv2pW3btpmu+OzZs9StWxdXV1fs7e1p06YNBw8+703uxo0b2Nvb07hxYwCGDBnywi95c6NIicDx4jgAtO5ViWl/GixsZU6VN7z1lgs7d3Zjz57ueHuLc/T52bRp06hWrRpnz57l7Nmz1KxZkylTprxymfzYXuRZCgVxlVcZJ51uf4HqmCqtAx59sozBhOzItNgPGTKExo0bU7lyZc6ePUuPHj345JNPMl1xeHg4KtXz87FeXl7pft0HBwfj6enJuHHj6Ny5M5MnTzaOtGeOlPFBeG4tbZyOaRcoX5g84tSpYP73v4vG6TJl3HB1FT+O8rugoCCGDx+Os7Mzbm5ujBgxguDg4Fcuk9/aizxPaYm62d8vzFYd98bp+hAZAgnZlek5+82bN+Pn50enTp1ea8UGgyFdL2j/7kAF0s7nnT9/nvXr11O5cmW+++47Zs+ezezZs7O8DQ+PXNpD/Ps32FXv+XS79ai8XV94mkplPh3oyJ1l585b9Oy5C41GT4MGb9GqVRlZ8/yX3O9PfqbT6UhNTcXGxgaA5OTkTHtEzFftRRaZ02fw5VmcoNc/V+WnqGGnFwC2Tzdia5EMjXfmch75mFueN5Fpsf/999/57rvvaN68Od27d6datWpZWnGhQoW4ePH5Xp1arcbLy8s4rVKpKFGiBJUrVwbAz8+PESNGvFb4yMgEDAYT3yKiT0W16Xmhj22yFo1nB1CnvxhPpXJCrTaPC/TkzrJhwzW+/PIoBoPERx9Vo0WL0mbz3oD878+/KZUKsytC2dWuXTv69etHly5dUCgU7NixgzZt2rxymXzTXmSROX0Gs5bFFlrFYRl7CbfzzeDxLuKvLCSl2ACZ8uQec8mT3bYi08P4CxYs4NChQ/j4+DB9+nT8/PxYs2ZNpiuuX78+586dIyoqiuTkZA4fPmw83wZQvXp1oqKiuH37NgDHjx/Hx8f87lH33PyW8e/o906hKdFJvjB5wOLFFxg58ggGg8RXX9VjxoxmKJWiR66C5JNPPqFr166cOXOGkydP0qVLF4YPH/7KZfJLe5Hf6VxqEFt9OwBOtz7H5ulWmRMJWZWl7nJdXFzo0aMHXl5eLF++nOXLl9O3b99XLuPt7c3IkSPp06cPWq2Wrl27UqVKFQYNGsSIESOoXLky//vf/5gwYQLJyckUKlSIOXPm5MiLyimWkX+g+OeCFK2qDjqPqjInMl+SJDF16kmWLLmEQgGzZjVnwIBqcscSclFCQgKOjo7ExMTQokULWrRoYXwsNjYWV1fXDJfND+1FQaHxbE1i6XE4PJyJ8/WPSEq4QeLbU+WOJWRCIUmv7irp5s2b7Nixg4MHD1KxYkW6detG8+bNsbSUv1t9Ux6WU2jj8dz0fOS/zMaiN5dDPSBPlqioZFq12sDTpwl8//27dOlSXtY8r2JOefLTYfzOnTuza9cuypcv/9Lz77du3ZIxnTiMn5E3zWIdtguXq893+hJLjSalSC8MdqVA8eY9sZvTewPmkye7bUWmFXvYsGH4+/uzbds2ihQpktnT8w37q8/Ho4/yO/3KQi+Au7sd27b5ExQUS/PmJeWOI8hg165dAMZD7UL+pvHuTHy5MJzufAWAw19zcfhrLgDqltGgEG2mOcn059eJEyf49NNPC1Sht3pyBPsbCwCIrzMfvbvoMvJl4uNT2bTpunG6dGk3UegFIiIiOHbsGADz5s2jb9++4gdAPpXy1hDUreKIrbo53XzVUTesIo7IlEp4mQyLfc+ePYG0QS3+/a969er4+vrmWsBcJRlwPDMM12P+xlkpZXrJGMh8qdVJdOq0jc8+O8yaNVfljiOYkTFjxhASEsK5c+c4efIkHTt2ZPr06XLHEkxI49UOdas41M2fD5LmesUfi/jrr1hKyE0ZHsZfuHAhAD///PMLj2Vymj/Psr2/AbsH643TUe3PgqXouOO/QkLi6NZtOw8fxlCqlCtNm5aQO5JgRmJiYujXrx/ffPMNfn5+dOnShQ0bNsgdS8gNFraoW8XhcG8K9kHzcf+tPrFV1qPx7iB3sgIvwz37Z/e4Tp48maJFi6b798UXX+RawNyiTHyM07nnPQOqez5F71ZJxkTm6fbtCPz8NvPwYQw+Pir27etBiRJiMBLhOa1Wi1ar5dSpU9SvX5/k5GSSkpLkjiXkosSyk0n1TOtW3eXqBzKnEeAVe/YjRozgr7/+IiQkJN0odzqdDmtr61wJl1uUSaF47KhonI54PxisHGRMZJ4uXvybXr12EROTSr16RVm3rhPOzjZyxxLMTIsWLahXrx4VKlSgUqVK+Pn54efnJ3csITcpFMRV35LWnz6gOuIs9vBlluGtd48fP+bJkydMnDgx3fk2CwsLypYtaxZDS+bIrTQGHar17sbJ6LZH0KnqvPZqzOX2DDBNFoNBomXL9Vy/rqZNm9IsW/YednZZG7rUnN4bMK88+enWu38LDQ3F29sbhULB7du3KV++fOYLmZi49e7lTJlFkRqG58m3jdPq5uGZDh5mTu8NmE8ek916V6xYMYoVK8ahQ4cy7dc6L7O/+rxv7bhGq96o0BcESqWC1as7sGLFH0ya1AhLyze/j1bIn/bs2UPHjh356aefXnjs3Llz9O/fX4ZUgpwkG2/ULWNxudwR66hAVMe9ULeKkztWgZRhse/ZsyebNm3C19f3pR1kXL58OVcCmpJl5J84XE3rhSu5bF9SS3WVOZH5OX/+b2rVKoxCoeCtt1yYNq2J3JEEM/Xo0SMA7t69K3MSwawoFMRW24TqeGEgrTMejXdnmUMVPBkexg8PD8fLy4snT568dMGiRYu+dH5uyu5hOecTPbEJ2Q+AulcYWNq98brM5VAP5EwWSZJYsOB3Zs8+y/jxDfnss9qy5slJ5pQnvx7Gv3DhArVq1SImJoaLFy/SsmVLuSOJw/gZyK0slrGXcTvf1DidWGYCSaW/ki1PVplLHpMNhPPsanx3d3fUajVFixbl5MmT/O9//8sfh/UlCYvYOwDE11ucrUKf3xgMEhMnBjJ79lkUCsQY9MJrWbBgAYsWLQIgJSWFZcuWsWTJEplTCXLTufhisHp+fZTDg+mojjhjEX9DxlQFR6YnXseOHcuxY8e4evUqK1asoHDhwkycODE3spmU9d9HsIy7j97Om5QyveWOYza0Wj3Dhx9k2bIrWFkpWb7cj759RQ+CQtYdO3aMVatWAWlD165fv54DBw7InEowB5FNg1C3iiOm+g7jPPff6qE64owy+ZGMyfK/TIt9SEgIX375JSdOnKBz5858+umnxMTE5EI0EzLocTmWdn5e81YHUMo/qI85SErS0q/fXrZvv4W9vRUbNnSmQ4d35I4l5DFarRYrq+d3alhZWeWPo4FCjtF6tkLdPIzYKmuN8zxOV8Yy5ncZU+VvmRZ7nU4HwOnTp6lbty56vT7Pd5DhHNjT+HdKqW4yJjEv48ef4MiRv3Bzs2Xnzq6iZzzhjfj6+vLll19y7tw5fvvtN8aOHUvVqmJ4aOE/LOzQeHdC3SoOnWNlANwutIKwEzIHy58yLfbVq1enXbt2pKSk4OvrS79+/ahfv35uZDMJZdLf2Dw+CEBKya7ovOrKnMh8jB5dj9q1i7BvXw98fQvLHUfIoyZOnIinpyezZs1izpw5eHh4MH78eLljCWYsut4ZtK7/tMXHmuN6viXKpAfyhspnMh3PXq/Xc+XKFcqXL4+joyOBgYE0btwYpVL++6zf5Opa52NdsXlyGMh8jPrXYS5XbMLrZVGrk/D0tDMeZn12a6VceXKDOeXJr1fjA8TGxppF51vPiKvxX86ssvzT494zcZWWkVr4fZnSpDGX98dkV+M/Y2FhQXh4ODNmzGDUqFFER0ebRaF/EzZBO7F5chgJBVEdLxb4Mepv3FDTrNk6vvnmrHGeOLcqZNfDhw9p164dfn5+hIWF0bZtWx48EHtpQubUreKgl0RK4bTRRp2vD8bm6RaZU+UPmVbtlStXsnTpUsqVK4ePjw+rV6/Os7fRWD/aC4CmaCv0LgX7wrPff39Cx45bCQ9P5Pz5v9Fq9XJHEvKJ6dOnM378eDw8PPD29uaDDz5g0qRJcscS8pD4Sj8S57MUAOfrg3C8NRIk0UZlR6bFfvfu3WzYsIF+/frRv39/1q9fz759+3IjW45SaGKxfbQTgETfqTKnkdfRow/p3n0HcXGptGtXlo0bO2NlVbCPcgg5JyYmhgYNGhine/fuTUJCgoyJhLwotUhPkosOAMDu8UpUR91AnyhzqrwrS8fjHR2fnydwcnLC0jLv3armeO4z4996Nx8Zk8hr+/Zb9Omzl+RkHb17V2LFCj9sbfPe/6dg3lJTU42nhNRqNQaDQeZEQl6UUPG7tMFz/qE6XljcnveGMi32RYsWZc2aNcYxqlevXk2RIkVyI1vOkSSUyaEA6B1LyptFRjt23GLYsF/Q6QwMH16T+fNbiQFthBzXs2dPBg4cSGRkJN9++y09evSgZ8+emS8oCC9jYYu6Zaxx0u1CK5yviHFMXlemu3RTp05l1KhRzJmTNmBM1apVmTdvnsmD5STrx79gHZ52EVpss40yp5FPw4ZvUaqUKx9+WJnhw2vJHUfIp7p160bJkiUJDAxEp9MREBCQ7rC+ILw2hQJ1qzgc7k7E/tFCbCIO43auPtH1zma+rABkodh7e3uzbt06kpOTMRgMODg45EauHGV/9RsAEnynoXerJHOa3GUwSCgUaVfZe3s7cPz4hzg4ZG0cekF4E3379mXNmjXUqiV+UAo5K/GdADSebXC91A7LhOtYh+9H4/We3LHyhAyP4QYFBeHv74+vry/Dhg0jKSkpTxZ6q79PYBV5BYCUsh/KnCZ3aTR6hg49wJQpJ3nWnYIo9IKpxcfH5/leNgXzpXVvSFLJLwBw+bMnNn9vgFd3FyPwimI/bdo0OnfuzLZt2yhRooTxMP7r2LdvH+3ataN169Zs2LAhw+cFBgbSvHnz115/VrgcT+sO12DjjmTrYZJtmKPERC0ffribXbvusHbtVR4/lr9TCKFgsLOzo1mzZvTp04chQ4YY/2XGHNoLIW9IfHsKKd6dAXC+MRTVURcso8/JnMq8ZXgYPyIigg8++ACAUaNG0bFjx9dacVhYGAsWLGDnzp1YW1vz/vvvU6dOHcqWLfvCdr755ps3iJ4F+lQUBg0A8Q2WmmYbZigqKplu3bZz8eJTPDzs2Ly5C8WLO2e+oCBk0927d2nRogUNGzakUKFCWV7OLNoLIU+Jr7KGlMg+uF5OK/puF9sAoG4eChb2ckYzSxnu2f/79joLC4vXvt3u7Nmz1K1bF1dXV+zt7WnTpg0HDx584XkTJkxg+PDhr7XurHL8/Uvj35pibUyyDXPz9Gk8jRv/xMWLTylWzIl9+3pQtaq33LGEAmDHjh188MEH/PLLL6xevRpPT086d+5s/Pcq5tBeCHmP1qMF6lZxxFZ73sue6ngh3M744nBvCuhE/w7PZFjB/9tl/ut2oxoeHo5KpTJOe3l5cfXq1XTPWbt2LRUrVnzjEbFe2U9w1B24/8/wiXUnolI5vdE2XkdubONVHj6MpmPHbQQFxVChgieHD39IsWLmsUcv93vzX+aWJz9Yt24d+/btw9vbmytXrrBgwQIaNWqUpWVlby9kYE6fQXPKAm+QR9Ud3mkP1ybBncVYJt3HMmg+9kHzodYSKPE+WLvlXh4zlGGxDw0NZfr06RlOT5gw4ZUrNhgM6X4g/HeAlbt373L48GFWr15NaGjoG4V/1cAWLkeGYv3P3+qyX4CJBzIwh8ESNBotVlZKatcuytq1HbCxUcieCczjvfk3c8qT3wbC8fZOO4pUvXp1oqOjs7yc3O1FbjOnz6A5ZYFs5ik2Ke2fPgnV8X9OI10YBheGYbBwJKrhNSTr17t2y1zen+y2FRkW+969e79yOjOFChXi4sWLxmm1Wo2Xl5dx+uDBg6jVavz9/dFqtYSHh9OrVy82bsyB++AlCeunaWMip5T0LzAD3nh42LFjR1dKlfIgOTlV7jhCAfPfo38WFln/3snaXgj5j4U96lZxWMb9iV3Qd9iG7UCpT8Dz11JmMZKeHDIs9tk9L1a/fn0WL15MVFQUdnZ2HD58mICAAOPjI0aMYMSIEQA8fvyYPn365NgX1zLykvHvhLoLcmSd5urgwQecORPCtGlNUCgUFCrkiKOjtSj2guxe59SfnO2FkH/pnKsSX+Un4qUV2D+cg8PDWThfH0xCaijJJT+XO16uMlmn6N7e3owcOZI+ffqg1Wrp2rUrVapUYdCgQYwYMYLKlSubatO4HXh+W45k7Wqy7chty5abfP75IfR6iQYNivPuu2XkjiQUYHfu3MHX19c4nZKSgq+vr/GQ/OXLlzNcVs72QigAFBYklRmLZOGI473xON6bhP2j74lscl/uZLlGIf33Srw85GXn4Kz+PoHr0bTbBOPrfkfKOwNyJUtun9f58cdLTJr0KwAjR9ZhzJj6xj0pcznH9IzIk7H8dM7+yZMnr3y8aNGiuZTk5cQ5+5czpyxg+jyK1DA8T74NQFTds+idXt2rqrm8PyY7Z59X2d1ZBkBShWG5VuhzkyRJzJp1hu++Ow/AtGlNGDKkhsypBEH+Yi4IWSHZeJNU8kvsg77F/bf66G2LE9XwKijy97VdmQ55ZjAYWLFiBV9//TUJCQksXboUvV6fG9neiE3IfgB0Hr6ZPDPv0esNjBp1lO++O4+FhYLFi98VhV4QBOE1Jb492XhvvkVKCKqjbiCZb13LCZkW+zlz5nD37l3jPa+nTp1i1qxZJg/2Jixi7hj/Ti3ZRcYkppGUpOXy5VBsbS1YvboDPXpUlDuSIAhCnqRRtUXdIsI4rTrqhv3Db/JtP/uZFvtz584xe/ZsbGxscHR0ZNWqVZw5cyY3sr02i7i7zyeU+e4MBU5ONmze3IVt27rSpo24GE8QBCFblNaoW8YYJx0ezEB11AXPY14otLHy5TKBTIu9paUlSuXzp1lbW79217m5xUqddh47pVR3mZPknMjIZBYtOm/s0dDb24E6dcS5UUEQhByhUKJuFUdUvfOkeKcdEVYYUvAMLI5l7KVMFs47Mq3a77zzDhs2bECv1/Pw4UNWr15N+fLlcyPba7MKOw2Awb6wzElyxpMn8XTvvoN796IAGDGitsyJBEEQ8ie9Y3niq6wmXlqF6qgrAG7nm8F5cCg5kqSSXyBZucgbMhsy3bMfP348N27cIDIykp49e5KYmMi4ceNyI9trUWhisYpI+xWmKdRE5jTZd+9eFH5+m7l3L4oKFTzo3l2cnxcEQTC5f/b04yqvxmCR1ie+fdACPAOLY/9gVp49p5/pnr2joyMzZ87MjSzZYvvgeW9a2qItZUySfX/8EUrPnruIjEymVq0ibNjQCVdXW7ljCYIgFBiphbqQWqgLKg97Un7tg+3TjTg8nIXDw1lE1z6OzrkGvOYAcXLKtNj/e/Cbf8tsIJzcZvU0rYMZraqOzEmy59SpYPr02UNiopbmzUuycmV7HBys5I4lCIJQMCktiK/0IwlvTzV2xuN2vjkGK3diq21F51IrTxT9TA/ju7q6Gv85ODhw/vz53Mj12hT6ZABS32ovc5I3J0kSc+acIzFRS5cu5Vi7tqMo9IIgCGZAsvFG3TLWeH++UhuF24WWqI664HjjE5nTZS7TPfv/DogzaNAghg4darJAb+rZKHeaIi1kTvLmFAoFq1d3YO3aq3z2WW2USvP/tSgIglBgKBRp9+e3jMUy9jzOf36AhSYMu7/XYff3OhJLjUbrWg+tp/mdSs50z/6/HB0dCQ8PN0WWN6dPef6nU2kZg7yZgwcfGPvs9vCwY+TIOqLQC4IgmCuFAp1rHaKa3COiyUPjbIe/5uJ6pQtuZ6qj0EbLGPBFme7ZBwQEGAdYkSSJGzduULq0eRVUK/WF5xOWdvIFeU2SJDFt2in+97+LfPRRNWbObJ75QoIgCILZkKw9UbeKA0mPVVQgrpc7Y5n0AM/AEgCkFH6feJ8fQfHa+9Y5KtNi7+bmlm66Q4cOdOjQwWSB3oQy8TEAkkXeKfQ6nYFRo46wceMNLC2V1KiRP/oGEARBKJAUFmg9WqBuFYfN0804Xx8MgO3Tzdg83Upk07+QrNwyWYnpZFrsg4ODmTNnTm5keWOWMbcB0LmaZ2c//5WSouPjj/fzyy8PsLOzZNWq9rRoUUruWIIgCEIOSC38PurC74MuAdWJIigw4BlYAo1HC2Kr75BlLz/TLd6+fRtzH/LeIv4BAHpH8y+Y8fGp9Oy5k19+eYCLiw3btnUVhV4QBCE/snRE3TKW5GIDAbCOPIbqqCuqI85Yq3/J3SiZPUGlUvHee+9RtWpVHBwcjPPN6T57g7UrAHrXcvIGyYJp005x5sxjvL0d2LKlCxUrquSOJAiCIJiKQkFChQUklpmAXfD/sA/6DoWkw+WPHgAkFf+YxHJzTH6vfobFXqPRYG1tTfXq1alevbpJQ2SX3f21QN64En/ChIao1UlMmdKYkiVd5Y4jCIIg5ALJ2oOkspNIKjsJi4Q7uPzRDYvkIOxDlmIfspTE0mNJLj4YydrDJNvPsNj36NGDXbt2vXCfvTmSLGxR6FMw2HrKHeWlHj2KpWhRJywtlbi42LJ6tXld4CgIgiDkHr1jOaIaXkWhicTlij9WcZeNXfHqbUsQW20TeqdKObrNDM/Zm/t5eiNJQvHPffY6lfmNCnfp0lNat97AqFFH8s57KgiCIJicZO1BTJ1AIpoGkfTWMAAsUh7h/lt9lClPc3RbGe7Zp6amcvPmzQwLlI+PT44GeVPPLs4DkKycZEzyosDAR/Trt5ekJC0REcloNHpsbDK9TEIQBEEoQCQrdxLLzSax3GxUR5wB8DhVDnWLKFDmTM3IcC0hISF8+umnLy32CoWCY8eO5UiA7LIM/x0AjXcDmZOkt3fvXYYOPYBWa6Bbtwp8911rrKws5I4lCIIgmDF1qzhjwVcdcyeq7tkcOaSfYbEvW7Ysu3fvzvYGTM0y7h4AOveqMid5bs2aq3z11VEkCT7+2JepU5uI7m8FQRCELFG3iER1LO1CPfff6qf10JdN8vbflwMsYu8CoHcuK3OSNDt23GL06LRCP3ZsA6ZNE4VeEARBeA1Kq3QF3uVCW5AM2VtlRg/UrFkzWysG2LdvH+3ataN169Zs2LDhhcePHj1Kx44d6dChA8OGDSM2Nva1t2GREASYz213bdqUwde3EHPmtGDkyDrGcQUEQXi13GgvBCEvia6ddrrcOuYMHmeqZGtdGRb77HaaExYWxoIFC9i4cSO7d+9my5Yt3L9/3/h4QkICU6ZMYdmyZezdu5dy5cqxePHi19uIJGEZfR0AvWvFbOXNDq1Wj1arB8DR0Zqff36ffv3M57SCIJi7XGkvBCGP0bnUIqJpUI6sy2SH8c+ePUvdunVxdXXF3t6eNm3acPDgQePjWq2WyZMn4+3tDUC5cuV4+vQ1bzXQJxv/NNh550ju15WcrKV//33067fHOEytpWWePzsiCLkqV9oLQciDJCv3tKvys8lk94GFh4ejUj3vCtbLy4urV68ap93c3GjVqhUAKSkpLFu2jA8//PC1tuFhm5j2h703Ki/n7Id+TTExKfj7b+fUqWDc3e0ICGhG6dLyjWr0byqVed2GKPIIr5Ir7YWHY86EzSHm9Bk0pywg8rxUx6BsLW6yYm8wGNKdr5Yk6aXnr+Pj4/nkk08oX748nTt3fq1txAVfxRnQ2r9FjDo+u5FfS1hYIu+/v5MbN9QULuzI0aN9cHKyRJ3LOV5GpXIyixzPiDwZUyoVZleE5JAb7UVkZILx6JvczOkzaE5ZQOTJSHbbCpMdby5UqBBqtdo4rVar8fLySvec8PBwevXqRbly5ZgxY8Zrb+PZbXeShW32wr6moKAY2rffzI0basqUcePnn98XA9oIQjbkRnshCAWZyYp9/fr1OXfuHFFRUSQnJ3P48GEaN25sfFyv1zNkyBDatm3L+PHj3+iqdYUuKW1duXjb3cOH0bRvv4WgoFiqVPFi794eFC+e+6cQBCE/yY32QhAKMpMdxvf29mbkyJH06dMHrVZL165dqVKlCoMGDWLEiBGEhoZy8+ZN9Ho9hw4dAqBSpUqv9YtdmRwGgFZVyySv4WUKFXLkrbdcKFvWjbVrO+LkZJNr2xaE/Co32gtBKMgUUh4enUWzqyvWD3cQ02In2qItc227sbEp2NhYYmv7/LeSuZzXAfPKAiLPq4hz9rlHnLN/OXPKAiJPRsz2nH1usI64CJj+trudO2/z8cf70evTejBycbFNV+gFQRAEwZzl6WL/jN6lnMnWvXLlHwwdeoBdu+5w4MD9zBcQBEEQBDOT53dPDVYuYGGd4+uVJIl5835j7txzAEyc2Ij27d/J8e0IgiAIgqnl6WKvtXTmUdVpJIc+ytH1SlJahzlNmzrStGkr3NzscHS0IvQV2wkPV2IwZG+ggpxiTllA5AGwtLTGzU2FhUWe/soJgpBH5emWJ6Tkp1i7l8HZ+60cuxXHYJAICYklNdUGa2sFJUq44Oqa+X38lpZKdDrzKGjmlAVEHkmSSEyMIzpajadn4VzbriAIwjN5+px9st1b2Du45vg9t1qtAaVSQenSrlkq9ILwKgqFAgcHZ3Q6jdxRBEEooPL0nj0KBQoLS3LyZhqlUkGpUq5oNHrs7KxycM1CQSY6gREEQU55es8eQMqBl6DV6nnyJN54D66FhVIUekEQBCHfyNt79gDK7BXl1FQdDx5Eo9HoUSqhcOE3G93o77//pnv3TpQsWRoASTKQmJhI27Z+DBz4MQBJSUn88MNizp8/h62tHQ4ODgwYMJiaNWsb13P27GnWrVtFUlIyBoOexo2bMXDgxyiV5vO77PTpX3n69Cndur0vdxQjrVbL7NnTuH37FjY2NkyePIMSJUqme85PPy3n119PoFCkXYQZHPyIjz4aQq9eH3Lo0AHWrVsNQN269Rk+/HMADhzYx8aNa1Eqlfj61mL48M+xtLTk1KlAVq5chiRJFClShLFjJ+Ps7ExAwCSGDBmOSpW+X3dBEAQ5mU8FeUOSpd0bL5uUpOXevSg0Gj329laoVA7ZyuLpqWL16o2sXr2RNWs28+OPq9i0aR1BQX8hSRJffz0SKytL1q/fxpo1m/j881EEBEzi8uW0zoF+++0sCxbMYezYyaxZs4kVK9Zy//5dVq5cmq1cOUmj0bB+/Ro6d+4qd5R0tm3bjK2tHRs2bGfEiC+ZMWPKC8/p338Qq1dvZN26zQwf/jklSpTA3787KSkpfPfdPL7/fhmrV2/kzz+vcOHC7wQHB7F8+Q98990PrF27BZ1Ox/btm0lMTGDevNnMnfsda9ZsokyZt1m1ahkAH3zQj0WL5ufyqxcEQXi1vL9n/4a/VxISNPz1Vwx6vQEnJxtKlnTBwiJnf/tEREQgSRL29vZcuXKJ0NBQFi360Xj+9p13ytO370DWrFmJr29N1q5dRZ8+A3jrrRIA2NjY8uWXY3j0KOiFdd+7d4c5c2aSmpqCs7MLkyYF8PhxCKtWLePHH1cAMGPGFKpXr0H16jX48stPcXFxxcbGhri4WL76agLly1dAr9fTtWt7Vq1aT2joUxYtmk9qagouLq6MHj2OIkWKptvu4cO/ULVqdSwt0z46S5f+j0uXLhAXF4enpyfTps3C3d0DP7+WlCtXkcjICFavXse6des4ceIIer2BOnXqMnToCBQKRYbLPxMWFsrXX3/xwutfsmQ59vbPf5ydO3eajz4aAkC1ar7ExEQTGhpKoUKFXlhWq9Uyb95sJk6cho2NDUlJiUiSgZSUZOzsbNHrddjY2HD//n18fCrj6ekJQIMGDVm/fg1t2/rxxRdfG/fey5Qpy+HDBwEoVao0oaF/8+TJY4oWLZaFT4kgCILp5f1ir0x7Cc7HumLz5HCWF1MBpf4940zGz00t2pq4FtszXWdEhJp+/Xqh0aQSGxtD+fI+zJw5Dy8vb44ePUT58hVeuFCrWrXq/Pjj90BaAf/ss1HpHvfy8sbL68XugKdOncjQoZ/SoEEjdu3azrZtm6lXr0GG2YKDH7Ft22IKFy7Cli0bjHkuX75A2bJv4+joxOzZn/DNNwsoVKgQv/9+jm++mcHChUvSrefMmZN06JA2jvjjxyEEBwfx44+rUCqVBARM4tChX+jZ8wNiYmLo3bsPvr41uXDhHHfu3GL58rUoFAoCAiZx+PAv+PhUznD5Z7y9C7F69cYsvfceHp7GaQ8PT9TqsJcW+wMHfqZs2bL4+FQCwN7e4Z/D+V2xtbWlWjVfKleuSkhIMN9/v4CwsFA8PVWcOHGMyMhIXFxcadKkGQCpqSmsX7+Grl17GNdfpUo1zpw5RffuPTPNLQiCkBvyfrFXmM+ZiGeH8Q0GA99/v4CgoL+oVavOP48q0Ov1Lyyj1ep4Vv8VCiXW1pn3BhgTE0NkZAQNGjQCMB5Sf3Y64GXc3NwpXLgIAC1btmHIkAF88slnHDlyiNat2xIS8oi//37MmDHP96ITExNfWE9ISAgqVdqPj2LFijN8+Ej27dtNcPAjbty4lm5v9lkxvXDhPDdvXmfgwA+BtALp7V2INm3avXJ5yPqevSRJ//khJaHI4LOxe/eOdD+q7t+/x/79e9mxYx8ODo5MmzaRTZvW0atXH4YMGc6YMV9gY2NL8+YtuXXrhnG5hIQExo0bRdmyb9O2rZ9xvrd3YR4/Dn7ptgVBEOSQx4u90ljss7Ln/YxebyA6OgUPDzuT3BKlVCoZNuwz+vfvxaZN6+jduy8+PpXYvn0zOp3OeAgc4Pr1q5QvXxGA8uUrcPv2TUqVKm18PDj4EWvWrGTixGnGeZaWlulyp6amEhGhfuG16HQ64982Ns+H4vXw8KR48RJcuXKJixfP88UXXxMSEkyRIkWNe9F6vZ7o6KgXXptCgTH/7du3mDJlPO+/34tmzVpgYaHk34Mo2tik9VFgMOjp3r0n77+ftsceHx+PhYVFpstD1vfsVSovIiIijD8WIiMj8fRUvfA8tTqc6OhoKlWqYpx3/vw5atSojZubOwDt2rVn165t+Pv3oEIFH376KW37x48fNa4/IiKCL78cjq9vLUaMSP9jxMLCwqwuqBQEQcjbLVIW9+olSSIyMjndrXWenvYmvffZ0tKSTz75nNWrVxIZGUHVqtUpVaoMCxd+ayzCt2/fYu3alfTtOxCAXr368NNPywkJSdsrTEpK4vvvF+Dtnf5QtKOjIyqVF+fP/wbAoUMHWLlyKS4urvz99xNSU1OJi4vlzz+vZJjv3Xfb8f33C/D1rYmtrS0lSpQkLi7OuMz+/XuZMmX8C8sVK1acp0//BuCPPy5RvXoNOnXqSvHib3H27OmXdkNbo0YtDh06QFJSEjqdjrFjvyQw8FiWl8+KevUacPDgfgD+/PMPrK1tXnoI//r1q1StWi3dvLJl3+HixfMkJycjSRJnzpykfPmKpKQk89lnQ0lKSkSr1bJjxxaaN2+NXq/n669H0qxZSz777MsXPkehoU8pWrT4G70OQRAEU8jTe/ZSFoq9JEk8fhxPZGQSCQmplCjhavpg/6hbtz6VKlVmxYof+frrCcycOYdly5bw4YfdsbCwwMnJmYkTA/D1rWl8/uDBw5g8eSx6vQG9XkezZi3p33/QC+ueNCmAefNmsWTJIlxcXJk4cRqenp7Uq9eAXr26UahQYapWrZ5htsaNmzF37iyGDv0UAGtrawICZrNw4Tw0Gg329g5MmDD1heUaNGjElSsXqVevAS1atGbcuNH06ZN2vrpcuQrGHwL/1qhRE+7cucPgwf0wGPTUqVOftm39iIhQZ2n5rPD378HcuTP54IPuWFtbGY+E3L59kxUrfmTevEUA/P33kxeugahduy53795m4MAPsLS0pEIFHz74oB82NjYMGDCIwYP7o9PpaNWqDa1bv8uvv57g7t3b6PV6AgOPA2lHZcaMmQik/QiaOnXWG70OQRAEU1BI/z1umodcvxCId8kaGT5uMEgEB8cSE5OCUpnWz72Li2m6vzWn/t9NmSU1NZVhwz5i6dKf0p2OkCvPmzBlnnv37rJmzUqmT//mhcdCQx9RqFCJdPOUSgUeHo4mySKkFxmZYDy6JzeVygm1Ol7uGIB5ZQGRJyPZbSvy9mF8Mj4Mr9cb+OuvGGJiUrCwUFK6tJvJCn1BYmNjQ58+A9i5c5vcUczSxo1rGT58pNwxBEEQ0snTh/FRWrx0tk5n4OHDaJKStFhaphV6e3vR/W1OeXbbmfCiyZOnyx1BEAThBXm72GcgLCyRpCQt1tYWlCnjho1NvnyZgiAIgpAlebsKKq1fcn81FC7siMEgUaiQA1ZWL9/7F4TclIcvjREEIR/I0+fsrUklMTEOSZJITtai16dddKVUKihe3FkUesEsSJJEYmIclpaZd5gkCIJgCnl6z97TViIsJYmIiEgiIpKwtbXAw8MeOYYOVyqVb3yPeE4zpywg8gBYWlrj5vZiJz+CIAi5IU8XewtrG86fTuTjj/eTmqqnc+dyLF78LtbWub9Hby63Z4B5ZQGRRxAEQW4mPYy/b98+2rVrR+vWrdmwYcMLj9+6dYsuXbrQpk0bxo8fn65716zYdSSeAQP2kZqqZ8CAqvzwQztZCr0gCNln6vZCEAoykxX7sLAwFixYwMaNG9m9ezdbtmzh/v376Z4zevRoJk2axKFDh5Akia1bt77WNqbMvo7BIDFqVF1mzWqOUinD8XtBELItN9oLQSjITHYY/+zZs9StWxdXV1cA2rRpw8GDBxk+fDgAT548ISUlhWrVqgHQpUsXFi1aRK9evbK8jWLFnPnqq3r06OGT0/HfiDn92DCnLCDyZMRccsgtN9oLc3uvzSmPOWUBkccUGUxW7MPDw1Gpnl+Q5OXlxdWrVzN8XKVSERYW9lrbOH16QPaD5iBz6vbUnLKAyCO8Wm60F25uDpk/KReZ02fQnLKAyGMKJjuMbzAY0t3//t/74TN7XBCEgkO0F4JgWiYr9oUKFUKtVhun1Wo1Xl5eGT4eERGR7nFBEAoO0V4IgmmZrNjXr1+fc+fOERUVRXJyMocPH6Zx48bGx4sWLYqNjQ2XLl0CYM+ePekeFwSh4BDthSCYlkmHuN23bx9Lly5Fq9XStWtXBg0axKBBgxgxYgSVK1fm9u3bTJgwgYSEBHx8fJg