credit_risk_deploy.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "credit_risk_deploy.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2vax0p7jbGkq"
      },
      "source": [
        "# Credit Risk Case"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bVXNtAuA4gyE"
      },
      "source": [
        "## Business Problems\n",
        "As a Financing Company, the user wants to build a credit scoring model to predict whether the client will default or not after their loan application."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZFQj_uhitr49"
      },
      "source": [
        "## Business Goals\n",
        "Research and develop the model to predict applicants whether the applicant will default or not, and also find the best metrics since this is an imbalance class dataset."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bMEDjQ4xgGVs"
      },
      "source": [
        "## Preprocessing\n",
        "\n",
        "![image.png](data:image/png;base64,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)\n",
        "\n",
        "![image.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wMudzvGqbKOx"
      },
      "source": [
        "# import package\n",
        "\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "Psav5S2Hcaxz",
        "outputId": "45a4171b-7f1e-43c3-eb90-92fd7c11b2a5"
      },
      "source": [
        "credit = pd.read_csv('C:\\\\Users\\\\brian\\\\Documents\\\\DataScience\\\\ODP BNI x Mars\\\\trial_deployment\\\\credit_risk_dataset.csv')\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "0          22          59000                  RENT              123.0   \n",
              "1          21           9600                   OWN                5.0   \n",
              "2          25           9600              MORTGAGE                1.0   \n",
              "3          23          65500                  RENT                4.0   \n",
              "4          24          54400                  RENT                8.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "0    PERSONAL          D      35000          16.02            1   \n",
              "1   EDUCATION          B       1000          11.14            0   \n",
              "2     MEDICAL          C       5500          12.87            1   \n",
              "3     MEDICAL          C      35000          15.23            1   \n",
              "4     MEDICAL          C      35000          14.27            1   \n",
              "\n",
              "   loan_percent_income cb_person_default_on_file  cb_person_cred_hist_length  \n",
              "0                 0.59                         Y                           3  \n",
              "1                 0.10                         N                           2  \n",
              "2                 0.57                         N                           3  \n",
              "3                 0.53                         N                           2  \n",
              "4                 0.55                         Y                           4  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>22</td>\n",
              "      <td>59000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>123.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>D</td>\n",
              "      <td>35000</td>\n",
              "      <td>16.02</td>\n",
              "      <td>1</td>\n",
              "      <td>0.59</td>\n",
              "      <td>Y</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 264
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EcjvSJn1sTfy",
        "outputId": "82591180-5916-4228-d9ee-2b6cc9627a3f"
      },
      "source": [
        "print(f\"This dataset will be divided into {0.2* len(credit)+1:.0f} for holdout sample, the rest {len(credit) - 0.2* len(credit):.0f} is for training and testing model.\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "This dataset will be divided into 6517 for holdout sample, the rest 26065 is for training and testing model.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "A9BzPM1EcpnI",
        "outputId": "3ab4a0ac-6b13-4a6c-c449-0b338e791d22"
      },
      "source": [
        "credit.info()\n",
        "\n",
        "\n",
        "# column which has null values:\n",
        "  # person_emp_length\n",
        "  # loan_amnt"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 32581 entries, 0 to 32580\n",
            "Data columns (total 12 columns):\n",
            " #   Column                      Non-Null Count  Dtype  \n",
            "---  ------                      --------------  -----  \n",
            " 0   person_age                  32581 non-null  int64  \n",
            " 1   person_income               32581 non-null  int64  \n",
            " 2   person_home_ownership       32581 non-null  object \n",
            " 3   person_emp_length           31686 non-null  float64\n",
            " 4   loan_intent                 32581 non-null  object \n",
            " 5   loan_grade                  32581 non-null  object \n",
            " 6   loan_amnt                   32581 non-null  int64  \n",
            " 7   loan_int_rate               29465 non-null  float64\n",
            " 8   loan_status                 32581 non-null  int64  \n",
            " 9   loan_percent_income         32581 non-null  float64\n",
            " 10  cb_person_default_on_file   32581 non-null  object \n",
            " 11  cb_person_cred_hist_length  32581 non-null  int64  \n",
            "dtypes: float64(3), int64(5), object(4)\n",
            "memory usage: 3.0+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "nn867bGxc5oV",
        "outputId": "509ce7c7-9fe1-4024-c2ee-b9c93cf3c8ac"
      },
      "source": [
        "sns.countplot(credit['loan_status'])\n",
        "plt.title('Loan Status')\n",
        "plt.show()\n",
        "\n",
        "# has more non default applicant (0) than the default (1)\n",
        "  # non default:78.183604%\t\n",
        "  # default: 21.816396%"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 79
        },
        "id": "7IMw2HG95yCK",
        "outputId": "dad7c227-cc67-47f8-c66a-e66eccf2c146"
      },
      "source": [
        "# percentage of loan_status\n",
        "\n",
        "(credit['loan_status'].value_counts() / len(credit) * 100).to_frame().T\n",
        "\n",
        "# with this proportion, it looks like we have imbalance target"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                     0          1\n",
              "loan_status  78.183604  21.816396"
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>loan_status</th>\n",
              "      <td>78.183604</td>\n",
              "      <td>21.816396</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 268
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y9rS5um5gL-v"
      },
      "source": [
        "## Descriptive Analysis"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "r5j_3Dmrd0l8",
        "outputId": "ef0179e3-d746-487a-e44b-238c8fee5d24"
      },
      "source": [
        "# Numerical\n",
        "credit.describe().T\n",
        "\n",
        "# The average of:\n",
        "#   age                   = 27 years old\n",
        "#   income                = $ 66,074\n",
        "#   employee length       = 4,7 years\n",
        "#   loan amount           = $ 9589\n",
        "#   loan interest rate    = 11,01 %\n",
        "#   loan percent income   = 0.17\n",
        "#   Credit history length = 5,8 years"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                              count          mean           std      min  \\\n",
              "person_age                  32581.0     27.734600      6.348078    20.00   \n",
              "person_income               32581.0  66074.848470  61983.119168  4000.00   \n",
              "person_emp_length           31686.0      4.789686      4.142630     0.00   \n",
              "loan_amnt                   32581.0   9589.371106   6322.086646   500.00   \n",
              "loan_int_rate               29465.0     11.011695      3.240459     5.42   \n",
              "loan_status                 32581.0      0.218164      0.413006     0.00   \n",
              "loan_percent_income         32581.0      0.170203      0.106782     0.00   \n",
              "cb_person_cred_hist_length  32581.0      5.804211      4.055001     2.00   \n",
              "\n",
              "                                 25%       50%       75%         max  \n",
              "person_age                     23.00     26.00     30.00      144.00  \n",
              "person_income               38500.00  55000.00  79200.00  6000000.00  \n",
              "person_emp_length               2.00      4.00      7.00      123.00  \n",
              "loan_amnt                    5000.00   8000.00  12200.00    35000.00  \n",
              "loan_int_rate                   7.90     10.99     13.47       23.22  \n",
              "loan_status                     0.00      0.00      0.00        1.00  \n",
              "loan_percent_income             0.09      0.15      0.23        0.83  \n",
              "cb_person_cred_hist_length      3.00      4.00      8.00       30.00  "
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>person_age</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>27.734600</td>\n",
              "      <td>6.348078</td>\n",
              "      <td>20.00</td>\n",
              "      <td>23.00</td>\n",
              "      <td>26.00</td>\n",
              "      <td>30.00</td>\n",
              "      <td>144.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_income</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>66074.848470</td>\n",
              "      <td>61983.119168</td>\n",
              "      <td>4000.00</td>\n",
              "      <td>38500.00</td>\n",
              "      <td>55000.00</td>\n",
              "      <td>79200.00</td>\n",
              "      <td>6000000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_emp_length</th>\n",
              "      <td>31686.0</td>\n",
              "      <td>4.789686</td>\n",
              "      <td>4.142630</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>7.00</td>\n",
              "      <td>123.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_amnt</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>9589.371106</td>\n",
              "      <td>6322.086646</td>\n",
              "      <td>500.00</td>\n",
              "      <td>5000.00</td>\n",
              "      <td>8000.00</td>\n",
              "      <td>12200.00</td>\n",
              "      <td>35000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_int_rate</th>\n",
              "      <td>29465.0</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>3.240459</td>\n",
              "      <td>5.42</td>\n",
              "      <td>7.90</td>\n",
              "      <td>10.99</td>\n",
              "      <td>13.47</td>\n",
              "      <td>23.22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_status</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>0.218164</td>\n",
              "      <td>0.413006</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_percent_income</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>0.170203</td>\n",
              "      <td>0.106782</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0.15</td>\n",
              "      <td>0.23</td>\n",
              "      <td>0.83</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>5.804211</td>\n",
              "      <td>4.055001</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.00</td>\n",
              "      <td>8.00</td>\n",
              "      <td>30.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 269
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 172
        },
        "id": "l-tDYZnyeMIN",
        "outputId": "e020d749-43c1-4a5a-95be-201b0cfd9b28"
      },
      "source": [
        "# Categorical\n",
        "credit.describe(include = 'object').T\n",
        "\n",
        "# The most of:\n",
        "#   Home Ownership          = Rent\n",
        "#   Loan intent             = Education\n",
        "#   Loan Grade              = A\n",
        "#   Historical default      = No"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                           count unique        top   freq\n",
              "person_home_ownership      32581      4       RENT  16446\n",
              "loan_intent                32581      6  EDUCATION   6453\n",
              "loan_grade                 32581      7          A  10777\n",
              "cb_person_default_on_file  32581      2          N  26836"
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>unique</th>\n",
              "      <th>top</th>\n",
              "      <th>freq</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>person_home_ownership</th>\n",
              "      <td>32581</td>\n",
              "      <td>4</td>\n",
              "      <td>RENT</td>\n",
              "      <td>16446</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_intent</th>\n",
              "      <td>32581</td>\n",
              "      <td>6</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>6453</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_grade</th>\n",
              "      <td>32581</td>\n",
              "      <td>7</td>\n",
              "      <td>A</td>\n",
              "      <td>10777</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <td>32581</td>\n",
              "      <td>2</td>\n",
              "      <td>N</td>\n",
              "      <td>26836</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 270
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fcdnBoELgSur"
      },
      "source": [
        "## Handling Missing Values"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BxUnWsZIfOkN",
        "outputId": "bb5654fe-30d7-4062-bd98-d584322d058e"
      },
      "source": [
        "# List Columns with NaN\n",
        "print('Total columns with nan:',len(credit.columns[credit.isna().any()]))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Total columns with nan: 2\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Pjbscj3vfQxF",
        "outputId": "23028d53-753b-4539-cece-7b0538d98be4"
      },
      "source": [
        "# Null Percentage\n",
        "credit.isnull().sum() / len(credit) * 100\n",
        "\n",
        "# column which has null values:\n",
        "  # person_emp_length   = 2.747%\n",
        "  # loan_amnt           = 9.564%"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "person_age                    0.000000\n",
              "person_income                 0.000000\n",
              "person_home_ownership         0.000000\n",
              "person_emp_length             2.747000\n",
              "loan_intent                   0.000000\n",
              "loan_grade                    0.000000\n",
              "loan_amnt                     0.000000\n",
              "loan_int_rate                 9.563856\n",
              "loan_status                   0.000000\n",
              "loan_percent_income           0.000000\n",
              "cb_person_default_on_file     0.000000\n",
              "cb_person_cred_hist_length    0.000000\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 272
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZaoYMEbxgaPi"
      },
      "source": [
        "### person_emp_length"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "kFPM2KnRfe8f",
        "outputId": "ff209d45-1799-4a8d-eefc-903fd6c8bb86"
      },
      "source": [
        "sns.distplot(credit['person_emp_length'], color = 'green')\n",
        "plt.title('Distribution of Employee Length')\n",
        "plt.show()\n",
        "\n",
        "# looks like we have outliers"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 79
        },
        "id": "Ihu62sBef-wf",
        "outputId": "00abde63-9756-4b27-86ba-091dc676ab96"
      },
      "source": [
        "credit.describe()['person_emp_length'].to_frame().T"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                     count      mean      std  min  25%  50%  75%    max\n",
              "person_emp_length  31686.0  4.789686  4.14263  0.0  2.0  4.0  7.0  123.0"
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>person_emp_length</th>\n",
              "      <td>31686.0</td>\n",
              "      <td>4.789686</td>\n",
              "      <td>4.14263</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>123.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 274
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5asmUdGDggGo",
        "outputId": "8f8132fe-3b6a-4184-b76a-8fd846225c63"
      },
      "source": [
        "# since it has outliers, we could fill the nulls with the median \n",
        "\n",
        "print(f\"There are {credit['person_emp_length'].isnull().sum()} null value in 'person_emp_length' column.\")\n",
        "print('='*15, \" ++ Imputing ++ \", \"=\"*15)\n",
        "\n",
        "credit['person_emp_length'] = credit['person_emp_length'].fillna(credit['person_emp_length'].median())\n",
        "print(f\"There are {credit['person_emp_length'].isnull().sum()} null value in 'person_emp_length' column.\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "There are 895 null value in 'person_emp_length' column.\n",
            "===============  ++ Imputing ++  ===============\n",
            "There are 0 null value in 'person_emp_length' column.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HxstPpgSg-Fq",
        "outputId": "f4248f42-492b-4f6d-cd33-965e410736e6"
      },
      "source": [
        "# Null percentage check\n",
        "credit.isnull().sum() / len(credit) * 100"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "person_age                    0.000000\n",
              "person_income                 0.000000\n",
              "person_home_ownership         0.000000\n",
              "person_emp_length             0.000000\n",
              "loan_intent                   0.000000\n",
              "loan_grade                    0.000000\n",
              "loan_amnt                     0.000000\n",
              "loan_int_rate                 9.563856\n",
              "loan_status                   0.000000\n",
              "loan_percent_income           0.000000\n",
              "cb_person_default_on_file     0.000000\n",
              "cb_person_cred_hist_length    0.000000\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 276
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 79
        },
        "id": "094xijL4hFOZ",
        "outputId": "71a9c87b-aba4-48c0-ae8f-eb0f2aaa1164"
      },
      "source": [
        "credit.describe()['person_emp_length'].to_frame().T\n",
        "\n",
        "# we will handle that outliers later"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                     count      mean       std  min  25%  50%  75%    max\n",
              "person_emp_length  32581.0  4.767994  4.087372  0.0  2.0  4.0  7.0  123.0"
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>person_emp_length</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>4.767994</td>\n",
              "      <td>4.087372</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>123.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 277
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dD26_A-ChrZG"
      },
      "source": [
        "### loan_int_rate"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "rODWjUfkhrZH",
        "outputId": "7580c9a0-70f6-4ea1-dc0d-681e240a373a"
      },
      "source": [
        "sns.distplot(credit['loan_int_rate'], color = 'maroon')\n",
        "plt.title('Distribution of Loan Interest Rate')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 79
        },
        "id": "O1puhSR_hrZK",
        "outputId": "5ed318d1-8720-41ed-e675-7ecd0b3f6ab5"
      },
      "source": [
        "credit.describe()['loan_int_rate'].to_frame().T"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                 count       mean       std   min  25%    50%    75%    max\n",
              "loan_int_rate  29465.0  11.011695  3.240459  5.42  7.9  10.99  13.47  23.22"
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>loan_int_rate</th>\n",
              "      <td>29465.0</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>3.240459</td>\n",
              "      <td>5.42</td>\n",
              "      <td>7.9</td>\n",
              "      <td>10.99</td>\n",
              "      <td>13.47</td>\n",
              "      <td>23.22</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 279
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QXaFNsI4hrZK",
        "outputId": "49d7d8ba-0704-4ac6-b5c5-e06b621d4064"
      },
      "source": [
        "# filling the nulls with the mean\n",
        "\n",
        "print(f\"There are {credit['loan_int_rate'].isnull().sum()} null value in 'loan_int_rate' column.\")\n",
        "print('='*15, \" ++ Imputing ++ \", \"=\"*15)\n",
        "\n",
        "credit['loan_int_rate'] = credit['loan_int_rate'].fillna(credit['loan_int_rate'].mean())\n",
        "print(f\"There are {credit['loan_int_rate'].isnull().sum()} null value in 'loan_int_rate' column.\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "There are 3116 null value in 'loan_int_rate' column.\n",
            "===============  ++ Imputing ++  ===============\n",
            "There are 0 null value in 'loan_int_rate' column.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NjkAgUzChrZL",
        "outputId": "2da1c179-6b49-4c56-cb79-db07d4452dee"
      },
      "source": [
        "# Null percentage check\n",
        "credit.isnull().sum() / len(credit) * 100"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "person_age                    0.0\n",
              "person_income                 0.0\n",
              "person_home_ownership         0.0\n",
              "person_emp_length             0.0\n",
              "loan_intent                   0.0\n",
              "loan_grade                    0.0\n",
              "loan_amnt                     0.0\n",
              "loan_int_rate                 0.0\n",
              "loan_status                   0.0\n",
              "loan_percent_income           0.0\n",
              "cb_person_default_on_file     0.0\n",
              "cb_person_cred_hist_length    0.0\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 281
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 79
        },
        "id": "t7qvBtVMhrZL",
        "outputId": "b2dce50a-1342-4c21-8447-bc53fc5461b2"
      },
      "source": [
        "credit.describe()['loan_int_rate'].to_frame().T"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                 count       mean       std   min   25%        50%    75%  \\\n",
              "loan_int_rate  32581.0  11.011695  3.081605  5.42  8.49  11.011695  13.11   \n",
              "\n",
              "                 max  \n",
              "loan_int_rate  23.22  "
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>loan_int_rate</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>3.081605</td>\n",
              "      <td>5.42</td>\n",
              "      <td>8.49</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>13.11</td>\n",
              "      <td>23.22</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 282
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zv5vbsyLn0Mz"
      },
      "source": [
        "## Outliers\n",
        "1. person_age, \n",
        "1. person_income, \n",
        "1. person_emp_length, \n",
        "1. loan_amnt, \n",
        "1. loan_int_rate, \n",
        "1. loan_percent_income, \n",
        "1. cb_person_cred_hist_length "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "kvVPL66R7fES",
        "outputId": "178795b3-ec23-4a26-945d-4ab070cde373"
      },
      "source": [
        "credit.describe().T\n",
        "\n",
        "# is there a guy which 144 years old?\n",
        "# is there a guy who work along 123 years?\n",
        "# applicants who has income 6M is quite make sense, becase when we see the data there are 9 applicants who has income more than 1M."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                              count          mean           std      min  \\\n",
              "person_age                  32581.0     27.734600      6.348078    20.00   \n",
              "person_income               32581.0  66074.848470  61983.119168  4000.00   \n",
              "person_emp_length           32581.0      4.767994      4.087372     0.00   \n",
              "loan_amnt                   32581.0   9589.371106   6322.086646   500.00   \n",
              "loan_int_rate               32581.0     11.011695      3.081605     5.42   \n",
              "loan_status                 32581.0      0.218164      0.413006     0.00   \n",
              "loan_percent_income         32581.0      0.170203      0.106782     0.00   \n",
              "cb_person_cred_hist_length  32581.0      5.804211      4.055001     2.00   \n",
              "\n",
              "                                 25%           50%       75%         max  \n",
              "person_age                     23.00     26.000000     30.00      144.00  \n",
              "person_income               38500.00  55000.000000  79200.00  6000000.00  \n",
              "person_emp_length               2.00      4.000000      7.00      123.00  \n",
              "loan_amnt                    5000.00   8000.000000  12200.00    35000.00  \n",
              "loan_int_rate                   8.49     11.011695     13.11       23.22  \n",
              "loan_status                     0.00      0.000000      0.00        1.00  \n",
              "loan_percent_income             0.09      0.150000      0.23        0.83  \n",
              "cb_person_cred_hist_length      3.00      4.000000      8.00       30.00  "
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>person_age</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>27.734600</td>\n",
              "      <td>6.348078</td>\n",
              "      <td>20.00</td>\n",
              "      <td>23.00</td>\n",
              "      <td>26.000000</td>\n",
              "      <td>30.00</td>\n",
              "      <td>144.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_income</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>66074.848470</td>\n",
              "      <td>61983.119168</td>\n",
              "      <td>4000.00</td>\n",
              "      <td>38500.00</td>\n",
              "      <td>55000.000000</td>\n",
              "      <td>79200.00</td>\n",
              "      <td>6000000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_emp_length</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>4.767994</td>\n",
              "      <td>4.087372</td>\n",
              "      <td>0.00</td>\n",
              "      <td>2.00</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>7.00</td>\n",
              "      <td>123.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_amnt</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>9589.371106</td>\n",
              "      <td>6322.086646</td>\n",
              "      <td>500.00</td>\n",
              "      <td>5000.00</td>\n",
              "      <td>8000.000000</td>\n",
              "      <td>12200.00</td>\n",
              "      <td>35000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_int_rate</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>3.081605</td>\n",
              "      <td>5.42</td>\n",
              "      <td>8.49</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>13.11</td>\n",
              "      <td>23.22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_status</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>0.218164</td>\n",
              "      <td>0.413006</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_percent_income</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>0.170203</td>\n",
              "      <td>0.106782</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0.150000</td>\n",
              "      <td>0.23</td>\n",
              "      <td>0.83</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <td>32581.0</td>\n",
              "      <td>5.804211</td>\n",
              "      <td>4.055001</td>\n",
              "      <td>2.00</td>\n",
              "      <td>3.00</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>8.00</td>\n",
              "      <td>30.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 283
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "r704FHXin1zI",
        "outputId": "93f19914-e585-45f9-84c0-253ab5656069"
      },
      "source": [
        "plt.figure(figsize = (20,20))\n",
        "\n",
        "n = 1\n",
        "for name in credit.describe().columns:\n",
        "  plt.subplot(3,3,n)\n",
        "  sns.boxplot(credit[name])\n",
        "  n += 1\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1440x1440 with 8 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KWKB6zAy4kJn"
      },
      "source": [
        "### cb_person_cred_hist_length"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 745
        },
        "id": "OtRcoPEOA0aP",
        "outputId": "22c96c1a-1fab-4ed1-9bdd-0a8460417413"
      },
      "source": [
        "credit[credit['cb_person_cred_hist_length'] > 29]\n",
        "\n",
        "# it seems fine so we will keep this outliers\n",
        "# because when you see the person_age, the average is around 50-60 years old which is make sense that applicants has 30 years credit history length."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "32320          55          50000                  RENT                2.0   \n",
              "32326          64         105000                  RENT                8.0   \n",
              "32353          51          37200                  RENT                4.0   \n",
              "32357          53          34000              MORTGAGE                0.0   \n",
              "32366          58          20000                  RENT                0.0   \n",
              "32395          51          46560              MORTGAGE                2.0   \n",
              "32403          60          49008                   OWN                5.0   \n",
              "32421          57          53000              MORTGAGE                1.0   \n",
              "32446          51          62046                  RENT                8.0   \n",
              "32448          52          83200                  RENT                0.0   \n",
              "32460          51          36000                  RENT                3.0   \n",
              "32463          53          43200                  RENT                0.0   \n",
              "32483          55          34000                  RENT                0.0   \n",
              "32495          58         100000                  RENT                6.0   \n",
              "32497          63        1782000                  RENT               13.0   \n",
              "32505          51          94368              MORTGAGE                4.0   \n",
              "32508          53          95000                   OWN               24.0   \n",
              "32510          55          97243              MORTGAGE               10.0   \n",
              "32541          52         163000              MORTGAGE                9.0   \n",
              "32543          52         720000              MORTGAGE                1.0   \n",
              "32576          57          53000              MORTGAGE                1.0   \n",
              "32580          66          42000                  RENT                2.0   \n",
              "\n",
              "             loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "32320    HOMEIMPROVEMENT          A      20000       7.510000            1   \n",
              "32326    HOMEIMPROVEMENT          E      20000      18.640000            0   \n",
              "32353            MEDICAL          B       3000      10.620000            1   \n",
              "32357            MEDICAL          B      10000      10.390000            1   \n",
              "32366          EDUCATION          B       4000       9.990000            0   \n",
              "32395           PERSONAL          A      11600       8.940000            0   \n",
              "32403           PERSONAL          C      12000      13.790000            0   \n",
              "32421           PERSONAL          C       5800      13.160000            0   \n",
              "32446            VENTURE          A       8000      11.011695            0   \n",
              "32448  DEBTCONSOLIDATION          C       8000      15.960000            0   \n",
              "32460            MEDICAL          D      10000      15.650000            0   \n",
              "32463            VENTURE          B      10000       9.880000            0   \n",
              "32483  DEBTCONSOLIDATION          A      12000       6.920000            1   \n",
              "32495          EDUCATION          A      12000       8.900000            0   \n",
              "32497          EDUCATION          C      12025      14.270000            0   \n",
              "32505           PERSONAL          A       5000      11.011695            0   \n",
              "32508           PERSONAL          A       6000       5.420000            0   \n",
              "32510           PERSONAL          E      17000      11.011695            0   \n",
              "32541           PERSONAL          D      25000      14.110000            1   \n",
              "32543           PERSONAL          A       7925       8.940000            0   \n",
              "32576           PERSONAL          C       5800      13.160000            0   \n",
              "32580            MEDICAL          B       6475       9.990000            0   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "32320                 0.40                         N   \n",
              "32326                 0.19                         N   \n",
              "32353                 0.08                         N   \n",
              "32357                 0.29                         N   \n",
              "32366                 0.20                         N   \n",
              "32395                 0.25                         N   \n",
              "32403                 0.24                         Y   \n",
              "32421                 0.11                         Y   \n",
              "32446                 0.13                         N   \n",
              "32448                 0.10                         Y   \n",
              "32460                 0.28                         N   \n",
              "32463                 0.23                         N   \n",
              "32483                 0.35                         N   \n",
              "32495                 0.12                         N   \n",
              "32497                 0.01                         N   \n",
              "32505                 0.05                         N   \n",
              "32508                 0.06                         N   \n",
              "32510                 0.17                         N   \n",
              "32541                 0.15                         N   \n",
              "32543                 0.01                         N   \n",
              "32576                 0.11                         N   \n",
              "32580                 0.15                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "32320                          30  \n",
              "32326                          30  \n",
              "32353                          30  \n",
              "32357                          30  \n",
              "32366                          30  \n",
              "32395                          30  \n",
              "32403                          30  \n",
              "32421                          30  \n",
              "32446                          30  \n",
              "32448                          30  \n",
              "32460                          30  \n",
              "32463                          30  \n",
              "32483                          30  \n",
              "32495                          30  \n",
              "32497                          30  \n",
              "32505                          30  \n",
              "32508                          30  \n",
              "32510                          30  \n",
              "32541                          30  \n",
              "32543                          30  \n",
              "32576                          30  \n",
              "32580                          30  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>32320</th>\n",
              "      <td>55</td>\n",
              "      <td>50000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>2.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>A</td>\n",
              "      <td>20000</td>\n",
              "      <td>7.510000</td>\n",
              "      <td>1</td>\n",
              "      <td>0.40</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32326</th>\n",
              "      <td>64</td>\n",
              "      <td>105000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>E</td>\n",
              "      <td>20000</td>\n",
              "      <td>18.640000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.19</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32353</th>\n",
              "      <td>51</td>\n",
              "      <td>37200</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>B</td>\n",
              "      <td>3000</td>\n",
              "      <td>10.620000</td>\n",
              "      <td>1</td>\n",
              "      <td>0.08</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32357</th>\n",
              "      <td>53</td>\n",
              "      <td>34000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>0.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>B</td>\n",
              "      <td>10000</td>\n",
              "      <td>10.390000</td>\n",
              "      <td>1</td>\n",
              "      <td>0.29</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32366</th>\n",
              "      <td>58</td>\n",
              "      <td>20000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>0.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>4000</td>\n",
              "      <td>9.990000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.20</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32395</th>\n",
              "      <td>51</td>\n",
              "      <td>46560</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>2.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>A</td>\n",
              "      <td>11600</td>\n",
              "      <td>8.940000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32403</th>\n",
              "      <td>60</td>\n",
              "      <td>49008</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>C</td>\n",
              "      <td>12000</td>\n",
              "      <td>13.790000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.24</td>\n",
              "      <td>Y</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32421</th>\n",
              "      <td>57</td>\n",
              "      <td>53000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5800</td>\n",
              "      <td>13.160000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.11</td>\n",
              "      <td>Y</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32446</th>\n",
              "      <td>51</td>\n",
              "      <td>62046</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>8000</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.13</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32448</th>\n",
              "      <td>52</td>\n",
              "      <td>83200</td>\n",
              "      <td>RENT</td>\n",
              "      <td>0.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>C</td>\n",
              "      <td>8000</td>\n",
              "      <td>15.960000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>Y</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32460</th>\n",
              "      <td>51</td>\n",
              "      <td>36000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>3.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>D</td>\n",
              "      <td>10000</td>\n",
              "      <td>15.650000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.28</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32463</th>\n",
              "      <td>53</td>\n",
              "      <td>43200</td>\n",
              "      <td>RENT</td>\n",
              "      <td>0.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>B</td>\n",
              "      <td>10000</td>\n",
              "      <td>9.880000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.23</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32483</th>\n",
              "      <td>55</td>\n",
              "      <td>34000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>0.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>A</td>\n",
              "      <td>12000</td>\n",
              "      <td>6.920000</td>\n",
              "      <td>1</td>\n",
              "      <td>0.35</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32495</th>\n",
              "      <td>58</td>\n",
              "      <td>100000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>6.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>A</td>\n",
              "      <td>12000</td>\n",
              "      <td>8.900000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.12</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32497</th>\n",
              "      <td>63</td>\n",
              "      <td>1782000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>13.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>C</td>\n",
              "      <td>12025</td>\n",
              "      <td>14.270000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.01</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32505</th>\n",
              "      <td>51</td>\n",
              "      <td>94368</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>4.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>A</td>\n",
              "      <td>5000</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.05</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32508</th>\n",
              "      <td>53</td>\n",
              "      <td>95000</td>\n",
              "      <td>OWN</td>\n",
              "      <td>24.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>A</td>\n",
              "      <td>6000</td>\n",
              "      <td>5.420000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.06</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32510</th>\n",
              "      <td>55</td>\n",
              "      <td>97243</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>10.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>E</td>\n",
              "      <td>17000</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.17</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32541</th>\n",
              "      <td>52</td>\n",
              "      <td>163000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>9.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>D</td>\n",
              "      <td>25000</td>\n",
              "      <td>14.110000</td>\n",
              "      <td>1</td>\n",
              "      <td>0.15</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32543</th>\n",
              "      <td>52</td>\n",
              "      <td>720000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>A</td>\n",
              "      <td>7925</td>\n",
              "      <td>8.940000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.01</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32576</th>\n",
              "      <td>57</td>\n",
              "      <td>53000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5800</td>\n",
              "      <td>13.160000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.11</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32580</th>\n",
              "      <td>66</td>\n",
              "      <td>42000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>2.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>B</td>\n",
              "      <td>6475</td>\n",
              "      <td>9.990000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.15</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 285
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "k9Xx1PzPQCeT",
        "outputId": "0809f2af-5e24-4cc2-eccb-6266fad02801"
      },
      "source": [
        "sns.distplot(credit['cb_person_cred_hist_length'], color = 'green')\n",
        "plt.title('Distribution of Credit History Length')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Apb7MN0h4bb6"
      },
      "source": [
        "### person_income"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 345
        },
        "id": "9gmrF_se_NEj",
        "outputId": "e11996cb-7304-4dad-dd44-76c0c14f91cf"
      },
      "source": [
        "credit[credit['person_income'] > 1000000]\n",
        "\n",
        "# applicants who has income 6M is quite make sense, \n",
        "  # becase when we see the data there are 9 applicants who has income more than 1M.\n",
        "# we will keep these rows."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "17833          32        1200000              MORTGAGE                1.0   \n",
              "29119          36        1200000                   OWN               16.0   \n",
              "29120          40        1200000              MORTGAGE                1.0   \n",
              "30049          42        2039784                  RENT                0.0   \n",
              "31922          47        1362000              MORTGAGE                9.0   \n",
              "31924          44        1440000              MORTGAGE                7.0   \n",
              "32297         144        6000000              MORTGAGE               12.0   \n",
              "32497          63        1782000                  RENT               13.0   \n",
              "32546          60        1900000              MORTGAGE                5.0   \n",
              "\n",
              "             loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "17833            VENTURE          A      12000       7.510000            0   \n",
              "29119            MEDICAL          A      10000       6.540000            0   \n",
              "29120            MEDICAL          A      10000      11.011695            0   \n",
              "30049            VENTURE          C       8450      12.290000            0   \n",
              "31922            VENTURE          A       6600       7.740000            0   \n",
              "31924  DEBTCONSOLIDATION          A       6400       7.400000            0   \n",
              "32297           PERSONAL          C       5000      12.730000            0   \n",
              "32497          EDUCATION          C      12025      14.270000            0   \n",
              "32546           PERSONAL          A       1500      11.011695            0   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "17833                 0.01                         N   \n",
              "29119                 0.01                         N   \n",
              "29120                 0.01                         N   \n",
              "30049                 0.00                         Y   \n",
              "31922                 0.00                         N   \n",
              "31924                 0.00                         N   \n",
              "32297                 0.00                         N   \n",
              "32497                 0.01                         N   \n",
              "32546                 0.00                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "17833                           8  \n",
              "29119                          11  \n",
              "29120                          16  \n",
              "30049                          15  \n",
              "31922                          17  \n",
              "31924                          15  \n",
              "32297                          25  \n",
              "32497                          30  \n",
              "32546                          21  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>17833</th>\n",
              "      <td>32</td>\n",
              "      <td>1200000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>12000</td>\n",
              "      <td>7.510000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.01</td>\n",
              "      <td>N</td>\n",
              "      <td>8</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29119</th>\n",
              "      <td>36</td>\n",
              "      <td>1200000</td>\n",
              "      <td>OWN</td>\n",
              "      <td>16.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>A</td>\n",
              "      <td>10000</td>\n",
              "      <td>6.540000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.01</td>\n",
              "      <td>N</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29120</th>\n",
              "      <td>40</td>\n",
              "      <td>1200000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>A</td>\n",
              "      <td>10000</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.01</td>\n",
              "      <td>N</td>\n",
              "      <td>16</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>30049</th>\n",
              "      <td>42</td>\n",
              "      <td>2039784</td>\n",
              "      <td>RENT</td>\n",
              "      <td>0.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>C</td>\n",
              "      <td>8450</td>\n",
              "      <td>12.290000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>Y</td>\n",
              "      <td>15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>31922</th>\n",
              "      <td>47</td>\n",
              "      <td>1362000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>9.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>6600</td>\n",
              "      <td>7.740000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>N</td>\n",
              "      <td>17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>31924</th>\n",
              "      <td>44</td>\n",
              "      <td>1440000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>7.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>A</td>\n",
              "      <td>6400</td>\n",
              "      <td>7.400000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>N</td>\n",
              "      <td>15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32297</th>\n",
              "      <td>144</td>\n",
              "      <td>6000000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>12.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5000</td>\n",
              "      <td>12.730000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>N</td>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32497</th>\n",
              "      <td>63</td>\n",
              "      <td>1782000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>13.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>C</td>\n",
              "      <td>12025</td>\n",
              "      <td>14.270000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.01</td>\n",
              "      <td>N</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32546</th>\n",
              "      <td>60</td>\n",
              "      <td>1900000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>5.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>A</td>\n",
              "      <td>1500</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>N</td>\n",
              "      <td>21</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 287
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "AZ2s3DOrQKGP",
        "outputId": "0caf14e7-baa5-4b8e-907a-5211e002319c"
      },
      "source": [
        "sns.distplot(credit['person_income'], color = 'green')\n",
        "plt.title('Distribution of Person Income')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HW-pefXD24QT"
      },
      "source": [
        "### person_emp_length"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "OcW6Kify9_2z",
        "outputId": "a0818bbf-0de0-487d-aac6-941128223c19"
      },
      "source": [
        "credit[(credit['person_emp_length'] == 123) | (credit['person_age'] == 144)]\n",
        "\n",
        "# index 0 person_emp_length = 123\n",
        "# index 81 person_age = 144\n",
        "# index 183 person_age = 144\n",
        "# index 210 person_emp_length = 123\n",
        "# index 32297 person_age = 144"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "0              22          59000                  RENT              123.0   \n",
              "81            144         250000                  RENT                4.0   \n",
              "183           144         200000              MORTGAGE                4.0   \n",
              "210            21         192000              MORTGAGE              123.0   \n",
              "32297         144        6000000              MORTGAGE               12.0   \n",
              "\n",
              "      loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "0        PERSONAL          D      35000          16.02            1   \n",
              "81        VENTURE          C       4800          13.57            0   \n",
              "183     EDUCATION          B       6000          11.86            0   \n",
              "210       VENTURE          A      20000           6.54            0   \n",
              "32297    PERSONAL          C       5000          12.73            0   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "0                     0.59                         Y   \n",
              "81                    0.02                         N   \n",
              "183                   0.03                         N   \n",
              "210                   0.10                         N   \n",
              "32297                 0.00                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "0                               3  \n",
              "81                              3  \n",
              "183                             2  \n",
              "210                             4  \n",
              "32297                          25  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>22</td>\n",
              "      <td>59000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>123.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>D</td>\n",
              "      <td>35000</td>\n",
              "      <td>16.02</td>\n",
              "      <td>1</td>\n",
              "      <td>0.59</td>\n",
              "      <td>Y</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>81</th>\n",
              "      <td>144</td>\n",
              "      <td>250000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>C</td>\n",
              "      <td>4800</td>\n",
              "      <td>13.57</td>\n",
              "      <td>0</td>\n",
              "      <td>0.02</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>183</th>\n",
              "      <td>144</td>\n",
              "      <td>200000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>4.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>6000</td>\n",
              "      <td>11.86</td>\n",
              "      <td>0</td>\n",
              "      <td>0.03</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>210</th>\n",
              "      <td>21</td>\n",
              "      <td>192000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>123.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>20000</td>\n",
              "      <td>6.54</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32297</th>\n",
              "      <td>144</td>\n",
              "      <td>6000000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>12.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5000</td>\n",
              "      <td>12.73</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>N</td>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 289
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 345
        },
        "id": "ZgYdWADhDNTC",
        "outputId": "62a01989-5aa5-48ab-dfd1-14009f2b9a19"
      },
      "source": [
        "credit[credit['person_emp_length'] > 30]\n",
        "\n",
        "# we prove that applicant who has employment length above 30 years is not make any sense\n",
        "# by distribution, there are only 2 applicants who has employment length above 30 years and their age is around 20 years old\n",
        "# we will delete these rows (index 0 and 210)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "0              22          59000                  RENT              123.0   \n",
              "210            21         192000              MORTGAGE              123.0   \n",
              "30914          48          70000              MORTGAGE               31.0   \n",
              "31866          47         178000                   OWN               31.0   \n",
              "31867          46         180000              MORTGAGE               31.0   \n",
              "32263          46         180000              MORTGAGE               31.0   \n",
              "32355          78          48000                  RENT               41.0   \n",
              "32428          58          49000              MORTGAGE               34.0   \n",
              "32515          53         106000              MORTGAGE               38.0   \n",
              "\n",
              "           loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "0             PERSONAL          D      35000          16.02            1   \n",
              "210            VENTURE          A      20000           6.54            0   \n",
              "30914  HOMEIMPROVEMENT          D       9000          14.54            0   \n",
              "31866          VENTURE          B       9000          10.99            0   \n",
              "31867          VENTURE          B      18000           9.91            0   \n",
              "32263          MEDICAL          B      18000           9.91            0   \n",
              "32355          MEDICAL          A       3000           7.51            0   \n",
              "32428          MEDICAL          D       7500          13.55            1   \n",
              "32515         PERSONAL          B      20000           9.88            0   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "0                     0.59                         Y   \n",
              "210                   0.10                         N   \n",
              "30914                 0.13                         N   \n",
              "31866                 0.05                         N   \n",
              "31867                 0.10                         N   \n",
              "32263                 0.10                         N   \n",
              "32355                 0.06                         N   \n",
              "32428                 0.15                         Y   \n",
              "32515                 0.19                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "0                               3  \n",
              "210                             4  \n",
              "30914                          17  \n",
              "31866                          17  \n",
              "31867                          11  \n",
              "32263                          12  \n",
              "32355                          25  \n",
              "32428                          24  \n",
              "32515                          23  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>22</td>\n",
              "      <td>59000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>123.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>D</td>\n",
              "      <td>35000</td>\n",
              "      <td>16.02</td>\n",
              "      <td>1</td>\n",
              "      <td>0.59</td>\n",
              "      <td>Y</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>210</th>\n",
              "      <td>21</td>\n",
              "      <td>192000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>123.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>20000</td>\n",
              "      <td>6.54</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>30914</th>\n",
              "      <td>48</td>\n",
              "      <td>70000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>31.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>D</td>\n",
              "      <td>9000</td>\n",
              "      <td>14.54</td>\n",
              "      <td>0</td>\n",
              "      <td>0.13</td>\n",
              "      <td>N</td>\n",
              "      <td>17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>31866</th>\n",
              "      <td>47</td>\n",
              "      <td>178000</td>\n",
              "      <td>OWN</td>\n",
              "      <td>31.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>B</td>\n",
              "      <td>9000</td>\n",
              "      <td>10.99</td>\n",
              "      <td>0</td>\n",
              "      <td>0.05</td>\n",
              "      <td>N</td>\n",
              "      <td>17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>31867</th>\n",
              "      <td>46</td>\n",
              "      <td>180000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>31.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>B</td>\n",
              "      <td>18000</td>\n",
              "      <td>9.91</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32263</th>\n",
              "      <td>46</td>\n",
              "      <td>180000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>31.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>B</td>\n",
              "      <td>18000</td>\n",
              "      <td>9.91</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>12</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32355</th>\n",
              "      <td>78</td>\n",
              "      <td>48000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>41.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>A</td>\n",
              "      <td>3000</td>\n",
              "      <td>7.51</td>\n",
              "      <td>0</td>\n",
              "      <td>0.06</td>\n",
              "      <td>N</td>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32428</th>\n",
              "      <td>58</td>\n",
              "      <td>49000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>34.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>D</td>\n",
              "      <td>7500</td>\n",
              "      <td>13.55</td>\n",
              "      <td>1</td>\n",
              "      <td>0.15</td>\n",
              "      <td>Y</td>\n",
              "      <td>24</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32515</th>\n",
              "      <td>53</td>\n",
              "      <td>106000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>38.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>B</td>\n",
              "      <td>20000</td>\n",
              "      <td>9.88</td>\n",
              "      <td>0</td>\n",
              "      <td>0.19</td>\n",
              "      <td>N</td>\n",
              "      <td>23</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 290
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1z6YnMtVNkJ1",
        "outputId": "8d80890e-233a-43c6-9287-a19e5b47f79f"
      },
      "source": [
        "print(f\"Before dropping the outliers, numbers of data is {len(credit['person_emp_length'])} rows.\")\n",
        "\n",
        "print(\"dropping . . .\")\n",
        "credit.drop(index = [0, 210], inplace = True)\n",
        "\n",
        "print(f\"After dropping the outliers, numbers of data is {len(credit['person_emp_length'])} rows.\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Before dropping the outliers, numbers of data is 32581 rows.\n",
            "dropping . . .\n",
            "After dropping the outliers, numbers of data is 32579 rows.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 284
        },
        "id": "MFthEe0iO0Ni",
        "outputId": "92e1782c-09f6-4c2f-f689-19d8265ed6b2"
      },
      "source": [
        "credit[credit['person_emp_length'] > 30]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "30914          48          70000              MORTGAGE               31.0   \n",
              "31866          47         178000                   OWN               31.0   \n",
              "31867          46         180000              MORTGAGE               31.0   \n",
              "32263          46         180000              MORTGAGE               31.0   \n",
              "32355          78          48000                  RENT               41.0   \n",
              "32428          58          49000              MORTGAGE               34.0   \n",
              "32515          53         106000              MORTGAGE               38.0   \n",
              "\n",
              "           loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "30914  HOMEIMPROVEMENT          D       9000          14.54            0   \n",
              "31866          VENTURE          B       9000          10.99            0   \n",
              "31867          VENTURE          B      18000           9.91            0   \n",
              "32263          MEDICAL          B      18000           9.91            0   \n",
              "32355          MEDICAL          A       3000           7.51            0   \n",
              "32428          MEDICAL          D       7500          13.55            1   \n",
              "32515         PERSONAL          B      20000           9.88            0   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "30914                 0.13                         N   \n",
              "31866                 0.05                         N   \n",
              "31867                 0.10                         N   \n",
              "32263                 0.10                         N   \n",
              "32355                 0.06                         N   \n",
              "32428                 0.15                         Y   \n",
              "32515                 0.19                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "30914                          17  \n",
              "31866                          17  \n",
              "31867                          11  \n",
              "32263                          12  \n",
              "32355                          25  \n",
              "32428                          24  \n",
              "32515                          23  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>30914</th>\n",
              "      <td>48</td>\n",
              "      <td>70000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>31.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>D</td>\n",
              "      <td>9000</td>\n",
              "      <td>14.54</td>\n",
              "      <td>0</td>\n",
              "      <td>0.13</td>\n",
              "      <td>N</td>\n",
              "      <td>17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>31866</th>\n",
              "      <td>47</td>\n",
              "      <td>178000</td>\n",
              "      <td>OWN</td>\n",
              "      <td>31.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>B</td>\n",
              "      <td>9000</td>\n",
              "      <td>10.99</td>\n",
              "      <td>0</td>\n",
              "      <td>0.05</td>\n",
              "      <td>N</td>\n",
              "      <td>17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>31867</th>\n",
              "      <td>46</td>\n",
              "      <td>180000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>31.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>B</td>\n",
              "      <td>18000</td>\n",
              "      <td>9.91</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32263</th>\n",
              "      <td>46</td>\n",
              "      <td>180000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>31.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>B</td>\n",
              "      <td>18000</td>\n",
              "      <td>9.91</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>12</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32355</th>\n",
              "      <td>78</td>\n",
              "      <td>48000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>41.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>A</td>\n",
              "      <td>3000</td>\n",
              "      <td>7.51</td>\n",
              "      <td>0</td>\n",
              "      <td>0.06</td>\n",
              "      <td>N</td>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32428</th>\n",
              "      <td>58</td>\n",
              "      <td>49000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>34.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>D</td>\n",
              "      <td>7500</td>\n",
              "      <td>13.55</td>\n",
              "      <td>1</td>\n",
              "      <td>0.15</td>\n",
              "      <td>Y</td>\n",
              "      <td>24</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32515</th>\n",
              "      <td>53</td>\n",
              "      <td>106000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>38.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>B</td>\n",
              "      <td>20000</td>\n",
              "      <td>9.88</td>\n",
              "      <td>0</td>\n",
              "      <td>0.19</td>\n",
              "      <td>N</td>\n",
              "      <td>23</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 292
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "Y10qIoWoP0pV",
        "outputId": "633f2072-2df3-4fb6-a668-735ae5228dcd"
      },
      "source": [
        "sns.distplot(credit['person_emp_length'], color = 'green')\n",
        "plt.title('Distribution of Employee Length')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5pxUGVf04YZe"
      },
      "source": [
        "### person_age"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 284
        },
        "id": "GA7_Kle9D3Ph",
        "outputId": "5dfe3d9c-68bd-484c-be8f-5fbec64b0508"
      },
      "source": [
        "credit[credit['person_age'] > 80]\n",
        "\n",
        "# applicant who has age until 123 years old seemse fine to me,\n",
        "# except they are who 144 years old.\n",
        "# when you search it through search engine, there's only one who lived around 144 years old.\n",
        "# we will drop these rows (index 81, 183, and 32297)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "81            144         250000                  RENT                4.0   \n",
              "183           144         200000              MORTGAGE                4.0   \n",
              "575           123          80004                  RENT                2.0   \n",
              "747           123          78000                  RENT                7.0   \n",
              "32297         144        6000000              MORTGAGE               12.0   \n",
              "32416          94          24000                  RENT                1.0   \n",
              "32506          84          94800              MORTGAGE                2.0   \n",
              "\n",
              "      loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "81        VENTURE          C       4800      13.570000            0   \n",
              "183     EDUCATION          B       6000      11.860000            0   \n",
              "575     EDUCATION          B      20400      10.250000            0   \n",
              "747       VENTURE          B      20000      11.011695            0   \n",
              "32297    PERSONAL          C       5000      12.730000            0   \n",
              "32416     MEDICAL          C       6500      11.011695            0   \n",
              "32506    PERSONAL          A      10000       7.510000            0   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "81                    0.02                         N   \n",
              "183                   0.03                         N   \n",
              "575                   0.25                         N   \n",
              "747                   0.26                         N   \n",
              "32297                 0.00                         N   \n",
              "32416                 0.27                         N   \n",
              "32506                 0.11                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "81                              3  \n",
              "183                             2  \n",
              "575                             3  \n",
              "747                             4  \n",
              "32297                          25  \n",
              "32416                          27  \n",
              "32506                          24  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>81</th>\n",
              "      <td>144</td>\n",
              "      <td>250000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>C</td>\n",
              "      <td>4800</td>\n",
              "      <td>13.570000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.02</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>183</th>\n",
              "      <td>144</td>\n",
              "      <td>200000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>4.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>6000</td>\n",
              "      <td>11.860000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.03</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>575</th>\n",
              "      <td>123</td>\n",
              "      <td>80004</td>\n",
              "      <td>RENT</td>\n",
              "      <td>2.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>20400</td>\n",
              "      <td>10.250000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>747</th>\n",
              "      <td>123</td>\n",
              "      <td>78000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>7.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>B</td>\n",
              "      <td>20000</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.26</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32297</th>\n",
              "      <td>144</td>\n",
              "      <td>6000000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>12.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5000</td>\n",
              "      <td>12.730000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>N</td>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32416</th>\n",
              "      <td>94</td>\n",
              "      <td>24000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>6500</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.27</td>\n",
              "      <td>N</td>\n",
              "      <td>27</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32506</th>\n",
              "      <td>84</td>\n",
              "      <td>94800</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>2.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>A</td>\n",
              "      <td>10000</td>\n",
              "      <td>7.510000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.11</td>\n",
              "      <td>N</td>\n",
              "      <td>24</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 294
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VZ-AVu8LPM-f",
        "outputId": "a533954f-73be-4f92-a58a-de47744d533d"
      },
      "source": [
        "print(f\"Before dropping the outliers, numbers of data is {len(credit)} rows.\")\n",
        "\n",
        "print(\"dropping . . .\")\n",
        "credit.drop(index = [81, 183, 32297], inplace = True)\n",
        "\n",
        "print(f\"After dropping the outliers, numbers of data is {len(credit)} rows.\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Before dropping the outliers, numbers of data is 32579 rows.\n",
            "dropping . . .\n",
            "After dropping the outliers, numbers of data is 32576 rows.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 192
        },
        "id": "ApzaVdVdPrzY",
        "outputId": "cb50dfc7-33d2-44c0-c729-e2e1e0d944e2"
      },
      "source": [
        "credit[credit['person_age'] > 80]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "575           123          80004                  RENT                2.0   \n",
              "747           123          78000                  RENT                7.0   \n",
              "32416          94          24000                  RENT                1.0   \n",
              "32506          84          94800              MORTGAGE                2.0   \n",
              "\n",
              "      loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "575     EDUCATION          B      20400      10.250000            0   \n",
              "747       VENTURE          B      20000      11.011695            0   \n",
              "32416     MEDICAL          C       6500      11.011695            0   \n",
              "32506    PERSONAL          A      10000       7.510000            0   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "575                   0.25                         N   \n",
              "747                   0.26                         N   \n",
              "32416                 0.27                         N   \n",
              "32506                 0.11                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "575                             3  \n",
              "747                             4  \n",
              "32416                          27  \n",
              "32506                          24  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>575</th>\n",
              "      <td>123</td>\n",
              "      <td>80004</td>\n",
              "      <td>RENT</td>\n",
              "      <td>2.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>20400</td>\n",
              "      <td>10.250000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>747</th>\n",
              "      <td>123</td>\n",
              "      <td>78000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>7.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>B</td>\n",
              "      <td>20000</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.26</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32416</th>\n",
              "      <td>94</td>\n",
              "      <td>24000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>6500</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>0</td>\n",
              "      <td>0.27</td>\n",
              "      <td>N</td>\n",
              "      <td>27</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32506</th>\n",
              "      <td>84</td>\n",
              "      <td>94800</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>2.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>A</td>\n",
              "      <td>10000</td>\n",
              "      <td>7.510000</td>\n",
              "      <td>0</td>\n",
              "      <td>0.11</td>\n",
              "      <td>N</td>\n",
              "      <td>24</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 296
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "9Sv6BAedQU9H",
        "outputId": "22ea5b60-791e-45ea-f4dd-cf3944778d5a"
      },
      "source": [
        "sns.distplot(credit['person_age'], color = 'green')\n",
        "plt.title('Distribution of Person Age')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\distributions.py:2619: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3m3LuAwk4err"
      },
      "source": [
        "### loan_amnt"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 435
        },
        "id": "pX2th-3cQwu2",
        "outputId": "1f8fd804-2680-409a-b3b6-8b108b5e7b99"
      },
      "source": [
        "credit[credit['loan_amnt'] > 34000.00]\n",
        "\n",
        "# loan amount around 30k seems fine to me.\n",
        "# and we will keep it for training materials."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "3              23          65500                  RENT                4.0   \n",
              "4              24          54400                  RENT                8.0   \n",
              "6              26          77100                  RENT                8.0   \n",
              "7              24          78956                  RENT                5.0   \n",
              "8              24          83000                  RENT                8.0   \n",
              "...           ...            ...                   ...                ...   \n",
              "32158          37          85000                  RENT                5.0   \n",
              "32299          65          76000                  RENT                3.0   \n",
              "32301          53          85000                  RENT               19.0   \n",
              "32302          58         108350                  RENT                5.0   \n",
              "32578          65          76000                  RENT                3.0   \n",
              "\n",
              "             loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "3                MEDICAL          C      35000          15.23            1   \n",
              "4                MEDICAL          C      35000          14.27            1   \n",
              "6              EDUCATION          B      35000          12.42            1   \n",
              "7                MEDICAL          B      35000          11.11            1   \n",
              "8               PERSONAL          A      35000           8.90            1   \n",
              "...                  ...        ...        ...            ...          ...   \n",
              "32158  DEBTCONSOLIDATION          A      35000           8.90            1   \n",
              "32299          EDUCATION          B      35000          10.99            1   \n",
              "32301           PERSONAL          D      35000          15.65            1   \n",
              "32302           PERSONAL          D      35000          14.91            1   \n",
              "32578    HOMEIMPROVEMENT          B      35000          10.99            1   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "3                     0.53                         N   \n",
              "4                     0.55                         Y   \n",
              "6                     0.45                         N   \n",
              "7                     0.44                         N   \n",
              "8                     0.42                         N   \n",
              "...                    ...                       ...   \n",
              "32158                 0.41                         N   \n",
              "32299                 0.46                         N   \n",
              "32301                 0.41                         N   \n",
              "32302                 0.32                         N   \n",
              "32578                 0.46                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "3                               2  \n",
              "4                               4  \n",
              "6                               3  \n",
              "7                               4  \n",
              "8                               2  \n",
              "...                           ...  \n",
              "32158                          17  \n",
              "32299                          27  \n",
              "32301                          19  \n",
              "32302                          19  \n",
              "32578                          28  \n",
              "\n",
              "[185 rows x 12 columns]"
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>26</td>\n",
              "      <td>77100</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>35000</td>\n",
              "      <td>12.42</td>\n",
              "      <td>1</td>\n",
              "      <td>0.45</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>24</td>\n",
              "      <td>78956</td>\n",
              "      <td>RENT</td>\n",
              "      <td>5.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>B</td>\n",
              "      <td>35000</td>\n",
              "      <td>11.11</td>\n",
              "      <td>1</td>\n",
              "      <td>0.44</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>24</td>\n",
              "      <td>83000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>A</td>\n",
              "      <td>35000</td>\n",
              "      <td>8.90</td>\n",
              "      <td>1</td>\n",
              "      <td>0.42</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32158</th>\n",
              "      <td>37</td>\n",
              "      <td>85000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>5.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>A</td>\n",
              "      <td>35000</td>\n",
              "      <td>8.90</td>\n",
              "      <td>1</td>\n",
              "      <td>0.41</td>\n",
              "      <td>N</td>\n",
              "      <td>17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32299</th>\n",
              "      <td>65</td>\n",
              "      <td>76000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>3.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>35000</td>\n",
              "      <td>10.99</td>\n",
              "      <td>1</td>\n",
              "      <td>0.46</td>\n",
              "      <td>N</td>\n",
              "      <td>27</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32301</th>\n",
              "      <td>53</td>\n",
              "      <td>85000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>19.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>D</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.65</td>\n",
              "      <td>1</td>\n",
              "      <td>0.41</td>\n",
              "      <td>N</td>\n",
              "      <td>19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32302</th>\n",
              "      <td>58</td>\n",
              "      <td>108350</td>\n",
              "      <td>RENT</td>\n",
              "      <td>5.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>D</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.91</td>\n",
              "      <td>1</td>\n",
              "      <td>0.32</td>\n",
              "      <td>N</td>\n",
              "      <td>19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32578</th>\n",
              "      <td>65</td>\n",
              "      <td>76000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>3.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>B</td>\n",
              "      <td>35000</td>\n",
              "      <td>10.99</td>\n",
              "      <td>1</td>\n",
              "      <td>0.46</td>\n",
              "      <td>N</td>\n",
              "      <td>28</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>185 rows × 12 columns</p>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 298
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z0RimLOl4f78"
      },
      "source": [
        "### loan_int_rate"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 837
        },
        "id": "NG-etSVwS5gj",
        "outputId": "166ebaa5-0428-46e1-fcc9-6875966a0ef5"
      },
      "source": [
        "credit[credit['loan_int_rate'] > 21]\n",
        "\n",
        "# we can't delete these rows, because it makes sense since the greater the interest rate, the bigger the risk of applicants could default their credit\n",
        "# and we just have 21.82% of default target (1), if we drop it, we could enlarge the gap between these imbalance class.\n",
        "# we keep these data."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "523            23          56000                  RENT                8.0   \n",
              "1827           24          30000                  RENT                5.0   \n",
              "2436           22          23000                  RENT                1.0   \n",
              "2897           22          35000              MORTGAGE                4.0   \n",
              "4899           24          33000                  RENT                1.0   \n",
              "6449           23          34255                  RENT                7.0   \n",
              "7470           24          30500                  RENT                2.0   \n",
              "13499          26          44000                  RENT                4.0   \n",
              "14487          26          96000              MORTGAGE                2.0   \n",
              "15223          23         113000              MORTGAGE                0.0   \n",
              "15744          24          44400                  RENT                4.0   \n",
              "17103          24          30500                  RENT                2.0   \n",
              "17989          29         162350                  RENT               12.0   \n",
              "18829          31          22000                  RENT                5.0   \n",
              "19272          34          28800                  RENT                0.0   \n",
              "24834          27          70800              MORTGAGE                0.0   \n",
              "25005          33          79985              MORTGAGE                6.0   \n",
              "25230          29          75840                   OWN                5.0   \n",
              "26083          31         144000                  RENT                0.0   \n",
              "27425          28         128000              MORTGAGE               11.0   \n",
              "27578          33         138000              MORTGAGE               18.0   \n",
              "27633          35          75000                  RENT               13.0   \n",
              "27851          28         288000              MORTGAGE               12.0   \n",
              "28228          34          28800                  RENT                0.0   \n",
              "29277          38          83000                  RENT                3.0   \n",
              "\n",
              "             loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "523              MEDICAL          G      21600          21.21            1   \n",
              "1827             MEDICAL          G       2000          21.27            1   \n",
              "2436             MEDICAL          F       2800          21.74            1   \n",
              "2897             VENTURE          F       5000          21.36            0   \n",
              "4899             MEDICAL          F       5000          21.36            1   \n",
              "6449             VENTURE          G       6000          22.11            1   \n",
              "7470           EDUCATION          F       6800          21.74            1   \n",
              "13499    HOMEIMPROVEMENT          G      12500          22.11            1   \n",
              "14487            VENTURE          G      10000          21.14            1   \n",
              "15223            VENTURE          G      19750          21.21            1   \n",
              "15744          EDUCATION          F      15650          21.74            1   \n",
              "17103           PERSONAL          F       6800          21.74            1   \n",
              "17989            MEDICAL          G      25000          21.64            1   \n",
              "18829            VENTURE          F       1400          22.06            0   \n",
              "19272           PERSONAL          F       2500          21.36            1   \n",
              "24834          EDUCATION          G      20000          21.21            1   \n",
              "25005          EDUCATION          F       8000          21.36            0   \n",
              "25230    HOMEIMPROVEMENT          G      25000          21.27            1   \n",
              "26083          EDUCATION          G      13000          21.21            1   \n",
              "27425  DEBTCONSOLIDATION          G      14400          22.11            1   \n",
              "27578    HOMEIMPROVEMENT          G      25000          21.27            1   \n",
              "27633            VENTURE          F      15175          21.74            0   \n",
              "27851            VENTURE          G      18000          22.48            1   \n",
              "28228  DEBTCONSOLIDATION          F       2500          21.36            1   \n",
              "29277    HOMEIMPROVEMENT          G      22250          23.22            1   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "523                   0.39                         Y   \n",
              "1827                  0.07                         Y   \n",
              "2436                  0.12                         N   \n",
              "2897                  0.14                         Y   \n",
              "4899                  0.15                         N   \n",
              "6449                  0.18                         Y   \n",
              "7470                  0.22                         Y   \n",
              "13499                 0.28                         Y   \n",
              "14487                 0.10                         N   \n",
              "15223                 0.17                         N   \n",
              "15744                 0.35                         N   \n",
              "17103                 0.22                         Y   \n",
              "17989                 0.13                         Y   \n",
              "18829                 0.06                         Y   \n",
              "19272                 0.09                         N   \n",
              "24834                 0.28                         Y   \n",
              "25005                 0.10                         N   \n",
              "25230                 0.33                         N   \n",
              "26083                 0.09                         N   \n",
              "27425                 0.11                         Y   \n",
              "27578                 0.18                         Y   \n",
              "27633                 0.20                         Y   \n",
              "27851                 0.06                         Y   \n",
              "28228                 0.09                         N   \n",
              "29277                 0.27                         Y   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "523                             4  \n",
              "1827                            3  \n",
              "2436                            4  \n",
              "2897                            2  \n",
              "4899                            2  \n",
              "6449                            3  \n",
              "7470                            3  \n",
              "13499                           2  \n",
              "14487                           3  \n",
              "15223                           4  \n",
              "15744                           4  \n",
              "17103                           4  \n",
              "17989                           9  \n",
              "18829                           7  \n",
              "19272                           9  \n",
              "24834                           8  \n",
              "25005                           5  \n",
              "25230                           6  \n",
              "26083                           6  \n",
              "27425                           8  \n",
              "27578                          10  \n",
              "27633                           8  \n",
              "27851                           9  \n",
              "28228                           7  \n",
              "29277                          12  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>523</th>\n",
              "      <td>23</td>\n",
              "      <td>56000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>G</td>\n",
              "      <td>21600</td>\n",
              "      <td>21.21</td>\n",
              "      <td>1</td>\n",
              "      <td>0.39</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1827</th>\n",
              "      <td>24</td>\n",
              "      <td>30000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>5.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>G</td>\n",
              "      <td>2000</td>\n",
              "      <td>21.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.07</td>\n",
              "      <td>Y</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2436</th>\n",
              "      <td>22</td>\n",
              "      <td>23000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>F</td>\n",
              "      <td>2800</td>\n",
              "      <td>21.74</td>\n",
              "      <td>1</td>\n",
              "      <td>0.12</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2897</th>\n",
              "      <td>22</td>\n",
              "      <td>35000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>4.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>F</td>\n",
              "      <td>5000</td>\n",
              "      <td>21.36</td>\n",
              "      <td>0</td>\n",
              "      <td>0.14</td>\n",
              "      <td>Y</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4899</th>\n",
              "      <td>24</td>\n",
              "      <td>33000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>F</td>\n",
              "      <td>5000</td>\n",
              "      <td>21.36</td>\n",
              "      <td>1</td>\n",
              "      <td>0.15</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6449</th>\n",
              "      <td>23</td>\n",
              "      <td>34255</td>\n",
              "      <td>RENT</td>\n",
              "      <td>7.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>G</td>\n",
              "      <td>6000</td>\n",
              "      <td>22.11</td>\n",
              "      <td>1</td>\n",
              "      <td>0.18</td>\n",
              "      <td>Y</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7470</th>\n",
              "      <td>24</td>\n",
              "      <td>30500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>2.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>F</td>\n",
              "      <td>6800</td>\n",
              "      <td>21.74</td>\n",
              "      <td>1</td>\n",
              "      <td>0.22</td>\n",
              "      <td>Y</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13499</th>\n",
              "      <td>26</td>\n",
              "      <td>44000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>G</td>\n",
              "      <td>12500</td>\n",
              "      <td>22.11</td>\n",
              "      <td>1</td>\n",
              "      <td>0.28</td>\n",
              "      <td>Y</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14487</th>\n",
              "      <td>26</td>\n",
              "      <td>96000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>G</td>\n",
              "      <td>10000</td>\n",
              "      <td>21.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15223</th>\n",
              "      <td>23</td>\n",
              "      <td>113000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>0.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>G</td>\n",
              "      <td>19750</td>\n",
              "      <td>21.21</td>\n",
              "      <td>1</td>\n",
              "      <td>0.17</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15744</th>\n",
              "      <td>24</td>\n",
              "      <td>44400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>F</td>\n",
              "      <td>15650</td>\n",
              "      <td>21.74</td>\n",
              "      <td>1</td>\n",
              "      <td>0.35</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17103</th>\n",
              "      <td>24</td>\n",
              "      <td>30500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>2.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>F</td>\n",
              "      <td>6800</td>\n",
              "      <td>21.74</td>\n",
              "      <td>1</td>\n",
              "      <td>0.22</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17989</th>\n",
              "      <td>29</td>\n",
              "      <td>162350</td>\n",
              "      <td>RENT</td>\n",
              "      <td>12.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>G</td>\n",
              "      <td>25000</td>\n",
              "      <td>21.64</td>\n",
              "      <td>1</td>\n",
              "      <td>0.13</td>\n",
              "      <td>Y</td>\n",
              "      <td>9</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18829</th>\n",
              "      <td>31</td>\n",
              "      <td>22000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>5.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>F</td>\n",
              "      <td>1400</td>\n",
              "      <td>22.06</td>\n",
              "      <td>0</td>\n",
              "      <td>0.06</td>\n",
              "      <td>Y</td>\n",
              "      <td>7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19272</th>\n",
              "      <td>34</td>\n",
              "      <td>28800</td>\n",
              "      <td>RENT</td>\n",
              "      <td>0.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>F</td>\n",
              "      <td>2500</td>\n",
              "      <td>21.36</td>\n",
              "      <td>1</td>\n",
              "      <td>0.09</td>\n",
              "      <td>N</td>\n",
              "      <td>9</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24834</th>\n",
              "      <td>27</td>\n",
              "      <td>70800</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>0.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>G</td>\n",
              "      <td>20000</td>\n",
              "      <td>21.21</td>\n",
              "      <td>1</td>\n",
              "      <td>0.28</td>\n",
              "      <td>Y</td>\n",
              "      <td>8</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25005</th>\n",
              "      <td>33</td>\n",
              "      <td>79985</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>6.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>F</td>\n",
              "      <td>8000</td>\n",
              "      <td>21.36</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>5</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25230</th>\n",
              "      <td>29</td>\n",
              "      <td>75840</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>G</td>\n",
              "      <td>25000</td>\n",
              "      <td>21.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.33</td>\n",
              "      <td>N</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>26083</th>\n",
              "      <td>31</td>\n",
              "      <td>144000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>0.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>G</td>\n",
              "      <td>13000</td>\n",
              "      <td>21.21</td>\n",
              "      <td>1</td>\n",
              "      <td>0.09</td>\n",
              "      <td>N</td>\n",
              "      <td>6</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27425</th>\n",
              "      <td>28</td>\n",
              "      <td>128000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>11.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>G</td>\n",
              "      <td>14400</td>\n",
              "      <td>22.11</td>\n",
              "      <td>1</td>\n",
              "      <td>0.11</td>\n",
              "      <td>Y</td>\n",
              "      <td>8</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27578</th>\n",
              "      <td>33</td>\n",
              "      <td>138000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>18.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>G</td>\n",
              "      <td>25000</td>\n",
              "      <td>21.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.18</td>\n",
              "      <td>Y</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27633</th>\n",
              "      <td>35</td>\n",
              "      <td>75000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>13.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>F</td>\n",
              "      <td>15175</td>\n",
              "      <td>21.74</td>\n",
              "      <td>0</td>\n",
              "      <td>0.20</td>\n",
              "      <td>Y</td>\n",
              "      <td>8</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27851</th>\n",
              "      <td>28</td>\n",
              "      <td>288000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>12.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>G</td>\n",
              "      <td>18000</td>\n",
              "      <td>22.48</td>\n",
              "      <td>1</td>\n",
              "      <td>0.06</td>\n",
              "      <td>Y</td>\n",
              "      <td>9</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>28228</th>\n",
              "      <td>34</td>\n",
              "      <td>28800</td>\n",
              "      <td>RENT</td>\n",
              "      <td>0.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>F</td>\n",
              "      <td>2500</td>\n",
              "      <td>21.36</td>\n",
              "      <td>1</td>\n",
              "      <td>0.09</td>\n",
              "      <td>N</td>\n",
              "      <td>7</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29277</th>\n",
              "      <td>38</td>\n",
              "      <td>83000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>3.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>G</td>\n",
              "      <td>22250</td>\n",
              "      <td>23.22</td>\n",
              "      <td>1</td>\n",
              "      <td>0.27</td>\n",
              "      <td>Y</td>\n",
              "      <td>12</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 299
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ggX3Pnmz4iLi"
      },
      "source": [
        "### loan_percent_income"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 435
        },
        "id": "LnjtURiiT3Q7",
        "outputId": "6961f9da-b4a2-4b8a-88d8-24845d2cd588"
      },
      "source": [
        "credit[credit['loan_percent_income'] > 0.4]\n",
        "\n",
        "# we can't delete these rows too, because it contains the default target, and we cant drop it again\n",
        "# since it could enlarge the gap the imbalance class."