1axbW1qKXMUEoiER7IQimk6fHsxcEQRAEIXN5um98QRAEQRAyJ4q9IAiCIORzotgLgiAIQj4nir0gCIIg5HOi2AuCIAhCPmf2xd7cRsLKLM/Ro0fp2LEjHTp0YNiwYcTGxsqW5ZnAwECaN29ushxZzfPw4UM+/PBDOnTowMCBA0363mQlz40bN/D396dDhw58/PHHxMXFmTRPQkICfn5+PH78+IXHxIhuOcOc2gtzaiuykucZ0V7k0/ZCMmOhoaFSs2bNpOjoaCkxMVFq3769dO/evXTPee+996QrV65IkiRJY8eOlTZs2CBbnvj4eKlBgwZSaGioJEmS9N1330kBAQGyZHlGrVZL7777rtSsWTOT5MhqHoPBILVu3Vr69ddfJUmSpLlz50pz5syRLY8kSVLPnj2lwMBASZIkadasWdL8+fNNluePP/6Q/Pz8JB8fHykkJOSFx3Pzc5xfmVN7YU5tRVbyPCPai/zbXpj1nv2/R8Kyt7c3joT1zMtGwvr347mdR6vVMnnyZLy9vQEoV64cT58+lSXLMxMmTDCOHGZKmeW5ceMG9vb2xl7PhgwZQu/evWXLA2n9rScmJgKQnJyMra2tyfJs3bqVyZMnv7SL19z+HOdX5tRemFNbkZU8z4j2Iv+2F2Zd7F82Eta/R7rKiZGwcjKPm5sbrVq1AiAlJYVly5bRsmVLWbIArF27looVK1K1alWTZHidPMHBwXh6ejJu3Dg6d+7M5MmTsbe3ly0PwJgxY5gwYQINGzbk7NmzvP/++ybLM2PGDGrWrJmlrKb+HOdX5tRemFNbkZU8INqL/N5emHWxN7eRsLK6vfj4eAYPHkz58uXp3LmzLFnu3r3L4cOHGTZsmEm2/7p5dDod58+fp2fPnuzatYvixYsze/Zs2fKkpKQwfvx4Vq9ezenTp+nVqxdff/21yfK8ihjRLWeYU3thTm1FVvKI9iL/txdmXezNbSSszPJA2q+uXr16Ua5cOWbMmCFbloMHD6JWq/H392fw4MHGXHLlUalUlChRgsqVKwPg5+eXbrzy3M5z9+5dbGxsqFKlCgA9evTg/PnzJsvzKmJEt5xhTu2FObUVWckj2ov8316YdbE3t5GwMsuj1+sZMmQIbdu2Zfz48SbdO8ssy4gRIzh06BB79uxh2bJleHl5sXHjRtnyVK9enaioKG7fvg3A8ePH8fHxkS1PiRIlCA0N5eHDhwAcO3bM2LDkNjGiW84wp/bCnNqKrOQR7UUBaC+yf92gae3du1d67733pNatW0vLli2TJEmSPvroI+nq1auSJEnSrVu3JH9/f6lNmzbSF198IaWmpsqW5/Dhw1K5cuWkDh06GP+NGzdOliz/FhISYvKra7OS548//pD8/f2ldu3aSQMGDJAiIiJkzRMYGCi1b99e8vPzk/r27SsFBwebNI8kSVKzZs2MV9fK+TnOr8ypvTCntiKzPP8m2ov82V6IUe8EQRAEIZ8z68P4giAIgiBknyj2giAIgpDPiWIvCIIgCPmcKPaCIAiCkM+JYi8IgiAI+Zyl3AHyo3LlyvHOO++gVD7/LVWpUqVXdpyxc+dODh06xNKlS7O9/cWLF7Nhwwa8vb1RKBTo9Xo8PDyYPHkypUqVeu31hYWF8dlnn7F582ZCQkKYM2cOixcvTjc/ux4/fkyrVq145513jPOSkpIoVKgQM2fOpHjx4q9c/vvvv6d8+fIm7XJUEPK7f7ddCoWC5ORkHB0dmTJlSo7fV/748WPat2/PlStXWLx4MdHR0UyaNClHtyE8J4q9iaxZswZ3d3fZtt+uXbt0X5x169bx5ZdfsnPnztdel7e3t7Gg//333/z1118vzM8Jtra27NmzxzgtSRLTp09nwYIFzJ8//5XL/v7775QtWzbHsghCQfXftmvlypVMnz6dLVu2yJhKyC5xGD+Xbd++nW7dutGpUyeaNWv20l6qDh8+TOfOnenSpQvdunXjwoULQFo/2mPGjKFLly60b9+emTNnZnk87nr16hmLdGhoKEOGDKF9+/b4+fmxYsUKIK0/6smTJ9O+fXu6dOnCiBEjSExM5PHjx1SvXh29Xs+ECRMIDg5m4MCB6eY3adKE69evG7f3+eefG1/bDz/8QOfOnenYsSPDhg3L8uAjqamphIeH4+LiAsBff/1F//796d69O82aNWPo0KGkpqayYcMGrl+/zpw5czhy5AgajYaZM2fSuXNnOnTowJgxY0hISMjSNgVBeE6n0/H06VPjdxAy/j6r1WqGDRvGu+++S7t27Vi7di0Af/zxB71796Zbt240bdqUcePGyfJaCjqxZ28iffv2TXcYf9WqVdja2rJt2zaWLVuGm5sbf/zxB/3793+hD+o5c+Ywb948qlWrxunTp/n999+pVasWM2fOxMfHh9mzZ6PX6xkzZgw//fQTgwYNemUWnU7H9u3bqVOnDgCjRo2iRYsW9O/fn/j4eHr37k3hwoXx9vbm/PnzHDhwAIVCwdy5c7lz546x32ULCwumT59OQEAAK1eu5PHjx8b5/v7+7Ny5k0qVKhEbG8u5c+cICAhg9+7d3L17l23btmFpacmWLVuYMGECy5cvfyFnSkoKHTt2xGAwEBkZiYuLC61bt2bw4MFA2rCPnTp1omPHjmi1Wrp06UJgYCC9e/fm4MGD9O7dm1atWvH9999jYWHBzp07USgUzJ8/n3nz5jFlypQ3/v8UhIKib9++AERHR2NjY0OzZs2YNWsWwCu/z1OnTqVkyZIsWbKE+Ph4evbsSZMmTVi7di0jRoygTp06JCYm0qJFC65fv46rq6uMr7LgEcXeRDI6jP/jjz/y66+/EhQUxO3bt0lKSnrhOe+99x7Dhw+nSZMmNGjQwFjMAwMDuXbtGtu3bwfSimNGDhw4YOw7WavV4uPjQ0BAAElJSVy+fJlVq1YB4OTkRJcuXTh58iTjx4/HwsKCbt260bBhQ9q0aUOVKlWMRf1V/P396dq1K2PGjOHnn3+mefPmODk5ceLECa5du4a/vz+QNmJTcnLyS9fx78P4p06dYvTo0TRr1gwHBwcARo8ezZkzZ1i+fDlBQUGEh4e/9P0LDAwkPj6es2fPGl+/h4dHpq9BEITnbdeNGzcYPHgwderUMX5/XvV9Pnv2LKNHjwbS2pWff/4ZgNmzZ3Py5El+/PFHHj58SGpqKklJSaLY5zJR7HNRaGgoPXr0oHv37tSoUYN3332XEydOvPC8kSNH4u/vz5kzZ9i5cyerVq1i+/btGAwGFi5cSJkyZQCIi4vLcACN/56zfyYhIYH/9pBsMBjQ6XQ4OzuzZ88eLl++zG+//cbnn3/OwIEDadKkSaavrWjRolSsWJHAwEB27txpPFRnMBj46KOPjEcvNBoNsbGxma6vUaNG9O/fn88++4z9+/fj6OjIF198gV6vp23btjRt2pSnT5++8FqebXPcuHHG3ImJiaSmpma6TUEQnvPx8WHs2LGMGTOGChUqUKxYsVd+ny0tLdO1RyEhIbi5uTFgwADKlStHo0aNaNu2LX/++edLv7eCaYlz9rno+vXruLu7M2zYMBo2bGgs9Hq93vgcnU5H8+bNSU5OpmfPnkyePJk7d+6g0Who2LAhq1evRpIkNBoNQ4cOZf369a+VwdHRkapVq7JhwwYg7TqA3bt3U79+fU6cOEG/fv2oXr06n376KZ06dUp3Hh7SDtlrtdqXrrt79+4sX76c5ORkatSoAUDDhg3Zvn278Zz5woUL+eqrr7KUdcCAATg4OLBo0SIATp8+zSeffEK7du0A+PPPP43vnYWFhfH6hYYNG7JhwwY0Gg0Gg4GJEydmeoGfIAgv8vPzo0qVKsbD+K/6PterV48dO3YAae1K3759CQoK4tq1a4waNYrWrVsTGhpKcHAwBoNBnhdUgIk9+1zUoEEDtm/fzrvvvotCoaB27dq4u7vz6NEj43MsLS0ZN24co0aNMv5SnjlzJtbW1owfP54ZM2bQvn17tFot9evX56OPPnrtHPPmzWPatGns3LkTjUZjvCDPYDBw8uRJ/Pz8sLe3x8XFhYCAgHTLli1bFhsbG7p27cqCBQvSPda8eXOmTp2a7hqCbt26ERYWRvfu3VEoFBQuXJjZs2dnKaeVlRUTJ07ko48+omvXrowcOZJPPvkEe3t7HB0dqVWrFsHBwcZtz58/H61Wy7Bhw/jmm2/o3Lkzer2eChUqMGbMmNd+nwRBgIkTJ9KhQwdOnTr1yu/zpEmTmDJlCu3bt0eSJD7++GMqVarE4MGD6dy5M/b29nh7e+Pr68ujR48yvZ1WyFli1DtBEARByOfEYXxBEARByOdEsRcEQRCEfE4Ue0EQBEHI50SxFwRBEIR8ThR7QRAEQcjnRLEXBEEQhHxOFHtBEARByOf+D4xhS8cF7vx8AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 576x216 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "probs = model.predict_proba(x_test)[:,1]\n",
    "sns.set(font_scale = 1)\n",
    "plot_roc_pre_recall_curve(y_test, probs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 7. Model Production"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "#prepare train & test data\n",
    "x_train = train_woe[features_use]\n",
    "y_train=train_woe['label']\n",
    "x_test =test_woe[features_use] \n",
    "y_test = test_woe['label']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(class_weight='balanced')"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Train LR\n",
    "#lr = LogisticRegression(random_state=42,C= 0.1, penalty='l2', solver='newton-cg')\n",
    "lr = LogisticRegression(class_weight = 'balanced')\n",
    "lr.fit(x_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x216 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#check AUC\n",
    "probs = lr.predict_proba(x_test)[:,1]\n",
    "sns.set(font_scale = 1)\n",
    "plot_roc_pre_recall_curve(y_test, probs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 8. Scorecard Generation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ScoreCard(base_score=1000,\n",
       "          combiner=<toad.transform.Combiner object at 0x0000021422E76CA0>,\n",
       "          pdo=80,\n",
       "          transer=<toad.transform.WOETransformer object at 0x0000021423D2D640>)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# scorecard tuning\n",
    "card = toad.ScoreCard(\n",
    "    combiner = combiner,\n",
    "    transer = t,\n",
    "    class_weight = 'balanced',\n",
    "    C=0.1,\n",
    "    base_score = 1000,\n",
    "    base_odds = 35 ,\n",
    "    pdo = 80,\n",
    "    rate = 2\n",
    ")\n",
    "\n",
    "card.fit(train_woe[features_use], train_woe['label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "# save 5: save the model to disk\n",
    "filename = 'CreditScore_save5_ScoreCard.pkl'\n",
    "pickle.dump(card, open(filename, 'wb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    7501.000000\n",
       "mean      627.437143\n",
       "std       113.055983\n",
       "min       173.540655\n",
       "25%       596.974091\n",
       "50%       659.897345\n",
       "75%       702.107780\n",
       "max       810.644058\n",
       "Name: CreditScore, dtype: float64"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#inference on test data\n",
    "test['CreditScore'] = card.predict(test)\n",
    "test['CreditScore'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    30000.000000\n",
       "mean       617.559238\n",
       "std        118.265235\n",
       "min        168.851703\n",
       "25%        578.204047\n",
       "50%        654.175758\n",
       "75%        696.461008\n",
       "max        818.460370\n",
       "Name: CreditScore, dtype: float64"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#inference on whole data\n",
    "data['CreditScore'] = card.predict(data)\n",
    "data['CreditScore'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#output the scorecard\n",
    "final_card_score=card.export()\n",
    "len(final_card_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "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>variable</th>\n",
       "      <th>binning</th>\n",
       "      <th>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>EDUCATION</td>\n",
       "      <td>[-inf ~ 2)</td>\n",
       "      <td>29.66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>EDUCATION</td>\n",
       "      <td>[2 ~ inf)</td>\n",
       "      <td>27.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BILL_AMT4</td>\n",
       "      <td>[-inf ~ 2501.0)</td>\n",
       "      <td>28.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BILL_AMT4</td>\n",
       "      <td>[2501.0 ~ inf)</td>\n",
       "      <td>27.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BILL_AMT6</td>\n",
       "      <td>[-inf ~ 19195.0)</td>\n",
       "      <td>27.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BILL_AMT6</td>\n",
       "      <td>[19195.0 ~ 50442.0)</td>\n",
       "      <td>31.37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BILL_AMT6</td>\n",
       "      <td>[50442.0 ~ inf)</td>\n",
       "      <td>26.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BILL_AMT5</td>\n",
       "      <td>[-inf ~ 2395.0)</td>\n",
       "      <td>28.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BILL_AMT5</td>\n",
       "      <td>[2395.0 ~ 8110.0)</td>\n",
       "      <td>32.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BILL_AMT5</td>\n",
       "      <td>[8110.0 ~ 35288.0)</td>\n",
       "      <td>25.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BILL_AMT5</td>\n",
       "      <td>[35288.0 ~ inf)</td>\n",
       "      <td>29.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BILL_AMT1</td>\n",
       "      <td>[-inf ~ 2501.0)</td>\n",
       "      <td>25.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BILL_AMT1</td>\n",
       "      <td>[2501.0 ~ inf)</td>\n",
       "      <td>28.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PAY_AMT5</td>\n",
       "      <td>[-inf ~ 1.0)</td>\n",
       "      <td>23.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PAY_AMT5</td>\n",
       "      <td>[1.0 ~ 1880.0)</td>\n",
       "      <td>26.88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PAY_AMT5</td>\n",
       "      <td>[1880.0 ~ 10000.0)</td>\n",
       "      <td>30.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>PAY_AMT5</td>\n",
       "      <td>[10000.0 ~ inf)</td>\n",