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "2              25           9600              MORTGAGE                1.0   \n",
              "3              23          65500                  RENT                4.0   \n",
              "4              24          54400                  RENT                8.0   \n",
              "6              26          77100                  RENT                8.0   \n",
              "7              24          78956                  RENT                5.0   \n",
              "...           ...            ...                   ...                ...   \n",
              "32317          66          43000                  RENT               17.0   \n",
              "32318          60          45600                  RENT                1.0   \n",
              "32369          54          39000              MORTGAGE                4.0   \n",
              "32571          60          45600                  RENT                1.0   \n",
              "32578          65          76000                  RENT                3.0   \n",
              "\n",
              "             loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "2                MEDICAL          C       5500          12.87            1   \n",
              "3                MEDICAL          C      35000          15.23            1   \n",
              "4                MEDICAL          C      35000          14.27            1   \n",
              "6              EDUCATION          B      35000          12.42            1   \n",
              "7                MEDICAL          B      35000          11.11            1   \n",
              "...                  ...        ...        ...            ...          ...   \n",
              "32317  DEBTCONSOLIDATION          D      22250          14.59            1   \n",
              "32318           PERSONAL          B      20000          10.00            1   \n",
              "32369           PERSONAL          B      20000          10.71            0   \n",
              "32571            VENTURE          B      20000          10.00            1   \n",
              "32578    HOMEIMPROVEMENT          B      35000          10.99            1   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "2                     0.57                         N   \n",
              "3                     0.53                         N   \n",
              "4                     0.55                         Y   \n",
              "6                     0.45                         N   \n",
              "7                     0.44                         N   \n",
              "...                    ...                       ...   \n",
              "32317                 0.52                         Y   \n",
              "32318                 0.44                         N   \n",
              "32369                 0.51                         N   \n",
              "32571                 0.44                         N   \n",
              "32578                 0.46                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "2                               3  \n",
              "3                               2  \n",
              "4                               4  \n",
              "6                               3  \n",
              "7                               4  \n",
              "...                           ...  \n",
              "32317                          23  \n",
              "32318                          20  \n",
              "32369                          19  \n",
              "32571                          26  \n",
              "32578                          28  \n",
              "\n",
              "[1119 rows x 12 columns]"
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>26</td>\n",
              "      <td>77100</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>35000</td>\n",
              "      <td>12.42</td>\n",
              "      <td>1</td>\n",
              "      <td>0.45</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>24</td>\n",
              "      <td>78956</td>\n",
              "      <td>RENT</td>\n",
              "      <td>5.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>B</td>\n",
              "      <td>35000</td>\n",
              "      <td>11.11</td>\n",
              "      <td>1</td>\n",
              "      <td>0.44</td>\n",
              "      <td>N</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32317</th>\n",
              "      <td>66</td>\n",
              "      <td>43000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>17.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>D</td>\n",
              "      <td>22250</td>\n",
              "      <td>14.59</td>\n",
              "      <td>1</td>\n",
              "      <td>0.52</td>\n",
              "      <td>Y</td>\n",
              "      <td>23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32318</th>\n",
              "      <td>60</td>\n",
              "      <td>45600</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>B</td>\n",
              "      <td>20000</td>\n",
              "      <td>10.00</td>\n",
              "      <td>1</td>\n",
              "      <td>0.44</td>\n",
              "      <td>N</td>\n",
              "      <td>20</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32369</th>\n",
              "      <td>54</td>\n",
              "      <td>39000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>4.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>B</td>\n",
              "      <td>20000</td>\n",
              "      <td>10.71</td>\n",
              "      <td>0</td>\n",
              "      <td>0.51</td>\n",
              "      <td>N</td>\n",
              "      <td>19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32571</th>\n",
              "      <td>60</td>\n",
              "      <td>45600</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>B</td>\n",
              "      <td>20000</td>\n",
              "      <td>10.00</td>\n",
              "      <td>1</td>\n",
              "      <td>0.44</td>\n",
              "      <td>N</td>\n",
              "      <td>26</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>32578</th>\n",
              "      <td>65</td>\n",
              "      <td>76000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>3.0</td>\n",
              "      <td>HOMEIMPROVEMENT</td>\n",
              "      <td>B</td>\n",
              "      <td>35000</td>\n",
              "      <td>10.99</td>\n",
              "      <td>1</td>\n",
              "      <td>0.46</td>\n",
              "      <td>N</td>\n",
              "      <td>28</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1119 rows × 12 columns</p>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 300
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SMT6ncyRiwpC"
      },
      "source": [
        "## Weight of Evidence (WoE) and Information Value (IV)\n",
        "trying to binning based on IQR."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LFX2_YyXhrZM",
        "outputId": "51a0a6d6-3c12-4875-8535-532c176b5bbc"
      },
      "source": [
        "credit.info()\n",
        "\n",
        "##############################################\n",
        "# # Binning method\n",
        "# bin_price =[0,250000,500000,750000, 1000000, 2000000, df['price'].max()]\n",
        "# label_price = ['upto 250k', 'upto 500k', 'upto 750k','upto 1mil','upto 2mil', 'more than 2 mil']\n",
        "\n",
        "# df['price_group'] = pd.cut(df['price'], bins = bin_price, labels = label_price)\n",
        "# df.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 32576 entries, 1 to 32580\n",
            "Data columns (total 12 columns):\n",
            " #   Column                      Non-Null Count  Dtype  \n",
            "---  ------                      --------------  -----  \n",
            " 0   person_age                  32576 non-null  int64  \n",
            " 1   person_income               32576 non-null  int64  \n",
            " 2   person_home_ownership       32576 non-null  object \n",
            " 3   person_emp_length           32576 non-null  float64\n",
            " 4   loan_intent                 32576 non-null  object \n",
            " 5   loan_grade                  32576 non-null  object \n",
            " 6   loan_amnt                   32576 non-null  int64  \n",
            " 7   loan_int_rate               32576 non-null  float64\n",
            " 8   loan_status                 32576 non-null  int64  \n",
            " 9   loan_percent_income         32576 non-null  float64\n",
            " 10  cb_person_default_on_file   32576 non-null  object \n",
            " 11  cb_person_cred_hist_length  32576 non-null  int64  \n",
            "dtypes: float64(3), int64(5), object(4)\n",
            "memory usage: 3.2+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U04I0REdjwOg"
      },
      "source": [
        "### person_age"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZdclEF1t8iPc",
        "outputId": "0c075508-9440-45f1-ae21-b7d365fa6749"
      },
      "source": [
        "credit['person_age'].describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    32576.000000\n",
              "mean        27.724276\n",
              "std          6.249549\n",
              "min         20.000000\n",
              "25%         23.000000\n",
              "50%         26.000000\n",
              "75%         30.000000\n",
              "max        123.000000\n",
              "Name: person_age, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 302
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "id": "n9tgxD3Uzc0q",
        "outputId": "ba10c8ed-4657-4512-c3a3-57b4b64f7e1d"
      },
      "source": [
        "feature = 'person_age'\n",
        "target = 'loan_status'\n",
        "\n",
        "# binning=[-float(\"inf\"), 30, 50, float(\"inf\")]\n",
        "binning=[-float(\"inf\"), 23,26,30, float(\"inf\")]\n",
        "# binning=[-float(\"inf\"), 30.75,61.5,92.25, float(\"inf\")]\n",
        "\n",
        "bin_feature = pd.cut(credit[feature], bins=binning).values.add_categories('Nan').fillna('Nan')\n",
        "woe = (pd.crosstab(bin_feature,\n",
        "                       credit[target],\n",
        "                       normalize='columns')\n",
        "             .assign(woe=lambda creditx: np.log(creditx[1] / creditx[0]))\n",
        "             .assign(iv=lambda creditx: np.sum(creditx['woe']*\n",
        "                                           (creditx[1]-creditx[0]))))\n",
        "\n",
        "woe"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "loan_status          0         1       woe        iv\n",
              "row_0                                               \n",
              "(-inf, 23.0]  0.262044  0.294076  0.115326  0.005673\n",
              "(23.0, 26.0]  0.278849  0.275925 -0.010540  0.005673\n",
              "(26.0, 30.0]  0.217284  0.205572 -0.055408  0.005673\n",
              "(30.0, inf]   0.241823  0.224427 -0.074659  0.005673"
            ],
            "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>loan_status</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>woe</th>\n",
              "      <th>iv</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>row_0</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>(-inf, 23.0]</th>\n",
              "      <td>0.262044</td>\n",
              "      <td>0.294076</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.005673</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(23.0, 26.0]</th>\n",
              "      <td>0.278849</td>\n",
              "      <td>0.275925</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.005673</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(26.0, 30.0]</th>\n",
              "      <td>0.217284</td>\n",
              "      <td>0.205572</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>0.005673</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(30.0, inf]</th>\n",
              "      <td>0.241823</td>\n",
              "      <td>0.224427</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.005673</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 303
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MFE7qnhwolEo"
      },
      "source": [
        "credit[f'{feature}_WOE'] = pd.cut(credit[feature], bins=binning, labels=woe['woe'].unique())\n",
        "credit[f'{feature}_WOE'] = credit[f'{feature}_WOE'].astype(float)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "ZYhv-kCvo3O0",
        "outputId": "2d68c83a-aec0-43f7-c790-e698b4b25c80"
      },
      "source": [
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "1          21           9600                   OWN                5.0   \n",
              "2          25           9600              MORTGAGE                1.0   \n",
              "3          23          65500                  RENT                4.0   \n",
              "4          24          54400                  RENT                8.0   \n",
              "5          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "1   EDUCATION          B       1000          11.14            0   \n",
              "2     MEDICAL          C       5500          12.87            1   \n",
              "3     MEDICAL          C      35000          15.23            1   \n",
              "4     MEDICAL          C      35000          14.27            1   \n",
              "5     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income cb_person_default_on_file  cb_person_cred_hist_length  \\\n",
              "1                 0.10                         N                           2   \n",
              "2                 0.57                         N                           3   \n",
              "3                 0.53                         N                           2   \n",
              "4                 0.55                         Y                           4   \n",
              "5                 0.25                         N                           2   \n",
              "\n",
              "   person_age_WOE  \n",
              "1        0.115326  \n",
              "2       -0.010540  \n",
              "3        0.115326  \n",
              "4       -0.010540  \n",
              "5        0.115326  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <th>person_age_WOE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "      <td>-0.010540</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "      <td>-0.010540</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 305
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NZYU4vUjpLAA"
      },
      "source": [
        "### person_emp_length"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "n4NtFV-29VB8",
        "outputId": "839d27a7-f45c-48fa-e632-d543ea5777c3"
      },
      "source": [
        "credit['person_emp_length'].describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    32576.000000\n",
              "mean         4.760560\n",
              "std          3.981108\n",
              "min          0.000000\n",
              "25%          2.000000\n",
              "50%          4.000000\n",
              "75%          7.000000\n",
              "max         41.000000\n",
              "Name: person_emp_length, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 306
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "id": "56pOUzOWj3Hc",
        "outputId": "a953e88a-cef1-42e6-c5ed-bbdabf44a345"
      },
      "source": [
        "feature = 'person_emp_length'\n",
        "target = 'loan_status'\n",
        "\n",
        "# binning=[-float(\"inf\"), 1, 5, 30, 50, float(\"inf\")]       \n",
        "  # 0.048\n",
        "# binning=[-float(\"inf\"), 10.25,20.5,30.75, float(\"inf\")]   \n",
        "  # 0.012\n",
        "binning=[-float(\"inf\"), 2,4,7, float(\"inf\")]                \n",
        "  # 0.057\n",
        "\n",
        "bin_feature = pd.cut(credit[feature], bins=binning).values.add_categories('Nan').fillna('Nan')\n",
        "woe_emplength = (pd.crosstab(bin_feature,\n",
        "                       credit[target],\n",
        "                       normalize='columns')\n",
        "             .assign(woe=lambda creditx: np.log(creditx[1] / creditx[0]))\n",
        "             .assign(iv=lambda creditx: np.sum(creditx['woe']*\n",
        "                                           (creditx[1]-creditx[0]))))\n",
        "\n",
        "woe_emplength"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "loan_status         0         1       woe        iv\n",
              "row_0                                              \n",
              "(-inf, 2.0]  0.311320  0.413677  0.284264  0.057403\n",
              "(2.0, 4.0]   0.221249  0.223442  0.009860  0.057403\n",
              "(4.0, 7.0]   0.249244  0.205431 -0.193322  0.057403\n",
              "(7.0, inf]   0.218187  0.157450 -0.326241  0.057403"
            ],
            "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>loan_status</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>woe</th>\n",
              "      <th>iv</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>row_0</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>(-inf, 2.0]</th>\n",
              "      <td>0.311320</td>\n",
              "      <td>0.413677</td>\n",
              "      <td>0.284264</td>\n",
              "      <td>0.057403</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(2.0, 4.0]</th>\n",
              "      <td>0.221249</td>\n",
              "      <td>0.223442</td>\n",
              "      <td>0.009860</td>\n",
              "      <td>0.057403</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(4.0, 7.0]</th>\n",
              "      <td>0.249244</td>\n",
              "      <td>0.205431</td>\n",
              "      <td>-0.193322</td>\n",
              "      <td>0.057403</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(7.0, inf]</th>\n",
              "      <td>0.218187</td>\n",
              "      <td>0.157450</td>\n",
              "      <td>-0.326241</td>\n",
              "      <td>0.057403</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 307
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "vAaDvWymlkGW",
        "outputId": "cef94e7f-8753-4d95-fc28-77882692cb0b"
      },
      "source": [
        "credit[f'{feature}_WOE'] = pd.cut(credit[feature], bins=binning, labels=woe['woe'].unique())\n",
        "credit[f'{feature}_WOE'] = credit[f'{feature}_WOE'].astype(float)\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "1          21           9600                   OWN                5.0   \n",
              "2          25           9600              MORTGAGE                1.0   \n",
              "3          23          65500                  RENT                4.0   \n",
              "4          24          54400                  RENT                8.0   \n",
              "5          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "1   EDUCATION          B       1000          11.14            0   \n",
              "2     MEDICAL          C       5500          12.87            1   \n",
              "3     MEDICAL          C      35000          15.23            1   \n",
              "4     MEDICAL          C      35000          14.27            1   \n",
              "5     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income cb_person_default_on_file  cb_person_cred_hist_length  \\\n",
              "1                 0.10                         N                           2   \n",
              "2                 0.57                         N                           3   \n",
              "3                 0.53                         N                           2   \n",
              "4                 0.55                         Y                           4   \n",
              "5                 0.25                         N                           2   \n",
              "\n",
              "   person_age_WOE  person_emp_length_WOE  \n",
              "1        0.115326              -0.055408  \n",
              "2       -0.010540               0.115326  \n",
              "3        0.115326              -0.010540  \n",
              "4       -0.010540              -0.074659  \n",
              "5        0.115326               0.115326  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <th>person_age_WOE</th>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.055408</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 308
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fP7HDLq8qI8X"
      },
      "source": [
        "### loan_amnt"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yLm0WDSDB5Uf",
        "outputId": "ffb51a00-3b44-4e95-cea7-dbf1c1af3041"
      },
      "source": [
        "credit['loan_amnt'].describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    32576.000000\n",
              "mean      9588.669573\n",
              "std       6320.602711\n",
              "min        500.000000\n",
              "25%       5000.000000\n",
              "50%       8000.000000\n",
              "75%      12200.000000\n",
              "max      35000.000000\n",
              "Name: loan_amnt, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 309
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "id": "_UGl910xpwy4",
        "outputId": "2dc19510-31a3-44bc-9444-0119eb588a83"
      },
      "source": [
        "feature = 'loan_amnt'\n",
        "target = 'loan_status'\n",
        "\n",
        "# binning=[-float(\"inf\"),6250,12500,18750, float(\"inf\")]\n",
        "  # 0.076385\n",
        "binning=[-float(\"inf\"),5000,8000,12200, float(\"inf\")]\n",
        "  # 0.076423\n",
        "\n",
        "bin_feature = pd.cut(credit[feature], \n",
        "                     bins=binning)\n",
        "\n",
        "woe_loan = (pd.crosstab(bin_feature,\n",
        "                       credit[target],\n",
        "                       normalize='columns')\n",
        "             .assign(woe=lambda creditx: np.log(creditx[1] / creditx[0]))\n",
        "             .assign(iv=lambda creditx: np.sum(creditx['woe']*\n",
        "                                           (creditx[1]-creditx[0]))))\n",
        "\n",
        "woe_loan"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "loan_status               0         1       woe        iv\n",
              "loan_amnt                                                \n",
              "(-inf, 5000.0]     0.296282  0.273111 -0.081433  0.076423\n",
              "(5000.0, 8000.0]   0.239428  0.171240 -0.335190  0.076423\n",
              "(8000.0, 12200.0]  0.239899  0.214577 -0.111551  0.076423\n",
              "(12200.0, inf]     0.224390  0.341072  0.418707  0.076423"
            ],
            "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>loan_status</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>woe</th>\n",
              "      <th>iv</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_amnt</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>(-inf, 5000.0]</th>\n",
              "      <td>0.296282</td>\n",
              "      <td>0.273111</td>\n",
              "      <td>-0.081433</td>\n",
              "      <td>0.076423</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(5000.0, 8000.0]</th>\n",
              "      <td>0.239428</td>\n",
              "      <td>0.171240</td>\n",
              "      <td>-0.335190</td>\n",
              "      <td>0.076423</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(8000.0, 12200.0]</th>\n",
              "      <td>0.239899</td>\n",
              "      <td>0.214577</td>\n",
              "      <td>-0.111551</td>\n",
              "      <td>0.076423</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(12200.0, inf]</th>\n",
              "      <td>0.224390</td>\n",
              "      <td>0.341072</td>\n",
              "      <td>0.418707</td>\n",
              "      <td>0.076423</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 310
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DiSAykd1CqW2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "outputId": "ee59d67a-245b-4da2-b299-69730cc47a4b"
      },
      "source": [
        "credit[f'{feature}_WOE'] = pd.cut(credit[feature], bins=binning, labels=woe['woe'].unique())\n",
        "credit[f'{feature}_WOE'] = credit[f'{feature}_WOE'].astype(float)\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "1          21           9600                   OWN                5.0   \n",
              "2          25           9600              MORTGAGE                1.0   \n",
              "3          23          65500                  RENT                4.0   \n",
              "4          24          54400                  RENT                8.0   \n",
              "5          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "1   EDUCATION          B       1000          11.14            0   \n",
              "2     MEDICAL          C       5500          12.87            1   \n",
              "3     MEDICAL          C      35000          15.23            1   \n",
              "4     MEDICAL          C      35000          14.27            1   \n",
              "5     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income cb_person_default_on_file  cb_person_cred_hist_length  \\\n",
              "1                 0.10                         N                           2   \n",
              "2                 0.57                         N                           3   \n",
              "3                 0.53                         N                           2   \n",
              "4                 0.55                         Y                           4   \n",
              "5                 0.25                         N                           2   \n",
              "\n",
              "   person_age_WOE  person_emp_length_WOE  loan_amnt_WOE  \n",
              "1        0.115326              -0.055408       0.115326  \n",
              "2       -0.010540               0.115326      -0.010540  \n",
              "3        0.115326              -0.010540      -0.074659  \n",
              "4       -0.010540              -0.074659      -0.074659  \n",
              "5        0.115326               0.115326       0.115326  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <th>person_age_WOE</th>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "      <th>loan_amnt_WOE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>0.115326</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.074659</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 311
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KwUvK9Pa5OXH"
      },
      "source": [
        "### loan_int_rate"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gEqdZI_25kDZ",
        "outputId": "7f10ce23-852d-4adc-ad47-c7a4b1a01c9b"
      },
      "source": [
        "credit['loan_int_rate'].describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    32576.000000\n",
              "mean        11.011521\n",
              "std          3.081566\n",
              "min          5.420000\n",
              "25%          8.490000\n",
              "50%         11.011695\n",
              "75%         13.110000\n",
              "max         23.220000\n",
              "Name: loan_int_rate, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 312
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "id": "qDw7X-9P5kDa",
        "outputId": "b3f50403-31dd-4ca9-a773-78bf08ab01cd"
      },
      "source": [
        "feature = 'loan_int_rate'\n",
        "target = 'loan_status'\n",
        "\n",
        "# binning=[-float(\"inf\"),5.5,11,16.5, float(\"inf\")]\n",
        "  # 0.422037\n",
        "binning=[-float(\"inf\"),8.5,11.012,13.11, float(\"inf\")]\n",
        "  # 0.509178\n",
        "\n",
        "bin_feature = pd.cut(credit[feature], \n",
        "                     bins=binning)\n",
        "\n",
        "woe = (pd.crosstab(bin_feature,\n",
        "                       credit[target],\n",
        "                       normalize='columns')\n",
        "             .assign(woe=lambda creditx: np.log(creditx[1] / creditx[0]))\n",
        "             .assign(iv=lambda creditx: np.sum(creditx['woe']*\n",
        "                                           (creditx[1]-creditx[0]))))\n",
        "\n",
        "woe"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "loan_status             0         1       woe        iv\n",
              "loan_int_rate                                          \n",
              "(-inf, 8.5]      0.300247  0.114957 -0.960048  0.509178\n",
              "(8.5, 11.012]    0.311555  0.230196 -0.302647  0.509178\n",
              "(11.012, 13.11]  0.205740  0.167863 -0.203469  0.509178\n",
              "(13.11, inf]     0.182457  0.486985  0.981718  0.509178"
            ],
            "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>loan_status</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>woe</th>\n",
              "      <th>iv</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>(-inf, 8.5]</th>\n",
              "      <td>0.300247</td>\n",
              "      <td>0.114957</td>\n",
              "      <td>-0.960048</td>\n",
              "      <td>0.509178</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(8.5, 11.012]</th>\n",
              "      <td>0.311555</td>\n",
              "      <td>0.230196</td>\n",
              "      <td>-0.302647</td>\n",
              "      <td>0.509178</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(11.012, 13.11]</th>\n",
              "      <td>0.205740</td>\n",
              "      <td>0.167863</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>0.509178</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(13.11, inf]</th>\n",
              "      <td>0.182457</td>\n",
              "      <td>0.486985</td>\n",
              "      <td>0.981718</td>\n",
              "      <td>0.509178</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 313
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "DR7iWZxw5kDa",
        "outputId": "cb1daa5d-c699-4e7a-fd8a-d835493ac713"
      },
      "source": [
        "credit[f'{feature}_WOE'] = pd.cut(\n",
        "                                  credit[feature], \n",
        "                                  bins=binning, \n",
        "                                  labels=woe['woe'].unique())\n",
        "credit[f'{feature}_WOE'] = credit[f'{feature}_WOE'].astype(float)\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "1          21           9600                   OWN                5.0   \n",
              "2          25           9600              MORTGAGE                1.0   \n",
              "3          23          65500                  RENT                4.0   \n",
              "4          24          54400                  RENT                8.0   \n",
              "5          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "1   EDUCATION          B       1000          11.14            0   \n",
              "2     MEDICAL          C       5500          12.87            1   \n",
              "3     MEDICAL          C      35000          15.23            1   \n",
              "4     MEDICAL          C      35000          14.27            1   \n",
              "5     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income cb_person_default_on_file  cb_person_cred_hist_length  \\\n",
              "1                 0.10                         N                           2   \n",
              "2                 0.57                         N                           3   \n",
              "3                 0.53                         N                           2   \n",
              "4                 0.55                         Y                           4   \n",
              "5                 0.25                         N                           2   \n",
              "\n",
              "   person_age_WOE  person_emp_length_WOE  loan_amnt_WOE  loan_int_rate_WOE  \n",
              "1        0.115326              -0.055408       0.115326          -0.203469  \n",
              "2       -0.010540               0.115326      -0.010540          -0.203469  \n",
              "3        0.115326              -0.010540      -0.074659           0.981718  \n",
              "4       -0.010540              -0.074659      -0.074659           0.981718  \n",
              "5        0.115326               0.115326       0.115326          -0.960048  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <th>person_age_WOE</th>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "      <th>loan_amnt_WOE</th>\n",
              "      <th>loan_int_rate_WOE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.203469</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.203469</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.981718</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.981718</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.960048</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 314
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o7QODRop5YN1"
      },
      "source": [
        "### loan_percent_income"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a5Qh2y0O65Mv",
        "outputId": "908f184e-9bfb-40c1-fe93-ec1089614f9a"
      },
      "source": [
        "credit['loan_percent_income'].describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    32576.000000\n",
              "mean         0.170207\n",
              "std          0.106754\n",
              "min          0.000000\n",
              "25%          0.090000\n",
              "50%          0.150000\n",
              "75%          0.230000\n",
              "max          0.830000\n",
              "Name: loan_percent_income, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 315
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "id": "ilHJ4usi65Mw",
        "outputId": "d839e22f-0550-452b-906c-88fe3d660b5d"
      },
      "source": [
        "feature = 'loan_percent_income'\n",
        "target = 'loan_status'\n",
        "\n",
        "# binning=[-float(\"inf\"),0.2075,0.415,0.6225, float(\"inf\")]\n",
        "  # 0.586803\n",
        "binning=[-float(\"inf\"), 0.09,0.15,0.23, float(\"inf\")]\n",
        "  # 0.611674\n",
        "\n",
        "bin_feature = pd.cut(credit[feature], \n",
        "                     bins=binning)\n",
        "\n",
        "woe = (pd.crosstab(bin_feature,\n",
        "                       credit[target],\n",
        "                       normalize='columns')\n",
        "             .assign(woe=lambda creditx: np.log(creditx[1] / creditx[0]))\n",
        "             .assign(iv=lambda creditx: np.sum(creditx['woe']*\n",
        "                                           (creditx[1]-creditx[0]))))\n",
        "\n",
        "woe"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "loan_status                 0         1       woe        iv\n",
              "loan_percent_income                                        \n",
              "(-inf, 0.09]         0.310770  0.145350 -0.759911  0.611674\n",
              "(0.09, 0.15]         0.282029  0.148445 -0.641795  0.611674\n",
              "(0.15, 0.23]         0.246064  0.201914 -0.197751  0.611674\n",
              "(0.23, inf]          0.161137  0.504292  1.140899  0.611674"
            ],
            "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>loan_status</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>woe</th>\n",