       "      <td>37.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PAY_AMT1</td>\n",
       "      <td>[-inf ~ 6.0)</td>\n",
       "      <td>11.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PAY_AMT1</td>\n",
       "      <td>[6.0 ~ 4902.0)</td>\n",
       "      <td>28.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PAY_AMT1</td>\n",
       "      <td>[4902.0 ~ inf)</td>\n",
       "      <td>42.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PAY_AMT4</td>\n",
       "      <td>[-inf ~ 1.0)</td>\n",
       "      <td>22.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PAY_AMT4</td>\n",
       "      <td>[1.0 ~ 1506.0)</td>\n",
       "      <td>27.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PAY_AMT4</td>\n",
       "      <td>[1506.0 ~ 3946.0)</td>\n",
       "      <td>29.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>PAY_AMT4</td>\n",
       "      <td>[3946.0 ~ inf)</td>\n",
       "      <td>33.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BILL_AMT2</td>\n",
       "      <td>[-inf ~ 2400.0)</td>\n",
       "      <td>27.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BILL_AMT2</td>\n",
       "      <td>[2400.0 ~ 8287.0)</td>\n",
       "      <td>24.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BILL_AMT2</td>\n",
       "      <td>[8287.0 ~ 35213.0)</td>\n",
       "      <td>35.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BILL_AMT2</td>\n",
       "      <td>[35213.0 ~ inf)</td>\n",
       "      <td>23.66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PAY_AMT6</td>\n",
       "      <td>[-inf ~ 28.0)</td>\n",
       "      <td>20.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PAY_AMT6</td>\n",
       "      <td>[28.0 ~ 480.0)</td>\n",
       "      <td>27.67</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    variable              binning  score\n",
       "0  EDUCATION           [-inf ~ 2)  29.66\n",
       "1  EDUCATION            [2 ~ inf)  27.20\n",
       "0  BILL_AMT4      [-inf ~ 2501.0)  28.21\n",
       "1  BILL_AMT4       [2501.0 ~ inf)  27.97\n",
       "0  BILL_AMT6     [-inf ~ 19195.0)  27.33\n",
       "1  BILL_AMT6  [19195.0 ~ 50442.0)  31.37\n",
       "2  BILL_AMT6      [50442.0 ~ inf)  26.23\n",
       "0  BILL_AMT5      [-inf ~ 2395.0)  28.16\n",
       "1  BILL_AMT5    [2395.0 ~ 8110.0)  32.85\n",
       "2  BILL_AMT5   [8110.0 ~ 35288.0)  25.20\n",
       "3  BILL_AMT5      [35288.0 ~ inf)  29.41\n",
       "0  BILL_AMT1      [-inf ~ 2501.0)  25.09\n",
       "1  BILL_AMT1       [2501.0 ~ inf)  28.87\n",
       "0   PAY_AMT5         [-inf ~ 1.0)  23.01\n",
       "1   PAY_AMT5       [1.0 ~ 1880.0)  26.88\n",
       "2   PAY_AMT5   [1880.0 ~ 10000.0)  30.60\n",
       "3   PAY_AMT5      [10000.0 ~ inf)  37.59\n",
       "0   PAY_AMT1         [-inf ~ 6.0)  11.44\n",
       "1   PAY_AMT1       [6.0 ~ 4902.0)  28.26\n",
       "2   PAY_AMT1       [4902.0 ~ inf)  42.70\n",
       "0   PAY_AMT4         [-inf ~ 1.0)  22.23\n",
       "1   PAY_AMT4       [1.0 ~ 1506.0)  27.31\n",
       "2   PAY_AMT4    [1506.0 ~ 3946.0)  29.40\n",
       "3   PAY_AMT4       [3946.0 ~ inf)  33.87\n",
       "0  BILL_AMT2      [-inf ~ 2400.0)  27.65\n",
       "1  BILL_AMT2    [2400.0 ~ 8287.0)  24.80\n",
       "2  BILL_AMT2   [8287.0 ~ 35213.0)  35.81\n",
       "3  BILL_AMT2      [35213.0 ~ inf)  23.66\n",
       "0   PAY_AMT6        [-inf ~ 28.0)  20.84\n",
       "1   PAY_AMT6       [28.0 ~ 480.0)  27.67"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#transform the scorecard into dataframe and save to csv\n",
    "keys = list(card.export().keys())\n",
    "score_card_df = pd.DataFrame()\n",
    "for n in keys:\n",
    "    temp = pd.DataFrame.from_dict(final_card_score[n], orient='index')\n",
    "    temp = temp.reset_index()\n",
    "    temp.columns= ['binning','score']\n",
    "    temp['variable'] = n\n",
    "    temp = temp[['variable','binning','score']]\n",
    "    score_card_df=score_card_df.append(temp)\n",
    "score_card_df.head(30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [],
   "source": [
    "#save 6: save the scorcard\n",
    "score_card_df.to_csv('CreditScore_save6_score_card_df.csv',index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 9. Distribution Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,10))\n",
    "import random\n",
    "import numpy\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "w = 40\n",
    "n = math.ceil((data['CreditScore'].max() - data['CreditScore'].min())/w)\n",
    "#bins = numpy.linspace(-10, 10, 100)\n",
    "\n",
    "plt.hist(data[data.label==1].CreditScore, alpha=0.5, label='Black',bins = n)\n",
    "plt.hist(data[data.label==0].CreditScore, alpha=0.5, label='White',bins = n)\n",
    "plt.legend(loc='upper left')\n",
    "plt.title('Credit Score Distribution: Test Set',size=15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,10))\n",
    "import random\n",
    "import numpy\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "w = 40\n",
    "n = math.ceil((test['CreditScore'].max() - test['CreditScore'].min())/w)\n",
    "#bins = numpy.linspace(-10, 10, 100)\n",
    "\n",
    "plt.hist(test[test.label==1].CreditScore, alpha=0.5, label='Black',bins = n)\n",
    "plt.hist(test[test.label==0].CreditScore, alpha=0.5, label='White',bins = n)\n",
    "plt.legend(loc='upper left')\n",
    "plt.title('Credit Score Distribution: Whole Dataset',size=15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 10. Threshold Tuning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_credit_level(\n",
    "    test,\n",
    "    target_score ='order_score',\n",
    "    out_col = 'order_level',\n",
    "    left_bound = -100,\n",
    "    level_0 = 100,\n",
    "    level_1 = 200,    \n",
    "    level_2 = 250,    \n",
    "    level_3 = 300,    \n",
    "    level_4 = 350,    \n",
    "    level_5 = 400,    \n",
    "    level_6 = 450,\n",
    "    level_7 = 500,\n",
    "    level_8 = 800):\n",
    "    level = []\n",
    "    for i in range(len(test)):\n",
    "        if (test[target_score][i]>left_bound) & (test[target_score][i]<=level_0):\n",
    "            level.append(0)\n",
    "        elif  (test[target_score][i]>level_0) & (test[target_score][i]<=level_1):\n",
    "            level.append(1)\n",
    "        elif  (test[target_score][i]>level_1) & (test[target_score][i]<=level_2):\n",
    "            level.append(2)\n",
    "        elif  (test[target_score][i]>level_2) & (test[target_score][i]<=level_3):\n",
    "            level.append(3)\n",
    "        elif  (test[target_score][i]>level_3) & (test[target_score][i]<=level_4):\n",
    "            level.append(4)\n",
    "        elif  (test[target_score][i]>level_4) & (test[target_score][i]<=level_5):\n",
    "            level.append(5)\n",
    "        elif  (test[target_score][i]>level_5) & (test[target_score][i]<=level_6):\n",
    "            level.append(6)\n",
    "        elif  (test[target_score][i]>level_6) & (test[target_score][i]<=level_7):\n",
    "            level.append(7)\n",
    "        elif  (test[target_score][i]>level_7 )& (test[target_score][i]<=level_8):\n",
    "            level.append(8)\n",
    "        \n",
    "    test[out_col] = level\n",
    "    return test\n",
    "\n",
    "def plot_bts_level_loss(test, target_col):\n",
    "    bts_level_df = test[target_col].value_counts()\n",
    "    bts_level_df=pd.DataFrame(bts_level_df)\n",
    "    df_label_level= test[test.label==1].groupby(target_col)['label'].count()/ test.groupby(target_col)['label'].count()\n",
    "    df_label_level = pd.DataFrame(df_label_level)\n",
    "    bts_level_df.sort_index().plot.bar(title='')\n",
    "    df_label_level.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    7501.000000\n",
       "mean      627.437143\n",
       "std       113.055983\n",
       "min       173.540655\n",
       "25%       596.974091\n",
       "50%       659.897345\n",
       "75%       702.107780\n",
       "max       810.644058\n",
       "Name: CreditScore, dtype: float64"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.CreditScore.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Threshold Tuning & Trade-off between loss & Coverage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "test = get_credit_level(test,\n",
    "                       target_score ='CreditScore',\n",
    "                       out_col = 'CreditScore_level',\n",
    "                      left_bound = -1000,\n",
    "    level_0 = 250,\n",
    "    level_1 = 300,    \n",
    "    level_2 = 400,    \n",
    "    level_3 = 500,    \n",
    "    level_4 = 580,    \n",
    "    level_5 = 630,    \n",
    "    level_6 = 690,\n",
    "    level_7 = 730,\n",
    "    level_8 = 1000\n",
    "                )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_bts_level_loss(test,target_col='CreditScore_level')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "120"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(test[test.label==1].CreditScore_level==7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_loss_coverage(test,target_level):\n",
    "    \n",
    "    #level 8 Loss\n",
    "    L8_loss=test[test[target_level]>=8 ].label.value_counts()/len(test[test[target_level]>=8 ])\n",
    "    #level 8 Coverage\n",
    "    L8_coverage=test[test[target_level]>=8].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 8: Loss is \",L8_loss[1], \"; Coverage is \",L8_coverage)\n",
    "    \n",
    "    #level 7-level 8 Loss\n",
    "    L7_loss=test[test[target_level]>=7 ].label.value_counts()/len(test[test[target_level]>=7 ])\n",
    "    #level 7-level 8 Coverage\n",
    "    L7_coverage=test[test[target_level]>=7].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 7-Level 8: Loss is \",L7_loss[1], \"; Coverage is \",L7_coverage)\n",
    "    \n",
    "    #level 6-level 8 Loss\n",
    "    L6_loss=test[test[target_level]>=6 ].label.value_counts()/len(test[test[target_level]>=6 ])\n",
    "    #level 6-level 8 Coverage\n",
    "    L6_coverage=test[test[target_level]>=6].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 6-Level 8: Loss is \",L6_loss[1], \"; Coverage is \",L6_coverage)\n",