              "      <th>iv</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>(-inf, 0.09]</th>\n",
              "      <td>0.310770</td>\n",
              "      <td>0.145350</td>\n",
              "      <td>-0.759911</td>\n",
              "      <td>0.611674</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(0.09, 0.15]</th>\n",
              "      <td>0.282029</td>\n",
              "      <td>0.148445</td>\n",
              "      <td>-0.641795</td>\n",
              "      <td>0.611674</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(0.15, 0.23]</th>\n",
              "      <td>0.246064</td>\n",
              "      <td>0.201914</td>\n",
              "      <td>-0.197751</td>\n",
              "      <td>0.611674</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(0.23, inf]</th>\n",
              "      <td>0.161137</td>\n",
              "      <td>0.504292</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.611674</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 316
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "QpvZ5irp65Mw",
        "outputId": "348fb10d-01d5-441a-cc0a-237cdf2f0a1b"
      },
      "source": [
        "credit[f'{feature}_WOE'] = pd.cut(\n",
        "                                  credit[feature], \n",
        "                                  bins=binning, \n",
        "                                  labels=woe['woe'].unique())\n",
        "credit[f'{feature}_WOE'] = credit[f'{feature}_WOE'].astype(float)\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "1          21           9600                   OWN                5.0   \n",
              "2          25           9600              MORTGAGE                1.0   \n",
              "3          23          65500                  RENT                4.0   \n",
              "4          24          54400                  RENT                8.0   \n",
              "5          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "1   EDUCATION          B       1000          11.14            0   \n",
              "2     MEDICAL          C       5500          12.87            1   \n",
              "3     MEDICAL          C      35000          15.23            1   \n",
              "4     MEDICAL          C      35000          14.27            1   \n",
              "5     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income cb_person_default_on_file  cb_person_cred_hist_length  \\\n",
              "1                 0.10                         N                           2   \n",
              "2                 0.57                         N                           3   \n",
              "3                 0.53                         N                           2   \n",
              "4                 0.55                         Y                           4   \n",
              "5                 0.25                         N                           2   \n",
              "\n",
              "   person_age_WOE  person_emp_length_WOE  loan_amnt_WOE  loan_int_rate_WOE  \\\n",
              "1        0.115326              -0.055408       0.115326          -0.203469   \n",
              "2       -0.010540               0.115326      -0.010540          -0.203469   \n",
              "3        0.115326              -0.010540      -0.074659           0.981718   \n",
              "4       -0.010540              -0.074659      -0.074659           0.981718   \n",
              "5        0.115326               0.115326       0.115326          -0.960048   \n",
              "\n",
              "   loan_percent_income_WOE  \n",
              "1                -0.641795  \n",
              "2                 1.140899  \n",
              "3                 1.140899  \n",
              "4                 1.140899  \n",
              "5                 1.140899  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <th>person_age_WOE</th>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "      <th>loan_amnt_WOE</th>\n",
              "      <th>loan_int_rate_WOE</th>\n",
              "      <th>loan_percent_income_WOE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>-0.641795</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>1.140899</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.981718</td>\n",
              "      <td>1.140899</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.981718</td>\n",
              "      <td>1.140899</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.960048</td>\n",
              "      <td>1.140899</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 317
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N7XHSo-F5bFI"
      },
      "source": [
        "### cb_person_cred_hist_length"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a9UF4gpi7m5I",
        "outputId": "ac5aa210-4416-4502-e2c1-f5fec22f18e9"
      },
      "source": [
        "credit['cb_person_cred_hist_length'].describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    32576.000000\n",
              "mean         5.803966\n",
              "std          4.053791\n",
              "min          2.000000\n",
              "25%          3.000000\n",
              "50%          4.000000\n",
              "75%          8.000000\n",
              "max         30.000000\n",
              "Name: cb_person_cred_hist_length, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 318
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "id": "V39XcMCq7m5J",
        "outputId": "5c6867f0-ef10-4c9f-afd4-c0f9d0a3d014"
      },
      "source": [
        "feature = 'cb_person_cred_hist_length'\n",
        "target = 'loan_status'\n",
        "\n",
        "# binning=[-float(\"inf\"),7.5,15,22.5, float(\"inf\")]\n",
        "binning=[-float(\"inf\"),3,4,8, float(\"inf\")]\n",
        "\n",
        "bin_feature = pd.cut(credit[feature], \n",
        "                     bins=binning)\n",
        "\n",
        "woe = (pd.crosstab(bin_feature,\n",
        "                       credit[target],\n",
        "                       normalize='columns')\n",
        "             .assign(woe=lambda creditx: np.log(creditx[1] / creditx[0]))\n",
        "             .assign(iv=lambda creditx: np.sum(creditx['woe']*\n",
        "                                           (creditx[1]-creditx[0]))))\n",
        "\n",
        "woe"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "loan_status                        0         1       woe        iv\n",
              "cb_person_cred_hist_length                                        \n",
              "(-inf, 3.0]                 0.360281  0.383988  0.063726  0.003566\n",
              "(3.0, 4.0]                  0.180690  0.186014  0.029037  0.003566\n",
              "(4.0, 8.0]                  0.234874  0.219361 -0.068329  0.003566\n",
              "(8.0, inf]                  0.224155  0.210637 -0.062199  0.003566"
            ],
            "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>loan_status</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>woe</th>\n",
              "      <th>iv</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>(-inf, 3.0]</th>\n",
              "      <td>0.360281</td>\n",
              "      <td>0.383988</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.003566</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(3.0, 4.0]</th>\n",
              "      <td>0.180690</td>\n",
              "      <td>0.186014</td>\n",
              "      <td>0.029037</td>\n",
              "      <td>0.003566</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(4.0, 8.0]</th>\n",
              "      <td>0.234874</td>\n",
              "      <td>0.219361</td>\n",
              "      <td>-0.068329</td>\n",
              "      <td>0.003566</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(8.0, inf]</th>\n",
              "      <td>0.224155</td>\n",
              "      <td>0.210637</td>\n",
              "      <td>-0.062199</td>\n",
              "      <td>0.003566</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 319
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "s_xc0sAi7m5J",
        "outputId": "79aacb1a-fff1-4180-fa1f-9d46dba3c35c"
      },
      "source": [
        "credit[f'{feature}_WOE'] = pd.cut(\n",
        "                                  credit[feature], \n",
        "                                  bins=binning, \n",
        "                                  labels=woe['woe'].unique())\n",
        "credit[f'{feature}_WOE'] = credit[f'{feature}_WOE'].astype(float)\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "1          21           9600                   OWN                5.0   \n",
              "2          25           9600              MORTGAGE                1.0   \n",
              "3          23          65500                  RENT                4.0   \n",
              "4          24          54400                  RENT                8.0   \n",
              "5          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "1   EDUCATION          B       1000          11.14            0   \n",
              "2     MEDICAL          C       5500          12.87            1   \n",
              "3     MEDICAL          C      35000          15.23            1   \n",
              "4     MEDICAL          C      35000          14.27            1   \n",
              "5     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income cb_person_default_on_file  cb_person_cred_hist_length  \\\n",
              "1                 0.10                         N                           2   \n",
              "2                 0.57                         N                           3   \n",
              "3                 0.53                         N                           2   \n",
              "4                 0.55                         Y                           4   \n",
              "5                 0.25                         N                           2   \n",
              "\n",
              "   person_age_WOE  person_emp_length_WOE  loan_amnt_WOE  loan_int_rate_WOE  \\\n",
              "1        0.115326              -0.055408       0.115326          -0.203469   \n",
              "2       -0.010540               0.115326      -0.010540          -0.203469   \n",
              "3        0.115326              -0.010540      -0.074659           0.981718   \n",
              "4       -0.010540              -0.074659      -0.074659           0.981718   \n",
              "5        0.115326               0.115326       0.115326          -0.960048   \n",
              "\n",
              "   loan_percent_income_WOE  cb_person_cred_hist_length_WOE  \n",
              "1                -0.641795                        0.063726  \n",
              "2                 1.140899                        0.063726  \n",
              "3                 1.140899                        0.063726  \n",
              "4                 1.140899                        0.029037  \n",
              "5                 1.140899                        0.063726  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <th>person_age_WOE</th>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "      <th>loan_amnt_WOE</th>\n",
              "      <th>loan_int_rate_WOE</th>\n",
              "      <th>loan_percent_income_WOE</th>\n",
              "      <th>cb_person_cred_hist_length_WOE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>-0.641795</td>\n",
              "      <td>0.063726</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.063726</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.981718</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.063726</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.981718</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.029037</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.960048</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.063726</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 320
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dPuqWK_QAzSG"
      },
      "source": [
        "### person_income"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 479
        },
        "id": "jwnQJea3AzSH",
        "outputId": "3bae84c0-9764-4a9f-9408-3112eca88a13"
      },
      "source": [
        "credit.describe().T"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                  count          mean           std  \\\n",
              "person_age                      32576.0     27.724276      6.249549   \n",
              "person_income                   32576.0  65879.286530  52530.427367   \n",
              "person_emp_length               32576.0      4.760560      3.981108   \n",
              "loan_amnt                       32576.0   9588.669573   6320.602711   \n",
              "loan_int_rate                   32576.0     11.011521      3.081566   \n",
              "loan_status                     32576.0      0.218167      0.413008   \n",
              "loan_percent_income             32576.0      0.170207      0.106754   \n",
              "cb_person_cred_hist_length      32576.0      5.803966      4.053791   \n",
              "person_age_WOE                  32576.0     -0.001574      0.074785   \n",
              "person_emp_length_WOE           32576.0      0.007561      0.079268   \n",
              "loan_amnt_WOE                   32576.0     -0.000420      0.077527   \n",
              "loan_int_rate_WOE               32576.0     -0.134198      0.704590   \n",
              "loan_percent_income_WOE         32576.0     -0.148534      0.746441   \n",
              "cb_person_cred_hist_length_WOE  32576.0     -0.001007      0.059775   \n",
              "\n",
              "                                        min           25%           50%  \\\n",
              "person_age                        20.000000     23.000000     26.000000   \n",
              "person_income                   4000.000000  38500.000000  55000.000000   \n",
              "person_emp_length                  0.000000      2.000000      4.000000   \n",
              "loan_amnt                        500.000000   5000.000000   8000.000000   \n",
              "loan_int_rate                      5.420000      8.490000     11.011695   \n",
              "loan_status                        0.000000      0.000000      0.000000   \n",
              "loan_percent_income                0.000000      0.090000      0.150000   \n",
              "cb_person_cred_hist_length         2.000000      3.000000      4.000000   \n",
              "person_age_WOE                    -0.074659     -0.055408     -0.010540   \n",
              "person_emp_length_WOE             -0.074659     -0.055408     -0.010540   \n",
              "loan_amnt_WOE                     -0.074659     -0.055408     -0.010540   \n",
              "loan_int_rate_WOE                 -0.960048     -0.960048     -0.302647   \n",
              "loan_percent_income_WOE           -0.759911     -0.759911     -0.641795   \n",
              "cb_person_cred_hist_length_WOE    -0.068329     -0.062199      0.029037   \n",
              "\n",
              "                                         75%           max  \n",
              "person_age                         30.000000  1.230000e+02  \n",
              "person_income                   79200.000000  2.039784e+06  \n",
              "person_emp_length                   7.000000  4.100000e+01  \n",
              "loan_amnt                       12200.000000  3.500000e+04  \n",
              "loan_int_rate                      13.110000  2.322000e+01  \n",
              "loan_status                         0.000000  1.000000e+00  \n",
              "loan_percent_income                 0.230000  8.300000e-01  \n",
              "cb_person_cred_hist_length          8.000000  3.000000e+01  \n",
              "person_age_WOE                      0.115326  1.153265e-01  \n",
              "person_emp_length_WOE               0.115326  1.153265e-01  \n",
              "loan_amnt_WOE                       0.115326  1.153265e-01  \n",
              "loan_int_rate_WOE                  -0.203469  9.817175e-01  \n",
              "loan_percent_income_WOE            -0.197751  1.140899e+00  \n",
              "cb_person_cred_hist_length_WOE      0.063726  6.372567e-02  "
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>person_age</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>27.724276</td>\n",
              "      <td>6.249549</td>\n",
              "      <td>20.000000</td>\n",
              "      <td>23.000000</td>\n",
              "      <td>26.000000</td>\n",
              "      <td>30.000000</td>\n",
              "      <td>1.230000e+02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_income</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>65879.286530</td>\n",
              "      <td>52530.427367</td>\n",
              "      <td>4000.000000</td>\n",
              "      <td>38500.000000</td>\n",
              "      <td>55000.000000</td>\n",
              "      <td>79200.000000</td>\n",
              "      <td>2.039784e+06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_emp_length</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>4.760560</td>\n",
              "      <td>3.981108</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>7.000000</td>\n",
              "      <td>4.100000e+01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_amnt</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>9588.669573</td>\n",
              "      <td>6320.602711</td>\n",
              "      <td>500.000000</td>\n",
              "      <td>5000.000000</td>\n",
              "      <td>8000.000000</td>\n",
              "      <td>12200.000000</td>\n",
              "      <td>3.500000e+04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_int_rate</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>11.011521</td>\n",
              "      <td>3.081566</td>\n",
              "      <td>5.420000</td>\n",
              "      <td>8.490000</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>13.110000</td>\n",
              "      <td>2.322000e+01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_status</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>0.218167</td>\n",
              "      <td>0.413008</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_percent_income</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>0.170207</td>\n",
              "      <td>0.106754</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.090000</td>\n",
              "      <td>0.150000</td>\n",
              "      <td>0.230000</td>\n",
              "      <td>8.300000e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>5.803966</td>\n",
              "      <td>4.053791</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>8.000000</td>\n",
              "      <td>3.000000e+01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_age_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.001574</td>\n",
              "      <td>0.074785</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>1.153265e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>0.007561</td>\n",
              "      <td>0.079268</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>1.153265e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_amnt_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.000420</td>\n",
              "      <td>0.077527</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>1.153265e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_int_rate_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.134198</td>\n",
              "      <td>0.704590</td>\n",
              "      <td>-0.960048</td>\n",
              "      <td>-0.960048</td>\n",
              "      <td>-0.302647</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>9.817175e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_percent_income_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.148534</td>\n",
              "      <td>0.746441</td>\n",
              "      <td>-0.759911</td>\n",
              "      <td>-0.759911</td>\n",
              "      <td>-0.641795</td>\n",
              "      <td>-0.197751</td>\n",
              "      <td>1.140899e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cb_person_cred_hist_length_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.001007</td>\n",
              "      <td>0.059775</td>\n",
              "      <td>-0.068329</td>\n",
              "      <td>-0.062199</td>\n",
              "      <td>0.029037</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>6.372567e-02</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 321
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "id": "Gu6P9E1nAzSH",
        "outputId": "be80cdcf-d2ed-4bae-c068-94f01c9293b8"
      },
      "source": [
        "feature = 'person_income'\n",
        "target = 'loan_status'\n",
        "\n",
        "# binning=[-float(\"inf\"), 100000,500000,700000, float(\"inf\")]\n",
        "  # 0.101265\n",
        "binning=[-float(\"inf\"), 38500,55000,79200, float(\"inf\")]\n",
        "  # 0.436687\n",
        "\n",
        "bin_feature = pd.cut(credit[feature], \n",
        "                     bins=binning)\n",
        "\n",
        "woe = (pd.crosstab(bin_feature,\n",
        "                       credit[target],\n",
        "                       normalize='columns')\n",
        "             .assign(woe=lambda creditx: np.log(creditx[1] / creditx[0]))\n",
        "             .assign(iv=lambda creditx: np.sum(creditx['woe']*\n",
        "                                           (creditx[1]-creditx[0]))))\n",
        "\n",
        "woe"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "loan_status                0         1       woe        iv\n",
              "person_income                                             \n",
              "(-inf, 38500.0]     0.193019  0.455326  0.858225  0.436687\n",
              "(38500.0, 55000.0]  0.254270  0.246940 -0.029252  0.436687\n",
              "(55000.0, 79200.0]  0.262280  0.193331 -0.305010  0.436687\n",
              "(79200.0, inf]      0.290432  0.104404 -1.023099  0.436687"
            ],
            "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>loan_status</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "      <th>woe</th>\n",
              "      <th>iv</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_income</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>(-inf, 38500.0]</th>\n",
              "      <td>0.193019</td>\n",
              "      <td>0.455326</td>\n",
              "      <td>0.858225</td>\n",
              "      <td>0.436687</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(38500.0, 55000.0]</th>\n",
              "      <td>0.254270</td>\n",
              "      <td>0.246940</td>\n",
              "      <td>-0.029252</td>\n",
              "      <td>0.436687</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(55000.0, 79200.0]</th>\n",
              "      <td>0.262280</td>\n",
              "      <td>0.193331</td>\n",
              "      <td>-0.305010</td>\n",
              "      <td>0.436687</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>(79200.0, inf]</th>\n",
              "      <td>0.290432</td>\n",
              "      <td>0.104404</td>\n",
              "      <td>-1.023099</td>\n",
              "      <td>0.436687</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 322
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "K0yN0MllAzSI",
        "outputId": "bc9fb1e9-7190-47bc-9201-c2474a5cc2b7"
      },
      "source": [
        "credit[f'{feature}_WOE'] = pd.cut(\n",
        "                                  credit[feature], \n",
        "                                  bins=binning, \n",
        "                                  labels=woe['woe'].unique())\n",
        "credit[f'{feature}_WOE'] = credit[f'{feature}_WOE'].astype(float)\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "1          21           9600                   OWN                5.0   \n",
              "2          25           9600              MORTGAGE                1.0   \n",
              "3          23          65500                  RENT                4.0   \n",
              "4          24          54400                  RENT                8.0   \n",
              "5          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "1   EDUCATION          B       1000          11.14            0   \n",
              "2     MEDICAL          C       5500          12.87            1   \n",
              "3     MEDICAL          C      35000          15.23            1   \n",
              "4     MEDICAL          C      35000          14.27            1   \n",
              "5     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income cb_person_default_on_file  cb_person_cred_hist_length  \\\n",
              "1                 0.10                         N                           2   \n",
              "2                 0.57                         N                           3   \n",
              "3                 0.53                         N                           2   \n",
              "4                 0.55                         Y                           4   \n",
              "5                 0.25                         N                           2   \n",
              "\n",
              "   person_age_WOE  person_emp_length_WOE  loan_amnt_WOE  loan_int_rate_WOE  \\\n",
              "1        0.115326              -0.055408       0.115326          -0.203469   \n",
              "2       -0.010540               0.115326      -0.010540          -0.203469   \n",
              "3        0.115326              -0.010540      -0.074659           0.981718   \n",
              "4       -0.010540              -0.074659      -0.074659           0.981718   \n",
              "5        0.115326               0.115326       0.115326          -0.960048   \n",
              "\n",
              "   loan_percent_income_WOE  cb_person_cred_hist_length_WOE  person_income_WOE  \n",
              "1                -0.641795                        0.063726           0.858225  \n",
              "2                 1.140899                        0.063726           0.858225  \n",
              "3                 1.140899                        0.063726          -0.305010  \n",
              "4                 1.140899                        0.029037          -0.029252  \n",
              "5                 1.140899                        0.063726           0.858225  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <th>person_age_WOE</th>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "      <th>loan_amnt_WOE</th>\n",
              "      <th>loan_int_rate_WOE</th>\n",
              "      <th>loan_percent_income_WOE</th>\n",
              "      <th>cb_person_cred_hist_length_WOE</th>\n",
              "      <th>person_income_WOE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>-0.641795</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.981718</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>-0.305010</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0.981718</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.029037</td>\n",
              "      <td>-0.029252</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>2</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.960048</td>\n",
              "      <td>1.140899</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 323
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e0x7hUeeAJA9"
      },
      "source": [
        "### Boxplot After WoE"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 510
        },
        "id": "M_KMrwti_5fu",
        "outputId": "aa723ee6-a20d-498f-f6ab-ed07115ff2ee"
      },
      "source": [
        "credit.describe().T"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                  count          mean           std  \\\n",
              "person_age                      32576.0     27.724276      6.249549   \n",
              "person_income                   32576.0  65879.286530  52530.427367   \n",
              "person_emp_length               32576.0      4.760560      3.981108   \n",
              "loan_amnt                       32576.0   9588.669573   6320.602711   \n",
              "loan_int_rate                   32576.0     11.011521      3.081566   \n",
              "loan_status                     32576.0      0.218167      0.413008   \n",
              "loan_percent_income             32576.0      0.170207      0.106754   \n",
              "cb_person_cred_hist_length      32576.0      5.803966      4.053791   \n",
              "person_age_WOE                  32576.0     -0.001574      0.074785   \n",
              "person_emp_length_WOE           32576.0      0.007561      0.079268   \n",
              "loan_amnt_WOE                   32576.0     -0.000420      0.077527   \n",
              "loan_int_rate_WOE               32576.0     -0.134198      0.704590   \n",
              "loan_percent_income_WOE         32576.0     -0.148534      0.746441   \n",
              "cb_person_cred_hist_length_WOE  32576.0     -0.001007      0.059775   \n",
              "person_income_WOE               32576.0     -0.123652      0.673633   \n",
              "\n",
              "                                        min           25%           50%  \\\n",
              "person_age                        20.000000     23.000000     26.000000   \n",
              "person_income                   4000.000000  38500.000000  55000.000000   \n",
              "person_emp_length                  0.000000      2.000000      4.000000   \n",
              "loan_amnt                        500.000000   5000.000000   8000.000000   \n",
              "loan_int_rate                      5.420000      8.490000     11.011695   \n",
              "loan_status                        0.000000      0.000000      0.000000   \n",
              "loan_percent_income                0.000000      0.090000      0.150000   \n",
              "cb_person_cred_hist_length         2.000000      3.000000      4.000000   \n",
              "person_age_WOE                    -0.074659     -0.055408     -0.010540   \n",
              "person_emp_length_WOE             -0.074659     -0.055408     -0.010540   \n",
              "loan_amnt_WOE                     -0.074659     -0.055408     -0.010540   \n",
              "loan_int_rate_WOE                 -0.960048     -0.960048     -0.302647   \n",
              "loan_percent_income_WOE           -0.759911     -0.759911     -0.641795   \n",
              "cb_person_cred_hist_length_WOE    -0.068329     -0.062199      0.029037   \n",
              "person_income_WOE                 -1.023099     -0.305010     -0.029252   \n",
              "\n",
              "                                         75%           max  \n",
              "person_age                         30.000000  1.230000e+02  \n",
              "person_income                   79200.000000  2.039784e+06  \n",
              "person_emp_length                   7.000000  4.100000e+01  \n",
              "loan_amnt                       12200.000000  3.500000e+04  \n",
              "loan_int_rate                      13.110000  2.322000e+01  \n",
              "loan_status                         0.000000  1.000000e+00  \n",
              "loan_percent_income                 0.230000  8.300000e-01  \n",
              "cb_person_cred_hist_length          8.000000  3.000000e+01  \n",
              "person_age_WOE                      0.115326  1.153265e-01  \n",
              "person_emp_length_WOE               0.115326  1.153265e-01  \n",
              "loan_amnt_WOE                       0.115326  1.153265e-01  \n",
              "loan_int_rate_WOE                  -0.203469  9.817175e-01  \n",
              "loan_percent_income_WOE            -0.197751  1.140899e+00  \n",
              "cb_person_cred_hist_length_WOE      0.063726  6.372567e-02  \n",
              "person_income_WOE                   0.858225  8.582246e-01  "
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>count</th>\n",
              "      <th>mean</th>\n",
              "      <th>std</th>\n",
              "      <th>min</th>\n",
              "      <th>25%</th>\n",
              "      <th>50%</th>\n",
              "      <th>75%</th>\n",
              "      <th>max</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>person_age</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>27.724276</td>\n",
              "      <td>6.249549</td>\n",
              "      <td>20.000000</td>\n",
              "      <td>23.000000</td>\n",
              "      <td>26.000000</td>\n",
              "      <td>30.000000</td>\n",
              "      <td>1.230000e+02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_income</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>65879.286530</td>\n",
              "      <td>52530.427367</td>\n",
              "      <td>4000.000000</td>\n",
              "      <td>38500.000000</td>\n",
              "      <td>55000.000000</td>\n",
              "      <td>79200.000000</td>\n",
              "      <td>2.039784e+06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_emp_length</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>4.760560</td>\n",
              "      <td>3.981108</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>7.000000</td>\n",
              "      <td>4.100000e+01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_amnt</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>9588.669573</td>\n",
              "      <td>6320.602711</td>\n",
              "      <td>500.000000</td>\n",
              "      <td>5000.000000</td>\n",
              "      <td>8000.000000</td>\n",
              "      <td>12200.000000</td>\n",
              "      <td>3.500000e+04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_int_rate</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>11.011521</td>\n",
              "      <td>3.081566</td>\n",
              "      <td>5.420000</td>\n",
              "      <td>8.490000</td>\n",
              "      <td>11.011695</td>\n",
              "      <td>13.110000</td>\n",
              "      <td>2.322000e+01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_status</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>0.218167</td>\n",
              "      <td>0.413008</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_percent_income</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>0.170207</td>\n",
              "      <td>0.106754</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.090000</td>\n",
              "      <td>0.150000</td>\n",
              "      <td>0.230000</td>\n",
              "      <td>8.300000e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>5.803966</td>\n",
              "      <td>4.053791</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>3.000000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>8.000000</td>\n",
              "      <td>3.000000e+01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_age_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.001574</td>\n",
              "      <td>0.074785</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>1.153265e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>0.007561</td>\n",
              "      <td>0.079268</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>1.153265e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_amnt_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.000420</td>\n",
              "      <td>0.077527</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>1.153265e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_int_rate_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.134198</td>\n",
              "      <td>0.704590</td>\n",
              "      <td>-0.960048</td>\n",
              "      <td>-0.960048</td>\n",
              "      <td>-0.302647</td>\n",
              "      <td>-0.203469</td>\n",
              "      <td>9.817175e-01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>loan_percent_income_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.148534</td>\n",
              "      <td>0.746441</td>\n",
              "      <td>-0.759911</td>\n",
              "      <td>-0.759911</td>\n",
              "      <td>-0.641795</td>\n",
              "      <td>-0.197751</td>\n",
              "      <td>1.140899e+00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>cb_person_cred_hist_length_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.001007</td>\n",
              "      <td>0.059775</td>\n",
              "      <td>-0.068329</td>\n",
              "      <td>-0.062199</td>\n",
              "      <td>0.029037</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>6.372567e-02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>person_income_WOE</th>\n",
              "      <td>32576.0</td>\n",
              "      <td>-0.123652</td>\n",
              "      <td>0.673633</td>\n",
              "      <td>-1.023099</td>\n",
              "      <td>-0.305010</td>\n",
              "      <td>-0.029252</td>\n",
              "      <td>0.858225</td>\n",
              "      <td>8.582246e-01</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 324
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "lDAFP0_C_x7B",
        "outputId": "c3be6528-77eb-4851-d080-c8135f03b4cd"
      },
      "source": [
        "plt.figure(figsize = (20,20))\n",
        "\n",
        "n = 1\n",
        "for name in credit.describe().columns:\n",
        "  plt.subplot(4,4,n)\n",
        "  sns.boxplot(credit[name], color = 'r')\n",
        "  n += 1\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1440x1440 with 15 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABHIAAARtCAYAAAA5y2KiAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAACTr0lEQVR4nOzdebgsZ1kv7N+T7MxhSEjQbQLsJDKIA1MORxTyBZk5Mp2gRJkinPgFOB5FwJCDHwdxIEwqyiRRDKiHzSRIUIQQwQgyJZKJKSQ7IRA3JBIGA0nI8H5/dK2dXmuvca+p37Xv+7r6WtVV1VVvVVc/6+1fV1dXay0AAAAATL491rsBAAAAACyOIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATm5Yy8yGHHNK2bNmySk0B1sK55577H621Q9e7HcuhFkH/1CJgUvRej9Qi2BiWUouWFORs2bIl55xzzq61CpgIVfWV9W7DcqlF0D+1CJgUvdcjtQg2hqXUIl+tAgAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADqxaS1Xdtppp2Xbtm2zTtu+fXuSZPPmzXM+/sgjj8yJJ564Km0DNr7TTjstH//4x5PcWmvUFWCSzNdXmmkxfaeFqIHAalpKTVvIStS8pVAfmWRrGuRs27Ytl55/fg67/vqdpn1v332TJNdfccWsj71ymA6wq7Zt25Zv/cd/ZO9bbsn1V1yhrgATZ76+0kwL9Z0WogYCq20pNW0hy615S6E+MunWNMhJksOuvz6/ftllO41/zRFHJMms08anAyzH3rfcksOHOqSuAJNorr7STAv1nRb7eIDVtNiatpDl1rxdWRdMKtfIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADoxKoEOaeddlpOO+201Vj0iuuprcDSLOf1rTYAK0U9WTz7ClaP1xfL4fiZLJtWY6Hbtm1bjcWuip7aCizNcl7fagOwUtSTxbOvYPV4fbEcjp/J4qtVAAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRi03o3YFI85jGPWbN1bdq0KXvvvXe+//3vzzrtpptu2nH/gAMOyPe+970cfPDBueaaa/KjP/qjuctd7pKzzjord7nLXfKVr3wlxxxzTD72sY/lgQ98YM4+++xs2bIll19+eTZv3pzt27fnmGOOyTe/+c2cfPLJOeigg5Ik//AP/5A3vOENuc997pPPfvaz+cVf/MU89alP3bHez372s3nJS16Sl770pbnXve41rY3XXHNNXvGKV+xY3kL3Z7OYeZZqNZbZg23btuWUU07JqaeemiOOOGK9m7NhXHTRRUlWtzZMvb5ns88+++T444/PW97yljznOc/Ju9/97nz961/Pc57znHzjG9/Iu971rhx88ME7jvXrr78+//7v/56TTjopb3nLW3LKKafktNNOy1e/+tVs2rQpe+65Z170ohdl69atedKTnpRTTz01p5xySrZu3ZqTTz45rbW89KUvzZVXXpnDDjss/+f//J+01vL7v//7ueGGG3LVVVfl53/+5/POd74zJ5xwQj7zmc/k5JNPzre+9a05j79rrrkm//t//+9ceeWVOfnkk/PABz5wx/i1eq3OV8uWYqE2T2L92ZU2LWc7JnEfsHGM1+IzzjhjSeNWa95nPOMZufrqq3PHO94xf/EXf7Fj+stf/vJ87GMfyzHHHJMXvOAFc4574xvfmL//+7/PYx7zmPzqr/7qjse/853vzFvf+taccMIJOe6445Lc2m97znOek0c+8pE75p2txs32Wpyrr3D22Wfnla985bQavRST2p9Tj2Dl3Xjjjbnssst2uW/8wz/8w/n617+eJNlrr71y44035phjjslxxx2XU045ZUe/9KCDDsrHPvaxJMmee+6Zm2++OSeccEJOP/30Hcv68R//8R391xe/+MX5yle+kiT5vd/7vdzpTnfKC1/4wmzfvn3a+s8444xcc801+YM/+IN873vfy5VXXpnf+q3fynve855UVZ7ylKfkxS9+cVprOeaYY/LMZz4zf/AHf5DWWn77t387H/7wh/PWt741d7zjHfOqV71qR22ZrWZP1aDPfe5z09a/kpyRsw5uuummWUOcqWnjpt7kXXPNNUmSSy65JGeddVaS7Dhgzz777Nxyyy05++yzkySXX355kuw4eM8+++x8/vOfz9atW3cs941vfGOSUQcgSd7xjndMW+/LX/7y3HLLLTn11FN3auPWrVunLW+h+7NZzDxLtRrL7MGrXvWqfP/738+rXvWq9W4KSzRXiJMkN9xwQ97ylrckSV7/+tfv+Mf3+te/Pu9617uSjOrCpZdemksvvTRXXnllWmt54xvfmO9///t5+ctfniuuuCKttdx44425/vrr8/KXvzyf//zn84pXvGLHPFOvma1bt+bSSy/N9ddfn0svvXTHuIsvvjhf+cpXct111+Wd73xnkuT000/f8bj5jr+tW7fmyiuvTJK8+tWvnjZ+rV6r89WypViozZNYf3alTcvZjkncB7Carr766iTJVVddNW381BuQqX7ZXOP+/u//PsnOnfu3vvWtSTLtTctUv+31r3/9tHlnq3GzvRbnqtV/9Ed/lGR6jV6KSe3PqUew8q6++ur84Ac/2OXHT/Vlk1EolIxq4lR9muqXTtXLJLn55puTTK+HSab1X6feEyfJqaeemq1bt+4U4kzZunVrvvSlL+VrX/taWmt59atfnYsvvjhf+tKXcuqpp6a1tqNdU/NefPHF2bp1647afNVVV02rLbPV7KkatJoEObn1k/eNrLWWD3/4w/nWt76Vf/iHf9hxkI77q7/6qySjcGfqDea1116b888/f8c811xzTc4666wdy9u2bdu0+5dddtm0+9/61rd2Ws/MZcw2z1KtxjJ7sG3btnz1q19NklxxxRW57LLL1rlFG8NanqG3GOOv19leu7PNe+211+407dprr01rbce0qftnnnlmzjzzzGnzfuhDH9pp3Mz1fOhDH5rz+LvmmmvyoQ99aMf9m266KR/72MfW9LU6Xy1bioXaPIn1Z1fatJztmMR9wMYxsyY/5jGPWfS4pTx+KfM+4xnPmDbumc98ZpJRsDLula985azjpoKZKW9605uSZEdgPuXd7373tH5bay3/+I//mGT2Gjfba3GuvsLZZ5+94wPEqRq9FJPan1OPYOVdc801O04sWGlT9WmqX7oYU/3X8b7m1DI++MEPzvqYxzzmMfnwhz88bdz4SRQzP2CdqrXJ6KzIcR/4wAfyrW99a9aaPV6DZq5/Ja3KV6u2b9+e6667Lqeccsq08du2bctee++9S8u8eu+9c+NwWuhK2rZt24oub5Ldcsst2bp1az7wgQ/MOv0d73hHnvrUp+7U4Tj11FPztre9LckoXbzlllt2LO/Vr371tPuvetWrpt3funVrnvWsZ01b3sxlzDbPUq3GMnsw85O1V73qVXnd6163Tq2ZPDNr0bZt23Jj1Y7pq1VXenPTTTft9M9m5tmBcz1u3Pjxt3Xr1h2fokx59atfnYc97GFr9lqdr5YtxUL1ZRLrz660aTnbMYn7YJLM1S+azXL6Sks1iTVw27Zt2W+//da7GQuaOhtnytRZOTPDkPEzcOYbd8YZZ+RXf/VXd3yyO+X0009Pjf3fSkZn5TzykY+ctcY96EEP2um1eOGFF06bb6pWT52NM+XVr371kr5eNan9OfVobkupRStlLWvaSprE+rie/v3f/329m7CT2fqvSXa8/ud6zGLNFyq11rJ169adAp7TTz893/jGN+Ztw0pZ8IycqvrVqjqnqs6Z+U+Lvtx00035yEc+smDSOTONHP9k/6Mf/ei0T2+uuOKKee9/5CMf2Wn5M5cx2zxLtRrL7MFUgj3liiuuWKeWrD61aPUs9tOPhYwffx/96Ed3mn7TTTet6Wt1vlq2FAu1eRLrz660aTnbMYn7YLWoRay1mTV66v5sNW621+JcfYWZb2iW8gYnmdz+3O5Sj9Qi1tK3v/3t9W7CTnal/7pSfd4kc9aW8Rq0mhY8I6e19qYkb0qSo48+elFbvnnz5iTJy172smnjTznllFz/qU8tuZFJcugPfpB9jzxyp2Uu1ymnnLJbfLUqGV1I+cEPfnA+8IEPzHsQz7wA64EHHrhj+Nhjj82ZZ56Zm266KZs2bcqP/MiP5N///d/nvP/gBz94p+XPXMZs8yzVaiyzB3e6052mddDufOc7r2NrVtdK1KJTTjkll4x9vWauujJpX61abVW1Iv/Yxo+/Y489dqez/zZt2rSmr9X5atlSLNTmSaw/u9Km5WzHJO6D1bKS/aLZLKevtFSr1bdaDp9+72xmjZ46Q2e2GvegBz1op9fihRdeOGtfYeYPbGzatLQT9Se1P7e71KPVrkUrZS1r2kqaxPq4nl7/+tfP+a2O9bIr/deV6vMmyYMf/OCdzshJpteg1eQaObuRPfbYI8cff3xOOumkWaf/4i/+YpLk5JNPnjb+hS984Y7h448/PnvssceO5T3vec+bdv/5z3/+tPvHH3/8TuuZuYzZ5lmq1VhmD57//OfPex8WY9OmTTt14GcbN9vjxo0ff8cff3z23HPPadOf97znrelrdb5athQLtXkS68+utGk52zGJ+wBW06GHHjrt/h3veMck2emrScccc8ys4/7bf/tv08ZNfYDwtKc9bdr4E044Yad+27Of/ewks9e42V6Lc/UVnvvc504b/7znPW/mZs5rUvtz6hGsvEl8HU39KutMU6//uR6zWDO/1jpz2vHHHz9rzR6vQatJkJPkJ37iJ9a7CauuqvLQhz40Bx10UB796EfPemBO/fz4fe5znxxwwAFJRp/ujP9k78EHH5yHPOQhO5Z35JFHTrt/xBFHTLs/208+zlzGSvws5GosswdHHnlk7nSnOyUZfcLm58dXxkr/POByjb9e5/unMj59trNPDjzwwFTVjmlT9x/2sIflYQ972LR5H/7wh+80buZ6Hv7wh895/B188MF5+MMfvuP+pk2b8sAHPnBNX6vz1bKlWKjNk1h/dqVNy9mOSdwHbBwza/IZZ5yx6HFLefxS5n3zm988bdzUz4/PDFde8IIXzDpuZjgz9fPjv/ALvzBt/HHHHTet31ZVO35+fLYaN9trca6+wjHHHLPjTc1UjV6KSe3PqUew8g4++OAcfPDBq7Lsqfo01S9djKn+63hfc2oZj3jEI2Z9zBlnnJGHPvSh08aNBztT9XTKVK1Nkkc/+tHTpj3qUY/KQQcdNGvNHq9BM9e/kgQ562DTpk3Zf//955w2buqAmnrh/OiP/mge8pCHJEnucpe7JBn9I95jjz1yzDHHJEm2bNmS5NbTJ4855pjc8573nJakTnUg7nOf+yS59WycKSeffHL22GOPWT/BPv7446ctb6H7s1nMPEu1GsvswfOf//zsv//+zsbp0Mx/GOP22WefPP3pT08y+vT1h3/4h3cMP/GJT0wyqgtHHXVUjjrqqBx22GGpqpx00knZf//9c/LJJ+fOd75zqip77bVX9t1335x88sm55z3vmd/6rd/aMc/Ua+b444/PUUcdlX333TdHHXXUjnF3u9vdcpe73CX77bffjn9WJ5xwwo7HzXf8HX/88TnssMOSTP+kdy1fq/PVsqVYqM2TWH92pU3L2Y5J3AewmqbOypk6G2fKVCAy1S+ba9zUWTkzv8479QnvCSecsGPcVL9t6mycKbPVuNlei3PV6qmzcpZ6Ns5861qulVimegQr79BDD83ey7hw9VRfNkn22muvJKOaOFWfpvql46Hy1Bk34/UwybT+69R74uTWMxOn3gfPdPzxx+fud797Dj/88FRVnve85+Vud7tb7n73u+eFL3zhjvDlmGOO2THv3e52t2ln39zxjnecVltmq9lTNWg11VK+I3b00Ue3c845Z8H5pr7fPNc1cn59lp9Ifs3w6cBs06am7/tf/+uqXCNntrbCRlVV57bWjl7vdizHrtaiqWvkHH799fn1yy6bt66oDbC6dudatNC8c/WVZlqo77SYx69G32o51F7WQ+/1aDVq0UpZSk1byHJr3lLXNWn1cb2pz6tvKbXIGTkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACd2LQaCz3yyCNXY7Groqe2AkuznNe32gCsFPVk8ewrWD1eXyyH42eyrEqQc+KJJ67GYldFT20FlmY5r2+1AVgp6sni2Vewery+WA7Hz2Tx1SoAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE5vWeoVX7rtvXnPEETuN/9q++ybJrNOmHnfUqrYM2B38YI898rWhDqkrwCSaq68000J9p8WsRw0EVttia9pCllvzlkJ9ZNKtaZBz5JFHzjntgO3bkyT7bt486/SjFng8wEKOPPLIbB+rNeoKMGmWUpMW6jstRA0EVttK1pjl1rylUB+ZdGsa5Jx44olruTqAaU488UR1CJhoahSwkahpsDpcIwcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOVGtt8TNXXZ3kK4uc/ZAk/7ErjVoHvbS1l3Ym2rpaVqKtd2mtHboSjVkvS6hFPT23S7VRt8129WU527U71aLEMdCbjbhdG3GbEn2jxdSi9X7urd/6rX9xFl2LlhTkLEVVndNaO3pVFr7CemlrL+1MtHW19NTWSbCR99dG3Tbb1ZeNul2rYaPuK9vVj424TcnG3a6VtN77yPqt3/pXfv2+WgUAAADQCUEOAAAAQCdWM8h50youe6X10tZe2plo62rpqa2TYCPvr426bbarLxt1u1bDRt1XtqsfG3Gbko27XStpvfeR9Vu/9a+wVbtGDgAAAAAry1erAAAAADohyAEAAADoxLKDnKq6U1V9pKo+X1Wfq6pfH8YfXFVnVtWXh78HLb+5K6Oq9qyqz1bV+4f7R1TVp6rqkqp6e1Xtvd5tTJKqun1VvauqvlhVX6iqB0zqfq2q5w7P/0VV9baq2ndS9mtVvbmqrqqqi8bGzbofa+RPhjZfUFX3Xed2vnJ4/i+oqvdU1e3Hpp0ytPNLVfWItWrnJKqqRw774ZKqeuEs0/cZjsFLhmNyyzo0c8kWsV0nVNXVVXXecPsf69HOpZrtWJ8xfd1eh8uxiO06tqq+M/Z8vXit27gr5vo/P2OeLp+ztbDQ67gnS/l/2ou5ju8NsF37VtWnq+r8Ybt+Zxg/EX2z5ahO+vGTYDXrT1VdXlUXDv/PzhnGLbl/XVVPH+b/clU9fZ71rUh/fq71VdX9hu25ZHhsLWL9L6mqK8f+rz96bNqs/fS5npOFjuOl1qqV3gfzrH9N9kEtsabVPH3/pbZrgfWfXlWXjW3/vVdj/8+qtbasW5LNSe47DN8mycVJ7pnkFUleOIx/YZKXL3ddK3VL8ptJ/m+S9w/335Hk+GH4jUmetd5tHNryliT/YxjeO8ntJ3G/JjksyWVJ9hvbnydMyn5NckyS+ya5aGzcrPsxyaOTfCBJJfnpJJ9a53Y+PMmmYfjlY+28Z5Lzk+yT5IgklybZc72PhXV6fvcctv/I4XVyfpJ7zpjn2UneOAwfn+Tt693uFdquE5K8dr3bugvbttOxPmP6ur0OV3m7jp36v9PTba7/8xvhOVuDfbfg67in21L+n/Zym+v43gDbVUkOHIb3SvKp4bU5EX2zZW5bF/349b6tdv1JcnmSQ2aMW1L/OsnBSbYNfw8ahg+aY33L7s/Pt74knx7mreGxj1rE+l+S5PmztHXWfvp8z8lCx/FSa9VK74N51r8m+yBLrGmZo++/K+1aYP2nJ3niLNu/4sfgzNuyz8hprW1vrf3bMPyfSb6Q0Rv7x2UURGT4+/jlrmslVNXhSf5bkj8f7leSn0vyrmGWiWhrVd0uo4LxF0nSWvtBa+3bmdD9mmRTkv2qalOS/ZNsz4Ts19ba2UmumTF6rv34uCRvbSOfTHL7qtq8Xu1srX2otXbTcPeTSQ4fa+fW1toNrbXLklyS5P5r0c4JdP8kl7TWtrXWfpBka0b7Z9z48/2uJA9ZMOVef4vZri7N8Zoct26vw+VYxHZ1aZ7/8+O6fM7WwIZ6HS/x/2kXeuvHLtbwWrx2uLvXcGuZkL7ZruqlHz8h1qP+LLV//YgkZ7bWrmmtfSvJmUkeOduCV6g/P+v6hmm3ba19so3eUb81M46jJf6Pn6ufPutzspjjeBdq1Yrug0X2BVZtH+xCTZur77+kdi1i/fNt/4oegzOt6DVyhlOW7pNRQvVDrbXtw6SvJ/mhlVzXMvxxkt9Kcstw/w5Jvj32Zvlrmf+gXCtHJLk6yV/W6PTRP6+qAzKB+7W1dmWSVyW5IqMA5ztJzs1k7tcpc+3Hw5J8dWy+SWr3MzJKZ5PJbudaW8y+2DHPcEx+J6PX/iRb7HN83HDK5ruq6k5r07RVt5GP7wcMp+V+oKp+fL0bs1Qz/s+P28jP2XLsDvtl4volu6qTfuyi1egrSOcluSqjNwuXZrL7Zovxx+mjHz8JVrv+tCQfqqpzq+pXh3FL7V8vt40rtb7DhuFdacf/HPphb65bv4K51PUv6TheZK1atX0wS19gTfbBEmvaXH3/XT4WZ66/tTa1/b8/bP8fVdU+u7j9Sz4GVyzIqaoDk7w7yW+01r47Pm1IleZLrNZEVf18kqtaa+eud1sWYVNGp++9obV2nyTfy+h0uR0maL8elFHqeESSH0lyQOZI0yfRpOzH+VTVi5LclORv1rstTJQzkmxprf1URv/Q3rLA/Kyvf0tyl9bavZL8aZL3rm9zlma+//OQ9PH/dC499GOXqrV2c2vt3hmdzXv/JPdY3xYtT2f9+N3BA1tr903yqCTPqapjxieu9etmnV6nb0hyVJJ7Z/Rh9qtXe4XrXatmWf+a7YP1rmkz119VP5HklKEd/yWjr0udvFbtWZEgp6r2yugJ/ZvW2t8Oo78xdVr18PeqlVjXMv1sksdW1eUZnS71c0lek9GpTpuGeQ5PcuX6NG+aryX52ljS966Mgp1J3K8PTXJZa+3q1tqNSf42o309ift1ylz78cok42c1rHu7q+qEJD+f5MlDgU4msJ3raDH7Ysc8wzF5uyTfXJPW7boFt6u19s3W2g3D3T9Pcr81attq25DHd2vtu1On5bbW/iHJXlV1yDo3a1Hm+D8/bkM+Zytgd9gvk9gvWZKO+rG7pI2+mv+RJA/IZPfNFtJTP34SrGr9Gc7IT2vtqiTvyeiN9VL718tt40qt78rcevmCRbejtfaN4c39LUlOy62XOVjq+r+ZRRzHS6xVK74PZlv/Wu+DYZ3fzsI1ba6+/7KPxbH1P3L4ylkb+uN/uYztX/IxuBK/WlUZXcflC621Pxyb9L4kTx+Gn57k75a7ruVqrZ3SWju8tbYlo4se/VNr7ckZPRFPHGablLZ+PclXq+ruw6iHJPl8JnC/ZvSVqp+uqv2H42GqrRO3X8fMtR/fl+Rpw5XGfzrJd8ZOV1xzVfXIjE4hfmxr7ftjk96X5PgaXZH9iCR3zegCWbujzyS5a42uWr93Rq/t982YZ/z5fmJGr/1J/3R1we2acQ2Sx2b0feWNYKJehyulqn546tpMVXX/jP4HT3qgON//+XEb8jlbAYupT72bxH7JovXUj12Kqjq0hl+6rKr9kjwso/8Rk9w3m1dP/fgJsWr1p6oOqKrbTA1n9OMcF2Xp/esPJnl4VR00nOH/8GHcYq3I+oZp362qnx5qwtOyiONoRj/sCcM+mFr/bP30WZ+ToU8673G8C7VqRffBXOtfq32wCzVtrr7/ktq1wPq/OBaiVUbXtBnf/tU9Btvyr1j+wIxO4bogyXnD7dEZfQftrCRfTvLhJAcvd10recvYr4dkdHXqT2d0saN3Jtlnvds3tOveSc4Z9u17M7qy9UTu1yS/k+SLw8H7VxldCXwi9muSt2V0qt+NGZ3p9My59mNGVwl/XUbfubwwydHr3M5LMvoe5dRr641j879oaOeXssBVzTf6bag5Fw/740XDuJdmFIAlyb7DMXjJcEweud5tXqHtelmSz2V0Zf2PJLnHerd5kds127F+UpKThunr9jpc5e36n2PP1yeT/Mx6t3mR2zXX//nun7M12n87vY57vS3l/2kvt3mO796366eSfHbYrouSvHgYPxF9sxXYvmMz4f34SbitVv0Z9vn5w+1zY32UJfevM7oG5CXD7VfmWeeK9OfnWl+So4fXyqVJXpukFrH+vxqWf0FGb9w3j80/az99rudkoeN4qbVqpffBPOtfk32QJda0zNP3X2q7Flj/Pw3bf1GSv86tv2y14sfgzFsNDwIAAABgwq3or1YBAAAAsHoEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AsCRV9eaquqqqLlp47qSqfrGqPl9Vn6uq/7va7QMA2MgEORtYVW1a7zYALEdV/UNV3X6928FOTk/yyMXMWFV3TXJKkp9trf14kt9YvWbBZKqqy6vqkBVe5paq+uWx+ydU1WtXch3Awqrq2vVuw3JU1eOr6p7r3Q6WRpAz4YZ/0l+sqr+pqi9U1buqav+qul9V/XNVnVtVH6yqzcP8H62qP66qc5L8elX9QlVdVFXnV9XZwzz7VtVfVtWFVfXZqnrwMP6EqvrbqvrHqvpyVb1igba9oarOGT5h/Z2x8Y8e2nxuVf1JVb1/GH/A8Cnup4f1Pm7VdhywJlY7MG6tPbq19u3VXAdL11o7O8k14+Oq6qjh/8e5VfUvVXWPYdKJSV7XWvvW8Nir1ri5dMoHUgvakuSXF5oJYAGPTyLI6Ywgpw93T/L61tqPJflukuck+dMkT2yt3S/Jm5P8/tj8e7fWjm6tvTrJi5M8orV2rySPHaY/J0lrrf1kkl9K8paq2neYdu8kT0ryk0meVFV3mqddL2qtHZ3kp5L8P1X1U8Ny/izJo4a2HTo+f5J/aq3dP8mDk7yyqg7YlR0CrJwJD4wvr6pDhjZ+oapOG8LjD1XVfsM8P1pVHx7W/29DoFBV9cqhXRdW1ZOGeY8dtunvqmpbVZ1aVU8eAuYLq+qoYb5Dq+rdVfWZ4fazq/gUbBRvSvJrQ+1/fpLXD+PvluRuVfXxqvpkVS3qTB42hgmvLw+vqk8MdeOdVXXgMP7yqnpZVZ1Xow+s7ju08dKqOmmY59iqOruq/r6qvlRVb6yqRfWrq+opQ805r6r+rKr2HMZfW1W/P2zrJ6vqh4bxRw33L6yq36tbP/0/NcmDhuU8dxj3I4vdfmBlzdP3OLCqzhpqzYU1fJg9X99mjuWfOPRJzh/6KPsP40+v0Qfsnxz6NsfW6MPzL1TV6WOP36nGVNXPZPQe8ZVDLTlqVXcSK6e15jbBt4w+bbli7P7PJflwRoHOecPtwiQfGqZ/NMn/Mzb/G5OcmdEnoncYxr0nyc+NzfMvGYUxJyQ5bWz8B5I8cJ62nZTk35JckOTqJMdnFAT989g8j03y/mH4nCQXjbX7iiQ/tt772M1td78NdaZl9NWXZBQOvyDJvyY5dBj3pCRvHoY/mlG4PPX4C5McNgzffvj7vLH57zG83vcd6sy2JLcb7n8lyZ3madvlSQ4Z2nhTknsP49+R5CnD8KeSPGEY3jfJ/kmOG2rfnkl+aFj/5iTHJvn2MLxPkiuT/M7w2F9P8sfD8P+dqn9J7pzkC+v9PE3abXhOLhqGD0xy3Vh9P29qnyV5//B/Z68kRyT56tRx4rbxb5NaX4a6cnaSA4b7Jyd58TB8eZJnDcN/lFE/5zYZfTj1jWH8sUmuT3LkUGfOzOgDtrn2w1Qt+7EkZyTZaxj/+iRPG4ZbkscMw69I8tvD8PuT/NIwfFKSa8fa8P6xdSx6+93c3FbuNvaanKvvsSnJbYd5DklySZLKPH2bOdZzh7Hh38vow5Nk9HXnrcMyH5fR+8SfzOikjXPHlj9XjTl9vvrlNpk3p6z2oc24/59JPtdae8Ac839vxwNbO6mq/muS/5bk3Kq63wLrumFs+OZk9mOkqo7I6BPX/9Ja+9aQ9u4727zjD0tyXGvtSwvMB6y9r7bWPj4M/3WS/53kJ5KcWVXJqFOyfWz+t48NfzzJ6VX1jiR/O4x7YEZnDqa19sWq+kpGZ2YkyVmtte8kSVV9PsldMnpzv5DLWmvnDcPnJtlSVbfJ6E3ee4Z1XT8s94FJ3tZauznJN6rqn5P8l4w6N59prW0f5rs0yYeGZV6Y0dmCSfLQJPcctj1JbltVB7bWuv4e/CraI8m3W2v3nmXa15J8qrV2Y5LLquriJHdN8pk1bB/raxLry09n9FWCjw9t2DvJJ8amv2/4e2GSA1tr/5nkP6vqhrr1ul2fbq1tG9b1tqFd71pgXzwkyf2SfGZY735Jpr5u+IOMQptkVOMeNgw/IKOvPiSjkPlV8yx/V+srsHxz9T0+kOQPquqYJLckOSyjoCeZpW8zz/J/oqp+L8ntM/oA5YNj085orbWqujCjwPnCJKmqzw3LPC9z1xg65KtVfbhzVU2FNr+c5JNJDp0aV1V7VdWPz/bAqjqqtfap1tqLMzpr5k4ZnYHz5GH63TL6tHmp4cptMwqMvjOc+vuoYfyXkhxZVVuG+08ae8wHk/xaDT2XqrrPEtcJrJ65AuN7D7efbK09fGz6tMA4yW9nVF/Orao7LLCuRQXGK/i4+ZZzy9j9W8aWuUeSnx7b/sOEOHNrrX03o5DmF5Idp5ffa5j83ozOHEiNLvZ6t4zOGmD3MYn1pZKcOdaGe7bWnjnLcsZrxNT9qWXO3K6Z9+da71vG1nv31tpLhmk3ttamlrGrNW6l6iSwcp6c0Rl99xs+8PhGbv0AfCmv2dOT/M82ujzG72T6h+iLqVkrUWOYEIKcPnwpyXOq6gtJDspwfZwkL6+q8zNKWH9mjse+cvgu5kUZncZ8fkan8e4xJLZvT3JCa+2GOR4/q9ba+Uk+m+SLGX069PFh/HVJnp3kH6vq3Iw6a98ZHva7GZ1af8GQDv/uUtYJrKpJDIwXNHxK/rWqevywrn2G74z/S0bX+dqzqg5NckySTy9h0R9K8mtTd6rq3ivW6A1gOPvgE0nuXlVfq6pnZvR8P3P4v/S5jE7vTkYh/jeHswM+kuQFrbVvrke7WTeTWF8+meRnq+pHh+UcMCxrKe5fVUfU6No4T0rysUU85qwkT6yqOw7rPbiq7rKIth43DB8/Nv4/M/rKFzAZ5up73C7JVa21G2t0Ta+FXvNzuU2S7VW1V4YauELUkg5J4fpwU2vtKTPGnZdRcZimtXbsjPv/fZblXZ/kV2Z57OkZJb1T939+vka11k6YY9JHWmv3GM68eV1G18aZCnn+3/mWCaybqcD4zUk+n1Fg/MEkf1JVt8vo/8UfZ/QGfaZX1ugnpiujNynnZxTyvmEIjG/KEBiPfVVpJT01yZ9V1UuT3JjkFzK6JssDhra0JL/VWvt63fpLSgv5X0leV1UXZLTtZ2d0bQqStNZ+aY5JO13IePj07zeHG7uniasvrbWrq+qEJG+rqn2G0b+d5OIlbNdnkrw2yY9mFFK+ZxHr/XxV/XaSDw0B0I0Z/QjFV+Z52G8k+euqelGSf8ytH5BdkOTmITw9Pcm3ltB2YOXN1ff4myRnDDXrnIxq2K74/zK6LuDVw9+VCl+2Jjmtqv5XRtfKuXSFlssqqlvPrmISDV9Ren9r7SfWuy2LVaNfTnh6Rt83/2ySE1tr31/fVgFz6bHOAH3YqPWlqo5N8vyFPvRaoXXtn+S64foXx2d04ePHLfQ4ADYuZ+RMuNba5RldEHDdVNWnMvp1l3FPnbqI1kyttT/K6FceAABYnvslee1wpvO3kzxjfZsDwHpzRg4A626pgTHAYq1lfVHLgOWqqtcl+dkZo1/TWvvL9WgPk0mQAwAAANAJv1oFAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0IlNS5n5kEMOaVu2bFmlpgBr4dxzz/2P1tqh692O5VCLoH9qETApeq9HahFsDEupRUsKcrZs2ZJzzjln11oFTISq+sp6t2G51CLon1oETIre65FaBBvDUmqRr1YBAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRi03o3YC2ddtpp2bZt27qse/v27UmSzZs3r/q6jjzyyJx44omrvh6YdKeddlqSeD0AG9py+zer1UfRH4G+LFRL1vL9zGzUFLjVbhXkbNu2LZeef34Ou/76NV/39/bdN0ly/RVXrOp6rhzWAyQf/vCHkwhygI1tuf2b1eij6I9AfxaqJWv1fmY2agpMt1sFOUly2PXX59cvu2zN1/uaI45IklVf99R6AIDdx3L6N6vRR9EfgT7NV0vW6v3MfOsGRlwjBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOjEptVY6GmnnZYkOfHEE1dj8bDmHNN9uu6669a7CQD+h2xAnlN65LidXJ4blmpVgpxt27atxmJh3Tim+9RaW+8mAPgfsgF5TumR43ZyeW5YKl+tAgAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATm9a7AdCLL3/5y3nMYx6z4/7mzZuz1157Zfv27bnxxhtz+OGH50d+5Efy6U9/Ovvvv39aa3nMYx6Td7zjHTse85znPCePfOQjc8011+SFL3xhtm/fPm3ci1/84lxxxRX53d/93dzrXvfKNddck1e84hU5+eSTc9BBB83arsXMA7DaxuvjGWecsY4tgbU1fuyvls2bN+c2t7lNLr744tzjHvfILbfckosvvjj7779/vv/97+eYY47Jj//4j+cNb3hDnvOc5+Qd73hHrr766uy111658cYbd5p+//vfP694xSty44035uKLL85P/uRP5oEPfOCO6fvvv39e+cpX5oQTTsg73vGOnHrqqfnqV7+aV77ylTn55JNzwAEH5CUveUlOOumknH766Tn11FPz7W9/Oy95yUvy0pe+NLfcckte8pKX5PnPf37+/u//flF9lPH+zLe+9a2ccsopOeWUU7J169Yl93H0jdhIFvr/utD0ZzzjGbn66qtzxzveMX/xF3+x0/SnPvWp+fa3v52DDjoob33rW3ea/vjHPz4333xz9txzz7z3ve/dafrTn/70XHPNNbnDHe6Q008/fafpT37yk/Pd7343t7vd7fLXf/3XS5r+hCc8ITfddFM2bdqU97znPTs99qSTTsqVV16Zww8/PG94wxt2mv6c5zwnV1xxRbZs2ZI//dM/XXLbnvjEJ+aGG27IPvvsk3e96107TX/e8563oy6/8pWvXPLyl0OQA4t0ww03TLu/ffv2afe/9rWv5Wtf+1qS5Pvf/36STAtxkuT1r399HvnIR2br1q07Hj8+7itf+UqS5NRTT83b3va2bN26NZ///OezdevWPOtZz5q1XYuZBwDo1/bt23f0G774xS/uGD/V3zj77LPzL//yL0lG/YrWWpLkxhtvnHX6tm3b8vnPf37HfBdeeGEuuuiiHdP33HPPJNnxpuxVr3pV/v3f/z1J8upXvzr77LNPbrnllrzhDW9Iay2vetWr8s1vfjO33HJLTj311LTWcsstt+QP//APc/PNNy+qjzLen7nwwgvz/e9/Py9/+cvzve99b8l9HH0juNXVV1+dJLnqqqtmnf7tb387SfKtb31r1uk333zztL8zXXPNNUmSb37zm7NO/+53v5sk+c53vrPk6TfddNO0vzNdeeWVSbLjPdhMV1xxRZLk8ssv36W2Tb3/m/k+cMrFF1+cZHpdXsryl8NXq2ARvvzlL6/Iclprefe7350PfehDO40788wzd4y79tpr87GPfSxnnXVWWmv58Ic/PGtxveaaaxacB2C1zTwjYS3OUIBJMEnH+lQoM/V3vukf/OAHd5pvfPrMN01XXHHFtDdU3/ve96Y95oorrtgx7tprr90xfNNNNy2qjzLenznzzDPz1a9+dceyltrH0TdiI1no/+tC95/xjGdMu//MZz5z2v2nPvWp0+4/7WlPm3b/8Y9//Lz3n/70p0+7f8IJJ0y7/+QnP3na/ac85SmLnv6EJzxh2rSZ90866aRp92eGts95znOm3f+1X/u1JbXtiU984rz3n/e85027/4IXvGBJy1+uVTkjZ/v27bnuuutyyimnrMbid9m2bduy1957r3czVtXVe++dG7dtm7h937u5UthdMdsph7ONe/WrX71j+JZbbpn1U6WtW7fmlltumXceANbXaveLJrF/s9H7I9u2bct+++233s3YZVN9h7Vc30JnF0+1aeosoqU8fq5l6RtNt1AtmsRaMkVN2TVTZ+NMmXlWztTZOFNmBp8zz8KZeX/qbJwpM8/KmTojZcrMM1Pmmz4zUJ55f+psnCkzz8qZOhtnysyzchZq28z3fzPvT52NM2XmWTkLLX+5Fjwjp6p+tarOqapzZh4IwOq56aabpn369ZGPfGSneT760Y8uOM9GoRYBk0AtgqVbqI8y3p/ZlcfPtayN3DdSi2D3tuAZOa21NyV5U5IcffTRs5+rOcPmzZuTJC972cuW07YVd8opp+T6T31qvZuxqg79wQ+y75FHTty+7916nDq9adPo5Tl1ga8HP/jBO81z7LHH5swzz5x3no1iV2oRwEqbxH7RJPZvNnp/ZOqsgLmuCcF0C/VRxvszu/L4uZa1kftGq1GLJrGWTNldagoslmvkwCLss88+K7asE044YcdFBMfHTQU3U573vOdljz1GL9E99tgjxx9//E7LOv744xecBwBg3FTfYS3XN18fZbw/s9deey358XMtS9+I3d2hhx467f4d73jHafdvf/vbT7s/81feZr5nmXn/4IMPnnb/Dne4w7T7t73tbafdv93tbrfo6TPfG828f9hhh027f/jhh0+7f+c733na/S1btiypbTPf/828f7e73W3a/Xvc4x5LWv5yCXJgEe5617uuyHKqKscdd1we/vCH7zTuYQ972I5xBx54YB74wAfmIQ95SKoqD33oQ2f9+cyDDz54wXkAVtvMnzv18+PsLibpWK+qaX/nm/6IRzxip/nGp898w3TnO995x7hNmzblgAMOmPaYO9/5zjvGHXjggTuGN23atKg+ynh/5mEPe1judKc77VjWUvs4+kZsJAv9f13o/pvf/OZp92f+/Phf/dVfTbs/8+fHZ/7c+Mz7b3nLW6bdn3ndz7/5m7+Zdn/mT3DPN33mz43PvP/GN75x2v2ZPz/+ute9btr9mT8/vlDbZv7c+Mz749czTbLTz48vtPzlEuTAIs1MYTdv3pw73/nOOz45Ovzww3P/+98/SbL//vtnv/32yy/+4i9Oe8yzn/3sJKNPi6ZObx0fd5e73CVVlRe+8IU7xt3znvdc8FOsheYBAPq1efPmHZ/+3uMe99gxvP/++ydJjjnmmB2/4PLsZz97x6fwU32UmdOn+g5Ty/nJn/zJadOf+9znJhmdMbz//vvn+c9//o5xz3ve83LyySdnjz32yLOe9awd06fGvfCFL9wx/Ju/+ZuL7qOM92ee//znZ//998/JJ5+8S30cfSO41VQ9mHk2zpSps3LmCj2nzsKZeTbOlKmzcmaejTNl6syUuc5ImW/6eIA8m6mzcmaejTNl6qycmWfjLLZtU+//5vp2xnhd3pXlL8eq/GoVbER3vetdd+l7uTN/1i8ZFbw3velNO4177Wtfu9O4U089dd7lL2YegNU2SWcmwFqapGP/0Y9+dJLkkY985LzTk8zadxiffswxxyRJjjvuuCTJEUccsWNckvzd3/1dkuRRj3rUTuPGhx/0oActqu3j/ZmDDjoob3/725Mk9773vRf1+LmWBb1bqMYsNH3mWTkzzTwrZ6aZZ+HMNPOsnJlmnpmylOkzz8KZaeZZOTPNPCtnKetOdj4LZ6aZZ+UsdfnL4YwcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADqxaTUWeuSRR67GYmHdOKb7VFXr3QQA/0M2IM8pPXLcTi7PDUu1KkHOiSeeuBqLhXXjmO7Tfvvtt95NAPA/ZAPynNIjx+3k8tywVL5aBQAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRi03o3YK1due++ec0RR6z5er+2775JsurrvnLffXPUqq4BAJg0y+nfrEYfRX8E+jRfLVmr9zOzUVNgut0qyDnyyCPXbd0HbN+eJNl38+ZVXc9RWd/thEny0Ic+dL2bALDqlvt/fzX6KPoj0J+FXrNr9X5mNmoKTLdbBTknnnjiejcBWENe88DuQK0DVoJaAv1wjRwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6Ua21xc9cdXWSr8wy6ZAk/7FSjVqi9Vr37rjN1r0x1n2X1tqhK7i8NTdPLZrNej53u0J7V5f2rq6ltFct6p9t6sNG3KZkZber63q0xFq02ib5eJvktiWT3T5t23Wr0jdaUpAz50KqzmmtHb3sBXW07t1xm61791v3RtDb/tPe1aW9q6u39q6ljbhvbFMfNuI2JRt3u3o3yc/LJLctmez2aduuW632+WoVAAAAQCcEOQAAAACdWKkg500rtJye1r07brN1737r3gh623/au7q0d3X11t61tBH3jW3qw0bcpmTjblfvJvl5meS2JZPdPm3bdavSvhW5Rg4AAAAAq89XqwAAAAA6seggp6our6oLq+q8qjpnlulVVX9SVZdU1QVVdd+VaGBV3X1Y59Ttu1X1GzPmObaqvjM2z4uXuc43V9VVVXXR2LiDq+rMqvry8PegOR779GGeL1fV01dgva+sqi8O+/Q9VXX7OR477/Ozi+t+SVVdObZfHz3HYx9ZVV8anvsXrtC63z623sur6rw5Hrvc7b5TVX2kqj5fVZ+rql8fxq/F8z3XutfkOd9oFjoOq2qf4bi6pKo+VVVb1qGZ4+1ZqL2/ORwbF1TVWVV1l/Vo51h7FvU6r6rjqqpV1br+esBi2ltVvzj2+vu/a93GGW1Z6Hi481AvPjscE7PW47UyW92eMb1qFfoEPeitFi1Gb/VqMXqraYvRW91bjN5q4+5s0vqls/2fWmz/fp3atqj3XWvQtmW9N1rH9q37/quqfavq01V1/tC23xnGHzH8v79k+P+/94qssLW2qFuSy5McMs/0Ryf5QJJK8tNJPrXYZS+hDXsm+XpGv68+Pv7YJO9fwfUck+S+SS4aG/eKJC8chl+Y5OWzPO7gJNuGvwcNwwctc70PT7JpGH75bOtdzPOzi+t+SZLnL+I5uTTJkUn2TnJ+knsud90zpr86yYtXabs3J7nvMHybJBcnuecaPd9zrXtNnvONdFvMcZjk2UneOAwfn+TtE97eByfZfxh+1qS3d5jvNknOTvLJJEdPcnuT3DXJZ6des0nuOOHtfVOSZw3D90xy+Xq1d2jDQnV71fsEk3jrrRat4DZNTL1aqW0a5puImraCz9PE1L0V3KaJqo278y0T1i+d7f9UFtG/X8e2vSQLvO9ao7bt8nujdW7fuu+/oc9z4DC8V5JPDX2gdyQ5fhj/xqmatdzbSn616nFJ3tpGPpnk9lW1eQWXnyQPSXJpa+0rK7zcaVprZye5ZsboxyV5yzD8liSPn+Whj0hyZmvtmtbat5KcmeSRy1lva+1DrbWbhrufTHL4Ype3FHNs82LcP8klrbVtrbUfJNma0b5akXVXVSX5xSRv24W2LWbd21tr/zYM/2eSLyQ5LGvzfM+67rV6zjeYxRyH48/pu5I8ZDi+1sOC7W2tfaS19v3h7nofB4t9nf9uRuHj9WvZuFkspr0nJnnd8NpNa+2qNW7juMW0tyW57TB8uyT/vobt28ki/mesRZ9gEvVWixajt3q1GL3VtMXore4tRne1kcmxjPdzq24Z77tW3TLfG61n+9bd0Oe5dri713BrSX4uo//3yQruu6UEOS3Jh6rq3Kr61VmmH5bkq2P3v5aV36nHZ+439A8YTmP6QFX9+AqvN0l+qLW2fRj+epIfmmWe1d4Hz8joE87ZLPT87Kr/OZyq+uY5TqFb7W1+UJJvtNa+PMf0Fdvu4fT2+2SUnq7p8z1j3ePW4znv0WKeix3zDEHZd5LcYU1at7OlHjvPzNzHwVpYsL3DV2fu1Fr7+7Vs2BwWs3/vluRuVfXxqvpkVS06hF0Fi2nvS5I8paq+luQfkvza2jRtl61Fn2AS9VaLFqO3erUYvdW0xeit7i3GRqyNG1kP/dLF9O/X00Lvu9bULrw3WlOzvH9a9/1XVXvW6JIgV2X0If+lSb499iH9ivWHlhLkPLC1dt8kj0rynKo6ZiUasFjDd8kem+Sds0z+t4y+bnWvJH+a5L2r2ZY2Oi9qTX/uq6pelOSmJH8zxyyr8fy8IclRSe6dZHtGX3Faa7+U+c/GWZHtrqoDk7w7yW+01r47Pm21n++51r1OzzkTpqqekuToJK9c77bMpar2SPKHSZ633m1Zgk0Zfc3g2IzqzGk1x/WoJsQvJTm9tXZ4Rl9b+qthv8PE6KFeLUanNW0xeqt7i6E2To6u+qXr8X5uAZPwvmuH9XxvtBiztG8i9l9r7ebW2r0zOjP1/knusVrrWnSha61dOfy9Ksl7MmrYuCuT3Gns/uHDuJXyqCT/1lr7xixt++7UaUyttX9IsldVHbKC606Sb0ydFj78ne101FXZB1V1QpKfT/Lk4YWzk0U8P0vWWvvGcDDekuS0OZa5as97VW1K8t+TvH2eNi57u6tqr4wKwd+01v52GL0mz/cc616357xji3kudswzHFu3S/LNNWndzhZ17FTVQ5O8KMljW2s3rFHbZrNQe2+T5CeSfLSqLs/o+8Dvq/W7OOhi9u/XkryvtXZja+2yjL5jfdc1at9Mi2nvMzP6jnVaa59Ism+Slf4/t5JWu08wqXqrRYvRW71ajN5q2mL0VvcWYyPWxg2rk37pYvr362KR77vWxDLeG61b+yZp/w3t+XaSjyR5QEZfL980TFqx/tCigpyqOqCqbjM1nNHFWGf+UsX7kjytRn46yXfGTr9aCXOemVFVPzz1/fKqun9G27XSnaL3JXn6MPz0JH83yzwfTPLwqjpoOJ3r4cO4XTac9vpbGXWMvj/HPIt5fnZl3ePXM3jCHMv8TJK7Dlfj3jujr7+9b7nrHjw0yRdba1+bo33L3u7huPmLJF9orf3h2KRVf77nWvd6PucdW8xxOP6cPjHJP80Vkq2BBdtbVfdJ8mcZHQfr3dGYt72tte+01g5prW1prW3J6BoZj22trdevVizmeHhvRp9KZwj+75bRBcvXw2Lae0VG14lLVf1YRm9Wrl7TVi7NavcJJlVvtWgxeqtXi9FbTVuM3ureYmzE2rghddQvXUz/fl0s8n3XWrRjOe+NVt0875/Wff9V1aFTZzlW1X5JHpbRNXw+ktH/+2Ql911b3BWYj8zoSvHnJ/lckhcN409KclK79SrNr8voe2AXZgWv7p/kgIyCmduNjRtf9/8c2nV+Rv9sf2aZ63tbRqdk3ZjRpxfPzOj762cl+XKSDyc5eJj36CR/PvbYZyS5ZLj9ygqs95KMvh983nCb+qWLH0nyD/M9Pyuw7r8anssLMnrxbp657uH+ozP6VOfSlVr3MP70qed4bN6V3u4HZnRq4AVj+/jRa/R8z7XuNXnON9pttuMwyUsz6nwno87dO4f9++kkR054ez+c5Btjx8H7Jrm9M+b9aNb5F14WsX8ro69OfD6jOnf8hLf3nkk+Przmz0vy8HVu72z/M9akTzDpt95q0Qpt00TVq5XYphnzrntNW6HnaaLq3gpt00TVxt31lgnsl87xf2rW/v2EtG3W913r0LYlvTeaoPat+/5L8lMZ/TLgBRkFSS8exh85/L+/ZPj/v89KrK+GhQMAAAAw4VwMDAAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHImUFVdu0breWlVPXSBeY6tqp9ZofXdu6oevRLLAtbeWtWmOda9qFq0kjULWLr1rBPrYTE1p6pOqqqnrVWbYKOqqtOr6onr3Y5JUlVbquqieaafUFWvnWPaP1TV7ed57G9U1f4LrP/yqjpk0Q1ehGGbfnns/pzbsDsT5OzGWmsvbq19eIHZjk2y6DdFVbVpnsn3TiLIAXbFsVlcLVrsfMBuaoG+ylIdmwVqTmvtja21t67gOoFVVlV7rvP6V7JOzaq19ujW2rfnmeU3kswb5KySLUl+eaGZdneCnAlWI6+sqouq6sKqetIw/sCqOquq/m0Y/7hh/Jaq+kJVnVZVn6uqD1XVfvMsf0eqPaSpvzO2zHtU1ZYkJyV5blWdV1UPmmc5b6yqTyV5RVXdv6o+UVWfrap/raq7V9XeSV6a5EnDsp5UVQdU1Zur6tPDvI9b2T0IrIY1qE3/q6o+X1UXVNXW2WpRVT2mqj411I4PV9UPzTHftE/vps4WqKrNVXX2MN9Fc9U3YNesQZ34aFW9Zuw1fP9h/Kx9i+ET3fdV1T8lOWtox18Obbigqo4b5nv40If5t6p6Z1UdOIxfTj/pJVX1/LF2v3xo38VTj6mqPavqVcO2XFBVvzaMf8iwHRcO27XPWHteNqz3nKq6b1V9sKouraqTxtb9gqr6zLDM31nOcwprraqeNhy751fVXw2jHzoc8xdX1c/P89gTqurvhtfcl6vq/4xNe8rwGjyvqv6shtCmqq6tqldX1flJHlBVp9at/ZFXDfNsqap/GsadVVV3HsafXlV/UqP3PttqgTOHqurk4XV9flWdOoz7aFX9cVWdk+TXq+p+VfXPVXXu8PrePMx3v+Fx5yd5ziJ25Y9U1T8O++EVY224vKoOGerm3w/LvKhG79P+V5IfSfKRqvrIItax0H79/WH5n6yqHxrGHzXcv7Cqfq9uPaPz1CQPGpbz3Pm2YbfWWnObsFuSa4e/xyU5M8meSX4oyRVJNifZlOS2wzyHJLkkSWWUXt6U5N7DtHckeco86zk9yROH4cuT/Now/Owkfz4MvyTJ8xdo7+lJ3p9kz+H+bZNsGoYfmuTdw/AJSV479rg/mGpfktsnuTjJAeu9/93c3Ga/rWFt+vck+wzDtx/+TqtFSQ5KUsPw/0jy6jnm21HnZmzD85K8aBjeM8lt1nv/urlthNsa1omPJjltGD4myUXD8Kx9i6EP8rUkBw/TXp7kj8eWd9DQnrOn+iJJTk7y4mH48ux6P2nHPEO7p+rVo5N8eBh+VpJ35db+08FJ9k3y1SR3G8a9NclvjLXnWcPwHyW5IMltkhya5BvD+IcnedOwf/fIqK92zHofI25ui7kl+fHh9XvIcP/g4X/6Pw7H812H1/S+czz+hCTbk9whyX5JLkpydJIfS3JGkr2G+V6f5GnDcEvyi8PwHZJ8Kbf2NW4//D0jydOH4Wckee8wfHqSdw5tu2eSS+bZtkcl+dck+09t2/D3o0lePwzvNcxz6HD/SUnePAxfMPVaTvLKDPVvnv2wLcnthprylSR3GqZdnlHdOy5DPR3G3258+gLP09QyFtqvjxmGX5Hkt4fh9yf5pWH4pNz6/+PYJO9fzDbszrdVP2WLZXlgkre11m5O8o2q+uck/yXJB5L8QVUdk+SWJIdl1ElKkstaa+cNw+dm1DFarL8de9x/X2Jb3zm0Mxm9yN5SVXfN6IW71xyPeXiSx059SpXRC/POSb6wxHUDa2u1a9MFSf6mqt6b5L1zzHN4krcPn07tneSyJW7DZ5K8uar2yqgTdt4C8wNLsxZ9mLclSWvt7Kq6bY2u9TBX3yJJzmytXTMMPzTJ8VMLaq19a/h0/55JPl5Vyai2fGJsfcvpJ40bX86Wsfa8sbV209Cea6rqXhntk4uHed6S0afvfzzcf9/w98IkB7bW/jPJf1bVDWP74uFJPjvMd2BGb37PXkbbYa38XEbvL/4j2fGaSJJ3tNZuSfLlqtqW5B5JzptjGWe21r6ZJFX1txnVpZuS3C/JZ4bl7ZfkqmH+m5O8exj+TpLrk/xFVb0/o9AhSR6QW1//f5VRMDHlvUPbPj911skcHprkL1tr35/atrFpbx/+3j3JTyQ5c2jnnkm2D6/t27fWpl7Hf5VRMDSfs1pr30mSqvp8krtkFBJPuTDJq6vq5RkFKP+ywPJm85DMvV9/kFv337lJHjYMPyDJ44fh/5vkVcvYht2OIKdPT87oE5f7tdZurKrLM+qoJMkNY/PdnNGLaLGmHntzln5sfG9s+HeTfKS19oQanXb80TkeU0mOa619aYnrAibTStWm/5bRJ+yPSfKiqvrJWeb50yR/2Fp7X1Udm9En3rO5KcPXiKtqj4zemE298TtmWNfpVfWHzTUsYC2sZB+mzXJ/1r5FVf3XTO+rzKYyeuP3S3NMX04/aTWXc0um77tbhuVWkpe11v5sGeuASTPb634p81aSt7TWTpll/uunPphurd1Uo69sPiTJE5P8z4zCpfmMvw5rgXnnMlWnKsnnWmsPGJ9Y81yceJHt2qnutNYurqr7ZnSW4O9V1VmttZcucR3z7dcb23BqzWzrX6R5t2F35Bo5k+1fMrqmzJ5VdWhGb2w+ndEZL1cNHaAHZ5RIrpb/zOhU3aW4XZIrh+ET5lnWB5P8Wg2xbVXdZxfbCKytVatNQ9hyp9baRzL6WsPtMvoUeWb9GK8zTx8bP3O+yzP6hChJHpvhDMGquktGXz84LcmfJ7nvUtsKzGst+jBT1915YJLvDJ/WLrZvcWbGri1RVQcl+WSSn62qHx3GHVBVd1ugDbvST5qrPf9vDRc4raqDM/pax5ap9iR5apJ/XsIyP5jkGXXrdX4Oq6o7rkBbYS38U5JfqKo7JDteExnG7VFVRyU5MqPXyVweVlUH1+h6W49P8vEkZyV54tRrYZi+Ux0aXje3a639Q5LnJrnXMOlfc+vZfE/OqNYt1ZlJfqWGX4Qa27ZxX0pyaFU9YJhnr6r68Ta6OPG3h7o31YZlqaofSfL91tpfZ/RVrak+0VLq26L26wyfzOhrXcnYGZJLXO9uS5Az2d6T0VcMzs+omP1Wa+3rSf4mydFVdWGSpyX54iq24YwkT6h5LuI3i1ckeVlVfTbT09KPJLnnsKwnZXTmzl5JLqiqzw33gcm3mrVpzyR/PSzjs0n+ZOi0zKxFL0nyzqo6N8l/jD1+5nynJfl/arhwYW79pOvYJOcPdepJSV6zC20F5rYWfZjrh9fwG5M8cxi32L7F7yU5qEYX9jw/yYNba1dn9AHU26rqgoy+VnWPBdqwK/2k2fx5RtcRumBozy+31q5P8isZ1boLMzrT5o2LXWBr7UMZfV3hE8Pj3xVvjuhEa+1zSX4/yT8Pr4k/HCZdkVEo/IEkJw2vk7l8OqOvSl2Q0TU7z2mtfT7Jbyf50PA6PzOj63fNdJsk7x/m+ViS3xzG/1pGIcwFGYWrv74L2/aPGX018pyqOi/J82eZ5wcZnQn08mH7z8utv5D3K0leNzx2V8/8GfeTST49LO//ZFQfk9E1tv6xFnGx4yXs13G/keQ3h/l/NKOvsyWj5+vmGl0c+blzPXh3V7ee5QQAAJOvqj6a0QWEz1nvtgCTp6pOSHJ0a+1/rndbmN1wRtJ1rbVWVcdndOHjx613u3qx23+3DAAAAFhT90vy2uGrsN/O6FfAWCRn5OwGqup1SX52xujXtNb+conLeVGSX5gx+p2ttd9fTvuA3dNK1SZg4+qpTugnwdqrqkckefmM0Ze11p6wHu0ZN/xYw1/NGH1Da+2/rsK6VnU/VNWnkuwzY/RTW2sXrsTyWTpBDgAAAEAnXOwYAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBObljLzIYcc0rZs2bJKTQHWwrnnnvsfrbVD17sdy6EWQf/UImBS9F6P1CLYGJZSi5YU5GzZsiXnnHPOrrUKmAhV9ZX1bsNyqUXQP7UImBS91yO1CDaGpdQiX60CAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOjEpvVuwFKcdtpp2bZt23o3Y8Vs3749SbJ58+Z1bgm9O/LII3PiiSeudzN2G2oRzE4tWltqEcxNPdrZRqsZq2mp9cjxxlrrKsjZtm1bLj3//Bx2/fXr3ZQV8b19902SXH/FFevcEnp25XAcsXbUItiZWrT21CKYnXo0u41WM1bTUuqR44310FWQkySHXX99fv2yy9a7GSviNUcckSQbZntYH1PHEWtLLYLp1KL1oRbBztSjuW2kmrGallKPHG+sB9fIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADqxaTUWetpppyVJTjzxxNVYPDDm6r33zqbt29e7GRNJLYK1oxbNTS2CtaUezW779u25ae+917sZG47jjfWwKkHOtm3bVmOxwCxu2GOP3HjddevdjImkFsHaUYvmphbB2lKPZnfdddfllj18IWOlOd5YD17JAAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRi03o3AAAAAGAjecxjHrNj+IwzzljRZTsjBwAAAKATghwAAACAFTJ+Ns5s95drVb5atX379lx33XU55ZRTVnS527Zty157772iywQ2LrUImARqEQCwkhY8I6eqfrWqzqmqc66++uq1aBPATtQiYBKoRcAkUItg97bgGTmttTcleVOSHH300W0xC928eXOS5GUve9ly2raTU045Jdd/6lMrukygD2oRMAnUImAS7EotAjYO18gBAAAA6IQgBwAAAGCFzPy5cT8/DgAAALCbWpVfrQIAAADYXa30WTjjnJEDAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQiU2rsdAjjzxyNRYLzGKfW27Jpv32W+9mTCS1CNaOWjQ3tQjWlno0u/322y83fec7692MDcfxxnpYlSDnxBNPXI3FArM49Ac/yL6bN693MyaSWgRrRy2am1oEa0s9mt3mzZtz/RVXrHczNhzHG+vBV6sAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADqxab0bsFRX7rtvXnPEEevdjBXxtX33TZINsz2sjyv33TdHrXcjdkNqEUynFq0PtQh2ph7NbSPVjNW0lHrkeGM9dBXkHHnkkevdhBV1wPbtSZJ9N29e55bQs6Oy8V4bk26j7W+1iJWgFq29jba/1SJWino0O/tk8ZZSjxxvrIeugpwTTzxxvZsAoBYBE0EtApZCzYCNwzVyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6IcgBAAAA6IQgBwAAAKATghwAAACATghyAAAAADohyAEAAADohCAHAAAAoBOCHAAAAIBOCHIAAAAAOiHIAQAAAOiEIAcAAACgE4IcAAAAgE4IcgAAAAA6Ua21xc9cdXWSr6xec5blkCT/sd6NWAO2c2NZj+28S2vt0DVe54pSiybG7rKttnN1qEWrz7G7sdjO1dN1PVqgFm3042ajb19iGzeKxWzjomvRkoKcSVZV57TWjl7vdqw227mx7C7buTvZnZ7T3WVbbSe92l2eU9u5sewu27lWNvr+3Ojbl9jGjWKlt9FXqwAAAAA6IcgBAAAA6MRGCnLetN4NWCO2c2PZXbZzd7I7Pae7y7baTnq1uzyntnNj2V22c61s9P250bcvsY0bxYpu44a5Rg4AAADARreRzsgBAAAA2NAEOQAAAACd6CrIqaqDq+rMqvry8PegOeZ7+jDPl6vq6WPjP1pVX6qq84bbHdeu9QurqkcO7bukql44y/R9qurtw/RPVdWWsWmnDOO/VFWPWNOGL9GubmdVbamq68aevzeueeOXYBHbeUxV/VtV3VRVT5wxbdZjmMmgFqlFahGTQC1Si9QiZqqqX6iqz1XVLVU1508dL/R8TKol1L2bx14b71vrdu6K5dS8XixiG0+oqqvHnrv/sR7t3FVV9eaquqqqLppjelXVnwzbf0FV3XeXV9Za6+aW5BVJXjgMvzDJy2eZ5+Ak24a/Bw3DBw3TPprk6PXejjm2bc8klyY5MsneSc5Pcs8Z8zw7yRuH4eOTvH0Yvucw/z5JjhiWs+d6b9MqbOeWJBet9zas4HZuSfJTSd6a5Ilj4+c8ht0m46YWqUVqkdsk3NQitUgtcptlX/9YkrvP9/pezPMxqbfF1L1h2rXr3dYlbtcu14JebovcxhOSvHa927qMbTwmyX3nqs1JHp3kA0kqyU8n+dSurqurM3KSPC7JW4bhtyR5/CzzPCLJma21a1pr30pyZpJHrk3zluX+SS5prW1rrf0gydaMtnfc+Pa/K8lDqqqG8Vtbaze01i5LcsmwvEm0nO3syYLb2Vq7vLV2QZJbZjy212N4d6IWqUW9UIs2NrVILeqFWrRGWmtfaK19aYHZFnPcTarF1L0e7Q61oOfjblFaa2cnuWaeWR6X5K1t5JNJbl9Vm3dlXb0FOT/UWts+DH89yQ/NMs9hSb46dv9rw7gpfzmcpvX/TdiBv1C7p83TWrspyXeS3GGRj50Uy9nOJDmiqj5bVf9cVQ9a7cYuw3Kek56ez92VWqQWqUVMArVILVKL2BU979PF1L0k2beqzqmqT1bV49emacuy3FrQg8Ued8cNXzt6V1XdaW2atmZW7LW3aUWas4Kq6sNJfniWSS8av9Naa1W11N9Of3Jr7cqquk2Sdyd5akanb9KH7Unu3Fr7ZlXdL8l7q+rHW2vfXe+GsfGoRcxDLWLNqEXMQy3aTc1XF1prf7fW7VlpK1T37jLUtyOT/FNVXdhau3Sl28qKOyPJ21prN1TV/5vRGUg/t85tmkgTF+S01h4617Sq+kZVbW6tbR9OQbpqltmuTHLs2P3DM/qOaFprVw5//7Oq/m9Gp3dNSoflyiTjiePhw7jZ5vlaVW1Kcrsk31zkYyfFLm9nG32x8IYkaa2dW1WXJrlbknNWvdVLt5znZM5jmLWjFu2gFqlF44/96Iq0ikVTi3ZQi9Si8cd+dEVa1an56sIiTfRrZAXq3nh921ZVH01yn4yuzzKpllPzerHgNrbWxrfnzzO6JtJGsmKvvd6+WvW+JFNXqn96ktkS5w8meXhVHVSjq5g/PMkHq2pTVR2SJFW1V5KfTzLr1aTXyWeS3LWqjqiqvTO6gNXMK6yPb/8Tk/zT8E/8fUmOr9GVzI9Ictckn16jdi/VLm9nVR1aVXsmyZCu3zWjC95NosVs51xmPYZXqZ3sGrVILVKLmARqkVqkFrErlvN8rLcF695wrOwzDB+S5GeTfH7NWrhrllPzerHgNtb068U8NskX1rB9a+F9SZ5WIz+d5DtjXxVcmjYBV3de7C2j7wCeleTLST6c5OBh/NFJ/nxsvmdkdGG7S5L8yjDugCTnJrkgyeeSvCYT9gsGGV3F+uKM0uIXDeNemuSxw/C+Sd45bNenkxw59tgXDY/7UpJHrfe2rMZ2JjlueO7OS/JvSR6z3tuyzO38Lxl9L/J7GaXpn5vvGHabnJtapBapRW6TcFOL1CK1yG2W/fyEYT/ekOQbST44jP+RJP8w3/PRw20xdS/JzyS5MKNfRbowyTPXu92L3LZdrnm93BaxjS8b6tr5ST6S5B7r3eYlbt/bMvra643D6/CZSU5KctIwvZK8btj+C7OMX46sYYEAAAAATLjevloFAAAAsNsS5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkMOqqqrPVtW9h+FNVXVtVT1lbPq5VXXfYfjxVXVBVX2hqi6sqsePzXd6VV1WVecNt39d620B1k9VXV5Vh6zwMrdU1S+P3T+hql67yMf+elX98dj9P6uqD4/d/7Wq+pNh+PCq+ruq+nJVXVpVr6mqvYdpx1bVd8Zq23lV9dAV20joQFVdu95tWI6h/3LPeabfq6rOG7v/S1V1XVXtNdz/yaq6YBjeu6r+uKouGWrG31XV4WOPvXlGvXjhKm4abEhrVXOq6qUL/U8f+gE/s0Lru3dVPXoXH+s9W2cEOR2oqk3r3YZl+HiSqeJ0ryQXT92vqgOSHJXk/Kq6V5JXJXlca+3Hkjw2yauq6qfGlvWC1tq9h9uKFDzYyDqvHWthS5JfXmimOYzXtmRU325XVXsO938myb9WVSX52yTvba3dNcndkhyY5PfHHvsvY7Xt3q21DwfoyeOTzBnkJLkwyZ2r6jbD/Z9J8oUk9xm7P/Vm5w+S3CbJ3Yea8d4kfzvUkiS5bka9OHXlNgNYSa21Fy/if/qxmd6fmNcCfbt7J9mlICfes3VHkLNGhk9+v1hVfzOkl++qqv2r6n5V9c9DyvnBqto8zP/R4ROZc5L8elX9QlVdVFXnV9XZwzz7VtVfDknoZ6vqwcP4E6rqb6vqH4dPc16xQNveUFXnVNXnqup3xsY/emjzuVX1J1X1/mH8AVX15qr69LDex82z+H/NrUXhZ5K8MaMikyT3T3Jua+3mJM9P8gettcuSZPj7siQvWMJuhg1nwmvHw6vqE1X1b1X1zqo6cBh/eVW9bPgk5pyquu/Qxkur6qRhnmOr6uyq+vuq+lJVvbGqFvU/qaqeMtSf82p0Jsyew/hrq+r3h239ZFX90DD+qOH+hVX1e3XrJ3GnJnnQsJznDuN+ZJHbf16Su1XVflV1uyTXDeN+cpj+Mxl1in4uyfWttb9MkqHePTfJM6pq/8VsL+wuauSVQ826sKqeNIw/sKrOGmrNhVP9jqE+fqGqThv6MB+qqv3mWf6JVfWZoUa8e+o1WKNPkN8w1IltQ31687Ds08cev1ONqdEn6Y9N8sqhlhw1c72ttVuSnJPkvw6j7pfkdZneP/r40J5fSfLcoVZkqB03ZFRLgBW0BjXn9Kp64jB8eVX9ztgy71FVW5KclOS5Q/140DzLeWNVfSrJK6rq/kP/67NV9a9Vdfcanen70iRPGpb1pPKebWNrrbmtwS2jT35bkp8d7r85owP+X5McOox7UpI3D8MfTfL6scdfmOSwYfj2w9/njc1/jyRXJNk3yQlJtiW53XD/K0nuNE/bDh7+7jms96eGx301yRHDtLclef8w/AdJnjLVlowS2wPmWPZdkmwbW8Y9knwko0+bXpTkd4dp/5bkXjMee68k/zYMn57ksozeKJ2X5G/W+zl1c1uL26TWjiSHJDl76rWf5OQkLx6GL0/yrGH4j5JcMLzmD03yjWH8sUmuT3LkUHvOTPLEefbD5cM6fyzJGUn2Gsa/PsnThuGW5DHD8CuS/PYw/P4kvzQMn5Tk2rE2vH9sHYve/mH+jyQ5JskjMgqFnpnk2UkOS3LFMM//SvJHszz2sxnV2mOTfGestp2X5Kj1Pu7c3NbyNvaaPG6oBXsm+aGhNm1OsinJbYd5DklySZLKqD7elOTew7R3ZOifzLGeO4wN/16SXxuGT0+ydVjm45J8N6NQdo8k544tf64ac/p89WuY5/8keXGSA5J8LKNPt98xTPvycP+nknx2lsf+UZL/NQzfPKNePGm9nz83t95ua1hzdtSGjPoxUzXn2Un+fBh+SZLnL9De0zPqy+w53L9tkk3D8EOTvHsYPiHJa8ce5z3bBr457X5tfbW19vFh+K+T/O8kP5HkzBqdMbtnku1j8799bPjjSU6vqndkdJp+kjwwyZ8mSWvti1X1lYxO20+Ss1pr30mSqvp8Ri/Or87Rrl+sql/NqGhtzuj04D0yejFfNszztiS/Ogw/PMljq+r5w/19k9w5o9OEp2mtfaVG3/f+4YwKwpeSfCajT6V+Zqr9i/SC1tq7ljA/bBSTWDt+OqNa8fGhDXsn+cTY9PcNfy9McmBr7T+T/GdV3VBVtx+mfbq1tm1Y19uGdi30Gn9IRp9mf2ZY735Jrhqm/SCjjk4yevP1sGH4ARl99SFJ/m9GpwTPZSm1c+rTq/0y2vYvZ/TcXJ1bvyaxGP/SWvv5JcwPG9UDk7ytjT71/UZV/XOS/5LkA0n+oKqOSXJLRmHpDw2Puay1dt4wfG5Gb7Tm8hNV9XsZvaE5MMkHx6ad0VprVXVhRoHzhUlSVZ8blnle5q4xi/GvGYXo/5LkM621S6vqR6vq0Ixq5KU1/asJc7mutXbvJawXmNtq15yZpvph5yb570ts6zuHdiajD5zeUlV3zShg3muOx3jPtoEJctZWm3H/P5N8rrX2gDnm/96OB7Z2UlX91yT/Lcm5VXW/BdZ1w9jwzZnjua6qIzI6Re6/tNa+NZxCvO8Cy64kx7XWvrTAfFP+NckvJNk+dJI+meRnMzpNb+qN3+czenN2/tjj7pfkc4tcB2xkE1c7MqoDZ7bWfmmB5dwyY5m3jC1z5nbNvD/Xet/SWjtllmk3tuHjoMzf9vksdvuTUUh2UkY183UZBTj3zPQg5/NJnjj+oKq6bUYdqUsyqoPA/J6c0Rl992ut3VhVl+fWvsrM1+ycX3PI6JPix7fWzq+qEzI6I27KYmrWcmrMJzN6g/izubXv87Ukx4/dvzTDtXSG8HvK/XJrgASsvpWqOTNNPXZX+ijfGxv+3SQfaa09Yfh61kfneIz3bBuYa+SsrTtX1dQbr1/O6J/6oVPjqmqvqvrx2R5YVUe11j7VWntxRm8S7pTRpzpPHqbfLaM3Bot9oU65bUaF4Ts1up7Eo4bxX0py5FAcktFXN6Z8MMmv1fBxeFXdJ/P71yS/kVsLwCeSPC3J16c++c7oE/JTptY3/P3fSV69xO2BjWgSa8cnk/xsVf3osJwDhmUtxf2r6ogaXRvnSRl93WAhZyV5YlXdcVjvwVV1l0W09bhh+Pix8f+Z0SnDu+oTGZ2ZdGhr7arhDd7VGX01Y+oMqrOS7F9VTxvau2dGde301tr3l7Fu2Ij+JaPrO+w5nKlyTJJPZ/Tp81XDG6oHZ3Sm3K64TZLtNfq1qCevSItHFqwlQzDz1YyugTPeH/qNDPWitfa9JG9J8od167W/npZk/yT/tILtBUZWu+Ysxq70RW6X5Mph+IR5luU92wYmyFlbX0rynKr6QpKDMjpF7YlJXl5V52d02u5cV/Z+5XBhrIsyepGdn9G1IfYYTgN+e5ITWms3zPH4WbXWzs/oWg1fzOgrB1Odiesy+v7mP1bVuRkVhqkX8O9mdArfBcMpx7+7wGo+ntF1MD4xLHt7Rl8F2fHVg+EUxZOTnFFVX8zoGhi/NXbq4tQ+GP/Jzb2Xsq3QsUmsHVdn1Hl4W41+NvcTGZ2KuxSfSfLajE7xvSzJexax3s8n+e0kHxrWe2ZGXwmdz28k+c1h/h/NrbXsgiQ31+jCpc+d68HztOVbGQU3459CfSLJHTN8UjWEO09I8gtV9eWMvp9+fUadnikPmlHbpp3BA7uR92T0ujw/o+Dit1prX0/yN0mOHmrW0zLqs+yK/y/JpzLql+zqMmazNckLhouJ7nSx4zEfT7JPa23q65qfyKh/NP5VzFMyqhEXDzXjF5I8YexMoP1m1Au/WgW7brVrzmKckeQJNc/FjmfxiiQvq6rPZvqZPR9Jcs9hWU+K92wbWt36f4HVNKSV72+t/cR6t2WxqurA1tq1Q4r7uiRfbq390Xq3C3YnPdaOxaiqYzO6uN+qXxumRr8Ec91wmvDxGV34+HGrvV4AAFgNrpHDfE6sqqdndBHTzyb5s3VuD8CuuF+S1w6h9LeTPGN9mwMAALvOGTm7kar6VJJ9Zox+6tQvMyxz2Y9I8vIZoy9rrT1hucsG1tdq1o71XNdiTWKbgLlV1esyukDnuNe01v5yI68bWB8r9bqvqhdl9HXKce9srf3+cto3Yx3es20QghwAAACATrjYMQAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnNi1l5kMOOaRt2bJllZoCrIVzzz33P1prh653O5ZDLYL+qUXApNgI9QjYvSwpyNmyZUvOOeec1WoLsAaq6ivr3YblUougf2oRMCk2Qj0Cdi++WgUAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAANAJQQ4AAABAJwQ5AAAAAJ0Q5AAAAAB0QpADAAAA0AlBDgAAAEAnBDkAAAAAnRDkAAAAAHRCkAMAAADQCUEOAAAAQCcEOQAAAACdEOQAAAAAdEKQAwAAAPD/t3fn8brd873AP18iTlCSmHpQImooVUraGltTad0qiktrCpo2HYxXS27dllZrrqH0qriGDlSj2oYqIgQ3EUQkOTEk4VDknhJTSiUx5Hf/WL+d8+Q4e589nD389n6/X6/12utZaz3r+a717PV71vNZwzMIQQ4AAADAIAQ5AAAAAIMQ5AAAAAAM4oDVmOmxxx6bnTt3Luk5u3btSpJs3759NUpaF4cffniOOuqo9S4D2IfltFmwWJvx820+PvdYCxuhzd5K2/WotEfAZrYqQc7OnTvzmTPPzPUvvnjRz/mvbduSJBd//vOrUdKaO78vD7DxLafNgsXabJ9v8/G5x1rZCG32VtmuR6U9Aja7VQlykuT6F1+cJ372s4ue/qU3vnGSLOk5G9nc8gBjWGqbBYu12T7f5uNzj7W03m32VtmuR6U9AjY798gBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgEIIcAAAAgEEIcgAAAAAGIcgBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgEIIcAAAAgEEIcgAAAAAGIcgBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgEIIcAAAAgEEIcgAAAAAGIcgBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABnHAasx0165d+d6BB67GrIdxwYEH5oBdu9a7DNjSjj322CTJUUcdtc6VwObnc29+2iJYW9ojYLNblSDnoosuyqVX2Non+1xyhSvkuxddtN5lwJa2c+fO9S4Btgyfe/PTFsHa0h4Bm93WTlsAAAAABiLIAQAAABiEIAcAAABgEIIcAAAAgEEIcgAAAAAGIcgBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgEIIcAAAAgEEIcgAAAAAGIcgBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgEIIcAAAAgEEIcgAAAAAGIcgBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgEIIcAAAAgEEIcgAAAAAGIcgBAAAAGIQgBwAAAGAQB6x3AcDK3O9+97us/61vfes6VgIAAMBqc0YOAAAAwCAEOTCw2bNx9vYYAACAzcWlVcCmtWvXrlx00UU55phjFpxu586dudKBB65RVcBWs9i2iMXRZgOw1e3zjJyq+o2qOq2qTrvgggvWoiaAH6AtAjYCbREAsN72eUZOa+1VSV6VJEcccURb9YoA9mI5bdH27duTJM95znMWnO6YY47JxR/60AorBLaC1WyLWBxtNgBbnXvkAAAAAAxCkAMD2/Pnxv38OAAAwOYmyAEAAAAYhF+tgsE5CwcAAGDrcEYOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADAIQQ4AAADAIAQ5AAAAAIMQ5AAAAAAMQpADAAAAMAhBDgAAAMAgBDkAAAAAgxDkAAAAAAxCkAMAAAAwCEEOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADAIQQ4AAADAIAQ5AAAAAIMQ5AAAAAAMQpADAAAAMAhBDgAAAMAgBDkAAAAAgxDkAAAAAAxCkAMAAAAwCEEOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADAIQQ4AAADAIAQ5AAAAAIMQ5AAAAAAMQpADAAAAMAhBDgAAAMAgBDkAAAAAgxDkAAAAAAxCkAMAAAAwCEEOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADAIQQ4AAADAIA5YjZkedNBB+d6FF67GrIdx5UsvzQEHHbTeZcCWdvjhh693CbBl+Nybn7YI1pb2CNjsViXI2b59ey7+/OdXY9bDuPZ3vpNt27evdxmwpR111FHrXQJsGT735qctgrWlPQI2O5dWAQAAAAxCkAMAAAAwCEEOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADAIQQ4AAADAIAQ5AAAAAIMQ5AAAAAAMQpADAAAAMAhBDgAAAMAgBDkAAAAAgxDkAAAAAAxCkAMAAAAwCEEOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADAIQQ4AAADAIAQ5AAAAAIMQ5AAAAAAMQpADAAAAMAhBDgAAAMAgBDkAAAAAgxDkAAAAAAxCkAMAAAAwCEEOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADAIQQ4AAADAIAQ5AAAAAIMQ5AAAAAAMQpADAAAAMAhBDgAAAMAgBDkAAAAAgxDkAAAAAAxCkAMAAAAwCEEOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADAIQQ4AAADAIAQ5AAAAAIM4YLVmfP62bXnpjW+86Om/uG1bkizpORvZ+du25SbrXQSwaEtts2CxNtvn23x87rGW1rvN3irb9ai0R8BmtypBzuGHH77k51x1164kybbt2/d3OeviJlneegDWnm2V1bTZPt/m43OPtbIR/s+2ynY9Ku0RsNmtSpBz1FFHrcZsAVaFNgtgHNpsALY698gBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgEIIcAAAAgEEIcgAAAAAGIcgBAAAAGIQgBwAAAGAQghwAAACAQQhyAAAAAAYhyAEAAAAYhCAHAAAAYBCCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgENVaW/zEVRck+ffVK2e/uFaSr6x3EWvAcm4ua7mcN2qtXXuNXmtVrFFbtBH+9zZCDcnGqEMNu22EOvZHDVu5LdoI7+Fa2CrLmVjW0Q3fHgFby5KCnBFU1WmttSPWu47VZjk3l62ynCPZCO/JRqhho9Shho1Vx0aoYWRbZf1tleVMLCsAa8ulVQAAAACDEOQAAAAADGIzBjmvWu8C1ojl3Fy2ynKOZCO8JxuhhmRj1KGG3TZCHRuhhpFtlfW3VZYzsawArKFNd48cAAAAgM1qM56RAwAAALApDR/kVNWhVXVCVZ3X/x4yz3Tfr6ozenf8Wte5XFX1C1V1TlV9uqqevpfxV66qN/XxH6qqw9ahzBVbxHIeWVUXzLyHv74eda5EVb2mqr5cVWfPM76q6mV9HZxVVbdb6xq3miW0H4/u05xXVY+eGX5gVb2qqs6tqk9V1YPWuoaZ8cfP97+1mjVU1VWq6l/78n+8qp67jNdfdjtXVcf04edU1X2W+torraGqfr6qPlpVO/rfe6x1DTPjb1hV36qqpy63hpXWUVU/UVUf7P8LO6pq20pq2Syq6iF9nVxaVfP+2s++1v1Gt4S2ZMh9ssR+2cz44ffLAIbWWhu6S/L8JE/v/U9P8rx5pvvWete6jGW7YpLPJDk8yYFJzkxyyz2m+e0kr+z9D0vypvWue5WW88gkL1/vWle4nD+b5HZJzp5n/H2T/FuSSnKHJB9a75o3e7eY9iPJoUl29r+H9P5D+rhnJXl2779CkmutdQ19/K8kecN8/1urWUOSqyS5e5/mwCQfSPKLS3jtZbdzSW7Zp79ykhv3+VxxGcu/khp+Msn1ev+PJzl/me/Bitv7JG9OclySp65gm1jJujggyVlJbtMfX3M578dm7JL8WJKbJzkpyRHLXfcbvVtMW9LHDbdPttj3aF/b6QjdIpfzyAy+X6bT6XQjd8OfkZPk/kle3/tfn+QB61fKfvfTST7dWtvZWvtOkr/PtLyzZpf/zUnuWVW1hjXuD4tZzuG11t6f5GsLTHL/JH/dJqcmObiqtq9NdVvWYtqP+yQ5obX2tdba15OckOQX+rjHJnlOkrTWLm2tfWWta6iqqyV5SpJnL+O1V1xDa+3brbX3Jknffk9PcoMlvPZK2rn7J/n71tolrbXPJvl0n99SLbuG1trHWmv/rw//eJKDqurKa1lDklTVA5J8ttewEiup495JzmqtnZkkrbWvtta+v8J6NoXW2idba+fsY7LN8Fm4mffJEvtlAGwQmyHIuW5rbVfv/48k151num1VdVpVndp3eEdw/SRfmHn8xT5sr9O01r6X5MJMR0FHspjlTJIH1XTJ0Zur6kfWprQ1tdj1wP6zmPZjr+9LVR3cH/9JVZ1eVcdV1Xztz6rUMPf6SV6U5NvLeO39VUOSpK+T+yU5cQmvvZJ2bn9tM/urrX1QktNba5esZQ09zHtapjPEVmol6+JmSVpVvbNvE7+/H+rZSjbDZ8Bm3idL7JftabPvlwFsWAesdwGLUVXvTvLDexn1B7MPWmutqub7Ga4btdbOr6rDk7ynqna01j6zv2tl1bw1yRtba5dU1W9mOtq17HtRsHXsp/Zjbw7IdObJKa21p1TVU5K8MMkj16qGqrptkpu01p68r/swrOJ6mJv/AUnemORlrbWdS33+6KrqVkmel+mslLX2zCQvbq19a50P/B+Q5C5JfipTsHhiVX20tbaUYG9YC21jrbV/Wet6Vot9Mjr7ZQDraIggp7V2r/nGVdWXqmp7a21Xvwzly/PM4/z+d2dVnZTpvgYbfafh/CSzRzhu0IftbZov9i9S10jy1bUpb7/Z53K21maX6dWZrsPfbBbzfrNE+6H9OD/J3WYe3yDTfS6+munL6lv68OOSPG6Na7hjkiOq6nOZ2vPrVNVJrbW77fH81axhzquSnNdae8l8rzOPlbRz+2ubWVFbW1U3SPJPSR61gi+jK6nhZ5I8uKqen+TgJJdW1cWttZevcR1fTPL+uUsMq+rtme4LtiWCnIW2sUUa4jNgC++TJfbLLrNF9ssANqzNcGnV8UnmfsHl0Ul+4KhXVR0yd8+CqrpWkjsn+cSaVbh8H0ly06q6cVUdmOmmeXv+usPs8j84yXtaa0s+mr7O9rmce9wr5peTfHIN61srxyd5VE3ukOTCmVPUWR37bD+SvDPJvXs7ckimMy7e2bezt2Z3uHHPLK9dWUkN/7u1dr3W2mGZzoQ4d28hzmrWkCRV9exMX1aetIzXXkk7d3ySh9X0KzE3TnLTJB9eyxr65WT/mukGrycv47VXXENr7a6ttcP6/8FLkvzZMkOcFdWR6f/h1jX9ktkBSX4uY3zWbhSLWfcb3WbeJ0vsl11mi+yXAWxcbQPccXklXabrjk9Mcl6Sdyc5tA8/Ismre/+dkuzIdNf9HUket951L2H57pvk3ExHqv6gD/vjJL/c+7dlOhPg05m+wBy+3jWv0nI+J9NNPM9M8t4kt1jvmpexjG9MsivJdzMduX5ckqOTHN3HV5JX9HWwI/P8soluv74n+2w/+uPH9m3s00keMzP8Rknen+mXek5McsO1rmFm/GFZ/q9WLbuGTEdqW6ad+DN69+tLfP1lt3OZLuf4TJJzsoRfy9pfNSR5RpL/mln2M5JcZy1r2GMez8wKfrVqP7wfj8jUVp+d5Pn7YzvdDF2SB2Zq9y9J8qVMQWySXC/J2xda9yN1i2lLMvA+2Xzv0VK30xG6RSzn8PtlOp1ON3JXrY12kAAAAABga9oMl1YBAAAAbAmCHAAAAIBBCHIAAAAABiHIAQAAABiEIAcAAABgEIIcAAAAgEEIcrqq+tZ617CWqupuVXWnfUxzdFU9aq1qWo6qenFVPWnm8Tur6tUzj19UVU/p/beqqvdU1TlVdV5V/a+qqj7uyKq6oKrOmOluueYLxKZTVa+rqgevdx0bSVUdVlVnLzD+yKp6+Tzj3l5VBy/w3CdV1VX28fqfq6prLbrgRejL9Gszj+ddhr0894lV9ZKZx39VVe+eefz4qnpZ779BVf1Lb8M+U1UvraoD+7i7VdWFe7Rj99pvCwmbwL7akPVWk69U1SH98faqalV1l5lpLqiqa/b+36iqT/Xuw3tMd1Lf55lrD9689ksEwGoQ5Aykqg7Yj7O7W5IFg5zW2itba3+9H19zNZycvhxVdYUk10pyq5nxd0pySlUdlOT4JM9trd08yW36uN+emfZNrbXbznSfWJMlgP2kqq64zq+/P9uovWqt3be19o0FJnlSkgWDnFVyWJJf29dE87isHetuk+QaM+/nXDtWSd6S5J9bazdNcrMkV0vypzPP/cAe7di7AwNZ7XZkEW3IumqttSSnJrljH3SnJB/L7n2dmyf5amvtq1X1S0l+M8ldWmu3SHJ0kjdU1Q/PzPLhM+2BgwoAm4QgZw/9SMgLqursqtpRVQ/tw69WVSdW1el9+P378MOq6pNVdWxVfbyq3tVDg/nmf1I/gnpGf42f7sOvWlWv6UdTPjYz/yOr6viqek+SE3sdr+01nFVVD+rT3buqPtjrO66qrtaHf66qnjVT9y2q6rBMH/ZP7nXcdZ5an1lVT52p+3m9vnPnnlNVV6yqF/ZlOauqHt+H37Mvx46+XFeeqec5/XVPq6rb1XQWzWeq6uiZ1/69qvpIn+ezFnjLTsnunZ1bJTk7yTer6pD+mj+W5PRMX7BObq29K0laa99O8rtJnr7AvGHJqupR/f/2zKr6mz74Xv3//dy+4z3fc4+s6WyLk2o64+KPZsY9om9/Z9R0xsYV+/Bv1XTm2ZlJ7lhVz62qT/QaXtinOayms9HO6u3YDfvw11XVy6rqlKraWfs4c6iqnta36TOr6rl92ElV9ZKqOi3JE6vq9lX1vqr6aN+2t/fpbt+fd2aS31nEqrxeVb2jr4fnz9Twuaq6Vm8z/7XP8+yqemhVPSHJ9ZK8t6reu4jX2Nd6/dM+/1Or6rp9+E364x1V9ezafTbnc5Pctc/nyQstw16ckeRmVXVQVV0jyUV92K37+DtlCnvukeTi1tprk6S19v0kT07y2NrHWUiwlnqb86mq+rua9pHeXFVXWaB92LMdeUjfrs+sqvf3abbV7v2fj1XV3fvwI6vqLYvc1mbbkHn336rqR6vq3f31T+/bfdXe9w/v1pfpX3o7+tyqenhvV3ZU1U36dNeuqn+sad/mI1V15wXKPCW7w907JXlxLh/snNz7n5bk91prX0mS1trpSV6fxbWxAIystaZrLUm+1f8+KMkJSa6Y5LpJPp9ke5IDkly9T3OtJJ9OUpmOwn4vyW37uH9I8ogFXuekJMf2/p9Ncnbv/7O55yU5OMm5Sa6a5MgkX0xyaB/3vCQvmZnfIb2e9ye5ah/2tCR/2Ps/l+Txvf+3k7y69z8zyVP3sU4um6bX/aLef98k7+79v5XkzUkO6I8PTbItyReS3KwP++skT5qp57d6/4uTnJXkh5JcO8mX+vB7J3lVX79XSPK2JD+7QJ2fTXLDTEeljk7yJ73GO2c6Op0kf57kiXt57teTXL2v5wsyfXma6w5a7/9L3VhdpjDx3CTX6o8PTfK6JO/o/8s37dvztnmef2SSXUmumeSgTMHkEZkCybcmuVKf7i+TPKr3tyT/vfdfM8k5Sao/Prj/fWuSR/f+x2Y6oyO9tuN6bbdM8ukFlu0XM325uMrcsvW/JyX5y95/pT7NtfvjhyZ5Te8/a247TvKC9LZvgfWwM8k1envy70l+pI/7XKY270HpbWkffo3Z8ft4n+bmsa/1er/e//wkz+j9b0vyq73/6Oz+7LhbkrctZhnmqem9mT4T7pMpFHpcpjb7+kk+36d5QpIX7+W5H0vyE72GC3P5duwm671d6LZel2nfqCW5c3/8miS/t0D7cFk70h/vSHL93n9w//s/Zqa/Rab9s23L2Nbmtv/DMs/+W5IPJXlg79+W6Sy/+fYP75bkG73/yknOT/Ks/twnpu+zJXlDpjNnkmmf5ZML1PhzSd7T+z+Q6cy70/rjY5M8rvd/Lb3tm3nu/ZO8ZWa9njPTHrxgvf83dDqdTrd/ulU/DX5Ad0nyxjYd6fxSVb0vyU8l+bckf1ZVP5vk0kw719ftz/lsa+2M3v/RTDsHC3ljkrTW3l9VV6/pWu17J/nl6mfAZNpxuGHvP6G19rXef68kD5ubUWvt6zUd4b9lkpNruuXLgUk+OPN6b5mp7Vf2tQIWMDufw2bqeWVr7Xu9nq9V1W0yrZNz+zRzR4de0h8f3//uSHK11to3M51Fc8nMurh3pi8nybQDc9NMYdXezB25ulOmwOb6vf/C7D5qtRhvaq397hKmhz3dI8lxbffR0a/1bfIfWmuXJjmvqnZm+hJyxjzzOKG19tUkqaq3ZGqTvpfk9kk+0ud3UJIv9+m/n+Qfe/+FSS5O8n+q6m2ZQodkOpI7t+3/TaZgYs4/99o+MXfWyTzuleS1bTqbLTNtUpK8qf+9eZIfT3JCr/OKSXb17frg1trcNvw3mYKhhZzYWrswSarqE0lulCkgnrMjyYuq6nmZApQP7GN+e3PPzL9ev5Pd6++jSX6+998xyQN6/xuSvHAFyzBrrh07KFP7fV6S/5kpYD5lCcv0gdbavGd9wRr6Qmtt7jP4bzP9P/9A+zAz/Ztm+k9O8rqq+ofs3ve4S5K/SJLW2qeq6t8zXV6YLG1bm/UD+29V9UOZQqR/6q91cZ/vfPuH/5nkI621XX26zyR5V5/njiR37/33SnLLvuxJcvWqulprbW/3aPxIkp+sqqtmCpq/1c/2+dFM7cSLFrFscx7eWjttCdMDMABBzuI9PNNZI7dvrX23qj6XKWxJkktmpvt+ph3xhbS9PK4kD2qtnTM7oqp+Jsl/7WN+lenL36/OM36uvu9nZe/5/p7Ppbn8uru0z7eSPKe19leLnN/c/SVunekMhi9kOnL3n0le26f5RKaj3ZepqsMzHU3/z5kdK1gNe9vmlzJtJXl9a+2YvUx/cf9ikdba92q6XPOeSR6c6fLBe+yjttltcLkbwlwbVUk+3lq74+zIWt6NRfdsVy/X5rTWzq2q22U6++7ZVXVia+2Pl/gaC63X77bW5t6L5bZ5Cy7DHk7OdIbPtiSvyBTg3DKXD3I+kel9vUxVXT1T6P/pJD+9jBphtezZln0ze2kfZly2r9NaO7rv//y3JB+tqtvv47WWsq0t9Lx97b8tZj6z+zZz+zXJdObjHeaCoYW01r5dVedlOoPy9D741Ezt3XUynWWTTG3C7ZO8Z+bpt0/y8WUsAwADcY+cH/SBJA+t6d4v18705f/DmU7Z/XIPce6e6WjPcs1dV32XJBf2o0jvTPL4qst+Rekn53nuCZm59rmmXzU4Ncmd+5Gaufvt3Gye58/5ZqZLmlbqhCS/Wf3mhFV1aKYdjMPm6knyyCTvW8I835npng9z9/m5flVdZ4HpT0nyS0m+1lr7fj9T4OBMR87nvgD9XZK7VP8Fl34d/Mty+TMTYKXek+QhtfvXRA7twx9SVVfo90o4PLt3wvfm56vq0P4/+oBMX/BPTPLgue2gj/+BNqhvM9dorb09071TbtNHnZLdZ/I9PFM7t1QnJHlM9XuxzCzbrHOSXLuq7tinuVJV3apNNxb9Ru3+NZWHL+P1L6eqrpfk2621v810qdbt+qiltG2LWq97ODXTJRbJzNmRS3zdvflgkjtkuuzkyz1EuiDTZRJzZzWcmOQq1X9NsKb7+bwoyevmzpSCDeSGc21BpvvUnZq9tA97e2JV3aS19qHW2h9m2g5+JFO79fA+/maZAsyF2tJl6WcJf7GqHtBf68q93Ztv/3Cx3pXk8XMPquq2+5j+lEw3b587w/qDmS7VOnUmZH5+kufNfObcNtOlZn+5hLoAGJAg5wf9U6Z7OZyZ6UvZ77fW/iNTEHBEVe1I8qgkn1rBa1xcVR9L8spM90FIpvu6XCnJWVX18f54b56d5JB+s70zk9y9tXZBpg/uN1bVWZk+7G+xjxremuSBtcDNjhfp1ZmuEz+r1/Nr/WjTY5Ic19fXpZmWdVHadEPiNyT5YH/+m7PwF6Qdma53P3WPYRfOXOJyUaYvRM+oqnP6+I8kmf154IfW5X+2d8Ff9YI9tdY+nukXhN7Xt4c/76M+n2mH/9+SHL2PI7IfznSp1FlJ/rG1dlqbfkHtGUne1bfxEzLdj2FPP5TkbX2a/5vkKX344zOFMGdlClafuIxle0emyyJPq6ozkjx1L9N8J9MZI8/ry39Gdt+w8zFJXtGfuz9Ogbt1kg/3+f1RprYxme6v9Y5axM2Ol7BeZz0pyVP69D+a6XK2ZHq/vl/TzVGfPN+TF6jl65m+sM4eSf9gpqPvZ/ZpWpIHZgoGz8t0P6aLM12yMueue7RjfqWG9XJOkt+pqk9mup/fX2T+9mFPL6jpRsFnZwo0zswUTlyh7xe8KcmRrbVL5nn+Sj0yyRP6dn5Kkh/O/PuHi/WETPuRZ/XLv47ex/QnZwr+54Kc05PcIDOXWrbWjs90/6FTqupTme6f84i5y7y6v5tpD/yKHcAmUbtDfdZCVZ2U6QbCrlcGLqeqjkxyhHs1bVz9yPxFrbVWVQ/LdOPj+693XbCR1PTrmG9rrf34etcCAJuRe+QAwOLdPsnL+2Ww38h0DwsAAFgzzshZJVX1ikw/fz3rpa211+5t+vVUVX+Q5CF7DD6utfan61HP3vTrv0/cy6h7zv3CD4yiqu6T5Hl7DP5sa+2B61HPrKq6daZflZp1SWvtZ1bhtVZ1PVTVhzL9HPCsR7bWduyP+S/HRqwJNqMRtrWqekx+8FLXk1trv7O36QFgjiAHAAAAYBBudgwAAAAwCEEOAAAAwCAEOQAAAACDEOQAAAAADEKQAwAAADCI/w+9NbMPRA5lnQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JvmBDL6muH_B"
      },
      "source": [
        "## Feature Engineering"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FSf9cFocuTH-",
        "outputId": "49849933-3a86-4a02-c3cd-f637a716159f"
      },
      "source": [
        "credit.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 32576 entries, 1 to 32580\n",
            "Data columns (total 19 columns):\n",
            " #   Column                          Non-Null Count  Dtype  \n",
            "---  ------                          --------------  -----  \n",
            " 0   person_age                      32576 non-null  int64  \n",
            " 1   person_income                   32576 non-null  int64  \n",
            " 2   person_home_ownership           32576 non-null  object \n",
            " 3   person_emp_length               32576 non-null  float64\n",
            " 4   loan_intent                     32576 non-null  object \n",
            " 5   loan_grade                      32576 non-null  object \n",
            " 6   loan_amnt                       32576 non-null  int64  \n",
            " 7   loan_int_rate                   32576 non-null  float64\n",
            " 8   loan_status                     32576 non-null  int64  \n",
            " 9   loan_percent_income             32576 non-null  float64\n",
            " 10  cb_person_default_on_file       32576 non-null  object \n",
            " 11  cb_person_cred_hist_length      32576 non-null  int64  \n",
            " 12  person_age_WOE                  32576 non-null  float64\n",
            " 13  person_emp_length_WOE           32576 non-null  float64\n",
            " 14  loan_amnt_WOE                   32576 non-null  float64\n",
            " 15  loan_int_rate_WOE               32576 non-null  float64\n",
            " 16  loan_percent_income_WOE         32576 non-null  float64\n",
            " 17  cb_person_cred_hist_length_WOE  32576 non-null  float64\n",
            " 18  person_income_WOE               32576 non-null  float64\n",
            "dtypes: float64(10), int64(5), object(4)\n",
            "memory usage: 5.0+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aUeUOu9aC96w"
      },
      "source": [
        "### Dummy Variable"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 251
        },
        "id": "n-Nnt_Hhpw2a",
        "outputId": "15cf7c53-e7ce-435c-ec22-49a07f166aa5"
      },
      "source": [
        "# dummy variable for categorical features\n",
        "\n",
        "credit = pd.concat([credit,pd.get_dummies(credit[['person_home_ownership', 'loan_intent']],\n",
        "                               drop_first=True)], axis=1).reset_index(drop=True)\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "0          21           9600                   OWN                5.0   \n",
              "1          25           9600              MORTGAGE                1.0   \n",
              "2          23          65500                  RENT                4.0   \n",
              "3          24          54400                  RENT                8.0   \n",
              "4          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "0   EDUCATION          B       1000          11.14            0   \n",
              "1     MEDICAL          C       5500          12.87            1   \n",
              "2     MEDICAL          C      35000          15.23            1   \n",
              "3     MEDICAL          C      35000          14.27            1   \n",
              "4     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income  ... cb_person_cred_hist_length_WOE  person_income_WOE  \\\n",
              "0                 0.10  ...                       0.063726           0.858225   \n",
              "1                 0.57  ...                       0.063726           0.858225   \n",
              "2                 0.53  ...                       0.063726          -0.305010   \n",
              "3                 0.55  ...                       0.029037          -0.029252   \n",
              "4                 0.25  ...                       0.063726           0.858225   \n",
              "\n",
              "   person_home_ownership_OTHER  person_home_ownership_OWN  \\\n",
              "0                            0                          1   \n",
              "1                            0                          0   \n",
              "2                            0                          0   \n",
              "3                            0                          0   \n",
              "4                            0                          1   \n",
              "\n",
              "   person_home_ownership_RENT  loan_intent_EDUCATION  \\\n",
              "0                           0                      1   \n",
              "1                           0                      0   \n",
              "2                           1                      0   \n",
              "3                           1                      0   \n",
              "4                           0                      0   \n",
              "\n",
              "   loan_intent_HOMEIMPROVEMENT  loan_intent_MEDICAL  loan_intent_PERSONAL  \\\n",
              "0                            0                    0                     0   \n",
              "1                            0                    1                     0   \n",
              "2                            0                    1                     0   \n",
              "3                            0                    1                     0   \n",
              "4                            0                    0                     0   \n",
              "\n",
              "   loan_intent_VENTURE  \n",
              "0                    0  \n",
              "1                    0  \n",
              "2                    0  \n",
              "3                    0  \n",
              "4                    1  \n",
              "\n",
              "[5 rows x 27 columns]"
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>...</th>\n",
              "      <th>cb_person_cred_hist_length_WOE</th>\n",
              "      <th>person_income_WOE</th>\n",
              "      <th>person_home_ownership_OTHER</th>\n",
              "      <th>person_home_ownership_OWN</th>\n",
              "      <th>person_home_ownership_RENT</th>\n",
              "      <th>loan_intent_EDUCATION</th>\n",
              "      <th>loan_intent_HOMEIMPROVEMENT</th>\n",
              "      <th>loan_intent_MEDICAL</th>\n",
              "      <th>loan_intent_PERSONAL</th>\n",
              "      <th>loan_intent_VENTURE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "      <td>0</td>\n",
              "      <td>1</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",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>-0.305010</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>...</td>\n",
              "      <td>0.029037</td>\n",
              "      <td>-0.029252</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</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>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 27 columns</p>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 327
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DdTp5jfJDScS"
      },
      "source": [
        "### Label Encoding"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-YkqrtYRB7Yh"
      },
      "source": [
        "##### cb_person_default_on_file"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 251
        },
        "id": "A6sIHi1bpw6a",
        "outputId": "e1557783-7d4e-4f73-902e-d9d11b4705ec"
      },
      "source": [
        "# transform categorical into binary value\n",
        "\n",
        "credit['cb_person_default_on_file'] = credit['cb_person_default_on_file'].map({\n",
        "    'Y' : 1,\n",
        "    'N' : 0\n",
        "})\n",
        "\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "0          21           9600                   OWN                5.0   \n",
              "1          25           9600              MORTGAGE                1.0   \n",
              "2          23          65500                  RENT                4.0   \n",
              "3          24          54400                  RENT                8.0   \n",
              "4          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "0   EDUCATION          B       1000          11.14            0   \n",
              "1     MEDICAL          C       5500          12.87            1   \n",
              "2     MEDICAL          C      35000          15.23            1   \n",
              "3     MEDICAL          C      35000          14.27            1   \n",
              "4     VENTURE          A       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income  ...  cb_person_cred_hist_length_WOE  \\\n",
              "0                 0.10  ...                        0.063726   \n",
              "1                 0.57  ...                        0.063726   \n",
              "2                 0.53  ...                        0.063726   \n",
              "3                 0.55  ...                        0.029037   \n",
              "4                 0.25  ...                        0.063726   \n",
              "\n",
              "   person_income_WOE  person_home_ownership_OTHER  person_home_ownership_OWN  \\\n",
              "0           0.858225                            0                          1   \n",
              "1           0.858225                            0                          0   \n",
              "2          -0.305010                            0                          0   \n",
              "3          -0.029252                            0                          0   \n",
              "4           0.858225                            0                          1   \n",
              "\n",
              "   person_home_ownership_RENT  loan_intent_EDUCATION  \\\n",
              "0                           0                      1   \n",
              "1                           0                      0   \n",
              "2                           1                      0   \n",
              "3                           1                      0   \n",
              "4                           0                      0   \n",
              "\n",
              "   loan_intent_HOMEIMPROVEMENT  loan_intent_MEDICAL  loan_intent_PERSONAL  \\\n",
              "0                            0                    0                     0   \n",
              "1                            0                    1                     0   \n",
              "2                            0                    1                     0   \n",
              "3                            0                    1                     0   \n",
              "4                            0                    0                     0   \n",
              "\n",
              "   loan_intent_VENTURE  \n",
              "0                    0  \n",
              "1                    0  \n",
              "2                    0  \n",
              "3                    0  \n",
              "4                    1  \n",
              "\n",
              "[5 rows x 27 columns]"
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>...</th>\n",
              "      <th>cb_person_cred_hist_length_WOE</th>\n",
              "      <th>person_income_WOE</th>\n",
              "      <th>person_home_ownership_OTHER</th>\n",
              "      <th>person_home_ownership_OWN</th>\n",
              "      <th>person_home_ownership_RENT</th>\n",
              "      <th>loan_intent_EDUCATION</th>\n",
              "      <th>loan_intent_HOMEIMPROVEMENT</th>\n",
              "      <th>loan_intent_MEDICAL</th>\n",
              "      <th>loan_intent_PERSONAL</th>\n",
              "      <th>loan_intent_VENTURE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>B</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "      <td>0</td>\n",
              "      <td>1</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",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>-0.305010</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>...</td>\n",
              "      <td>0.029037</td>\n",
              "      <td>-0.029252</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>A</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</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>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 27 columns</p>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 328
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fE2zpzhBB9wV"
      },
      "source": [
        "#### loan_grade"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "knJU3iw5nZZx",
        "outputId": "7335e1c7-dfb5-4465-8c34-e84657a3f318"
      },
      "source": [
        "# ordinal categorical\n",
        "\n",
        "grade_encode = {}\n",
        "v = 1\n",
        "\n",
        "for n in sorted(credit['loan_grade'].unique()):\n",
        "  grade_encode[n] = v\n",
        "  v += 1\n",
        "\n",
        "grade_encode"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'A': 1, 'B': 2, 'C': 3, 'D': 4, 'E': 5, 'F': 6, 'G': 7}"
            ]
          },
          "metadata": {},
          "execution_count": 329
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 251
        },
        "id": "WKTXly66vWhU",
        "outputId": "ce22b13b-ad8e-4fb6-eafa-706698d9e8df"
      },
      "source": [
        "credit['loan_grade'] = credit['loan_grade'].map(grade_encode)\n",
        "credit.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "0          21           9600                   OWN                5.0   \n",
              "1          25           9600              MORTGAGE                1.0   \n",
              "2          23          65500                  RENT                4.0   \n",
              "3          24          54400                  RENT                8.0   \n",
              "4          21           9900                   OWN                2.0   \n",
              "\n",
              "  loan_intent  loan_grade  loan_amnt  loan_int_rate  loan_status  \\\n",
              "0   EDUCATION           2       1000          11.14            0   \n",
              "1     MEDICAL           3       5500          12.87            1   \n",
              "2     MEDICAL           3      35000          15.23            1   \n",
              "3     MEDICAL           3      35000          14.27            1   \n",
              "4     VENTURE           1       2500           7.14            1   \n",
              "\n",
              "   loan_percent_income  ...  cb_person_cred_hist_length_WOE  \\\n",
              "0                 0.10  ...                        0.063726   \n",
              "1                 0.57  ...                        0.063726   \n",
              "2                 0.53  ...                        0.063726   \n",
              "3                 0.55  ...                        0.029037   \n",
              "4                 0.25  ...                        0.063726   \n",
              "\n",
              "   person_income_WOE  person_home_ownership_OTHER  person_home_ownership_OWN  \\\n",
              "0           0.858225                            0                          1   \n",
              "1           0.858225                            0                          0   \n",
              "2          -0.305010                            0                          0   \n",
              "3          -0.029252                            0                          0   \n",
              "4           0.858225                            0                          1   \n",
              "\n",
              "   person_home_ownership_RENT  loan_intent_EDUCATION  \\\n",
              "0                           0                      1   \n",
              "1                           0                      0   \n",
              "2                           1                      0   \n",
              "3                           1                      0   \n",
              "4                           0                      0   \n",
              "\n",
              "   loan_intent_HOMEIMPROVEMENT  loan_intent_MEDICAL  loan_intent_PERSONAL  \\\n",
              "0                            0                    0                     0   \n",
              "1                            0                    1                     0   \n",
              "2                            0                    1                     0   \n",
              "3                            0                    1                     0   \n",
              "4                            0                    0                     0   \n",
              "\n",
              "   loan_intent_VENTURE  \n",
              "0                    0  \n",
              "1                    0  \n",
              "2                    0  \n",
              "3                    0  \n",
              "4                    1  \n",
              "\n",
              "[5 rows x 27 columns]"
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_status</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>...</th>\n",
              "      <th>cb_person_cred_hist_length_WOE</th>\n",
              "      <th>person_income_WOE</th>\n",
              "      <th>person_home_ownership_OTHER</th>\n",
              "      <th>person_home_ownership_OWN</th>\n",
              "      <th>person_home_ownership_RENT</th>\n",
              "      <th>loan_intent_EDUCATION</th>\n",
              "      <th>loan_intent_HOMEIMPROVEMENT</th>\n",
              "      <th>loan_intent_MEDICAL</th>\n",
              "      <th>loan_intent_PERSONAL</th>\n",
              "      <th>loan_intent_VENTURE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>21</td>\n",
              "      <td>9600</td>\n",
              "      <td>OWN</td>\n",
              "      <td>5.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>2</td>\n",
              "      <td>1000</td>\n",
              "      <td>11.14</td>\n",
              "      <td>0</td>\n",
              "      <td>0.10</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "      <td>0</td>\n",
              "      <td>1</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",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>25</td>\n",
              "      <td>9600</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>3</td>\n",
              "      <td>5500</td>\n",
              "      <td>12.87</td>\n",
              "      <td>1</td>\n",
              "      <td>0.57</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>23</td>\n",
              "      <td>65500</td>\n",
              "      <td>RENT</td>\n",
              "      <td>4.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>3</td>\n",
              "      <td>35000</td>\n",
              "      <td>15.23</td>\n",
              "      <td>1</td>\n",
              "      <td>0.53</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>-0.305010</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>24</td>\n",
              "      <td>54400</td>\n",
              "      <td>RENT</td>\n",
              "      <td>8.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>3</td>\n",
              "      <td>35000</td>\n",
              "      <td>14.27</td>\n",
              "      <td>1</td>\n",
              "      <td>0.55</td>\n",
              "      <td>...</td>\n",
              "      <td>0.029037</td>\n",
              "      <td>-0.029252</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>21</td>\n",
              "      <td>9900</td>\n",
              "      <td>OWN</td>\n",
              "      <td>2.0</td>\n",
              "      <td>VENTURE</td>\n",
              "      <td>1</td>\n",
              "      <td>2500</td>\n",
              "      <td>7.14</td>\n",
              "      <td>1</td>\n",
              "      <td>0.25</td>\n",
              "      <td>...</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0.858225</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>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 27 columns</p>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 330
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CqvFKTdJrCil"
      },
      "source": [
        "## Splitting Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XzjZF_aHrDi9"
      },
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.model_selection import train_test_split"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SL4Q1AtyCGN1",
        "outputId": "a2e6b3ad-3673-400c-e8e1-4291a5acdd13"
      },
      "source": [
        "print(f\"After all these processes, our dataset has {len(credit)} rows.\")\n",
        "print(f\"80% of this dataset is: {len(credit)*0.8 :.0f} rows for modeling.\")\n",
        "print(f\"20% of this dataset is: {len(credit)*0.2 :.0f} rows for holdout sample.\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "After all these processes, our dataset has 32576 rows.\n",
            "80% of this dataset is: 26061 rows for modeling.\n",
            "20% of this dataset is: 6515 rows for holdout sample.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HUgqrUkCtZlw"
      },
      "source": [
        "# dataset for modelling - 80%\n",
        "credit_model = credit.loc[:26060]\n",
        "\n",
        "# dataset for holdout sample - 20%\n",
        "credit_holdout = credit.loc[26060:]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "S0tKcqcotra2",
        "outputId": "2e412467-a531-47fb-c815-55e67c637f09"
      },
      "source": [
        "print(f\"Dataset for modeling has {len(credit_model)} rows.\")\n",
        "print(f\"Dataset for holdout sample has {len(credit_holdout)} rows.\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Dataset for modeling has 26061 rows.\n",
            "Dataset for holdout sample has 6516 rows.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "T7WNhITXDdFt",
        "outputId": "9a856930-ce1c-4976-ea26-ba95942d83c6"
      },
      "source": [
        "credit_model.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 26061 entries, 0 to 26060\n",
            "Data columns (total 27 columns):\n",
            " #   Column                          Non-Null Count  Dtype  \n",
            "---  ------                          --------------  -----  \n",
            " 0   person_age                      26061 non-null  int64  \n",
            " 1   person_income                   26061 non-null  int64  \n",
            " 2   person_home_ownership           26061 non-null  object \n",
            " 3   person_emp_length               26061 non-null  float64\n",
            " 4   loan_intent                     26061 non-null  object \n",
            " 5   loan_grade                      26061 non-null  int64  \n",
            " 6   loan_amnt                       26061 non-null  int64  \n",
            " 7   loan_int_rate                   26061 non-null  float64\n",
            " 8   loan_status                     26061 non-null  int64  \n",
            " 9   loan_percent_income             26061 non-null  float64\n",
            " 10  cb_person_default_on_file       26061 non-null  int64  \n",
            " 11  cb_person_cred_hist_length      26061 non-null  int64  \n",
            " 12  person_age_WOE                  26061 non-null  float64\n",
            " 13  person_emp_length_WOE           26061 non-null  float64\n",
            " 14  loan_amnt_WOE                   26061 non-null  float64\n",
            " 15  loan_int_rate_WOE               26061 non-null  float64\n",
            " 16  loan_percent_income_WOE         26061 non-null  float64\n",
            " 17  cb_person_cred_hist_length_WOE  26061 non-null  float64\n",
            " 18  person_income_WOE               26061 non-null  float64\n",
            " 19  person_home_ownership_OTHER     26061 non-null  uint8  \n",
            " 20  person_home_ownership_OWN       26061 non-null  uint8  \n",
            " 21  person_home_ownership_RENT      26061 non-null  uint8  \n",
            " 22  loan_intent_EDUCATION           26061 non-null  uint8  \n",
            " 23  loan_intent_HOMEIMPROVEMENT     26061 non-null  uint8  \n",
            " 24  loan_intent_MEDICAL             26061 non-null  uint8  \n",