    "    \n",
    "     #level 5-Leve 8 Loss (percentage of default people)\n",
    "    L5_loss = test[test[target_level]>=5 ].label.value_counts()/len(test[test[target_level]>=5 ])\n",
    "    #level 5- level 8 Coverage (percentage of good people)\n",
    "    L5_coverage=test[test[target_level]>=5 ].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 5-Level 8: Loss is \",L5_loss[1], \"; Coverage is \",L5_coverage)\n",
    "    \n",
    "    #level 4-level 8 Loss\n",
    "    L4_loss=test[test[target_level]>=4 ].label.value_counts()/len(test[test[target_level]>=4 ])\n",
    "    #level 4-level 8 Coverage\n",
    "    L4_coverage=test[test[target_level]>=4].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 4-Level 8: Loss is \",L4_loss[1], \"; Coverage is \",L4_coverage)\n",
    "    \n",
    "    \n",
    "    #level 3-level 8 Loss\n",
    "    L3_loss=test[test[target_level]>=3].label.value_counts()/len(test[test[target_level]>=3 ])\n",
    "    #level 3-level 8 Coverage\n",
    "    L3_coverage=test[test[target_level]>=3].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 3-Level 8: Loss is \",L3_loss[1], \"; Coverage is \",L3_coverage)\n",
    "    \n",
    "    #level 2-level 8 Loss\n",
    "    L2_loss=test[test[target_level]>=2].label.value_counts()/len(test[test[target_level]>=2 ])\n",
    "    #level 2-level 8 Coverage\n",
    "    L2_coverage=test[test[target_level]>=2].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 2-Level 8: Loss is \",L2_loss[1], \"; Coverage is \",L2_coverage)\n",
    "    \n",
    "    #level 1-level 8 Loss\n",
    "    L1_loss=test[test[target_level]>=1].label.value_counts()/len(test[test[target_level]>=1 ])\n",
    "    #level 1-level 8 Coverage\n",
    "    L1_coverage=test[test[target_level]>=1].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 1-Level 8: Loss is \",L1_loss[1], \"; Coverage is \",L1_coverage)\n",
    "    \n",
    "    #level 0-level 8 Loss\n",
    "    L0_loss=test[test[target_level]>=0].label.value_counts()/len(test[test[target_level]>=0 ])\n",
    "    #level 0-level 8 Coverage\n",
    "    L0_coverage=test[test[target_level]>=0].label.value_counts()[0]/test[test.label==0].shape[0]\n",
    "    print(\"Level 0-Level 8: Loss is \",L0_loss[1], \"; Coverage is \",L0_coverage)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Level 8: Loss is  0.04271356783919598 ; Coverage is  0.12800268772047707\n",
      "Level 7-Level 8: Loss is  0.0650887573964497 ; Coverage is  0.3715773559549807\n",
      "Level 6-Level 8: Loss is  0.10143753796315043 ; Coverage is  0.745506467327398\n",
      "Level 5-Level 8: Loss is  0.1204323211528564 ; Coverage is  0.8612464303712414\n",
      "Level 4-Level 8: Loss is  0.13876040703052728 ; Coverage is  0.938350411557198\n",
      "Level 3-Level 8: Loss is  0.1684751570531125 ; Coverage is  0.97833025365362\n",
      "Level 2-Level 8: Loss is  0.19373723621511232 ; Coverage is  0.9947925415756761\n",
      "Level 1-Level 8: Loss is  0.20193340494092374 ; Coverage is  0.9984881572316479\n",
      "Level 0-Level 8: Loss is  0.20637248366884414 ; Coverage is  1.0\n"
     ]
    }
   ],
   "source": [
    "get_loss_coverage(test,target_level='CreditScore_level')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "#save 7: save the scorcard\n",
    "data.to_csv('OrderScore_save7_whole_data.csv',index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ID                    22500.000000\n",
       "LIMIT_BAL            420000.000000\n",
       "SEX                       2.000000\n",
       "EDUCATION                 2.000000\n",
       "MARRIAGE                  1.000000\n",
       "AGE                      37.000000\n",
       "PAY_0                     0.000000\n",
       "PAY_2                     0.000000\n",
       "PAY_3                     0.000000\n",
       "PAY_4                    -1.000000\n",
       "PAY_5                     0.000000\n",
       "PAY_6                     0.000000\n",
       "BILL_AMT1             36032.000000\n",
       "BILL_AMT2             41932.000000\n",
       "BILL_AMT3              9778.000000\n",
       "BILL_AMT4            158901.000000\n",
       "BILL_AMT5            161876.000000\n",
       "BILL_AMT6            165378.000000\n",
       "PAY_AMT1               7022.000000\n",
       "PAY_AMT2               1846.000000\n",
       "PAY_AMT3             163862.000000\n",
       "PAY_AMT4               6000.000000\n",
       "PAY_AMT5               6000.000000\n",
       "PAY_AMT6               6000.000000\n",
       "label                     0.000000\n",
       "CreditScore             751.330191\n",
       "CreditScore_level         8.000000\n",
       "Name: 0, dtype: float64"
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.loc[0,:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "751.3301912709491"
      ]
     },
     "execution_count": 180,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "card.predict(test)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}