            " 25  loan_intent_PERSONAL            26061 non-null  uint8  \n",
            " 26  loan_intent_VENTURE             26061 non-null  uint8  \n",
            "dtypes: float64(10), int64(7), object(2), uint8(8)\n",
            "memory usage: 4.0+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZJ3JFLxoDe8F",
        "outputId": "d5d8acc4-1a26-41f8-c21d-09311f48467a"
      },
      "source": [
        "credit_holdout.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 6516 entries, 26060 to 32575\n",
            "Data columns (total 27 columns):\n",
            " #   Column                          Non-Null Count  Dtype  \n",
            "---  ------                          --------------  -----  \n",
            " 0   person_age                      6516 non-null   int64  \n",
            " 1   person_income                   6516 non-null   int64  \n",
            " 2   person_home_ownership           6516 non-null   object \n",
            " 3   person_emp_length               6516 non-null   float64\n",
            " 4   loan_intent                     6516 non-null   object \n",
            " 5   loan_grade                      6516 non-null   int64  \n",
            " 6   loan_amnt                       6516 non-null   int64  \n",
            " 7   loan_int_rate                   6516 non-null   float64\n",
            " 8   loan_status                     6516 non-null   int64  \n",
            " 9   loan_percent_income             6516 non-null   float64\n",
            " 10  cb_person_default_on_file       6516 non-null   int64  \n",
            " 11  cb_person_cred_hist_length      6516 non-null   int64  \n",
            " 12  person_age_WOE                  6516 non-null   float64\n",
            " 13  person_emp_length_WOE           6516 non-null   float64\n",
            " 14  loan_amnt_WOE                   6516 non-null   float64\n",
            " 15  loan_int_rate_WOE               6516 non-null   float64\n",
            " 16  loan_percent_income_WOE         6516 non-null   float64\n",
            " 17  cb_person_cred_hist_length_WOE  6516 non-null   float64\n",
            " 18  person_income_WOE               6516 non-null   float64\n",
            " 19  person_home_ownership_OTHER     6516 non-null   uint8  \n",
            " 20  person_home_ownership_OWN       6516 non-null   uint8  \n",
            " 21  person_home_ownership_RENT      6516 non-null   uint8  \n",
            " 22  loan_intent_EDUCATION           6516 non-null   uint8  \n",
            " 23  loan_intent_HOMEIMPROVEMENT     6516 non-null   uint8  \n",
            " 24  loan_intent_MEDICAL             6516 non-null   uint8  \n",
            " 25  loan_intent_PERSONAL            6516 non-null   uint8  \n",
            " 26  loan_intent_VENTURE             6516 non-null   uint8  \n",
            "dtypes: float64(10), int64(7), object(2), uint8(8)\n",
            "memory usage: 1018.3+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5fFqFrEjrfv6"
      },
      "source": [
        "# Splitting dataset for feature selection\n",
        "\n",
        "X_init = credit_model.drop(['loan_status','person_home_ownership', 'loan_intent'], axis = 1)\n",
        "y_init = credit_model['loan_status']\n",
        "\n",
        "X_train_init, X_test_init, y_train_init, y_test_init = train_test_split(X_init, \n",
        "                                                                        y_init, \n",
        "                                                                        test_size = 0.2, \n",
        "                                                                        random_state = 88, \n",
        "                                                                        stratify = y_init)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CLHiawG04dQI"
      },
      "source": [
        "## Feature Selection"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RCrfIrNht81K",
        "outputId": "08572d3f-4730-4c4e-a5cf-dd6bb2d88dbd"
      },
      "source": [
        "# using LogReg with L1 regularization\n",
        "\n",
        "initial = LogisticRegression(penalty= 'l1',\n",
        "                             C = 1,\n",
        "                             solver = 'liblinear')\n",
        "initial.fit(X_train_init, y_train_init)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression(C=1, penalty='l1', solver='liblinear')"
            ]
          },
          "metadata": {},
          "execution_count": 338
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 786
        },
        "id": "aMyacOcpt-jV",
        "outputId": "9f056229-7d27-47c9-ac94-6d44f7fb9758"
      },
      "source": [
        "feature_importance_df = pd.DataFrame(\n",
        "    {\n",
        "        'feature': X_init.columns,\n",
        "        'importance': abs(initial.coef_)[0],        \n",
        "    }\n",
        ")\n",
        "select_from_model = feature_importance_df.sort_values(by='importance', ascending=False)\n",
        "select_from_model"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                           feature  importance\n",
              "6              loan_percent_income   14.002825\n",
              "11                   loan_amnt_WOE    7.655567\n",
              "10           person_emp_length_WOE    1.789665\n",
              "17       person_home_ownership_OWN    1.743541\n",
              "14  cb_person_cred_hist_length_WOE    1.173597\n",
              "9                   person_age_WOE    1.079938\n",
              "3                       loan_grade    1.028083\n",
              "23             loan_intent_VENTURE    1.002669\n",
              "19           loan_intent_EDUCATION    0.830802\n",
              "18      person_home_ownership_RENT    0.781119\n",
              "22            loan_intent_PERSONAL    0.499078\n",
              "20     loan_intent_HOMEIMPROVEMENT    0.366929\n",
              "15               person_income_WOE    0.186421\n",
              "21             loan_intent_MEDICAL    0.152930\n",
              "16     person_home_ownership_OTHER    0.148486\n",
              "7        cb_person_default_on_file    0.126560\n",
              "0                       person_age    0.072516\n",
              "12               loan_int_rate_WOE    0.070198\n",
              "5                    loan_int_rate    0.066406\n",
              "13         loan_percent_income_WOE    0.065443\n",
              "2                person_emp_length    0.031474\n",
              "8       cb_person_cred_hist_length    0.000436\n",
              "4                        loan_amnt    0.000054\n",
              "1                    person_income    0.000005"
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>feature</th>\n",
              "      <th>importance</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>loan_percent_income</td>\n",
              "      <td>14.002825</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>loan_amnt_WOE</td>\n",
              "      <td>7.655567</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>person_emp_length_WOE</td>\n",
              "      <td>1.789665</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>person_home_ownership_OWN</td>\n",
              "      <td>1.743541</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>cb_person_cred_hist_length_WOE</td>\n",
              "      <td>1.173597</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>person_age_WOE</td>\n",
              "      <td>1.079938</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>loan_grade</td>\n",
              "      <td>1.028083</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>23</th>\n",
              "      <td>loan_intent_VENTURE</td>\n",
              "      <td>1.002669</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>loan_intent_EDUCATION</td>\n",
              "      <td>0.830802</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>person_home_ownership_RENT</td>\n",
              "      <td>0.781119</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>22</th>\n",
              "      <td>loan_intent_PERSONAL</td>\n",
              "      <td>0.499078</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20</th>\n",
              "      <td>loan_intent_HOMEIMPROVEMENT</td>\n",
              "      <td>0.366929</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>person_income_WOE</td>\n",
              "      <td>0.186421</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>loan_intent_MEDICAL</td>\n",
              "      <td>0.152930</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>person_home_ownership_OTHER</td>\n",
              "      <td>0.148486</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>cb_person_default_on_file</td>\n",
              "      <td>0.126560</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>person_age</td>\n",
              "      <td>0.072516</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>loan_int_rate_WOE</td>\n",
              "      <td>0.070198</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>loan_int_rate</td>\n",
              "      <td>0.066406</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>loan_percent_income_WOE</td>\n",
              "      <td>0.065443</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>person_emp_length</td>\n",
              "      <td>0.031474</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>cb_person_cred_hist_length</td>\n",
              "      <td>0.000436</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>loan_amnt</td>\n",
              "      <td>0.000054</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>person_income</td>\n",
              "      <td>0.000005</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 339
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "i09iUODmEd2A",
        "outputId": "fdaea92f-8cb4-4167-dcc0-7741cca8414d"
      },
      "source": [
        "# importance > 0.30\n",
        "select_from_model[select_from_model['importance'] > 0.3]['feature'].unique()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array(['loan_percent_income', 'loan_amnt_WOE', 'person_emp_length_WOE',\n",
              "       'person_home_ownership_OWN', 'cb_person_cred_hist_length_WOE',\n",
              "       'person_age_WOE', 'loan_grade', 'loan_intent_VENTURE',\n",
              "       'loan_intent_EDUCATION', 'person_home_ownership_RENT',\n",
              "       'loan_intent_PERSONAL', 'loan_intent_HOMEIMPROVEMENT'],\n",
              "      dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 340
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ShYS5PP-wYB4"
      },
      "source": [
        "# selecting importance > 0.30\n",
        "\n",
        "# selected_feat =   ['loan_percent_income', 'person_age_WOE',\n",
        "#          'person_home_ownership_OWN', 'loan_intent_VENTURE', 'loan_grade',\n",
        "#          'loan_intent_EDUCATION', 'person_home_ownership_RENT',\n",
        "#          'loan_intent_PERSONAL', 'person_emp_length_WOE',\n",
        "#          'loan_intent_HOMEIMPROVEMENT', \n",
        "#          'loan_status']\n",
        "\n",
        "selected_feat = ['loan_percent_income', 'loan_amnt_WOE', 'person_emp_length_WOE',\n",
        "       'person_home_ownership_OWN', 'loan_grade', 'loan_intent_VENTURE',\n",
        "       'loan_intent_EDUCATION', 'person_home_ownership_RENT',\n",
        "       'person_age_WOE', 'cb_person_cred_hist_length_WOE',\n",
        "       'loan_intent_PERSONAL', 'loan_intent_HOMEIMPROVEMENT',\n",
        "       'loan_status']\n",
        "\n",
        "# masking dataset for modelling purposes\n",
        "credit_selected = credit_model[selected_feat]\n",
        "\n",
        "# masking dataset for holdout purposes\n",
        "credit_holdout  = credit_holdout[selected_feat]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "7PdSkAIVwodI",
        "outputId": "7d89c14e-341f-4219-f23f-205bba904034"
      },
      "source": [
        "# from 27 features, 13 features selected.\n",
        "\n",
        "credit_selected.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   loan_percent_income  loan_amnt_WOE  person_emp_length_WOE  \\\n",
              "0                 0.10       0.115326              -0.055408   \n",
              "1                 0.57      -0.010540               0.115326   \n",
              "2                 0.53      -0.074659              -0.010540   \n",
              "3                 0.55      -0.074659              -0.074659   \n",
              "4                 0.25       0.115326               0.115326   \n",
              "\n",
              "   person_home_ownership_OWN  loan_grade  loan_intent_VENTURE  \\\n",
              "0                          1           2                    0   \n",
              "1                          0           3                    0   \n",
              "2                          0           3                    0   \n",
              "3                          0           3                    0   \n",
              "4                          1           1                    1   \n",
              "\n",
              "   loan_intent_EDUCATION  person_home_ownership_RENT  person_age_WOE  \\\n",
              "0                      1                           0        0.115326   \n",
              "1                      0                           0       -0.010540   \n",
              "2                      0                           1        0.115326   \n",
              "3                      0                           1       -0.010540   \n",
              "4                      0                           0        0.115326   \n",
              "\n",
              "   cb_person_cred_hist_length_WOE  loan_intent_PERSONAL  \\\n",
              "0                        0.063726                     0   \n",
              "1                        0.063726                     0   \n",
              "2                        0.063726                     0   \n",
              "3                        0.029037                     0   \n",
              "4                        0.063726                     0   \n",
              "\n",
              "   loan_intent_HOMEIMPROVEMENT  loan_status  \n",
              "0                            0            0  \n",
              "1                            0            1  \n",
              "2                            0            1  \n",
              "3                            0            1  \n",
              "4                            0            1  "
            ],
            "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>loan_percent_income</th>\n",
              "      <th>loan_amnt_WOE</th>\n",
              "      <th>person_emp_length_WOE</th>\n",
              "      <th>person_home_ownership_OWN</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_intent_VENTURE</th>\n",
              "      <th>loan_intent_EDUCATION</th>\n",
              "      <th>person_home_ownership_RENT</th>\n",
              "      <th>person_age_WOE</th>\n",
              "      <th>cb_person_cred_hist_length_WOE</th>\n",
              "      <th>loan_intent_PERSONAL</th>\n",
              "      <th>loan_intent_HOMEIMPROVEMENT</th>\n",
              "      <th>loan_status</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.10</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>-0.055408</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.57</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.53</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.55</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>-0.074659</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>-0.010540</td>\n",
              "      <td>0.029037</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.25</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.115326</td>\n",
              "      <td>0.063726</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 342
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "P5kTNPgED-IX",
        "outputId": "bf3a6941-be3b-4899-ea16-037b9dc9cdf2"
      },
      "source": [
        "credit_selected.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 26061 entries, 0 to 26060\n",
            "Data columns (total 13 columns):\n",
            " #   Column                          Non-Null Count  Dtype  \n",
            "---  ------                          --------------  -----  \n",
            " 0   loan_percent_income             26061 non-null  float64\n",
            " 1   loan_amnt_WOE                   26061 non-null  float64\n",
            " 2   person_emp_length_WOE           26061 non-null  float64\n",
            " 3   person_home_ownership_OWN       26061 non-null  uint8  \n",
            " 4   loan_grade                      26061 non-null  int64  \n",
            " 5   loan_intent_VENTURE             26061 non-null  uint8  \n",
            " 6   loan_intent_EDUCATION           26061 non-null  uint8  \n",
            " 7   person_home_ownership_RENT      26061 non-null  uint8  \n",
            " 8   person_age_WOE                  26061 non-null  float64\n",
            " 9   cb_person_cred_hist_length_WOE  26061 non-null  float64\n",
            " 10  loan_intent_PERSONAL            26061 non-null  uint8  \n",
            " 11  loan_intent_HOMEIMPROVEMENT     26061 non-null  uint8  \n",
            " 12  loan_status                     26061 non-null  int64  \n",
            "dtypes: float64(5), int64(2), uint8(6)\n",
            "memory usage: 1.5 MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3rgtFqeaEAKs",
        "outputId": "db7f873f-e845-4e20-f0d5-092969780ce9"
      },
      "source": [
        "credit_holdout.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 6516 entries, 26060 to 32575\n",
            "Data columns (total 13 columns):\n",
            " #   Column                          Non-Null Count  Dtype  \n",
            "---  ------                          --------------  -----  \n",
            " 0   loan_percent_income             6516 non-null   float64\n",
            " 1   loan_amnt_WOE                   6516 non-null   float64\n",
            " 2   person_emp_length_WOE           6516 non-null   float64\n",
            " 3   person_home_ownership_OWN       6516 non-null   uint8  \n",
            " 4   loan_grade                      6516 non-null   int64  \n",
            " 5   loan_intent_VENTURE             6516 non-null   uint8  \n",
            " 6   loan_intent_EDUCATION           6516 non-null   uint8  \n",
            " 7   person_home_ownership_RENT      6516 non-null   uint8  \n",
            " 8   person_age_WOE                  6516 non-null   float64\n",
            " 9   cb_person_cred_hist_length_WOE  6516 non-null   float64\n",
            " 10  loan_intent_PERSONAL            6516 non-null   uint8  \n",
            " 11  loan_intent_HOMEIMPROVEMENT     6516 non-null   uint8  \n",
            " 12  loan_status                     6516 non-null   int64  \n",
            "dtypes: float64(5), int64(2), uint8(6)\n",
            "memory usage: 394.7 KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_eDDYGZ2w9xt"
      },
      "source": [
        "## Modelling"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "DtHWt6XGHPQ2",
        "outputId": "0eac1f89-2627-4803-c9aa-85349e740bad"
      },
      "source": [
        "sns.countplot(credit_selected['loan_status'])\n",
        "plt.title('Loan Status')\n",
        "plt.show()\n",
        "\n",
        "# has more non default applicant (0) than the default (1)\n",
        "  # non default   : 77.7215%\t\n",
        "  # default       : 22.2785%\n",
        "\n",
        "# Error Type:\n",
        "#   Type 1: predict default (1), actual non default (0)\n",
        "#   Type 2: predict non-default (0), actual default (1) <-- riskier in business decision\n",
        "\n",
        "# For business purposes, we assume to minimize Type 2 (minimize False Negative).\n",
        "  # Hence, we use Recall metrics for optimum result"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\seaborn\\_decorators.py:36: FutureWarning: Pass the following variable as a keyword arg: x. 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"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZEAAAEXCAYAAABsyHmSAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAZ8klEQVR4nO3df7RdZX3n8ffH8EMtUEK5TZGgQRtdg7aNcguMLVaLhWBbox1HYVqJlhodpVOnthVtp1AsXUyrVVGHDtYIVAWxVE1dKKaMSp0RzQ1GfmoJCCUxJFcQgz9KDX7nj/NcOcR7w2Un5xyu9/1a66yzz3c/e+9nZ4V82M/e5zypKiRJ6uJRo+6AJGnuMkQkSZ0ZIpKkzgwRSVJnhogkqTNDRJLUmSEiSerMENG8l+S2JM8dcR/emOSrSb6VZFOSD/at+3SS33kY+zozyfsG01PpwQwRacSSrAReCjy3qvYDxoErR9sraXYMEWkGSfZN8rYkX2uvtyXZt61bmORjSSaTfKMtL+7b9tNJ3pTk/ya5N8knkxw8w6F+Hriiqm4BqKo7q+r8tp+zgWOBd7arlHe2+tuT3JFke5L1SY5t9eXAG4GXtPZfavUHXW31X60keXSS9yW5K8k9SdYlWbRn/zT1o8oQkWb2x8AxwDLg54CjgD9p6x4FvBd4AvB44LvAO3fa/r8ALwd+EtgH+IMZjnM1cEqSP0wynmTB1Iqq+mPgn4HTqmq/qjqtrVrX+nUQ8AHgQ0keXVWfAP4C+GBr/3OzOM+VwI8DhwE/AbyqnY/0kAwRaWa/CZxVVduqahL4M3rDTlTVXVV1WVV9p6ruBc4Gfmmn7d9bVf9SVd8FLqX3j/4Pqar3Ab8LnAB8BtiW5PW76lhVva/1YUdVvQXYF3hKx/P8Hr3w+Omqur+q1lfV9o770jxjiEgzexxwe9/n21uNJI9N8r+T3J5kO3AVcGD/VQRwZ9/yd4D9ZjpQVb2/qp4LHEjvSuBNSU6YqX2SP0hyU5JvJrmH3pXETMNlD+XvgCuAS9qw3V8m2bvjvjTPGCLSzL5Gb7hqyuNbDeB19P7P/+iqOgB4Vqtndw5YVd+rqg8B1wJPmyr3t2n3P/4IeDGwsKoOBL7Zd+zpfpr728Bj+z7/1E7H/LOqOgJ4JvBrwCm7cx6aPwwRqWfvdoN56rUXcDHwJ0nG2k3xPwWmHp3dn959g3uSHASc0fXASV6W5FeT7J/kUUlOBJ4KfL412Qo8sW+T/YEdwCSwV5I/BQ7oW78VWJKk/7/vDcBJSfZOMg68qO/4z0nyM+0qaju94a3vdz0fzS+GiNRzOb1QmHqdCfw5MEHvquA64JpWA3gb8Bjg6/RujH9iN469nd4TVf8K3AP8JfBfq+qzbf3bgRe1p8DOpTf09AngX+gNsf0bcEff/j7U3u9Kck1b/h/Ak4Bv0Lu384G+9j8F/H3rx0307sv83W6cj+aROCmVJKkrr0QkSZ0ZIpKkzgwRSVJnhogkqbO9Rt2BYTv44INryZIlo+6GJM0p69ev/3pVje1cn3chsmTJEiYmJkbdDUmaU5LcPl3d4SxJUmeGiCSpM0NEktSZISJJ6swQkSR1ZohIkjozRCRJnRkikqTOBhYiSQ5L8qkkNya5IcnvtfpBSdYmubm9L2z1JDk3ycYk1yZ5Rt++Vrb2NydZ2Vc/Msl1bZtzk+zWrHKSpIdnkN9Y3wG8rqquSbI/sD7JWuBlwJVVdU6S04HTgdcDJwJL2+to4Dzg6L5Z48bpTfu5PsmaqvpGa/MKejPAXQ4sBz4+wHMC4Mg/vGjQh9Acs/6vnE1W89PArkSqaktVXdOW76U3Y9qhwArgwtbsQuAFbXkFcFH1XA0cmOQQ4ARgbVXd3YJjLbC8rTugqq6u3sxaF/XtS5I0BEO5J5JkCfB0elcMi6pqS1t1J7CoLR/Kg6f43NRqu6pvmqY+3fFXJZlIMjE5Obl7JyNJ+oGBh0iS/YDLgNdW1fb+de0KYuDz81bV+VU1XlXjY2M/9COUkqSOBhoiSfamFyDvr6p/aOWtbSiK9r6t1TcDh/VtvrjVdlVfPE1dkjQkg3w6K8B7gJuq6q/7Vq0Bpp6wWgl8tK9+SntK6xjgm23Y6wrg+CQL25NcxwNXtHXbkxzTjnVK374kSUMwyKezfgF4KXBdkg2t9kbgHODSJKcCtwMvbusuB54HbAS+A7wcoKruTvImYF1rd1ZV3d2WXw1cADyG3lNZA38yS5L0gIGFSFV9FpjpexvHTdO+gNfMsK/VwOpp6hPA03ajm5Kk3eA31iVJnRkikqTODBFJUmeGiCSpM0NEktSZISJJ6swQkSR1ZohIkjozRCRJnRkikqTODBFJUmeGiCSpM0NEktSZISJJ6swQkSR1ZohIkjozRCRJnQ1yjvXVSbYlub6v9sEkG9rrtqlpc5MsSfLdvnV/07fNkUmuS7IxybltPnWSHJRkbZKb2/vCQZ2LJGl6g7wSuQBY3l+oqpdU1bKqWgZcBvxD3+pbptZV1av66ucBrwCWttfUPk8HrqyqpcCV7bMkaYgGFiJVdRVw93Tr2tXEi4GLd7WPJIcAB1TV1W0O9ouAF7TVK4AL2/KFfXVJ0pCM6p7IscDWqrq5r3Z4ki8m+UySY1vtUGBTX5tNrQawqKq2tOU7gUUzHSzJqiQTSSYmJyf30ClIkkYVIifz4KuQLcDjq+rpwO8DH0hywGx31q5Sahfrz6+q8aoaHxsb69pnSdJO9hr2AZPsBfwGcORUraruA+5ry+uT3AI8GdgMLO7bfHGrAWxNckhVbWnDXtuG0X9J0gNGcSXyXODLVfWDYaokY0kWtOUn0ruBfmsbrtqe5Jh2H+UU4KNtszXAyra8sq8uSRqSQT7iezHwOeApSTYlObWtOokfvqH+LODa9sjv3wOvqqqpm/KvBv4W2AjcAny81c8BfiXJzfSC6ZxBnYskaXoDG86qqpNnqL9smtpl9B75na79BPC0aep3AcftXi8lSbvDb6xLkjozRCRJnRkikqTODBFJUmeGiCSpM0NEktSZISJJ6swQkSR1ZohIkjozRCRJnRkikqTODBFJUmeGiCSpM0NEktSZISJJ6swQkSR1NsiZDVcn2Zbk+r7amUk2J9nQXs/rW/eGJBuTfCXJCX315a22McnpffXDk3y+1T+YZJ9BnYskaXqDvBK5AFg+Tf2tVbWsvS4HSHIEvWlzn9q2+V9JFrR5198FnAgcAZzc2gL8z7avnwa+AZy684EkSYM1sBCpqquAux+yYc8K4JKquq+qvkpvPvWj2mtjVd1aVf8OXAKsSBLgl+nNxw5wIfCCPdl/SdJDG8U9kdOSXNuGuxa22qHAHX1tNrXaTPWfAO6pqh071SVJQzTsEDkPeBKwDNgCvGUYB02yKslEkonJyclhHFKS5oWhhkhVba2q+6vq+8C76Q1XAWwGDutrurjVZqrfBRyYZK+d6jMd9/yqGq+q8bGxsT1zMpKk4YZIkkP6Pr4QmHpyaw1wUpJ9kxwOLAW+AKwDlrYnsfahd/N9TVUV8CngRW37lcBHh3EOkqQH7PXQTbpJcjHwbODgJJuAM4BnJ1kGFHAb8EqAqrohyaXAjcAO4DVVdX/bz2nAFcACYHVV3dAO8XrgkiR/DnwReM+gzkWSNL2BhUhVnTxNecZ/6KvqbODsaeqXA5dPU7+VB4bDJEkj4DfWJUmdGSKSpM4MEUlSZ4aIJKkzQ0SS1JkhIknqzBCRJHVmiEiSOjNEJEmdGSKSpM4MEUlSZ4aIJKkzQ0SS1JkhIknqzBCRJHVmiEiSOjNEJEmdDSxEkqxOsi3J9X21v0ry5STXJvlwkgNbfUmS7ybZ0F5/07fNkUmuS7IxyblJ0uoHJVmb5Ob2vnBQ5yJJmt4gr0QuAJbvVFsLPK2qfhb4F+ANfetuqapl7fWqvvp5wCuApe01tc/TgSurailwZfssSRqigYVIVV0F3L1T7ZNVtaN9vBpYvKt9JDkEOKCqrq6qAi4CXtBWrwAubMsX9tUlSUMyynsivw18vO/z4Um+mOQzSY5ttUOBTX1tNrUawKKq2tKW7wQWzXSgJKuSTCSZmJyc3EPdlySNJESS/DGwA3h/K20BHl9VTwd+H/hAkgNmu792lVK7WH9+VY1X1fjY2Nhu9FyS1G+vYR8wycuAXwOOa//4U1X3Afe15fVJbgGeDGzmwUNei1sNYGuSQ6pqSxv22jakU5AkNUO9EkmyHPgj4PlV9Z2++liSBW35ifRuoN/ahqu2JzmmPZV1CvDRttkaYGVbXtlXlyQNycCuRJJcDDwbODjJJuAMek9j7QusbU/qXt2exHoWcFaS7wHfB15VVVM35V9N70mvx9C7hzJ1H+Uc4NIkpwK3Ay8e1LlIkqY3sBCpqpOnKb9nhraXAZfNsG4CeNo09buA43anj5Kk3eM31iVJnRkikqTOZhUiSa6cTU2SNL/s8p5IkkcDj6V3c3whkLbqAB740p8kaZ56qBvrrwReCzwOWM8DIbIdeOfguiVJmgt2GSJV9Xbg7Ul+t6reMaQ+SZLmiFk94ltV70jyTGBJ/zZVddGA+iVJmgNmFSJJ/g54ErABuL+Vp35VV5I0T832y4bjwBFTv3UlSRLM/nsi1wM/NciOSJLmntleiRwM3JjkC7Rf2wWoqucPpFeSpDlhtiFy5iA7IUmam2b7dNZnBt0RSdLcM9uns+7lgZkD9wH2Br5dVbOefVCS9KNntlci+08tt8mhVgDHDKpTkqS54WH/im/1fAQ4Yc93R5I0l8z2V3x/o+/1oiTnAP82i+1WJ9mW5Pq+2kFJ1ia5ub0vbPUkOTfJxiTXJnlG3zYrW/ubk6zsqx+Z5Lq2zbntKkmSNCSzvRL59b7XCcC99Ia0HsoFwPKdaqcDV1bVUuDK9hngRHpzqy8FVgHnQS906E2tezRwFHDGVPC0Nq/o227nY0mSBmi290Re3mXnVXVVkiU7lVfQm3sd4ELg08DrW/2i9q34q5McmOSQ1nbt1JzrSdYCy5N8Gjigqq5u9YuAF/DAHOySpAGb7XDW4iQfbkNT25JclmRxx2MuqqotbflOYFFbPhS4o6/dplbbVX3TNHVJ0pDMdjjrvcAaevOKPA74x1bbLe2qY+C/x5VkVZKJJBOTk5ODPpwkzRuzDZGxqnpvVe1orwuAsY7H3NqGqWjv21p9M3BYX7vFrbar+uJp6j+kqs6vqvGqGh8b69ptSdLOZhsidyX5rSQL2uu3gLs6HnMNMPWE1Urgo331U9pTWscA32zDXlcAxydZ2G6oHw9c0dZtT3JMeyrrlL59SZKGYLa/nfXbwDuAt9Ibfvp/wMseaqMkF9O7MX5wkk30nrI6B7g0yanA7cCLW/PLgecBG4HvAC8HqKq7k7wJWNfanTV1kx14Nb0nwB5D74a6N9UlaYhmGyJnASur6hvwg8du30wvXGZUVSfPsOq4adoW8JoZ9rMaWD1NfQJ42i57LkkamNkOZ/3sVIBA7+oAePpguiRJmitmGyKP6vuC39SVyGyvYiRJP6JmGwRvAT6X5EPt838Gzh5MlyRJc8Vsv7F+UZIJ4Jdb6Teq6sbBdUuSNBfMekiqhYbBIUn6gYf9U/CSJE0xRCRJnRkikqTODBFJUmeGiCSpM0NEktSZISJJ6swQkSR1ZohIkjozRCRJnRkikqTODBFJUmdDD5EkT0myoe+1Pclrk5yZZHNf/Xl927whycYkX0lyQl99eattTHL6sM9Fkua7oU8sVVVfAZYBJFkAbAY+TG9O9bdW1Zv72yc5AjgJeCrwOOCfkjy5rX4X8CvAJmBdkjX+RL0kDc+oZyc8Drilqm5PMlObFcAlVXUf8NUkG4Gj2rqNVXUrQJJLWltDRJKGZNT3RE4CLu77fFqSa5Os7puO91Dgjr42m1ptpvoPSbIqyUSSicnJyT3Xe0ma50YWIkn2AZ4PTE25ex7wJHpDXVvoTcm7R1TV+VU1XlXjY2Nje2q3kjTvjXI460TgmqraCjD1DpDk3cDH2sfNwGF92y1uNXZRlyQNwSiHs06mbygrySF9614IXN+W1wAnJdk3yeHAUuALwDpgaZLD21XNSa2tJGlIRnIlkuTH6D1V9cq+8l8mWQYUcNvUuqq6Icml9G6Y7wBeU1X3t/2cBlwBLABWV9UNwzoHSdKIQqSqvg38xE61l+6i/dnA2dPULwcu3+MdlCTNyqifzpIkzWGGiCSpM0NEktSZISJJ6swQkSR1ZohIkjozRCRJnRkikqTODBFJUmeGiCSpM0NEktSZISJJ6mzU0+NK2oP+9ayfGXUX9Aj0+D+9bmD79kpEktSZISJJ6swQkSR1ZohIkjobWYgkuS3JdUk2JJlotYOSrE1yc3tf2OpJcm6SjUmuTfKMvv2sbO1vTrJyVOcjSfPRqK9EnlNVy6pqvH0+HbiyqpYCV7bPACcCS9trFXAe9EIHOAM4GjgKOGMqeCRJgzfqENnZCuDCtnwh8IK++kXVczVwYJJDgBOAtVV1d1V9A1gLLB9ynyVp3hpliBTwySTrk6xqtUVVtaUt3wksasuHAnf0bbup1WaqP0iSVUkmkkxMTk7uyXOQpHltlF82/MWq2pzkJ4G1Sb7cv7KqKkntiQNV1fnA+QDj4+N7ZJ+SpBFeiVTV5va+DfgwvXsaW9swFe19W2u+GTisb/PFrTZTXZI0BCMJkSQ/lmT/qWXgeOB6YA0w9YTVSuCjbXkNcEp7SusY4Jtt2OsK4PgkC9sN9eNbTZI0BKMazloEfDjJVB8+UFWfSLIOuDTJqcDtwItb+8uB5wEbge8ALweoqruTvAlY19qdVVV3D+80JGl+G0mIVNWtwM9NU78LOG6aegGvmWFfq4HVe7qPkqSH9kh7xFeSNIcYIpKkzgwRSVJnhogkqTNDRJLUmSEiSerMEJEkdWaISJI6M0QkSZ0ZIpKkzgwRSVJnhogkqTNDRJLUmSEiSerMEJEkdWaISJI6G3qIJDksyaeS3JjkhiS/1+pnJtmcZEN7Pa9vmzck2ZjkK0lO6Ksvb7WNSU4f9rlI0nw3ipkNdwCvq6pr2jzr65OsbeveWlVv7m+c5AjgJOCpwOOAf0ry5Lb6XcCvAJuAdUnWVNWNQzkLSdLwQ6SqtgBb2vK9SW4CDt3FJiuAS6rqPuCrSTYCR7V1G9tUuyS5pLU1RCRpSEZ6TyTJEuDpwOdb6bQk1yZZnWRhqx0K3NG32aZWm6k+3XFWJZlIMjE5ObknT0GS5rWRhUiS/YDLgNdW1XbgPOBJwDJ6Vypv2VPHqqrzq2q8qsbHxsb21G4lad4bxT0RkuxNL0DeX1X/AFBVW/vWvxv4WPu4GTisb/PFrcYu6pKkIRjF01kB3gPcVFV/3Vc/pK/ZC4Hr2/Ia4KQk+yY5HFgKfAFYByxNcniSfejdfF8zjHOQJPWM4krkF4CXAtcl2dBqbwROTrIMKOA24JUAVXVDkkvp3TDfAbymqu4HSHIacAWwAFhdVTcM7zQkSaN4OuuzQKZZdfkutjkbOHua+uW72k6SNFh+Y12S1JkhIknqzBCRJHVmiEiSOjNEJEmdGSKSpM4MEUlSZ4aIJKkzQ0SS1JkhIknqzBCRJHVmiEiSOjNEJEmdGSKSpM4MEUlSZ4aIJKkzQ0SS1NmcD5Eky5N8JcnGJKePuj+SNJ/M6RBJsgB4F3AicAS9edqPGG2vJGn+mNMhAhwFbKyqW6vq34FLgBUj7pMkzRt7jboDu+lQ4I6+z5uAo3dulGQVsKp9/FaSrwyhb/PFwcDXR92JUcubV466C/ph/t2cckb2xF6eMF1xrofIrFTV+cD5o+7Hj6IkE1U1Pup+SDvz7+ZwzPXhrM3AYX2fF7eaJGkI5nqIrAOWJjk8yT7AScCaEfdJkuaNOT2cVVU7kpwGXAEsAFZX1Q0j7tZ84zChHqn8uzkEqapR90GSNEfN9eEsSdIIGSKSpM4MEXXiz83okSrJ6iTbklw/6r7MB4aIHjZ/bkaPcBcAy0fdifnCEFEX/tyMHrGq6irg7lH3Y74wRNTFdD83c+iI+iJphAwRSVJnhoi68OdmJAGGiLrx52YkAYaIOqiqHcDUz83cBFzqz83okSLJxcDngKck2ZTk1FH36UeZP3siSerMKxFJUmeGiCSpM0NEktSZISJJ6swQkSR1ZohIkjozRKRdSPKtER772UmeuafaSYNgiEiPXM8GZhMOs20n7XGGiDQL6fmrJNcnuS7JS1p9vyRXJrmm1Ve0+pIkNyV5d5IbknwyyWN2sf//luTGJNcmuSTJEuBVwH9PsiHJsUl+Pcnnk3wxyT8lWTRDuwuSvKhv399q74ckuaq1uz7JsYP7E9N84TfWpV1I8q2q2i/Jf6L3j/Vy4GB6vx92NDAJPLaqtic5GLgaWAo8AdgIjFfVhiSXAmuq6n0zHOdrwOFVdV+SA6vqniRnAt+qqje3NguBe6qqkvwO8B+q6nXTtLsA+FhV/f1O5/A64NFVdXabWOyxVXXvnv9T03yy16g7IM0RvwhcXFX3A1uTfAb4eeDjwF8keRbwfXrzqixq23y1qja05fXAkl3s/1rg/Uk+AnxkhjaLgQ8mOQTYB/jqwzyHdcDqJHsDH+nrm9SZw1nS7vlNYAw4sqqWAVuBR7d19/W1u59d/0/br9KbcvgZwLok07V9B/DOqvoZ4JV9x9nZDtp/20keRS9wpmb8exa9n+2/IMkpD3Vy0kMxRKTZ+WfgJUkWJBmj94/xF4AfB7ZV1feSPIfeMNbD0v6hP6yqPgW8vu1zP+BeYP++pj/OA/O2rOyr79zuNuDItvx8YO92nCcAW6vq3cDf0gssabcYItLsfJjekNOXgP8D/FFV3Qm8HxhPch1wCvDlDvteALyv7eOLwLlVdQ/wj8ALp26YA2cCH0qyHvh63/Y7t3s38EtJvgT8R+Dbrd2zgS8l+SLwEuDtHfoqPYg31iVJnXklIknqzKezpCFK8i7gF3Yqv72q3juK/ki7y+EsSVJnDmdJkjozRCRJnRkikqTODBFJUmf/H+8zrzB+L3caAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RagO9EsvEFTp"
      },
      "source": [
        "from sklearn.preprocessing import RobustScaler\n",
        "from sklearn.metrics import confusion_matrix, classification_report, roc_auc_score, f1_score, recall_score"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b0yeLChnw_8f"
      },
      "source": [
        "def evaluate(true, predicted):\n",
        "\n",
        "    rec = recall_score(true, predicted)\n",
        "    f1s = f1_score(true, predicted)\n",
        "    auc = roc_auc_score(true, predicted) \n",
        "    return rec,f1s,auc\n",
        "\n",
        "def print_evaluate(true, predicted):\n",
        "    \n",
        "    rec = recall_score(true, predicted)\n",
        "    f1s = f1_score(true, predicted)\n",
        "    auc = roc_auc_score(true, predicted)\n",
        "\n",
        "    print(' Recall:', rec)\n",
        "    print(' F1-score:', f1s)\n",
        "    print(' AUC:', auc)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UZVcdFbSxIli"
      },
      "source": [
        "# Splitting into Train and Test set\n",
        "\n",
        "X = credit_selected.drop('loan_status', axis = 1)\n",
        "y = credit_selected['loan_status']\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, \n",
        "                                                    y, \n",
        "                                                    test_size = 0.2, \n",
        "                                                    random_state = 88, \n",
        "                                                    stratify = y)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tzqhXd1TyEKB"
      },
      "source": [
        "# Scaling for distance based algorithm\n",
        "\n",
        "scaler_robust = RobustScaler()\n",
        "\n",
        "X_train_sc = scaler_robust.fit_transform(X_train)\n",
        "X_test_sc = scaler_robust.transform(X_test)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3ebEJGmsykzO"
      },
      "source": [
        "from sklearn.model_selection import GridSearchCV, RandomizedSearchCV\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from xgboost import XGBClassifier"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ivZFlnwXySBP"
      },
      "source": [
        "### Logistic Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KGnND1GwyT2B"
      },
      "source": [
        "lr_t = LogisticRegression()\n",
        "\n",
        "param_lr = {\n",
        "    \"penalty\": ['l1', 'l2', 'elasticnet', 'none'], \n",
        "    \"C\"      : np.logspace(-5,5,20),\n",
        "    \"max_iter\" : [100, 500, 1000]\n",
        "}\n",
        "\n",
        "grid_lr = GridSearchCV(\n",
        "    estimator = lr_t,\n",
        "    param_grid = param_lr,\n",
        "    cv = 5,\n",
        "    scoring = 'roc_auc',\n",
        "    refit = True,\n",
        "    n_jobs = -1,\n",
        "    verbose = 3\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_ppnUnofzIW-",
        "outputId": "35837693-548e-4209-f0ef-3aff2a50fb4a"
      },
      "source": [
        "grid_lr.fit(X_train_sc, y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fitting 5 folds for each of 240 candidates, totalling 1200 fits\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\sklearn\\model_selection\\_search.py:918: UserWarning: One or more of the test scores are non-finite: [       nan 0.82231169        nan 0.86884023        nan 0.82231169\n",
            "        nan 0.86884023        nan 0.82231169        nan 0.86884023\n",
            "        nan 0.82373809        nan 0.86884023        nan 0.82373809\n",
            "        nan 0.86884023        nan 0.82373809        nan 0.86884023\n",
            "        nan 0.82797459        nan 0.86884023        nan 0.82797459\n",
            "        nan 0.86884023        nan 0.82797459        nan 0.86884023\n",
            "        nan 0.83773898        nan 0.86884023        nan 0.83773898\n",
            "        nan 0.86884023        nan 0.83773898        nan 0.86884023\n",
            "        nan 0.85073191        nan 0.86884023        nan 0.85073191\n",
            "        nan 0.86884023        nan 0.85073191        nan 0.86884023\n",
            "        nan 0.86088443        nan 0.86884023        nan 0.86088443\n",
            "        nan 0.86884023        nan 0.86088443        nan 0.86884023\n",
            "        nan 0.86623416        nan 0.86884023        nan 0.86623416\n",
            "        nan 0.86884023        nan 0.86623416        nan 0.86884023\n",
            "        nan 0.8681728         nan 0.86884023        nan 0.8681728\n",
            "        nan 0.86884023        nan 0.8681728         nan 0.86884023\n",
            "        nan 0.86868632        nan 0.86884023        nan 0.86868632\n",
            "        nan 0.86884023        nan 0.86868632        nan 0.86884023\n",
            "        nan 0.86880363        nan 0.86884023        nan 0.86880363\n",
            "        nan 0.86884023        nan 0.86880363        nan 0.86884023\n",
            "        nan 0.86882834        nan 0.86884023        nan 0.86882834\n",
            "        nan 0.86884023        nan 0.86882834        nan 0.86884023\n",
            "        nan 0.86883565        nan 0.86884023        nan 0.86883565\n",
            "        nan 0.86884023        nan 0.86883565        nan 0.86884023\n",
            "        nan 0.8688393         nan 0.86884023        nan 0.8688393\n",
            "        nan 0.86884023        nan 0.8688393         nan 0.86884023\n",
            "        nan 0.86883983        nan 0.86884023        nan 0.86883983\n",
            "        nan 0.86884023        nan 0.86883983        nan 0.86884023\n",
            "        nan 0.86884003        nan 0.86884023        nan 0.86884003\n",
            "        nan 0.86884023        nan 0.86884003        nan 0.86884023\n",
            "        nan 0.8688401         nan 0.86884023        nan 0.8688401\n",
            "        nan 0.86884023        nan 0.8688401         nan 0.86884023\n",
            "        nan 0.86884023        nan 0.86884023        nan 0.86884023\n",
            "        nan 0.86884023        nan 0.86884023        nan 0.86884023\n",
            "        nan 0.86884017        nan 0.86884023        nan 0.86884017\n",
            "        nan 0.86884023        nan 0.86884017        nan 0.86884023\n",
            "        nan 0.86884023        nan 0.86884023        nan 0.86884023\n",
            "        nan 0.86884023        nan 0.86884023        nan 0.86884023\n",
            "        nan 0.86884023        nan 0.86884023        nan 0.86884023\n",
            "        nan 0.86884023        nan 0.86884023        nan 0.86884023]\n",
            "  warnings.warn(\n",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:1322: UserWarning: Setting penalty='none' will ignore the C and l1_ratio parameters\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "GridSearchCV(cv=5, estimator=LogisticRegression(), n_jobs=-1,\n",
              "             param_grid={'C': array([1.00000000e-05, 3.35981829e-05, 1.12883789e-04, 3.79269019e-04,\n",
              "       1.27427499e-03, 4.28133240e-03, 1.43844989e-02, 4.83293024e-02,\n",
              "       1.62377674e-01, 5.45559478e-01, 1.83298071e+00, 6.15848211e+00,\n",
              "       2.06913808e+01, 6.95192796e+01, 2.33572147e+02, 7.84759970e+02,\n",
              "       2.63665090e+03, 8.85866790e+03, 2.97635144e+04, 1.00000000e+05]),\n",
              "                         'max_iter': [100, 500, 1000],\n",
              "                         'penalty': ['l1', 'l2', 'elasticnet', 'none']},\n",
              "             scoring='roc_auc', verbose=3)"
            ]
          },
          "metadata": {},
          "execution_count": 352
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MXPpdQ6OzQfA",
        "outputId": "77081f72-19af-45d5-bdc4-385466567b1d"
      },
      "source": [
        "grid_lr.best_params_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'C': 1e-05, 'max_iter': 100, 'penalty': 'none'}"
            ]
          },
          "metadata": {},
          "execution_count": 353
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I4t40NXdzTlS"
      },
      "source": [
        "pred_lrt = grid_lr.predict(X_test_sc)\n",
        "train_lrt = grid_lr.predict(X_train_sc)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qjiayEzLG9fL",
        "outputId": "fd08ca3e-bfe9-499a-e852-b9c33ad1b60c"
      },
      "source": [
        "print(classification_report(y_test, pred_lrt))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.87      0.95      0.91      4052\n",
            "           1       0.76      0.52      0.62      1161\n",
            "\n",
            "    accuracy                           0.86      5213\n",
            "   macro avg       0.82      0.74      0.77      5213\n",
            "weighted avg       0.85      0.86      0.85      5213\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2lYjCclyzWZT",
        "outputId": "c4797ae3-a253-4e85-a76a-b01647689063"
      },
      "source": [
        "print('Train set evaluation:')\n",
        "print_evaluate(y_train, train_lrt)\n",
        "\n",
        "print('\\n')\n",
        "\n",
        "print('Test set evaluation:')\n",
        "print_evaluate(y_test, pred_lrt)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train set evaluation:\n",
            " Recall: 0.5252960172228203\n",
            " F1-score: 0.6187396982376061\n",
            " AUC: 0.7378995052478354\n",
            "\n",
            "\n",
            "Test set evaluation:\n",
            " Recall: 0.524547803617571\n",
            " F1-score: 0.6211116777154513\n",
            " AUC: 0.7387052937140175\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 110
        },
        "id": "fZrKNk5_0pEC",
        "outputId": "e65db059-24ef-4b34-9003-63e1bfb5f1fc"
      },
      "source": [
        "index = pd.MultiIndex.from_product([\n",
        "                                    ['Train', 'Test'], \n",
        "                                    ['Recall', 'F1-Score','AUC']\n",
        "                                    ])\n",
        "\n",
        "result_lrt = pd.DataFrame(\n",
        "    [pd.DataFrame(\n",
        "        {'Train' : list(evaluate(y_train, train_lrt)),\n",
        "         'Test' : list(evaluate(y_test, pred_lrt))}\n",
        "         ).unstack().values], \n",
        "         columns = index)\n",
        "\n",
        "result_lrt.insert(loc=0, column='Model', value = 'LogisticRegression')\n",
        "result_lrt"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                Model     Train                       Test                    \n",
              "                         Recall F1-Score     AUC    Recall  F1-Score       AUC\n",
              "0  LogisticRegression  0.525296  0.61874  0.7379  0.524548  0.621112  0.738705"
            ],
            "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 tr th {\n",
              "        text-align: left;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Train</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Test</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>LogisticRegression</td>\n",
              "      <td>0.525296</td>\n",
              "      <td>0.61874</td>\n",
              "      <td>0.7379</td>\n",
              "      <td>0.524548</td>\n",
              "      <td>0.621112</td>\n",
              "      <td>0.738705</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 357
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NHJzcwbN0lUu"
      },
      "source": [
        "### Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vxbWK5f-zyv2"
      },
      "source": [
        "rf_t = RandomForestClassifier()\n",
        "param_rf = {\n",
        "    \"n_estimators\" : [10,25,30,50,100,500],\n",
        "    \"max_depth\" : [1,2,4,5,6],\n",
        "    \"min_samples_leaf\" : np.linspace(0.07, 0.2, 30),\n",
        "}\n",
        "\n",
        "grid_rf = RandomizedSearchCV(\n",
        "    estimator = rf_t,\n",
        "    param_distributions = param_rf,\n",
        "    cv = 5,\n",
        "    scoring = 'roc_auc',\n",
        "    refit = True,\n",
        "    n_jobs = -1,\n",
        "    verbose = 3\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PnZ-127Y1WZ8",
        "outputId": "8c92d0ec-661a-42c1-ef23-27cc39dc89ac"
      },
      "source": [
        "grid_rf.fit(X_train, y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fitting 5 folds for each of 10 candidates, totalling 50 fits\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomizedSearchCV(cv=5, estimator=RandomForestClassifier(), n_jobs=-1,\n",
              "                   param_distributions={'max_depth': [1, 2, 4, 5, 6],\n",
              "                                        'min_samples_leaf': array([0.07      , 0.07448276, 0.07896552, 0.08344828, 0.08793103,\n",
              "       0.09241379, 0.09689655, 0.10137931, 0.10586207, 0.11034483,\n",
              "       0.11482759, 0.11931034, 0.1237931 , 0.12827586, 0.13275862,\n",
              "       0.13724138, 0.14172414, 0.1462069 , 0.15068966, 0.15517241,\n",
              "       0.15965517, 0.16413793, 0.16862069, 0.17310345, 0.17758621,\n",
              "       0.18206897, 0.18655172, 0.19103448, 0.19551724, 0.2       ]),\n",
              "                                        'n_estimators': [10, 25, 30, 50, 100,\n",
              "                                                         500]},\n",
              "                   scoring='roc_auc', verbose=3)"
            ]
          },
          "metadata": {},
          "execution_count": 359
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "htnDLBDK1ezZ",
        "outputId": "48ff671f-1104-479d-f6bd-856fb514de60"
      },
      "source": [
        "grid_rf.best_params_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'n_estimators': 25, 'min_samples_leaf': 0.07, 'max_depth': 1}"
            ]
          },
          "metadata": {},
          "execution_count": 360
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nJ16X_431WZ-"
      },
      "source": [
        "pred_rft = grid_rf.predict(X_test)\n",
        "train_rft = grid_rf.predict(X_train)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "h4c_CfJC1WZ-",
        "outputId": "9ba5381c-cc5d-43bc-f54e-2eda38287ce1"
      },
      "source": [
        "print('Train set evaluation:')\n",
        "print_evaluate(y_train, train_rft)\n",
        "\n",
        "print('\\n')\n",
        "\n",
        "print('Test set evaluation:')\n",
        "print_evaluate(y_test, pred_rft)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train set evaluation:\n",
            " Recall: 0.0\n",
            " F1-score: 0.0\n",
            " AUC: 0.5\n",
            "\n",
            "\n",
            "Test set evaluation:\n",
            " Recall: 0.0\n",
            " F1-score: 0.0\n",
            " AUC: 0.5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 110
        },
        "id": "FfsadoIa1WZ_",
        "outputId": "3b98ecdc-f300-45c8-d3af-02de59e71233"
      },
      "source": [
        "index = pd.MultiIndex.from_product([\n",
        "                                    ['Train', 'Test'], \n",
        "                                    ['Recall', 'F1-Score','AUC']\n",
        "                                    ])\n",
        "result_rft = pd.DataFrame(\n",
        "    [pd.DataFrame(\n",
        "        {'Train' : list(evaluate(y_train, train_rft)),\n",
        "         'Test' : list(evaluate(y_test, pred_rft))}\n",
        "         ).unstack().values], \n",
        "         columns=index)\n",
        "\n",
        "result_rft.insert(loc=0, column='Model', value = 'RandomForest')\n",
        "result_rft"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "          Model  Train                 Test              \n",
              "                Recall F1-Score  AUC Recall F1-Score  AUC\n",
              "0  RandomForest    0.0      0.0  0.5    0.0      0.0  0.5"
            ],
            "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 tr th {\n",
              "        text-align: left;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Train</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Test</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>RandomForest</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.5</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 363
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kaY8c6oS1-Pe"
      },
      "source": [
        "### XGBoost"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ihMkDCXZ2AHS"
      },
      "source": [
        "xg_t = XGBClassifier()\n",
        "\n",
        "param_xg = {\n",
        "    \"n_estimators\" : [10, 20, 25, 30, 50, 100, 500],\n",
        "    \"learning_rate\" : np.linspace(0.08, 4, 20),\n",
        "    \"subsample\" : [0.1, 0.5, 1, 4, 7]\n",
        "}\n",
        "\n",
        "grid_xg = RandomizedSearchCV(\n",
        "    estimator = xg_t,\n",
        "    param_distributions = param_xg,\n",
        "    cv = 10,\n",
        "    scoring = 'roc_auc',\n",
        "    refit = True,\n",
        "    n_jobs = -1,\n",
        "    verbose = 3,\n",
        "    n_iter = 100\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OnApWZkH2AHT",
        "outputId": "6fb55689-63b0-4c0e-f1f6-c588af1295b2"
      },
      "source": [
        "grid_xg.fit(X_train, y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fitting 10 folds for each of 100 candidates, totalling 1000 fits\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\sklearn\\model_selection\\_search.py:918: UserWarning: One or more of the test scores are non-finite: [0.73444283        nan 0.63391429        nan 0.594017   0.85999225\n",
            " 0.59119413 0.77992296 0.68361949 0.87442117 0.82936762        nan\n",
            " 0.69567771        nan 0.721392   0.890986   0.89187455 0.67886208\n",
            " 0.89969896        nan 0.8782206  0.594017          nan        nan\n",
            " 0.8850667  0.87007578 0.90765371 0.62478548 0.67311065        nan\n",
            "        nan 0.89242497        nan        nan 0.64619748 0.68574524\n",
            " 0.63842286        nan        nan        nan 0.86750173 0.8943435\n",
            "        nan 0.90780362 0.88921908        nan        nan 0.91345815\n",
            "        nan 0.64405128 0.87623741        nan        nan 0.88466707\n",
            " 0.89374772        nan 0.70298995 0.71306105        nan 0.89385238\n",
            " 0.73633835 0.63053214        nan        nan        nan 0.68085077\n",
            " 0.87952831 0.75972067 0.70960299 0.6882299         nan 0.67311065\n",
            " 0.84547411 0.69567771 0.85483582 0.87384005        nan 0.86101617\n",
            " 0.90327784 0.8886836         nan        nan        nan 0.91908677\n",
            " 0.91424173        nan 0.66350504 0.90899603        nan        nan\n",
            "        nan 0.84460769 0.67682628 0.68932612 0.67808555 0.65597094\n",
            "        nan 0.62478548 0.68946608 0.65711057]\n",
            "  warnings.warn(\n",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\xgboost\\sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
            "  warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[02:15:41] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.4.0/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomizedSearchCV(cv=10,\n",
              "                   estimator=XGBClassifier(base_score=None, booster=None,\n",
              "                                           colsample_bylevel=None,\n",
              "                                           colsample_bynode=None,\n",
              "                                           colsample_bytree=None, gamma=None,\n",
              "                                           gpu_id=None, importance_type='gain',\n",
              "                                           interaction_constraints=None,\n",
              "                                           learning_rate=None,\n",
              "                                           max_delta_step=None, max_depth=None,\n",
              "                                           min_child_weight=None, missing=nan,\n",
              "                                           monotone_constraints=None,\n",
              "                                           n_estimators=100...\n",
              "                   n_iter=100, n_jobs=-1,\n",
              "                   param_distributions={'learning_rate': array([0.08      , 0.28631579, 0.49263158, 0.69894737, 0.90526316,\n",
              "       1.11157895, 1.31789474, 1.52421053, 1.73052632, 1.93684211,\n",
              "       2.14315789, 2.34947368, 2.55578947, 2.76210526, 2.96842105,\n",
              "       3.17473684, 3.38105263, 3.58736842, 3.79368421, 4.        ]),\n",
              "                                        'n_estimators': [10, 20, 25, 30, 50,\n",
              "                                                         100, 500],\n",
              "                                        'subsample': [0.1, 0.5, 1, 4, 7]},\n",
              "                   scoring='roc_auc', verbose=3)"
            ]
          },
          "metadata": {},
          "execution_count": 365
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mcQWNmea2AHT",
        "outputId": "11ce457e-c6e2-429b-c9bc-5c985a4eab3d"
      },
      "source": [
        "grid_xg.best_params_"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'subsample': 1, 'n_estimators': 30, 'learning_rate': 0.2863157894736842}"
            ]
          },
          "metadata": {},
          "execution_count": 366
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5xbc9vA_2AHT"
      },
      "source": [
        "pred_xgt = grid_xg.predict(X_test)\n",
        "train_xgt = grid_xg.predict(X_train)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "B0eb0JZE2AHT",
        "outputId": "348087ea-cd3e-44de-9a71-93d66994ad38"
      },
      "source": [
        "print('Train set evaluation:')\n",
        "print_evaluate(y_train, train_xgt)\n",
        "\n",
        "print('\\n')\n",
        "\n",
        "print('Test set evaluation:')\n",
        "print_evaluate(y_test, pred_xgt)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train set evaluation:\n",
            " Recall: 0.6955866523143165\n",
            " F1-score: 0.8132393657186004\n",
            " AUC: 0.8456332323473699\n",
            "\n",
            "\n",
            "Test set evaluation:\n",
            " Recall: 0.6890611541774333\n",
            " F1-score: 0.7984031936127746\n",
            " AUC: 0.8392245553710462\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 110
        },
        "id": "Q6LK0PSo2AHT",
        "outputId": "e29d391e-9ee9-4130-f18e-f379f3581916"
      },
      "source": [
        "index = pd.MultiIndex.from_product([\n",
        "                                    ['Train', 'Test'], \n",
        "                                    ['Recall', 'F1-Score', 'AUC']\n",
        "                                    ])\n",
        "result_xgt = pd.DataFrame(\n",
        "    [pd.DataFrame(\n",
        "        {'Train' : list(evaluate(y_train, train_xgt)),\n",
        "         'Test' : list(evaluate(y_test, pred_xgt))}\n",
        "         ).unstack().values], \n",
        "         columns=index)\n",
        "\n",
        "result_xgt.insert(loc=0, column='Model', value = 'XGBoost')\n",
        "result_xgt"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     Model     Train                          Test                    \n",
              "              Recall  F1-Score       AUC    Recall  F1-Score       AUC\n",
              "0  XGBoost  0.695587  0.813239  0.845633  0.689061  0.798403  0.839225"
            ],
            "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 tr th {\n",
              "        text-align: left;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Train</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Test</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>XGBoost</td>\n",
              "      <td>0.695587</td>\n",
              "      <td>0.813239</td>\n",
              "      <td>0.845633</td>\n",
              "      <td>0.689061</td>\n",
              "      <td>0.798403</td>\n",
              "      <td>0.839225</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 369
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ym4KQ1sAgrXH"
      },
      "source": [
        "tuned_xg = grid_xg.best_estimator_"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IAZ9TpXH3n2a"
      },
      "source": [
        "## Model Results"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j0fuSvvC1_fR"
      },
      "source": [
        "result_all = pd.concat(\n",
        "    [result_lrt, \n",
        "     result_rft, \n",
        "     result_xgt], \n",
        "     axis=0)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 172
        },
        "id": "R8jPGJJ03RYi",
        "outputId": "3f5d983e-d2dc-4331-d71f-f8ef42ab8fef"
      },
      "source": [
        "result_all"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                Model     Train                          Test            \\\n",
              "                         Recall  F1-Score       AUC    Recall  F1-Score   \n",
              "0  LogisticRegression  0.525296  0.618740  0.737900  0.524548  0.621112   \n",
              "0        RandomForest  0.000000  0.000000  0.500000  0.000000  0.000000   \n",
              "0             XGBoost  0.695587  0.813239  0.845633  0.689061  0.798403   \n",
              "\n",
              "             \n",
              "        AUC  \n",
              "0  0.738705  \n",
              "0  0.500000  \n",
              "0  0.839225  "
            ],
            "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 tr th {\n",
              "        text-align: left;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Train</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Test</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>LogisticRegression</td>\n",
              "      <td>0.525296</td>\n",
              "      <td>0.618740</td>\n",
              "      <td>0.737900</td>\n",
              "      <td>0.524548</td>\n",
              "      <td>0.621112</td>\n",
              "      <td>0.738705</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>RandomForest</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.500000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.500000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>XGBoost</td>\n",
              "      <td>0.695587</td>\n",
              "      <td>0.813239</td>\n",
              "      <td>0.845633</td>\n",
              "      <td>0.689061</td>\n",
              "      <td>0.798403</td>\n",
              "      <td>0.839225</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 371
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IDysAU-pfYt1"
      },
      "source": [
        "## Holdout Sample"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VbYjPfVQgnFm",
        "outputId": "6b03bcaf-f604-4889-b0d1-2c667a55d118"
      },
      "source": [
        "credit_holdout.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 6516 entries, 26060 to 32575\n",
            "Data columns (total 13 columns):\n",
            " #   Column                          Non-Null Count  Dtype  \n",
            "---  ------                          --------------  -----  \n",
            " 0   loan_percent_income             6516 non-null   float64\n",
            " 1   loan_amnt_WOE                   6516 non-null   float64\n",
            " 2   person_emp_length_WOE           6516 non-null   float64\n",
            " 3   person_home_ownership_OWN       6516 non-null   uint8  \n",
            " 4   loan_grade                      6516 non-null   int64  \n",
            " 5   loan_intent_VENTURE             6516 non-null   uint8  \n",
            " 6   loan_intent_EDUCATION           6516 non-null   uint8  \n",
            " 7   person_home_ownership_RENT      6516 non-null   uint8  \n",
            " 8   person_age_WOE                  6516 non-null   float64\n",
            " 9   cb_person_cred_hist_length_WOE  6516 non-null   float64\n",
            " 10  loan_intent_PERSONAL            6516 non-null   uint8  \n",
            " 11  loan_intent_HOMEIMPROVEMENT     6516 non-null   uint8  \n",
            " 12  loan_status                     6516 non-null   int64  \n",
            "dtypes: float64(5), int64(2), uint8(6)\n",
            "memory usage: 394.7 KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ejBbYKwNhXqH",
        "outputId": "53e4b8ce-f118-49be-9cc9-2684f861861e"
      },
      "source": [
        "# Imbalanced target too\n",
        "\n",
        "credit_holdout['loan_status'].value_counts() / len(credit_holdout) * 100"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    80.033763\n",
              "1    19.966237\n",
              "Name: loan_status, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 373
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fbAchU7ufJOS"
      },
      "source": [
        "# Define X and y for predict the holdout sample\n",
        "\n",
        "X_holdout = credit_holdout.drop('loan_status', \n",
        "                                axis = 1)\n",
        "y_holdout = credit_holdout['loan_status']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "z0gjppaShCET"
      },
      "source": [
        "### Logistic Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X_j2lzG_fJTA",
        "outputId": "4820c1a7-ccdd-4435-e817-9b0c371435b3"
      },
      "source": [
        "# without scaler\n",
        "\n",
        "holdout_lr = grid_lr.predict(X_holdout)\n",
        "print_evaluate(y_holdout, holdout_lr)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            " Recall: 0.9800153727901614\n",
            " F1-score: 0.3702090592334495\n",
            " AUC: 0.5765848676031344\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9fw9T3b_gvgc",
        "outputId": "54fde8b6-24e0-4c41-a5f5-1e775817a049"
      },
      "source": [
        "# with scaler\n",
        "\n",
        "holdout_lr = grid_lr.predict(scaler_robust.transform(X_holdout))\n",
        "print_evaluate(y_holdout, holdout_lr)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            " Recall: 0.4704073789392775\n",
            " F1-score: 0.5842482100238664\n",
            " AUC: 0.7177540250401084\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 110
        },
        "id": "pEJzGdECiKE5",
        "outputId": "3d259d83-8214-4818-d114-8330f9ad8d34"
      },
      "source": [
        "index = pd.MultiIndex.from_product([\n",
        "                                    ['Train', 'Test', 'Holdout Sample'], \n",
        "                                    ['Recall', 'F1-Score','AUC']\n",
        "                                    ])\n",
        "\n",
        "result_holdout_lr = pd.DataFrame(\n",
        "    [pd.DataFrame(\n",
        "        {'Train' : list(evaluate(y_train, train_lrt)),\n",
        "         'Test' : list(evaluate(y_test, pred_lrt)),\n",
        "         'Holdout Sample' : list(evaluate(y_holdout, holdout_lr))}\n",
        "         ).unstack().values], \n",
        "         columns = index)\n",
        "\n",
        "result_holdout_lr.insert(loc=0, column='Model', value = 'LogisticRegression')\n",
        "result_holdout_lr"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                Model     Train                       Test            \\\n",
              "                         Recall F1-Score     AUC    Recall  F1-Score   \n",
              "0  LogisticRegression  0.525296  0.61874  0.7379  0.524548  0.621112   \n",
              "\n",
              "            Holdout Sample                      \n",
              "        AUC         Recall  F1-Score       AUC  \n",
              "0  0.738705       0.470407  0.584248  0.717754  "
            ],
            "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 tr th {\n",
              "        text-align: left;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Train</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Test</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Holdout Sample</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>LogisticRegression</td>\n",
              "      <td>0.525296</td>\n",
              "      <td>0.61874</td>\n",
              "      <td>0.7379</td>\n",
              "      <td>0.524548</td>\n",
              "      <td>0.621112</td>\n",
              "      <td>0.738705</td>\n",
              "      <td>0.470407</td>\n",
              "      <td>0.584248</td>\n",
              "      <td>0.717754</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 377
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ujg6v-NnhPwz"
      },
      "source": [
        "### Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bCHIcFKqfJXa",
        "outputId": "00e23380-0885-4634-e19d-3d882f7bb5da"
      },
      "source": [
        "holdout_rf = grid_rf.predict(X_holdout)\n",
        "print_evaluate(y_holdout, holdout_rf)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            " Recall: 0.0\n",
            " F1-score: 0.0\n",
            " AUC: 0.5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 110
        },
        "id": "XeKZlvz3iiCn",
        "outputId": "b18f02e3-a4cc-4829-e5f8-8185b9e0be43"
      },
      "source": [
        "index = pd.MultiIndex.from_product([\n",
        "                                    ['Train', 'Test', 'Holdout Sample'], \n",
        "                                    ['Recall', 'F1-Score','AUC']\n",
        "                                    ])\n",
        "\n",
        "result_holdout_rf = pd.DataFrame(\n",
        "    [pd.DataFrame(\n",
        "        {'Train' : list(evaluate(y_train, train_rft)),\n",
        "         'Test' : list(evaluate(y_test, pred_rft)),\n",
        "         'Holdout Sample' : list(evaluate(y_holdout, holdout_rf)) }\n",
        "         ).unstack().values], \n",
        "         columns = index)\n",
        "\n",
        "result_holdout_rf.insert(loc=0, column='Model', value = 'RandomForest')\n",
        "result_holdout_rf"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "          Model  Train                 Test               Holdout Sample  \\\n",
              "                Recall F1-Score  AUC Recall F1-Score  AUC         Recall   \n",
              "0  RandomForest    0.0      0.0  0.5    0.0      0.0  0.5            0.0   \n",
              "\n",
              "                 \n",
              "  F1-Score  AUC  \n",
              "0      0.0  0.5  "
            ],
            "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 tr th {\n",
              "        text-align: left;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Train</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Test</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Holdout Sample</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>RandomForest</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.5</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 379
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JXzTAFbHhRRX"
      },
      "source": [
        "### XGBoost"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FP4IYML3fJa-",
        "outputId": "c928df98-55f7-495a-cae5-458cb84501c8"
      },
      "source": [
        "holdout_xg = grid_xg.predict(X_holdout)\n",
        "print_evaluate(y_holdout, holdout_xg)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            " Recall: 0.6963873943120676\n",
            " F1-score: 0.8024800708591674\n",
            " AUC: 0.8433039560246819\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 110
        },
        "id": "SkRet25DiqYp",
        "outputId": "a42eb08f-152d-4d4b-af26-b2662ceb82d0"
      },
      "source": [
        "index = pd.MultiIndex.from_product([\n",
        "                                    ['Train', 'Test', 'Holdout Sample'], \n",
        "                                    ['Recall', 'F1-Score','AUC']\n",
        "                                    ])\n",
        "\n",
        "result_holdout_xg = pd.DataFrame(\n",
        "    [pd.DataFrame(\n",
        "        {'Train' : list(evaluate(y_train, train_xgt)),\n",
        "         'Test' : list(evaluate(y_test, pred_xgt)),\n",
        "         'Holdout Sample' : list(evaluate(y_holdout, holdout_xg))}\n",
        "         ).unstack().values], \n",
        "         columns = index)\n",
        "\n",
        "result_holdout_xg.insert(loc=0, column='Model', value = 'XGBoost')\n",
        "result_holdout_xg"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     Model     Train                          Test                      \\\n",
              "              Recall  F1-Score       AUC    Recall  F1-Score       AUC   \n",
              "0  XGBoost  0.695587  0.813239  0.845633  0.689061  0.798403  0.839225   \n",
              "\n",
              "  Holdout Sample                     \n",
              "          Recall F1-Score       AUC  \n",
              "0       0.696387  0.80248  0.843304  "
            ],
            "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 tr th {\n",
              "        text-align: left;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Train</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Test</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Holdout Sample</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>XGBoost</td>\n",
              "      <td>0.695587</td>\n",
              "      <td>0.813239</td>\n",
              "      <td>0.845633</td>\n",
              "      <td>0.689061</td>\n",
              "      <td>0.798403</td>\n",
              "      <td>0.839225</td>\n",
              "      <td>0.696387</td>\n",
              "      <td>0.80248</td>\n",
              "      <td>0.843304</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 381
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6dAEQqsgjRqU"
      },
      "source": [
        "## Holdout Sample Result"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 172
        },
        "id": "grxFgkTHjCoa",
        "outputId": "d8983b6d-d1f1-459d-cb33-00d97ee0e57f"
      },
      "source": [
        "holdout_all = pd.concat(\n",
        "                        [result_holdout_lr, \n",
        "                        result_holdout_rf, \n",
        "                        result_holdout_xg], \n",
        "                        axis=0)\n",
        "holdout_all"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                Model     Train                          Test            \\\n",
              "                         Recall  F1-Score       AUC    Recall  F1-Score   \n",
              "0  LogisticRegression  0.525296  0.618740  0.737900  0.524548  0.621112   \n",
              "0        RandomForest  0.000000  0.000000  0.500000  0.000000  0.000000   \n",
              "0             XGBoost  0.695587  0.813239  0.845633  0.689061  0.798403   \n",
              "\n",
              "            Holdout Sample                      \n",
              "        AUC         Recall  F1-Score       AUC  \n",
              "0  0.738705       0.470407  0.584248  0.717754  \n",
              "0  0.500000       0.000000  0.000000  0.500000  \n",
              "0  0.839225       0.696387  0.802480  0.843304  "
            ],
            "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 tr th {\n",
              "        text-align: left;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th>Model</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Train</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Test</th>\n",
              "      <th colspan=\"3\" halign=\"left\">Holdout Sample</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "      <th>Recall</th>\n",
              "      <th>F1-Score</th>\n",
              "      <th>AUC</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>LogisticRegression</td>\n",
              "      <td>0.525296</td>\n",
              "      <td>0.618740</td>\n",
              "      <td>0.737900</td>\n",
              "      <td>0.524548</td>\n",
              "      <td>0.621112</td>\n",
              "      <td>0.738705</td>\n",
              "      <td>0.470407</td>\n",
              "      <td>0.584248</td>\n",
              "      <td>0.717754</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>RandomForest</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.500000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.500000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.500000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>XGBoost</td>\n",
              "      <td>0.695587</td>\n",
              "      <td>0.813239</td>\n",
              "      <td>0.845633</td>\n",
              "      <td>0.689061</td>\n",
              "      <td>0.798403</td>\n",
              "      <td>0.839225</td>\n",
              "      <td>0.696387</td>\n",
              "      <td>0.802480</td>\n",
              "      <td>0.843304</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 382
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Iz_C5ozrfpAr"
      },
      "source": [
        "## Export Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AIPYsqSHfm0m"
      },
      "source": [
        "import pickle"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wFnojhB5gdcy"
      },
      "source": [
        "with open('xgb_without_pipe', 'wb') as name:\n",
        "  pickle.dump(tuned_xg, name)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ezlaaIJW3xxs"
      },
      "source": [
        "## Conclusions\n",
        "Since this case is an imbalanced dataset `(non-default:77.7% ; default:22.3%)`, it's worth looking at the AUC and Recall metrics instead. Why? Especially for Recall metrics. For business purposes, we assume to minimize Type 2 (minimize False Negative -- *predict non-default (0), actual default (1)*). Hence, we use Recall metrics for optimum result.\n",
        "\n",
        "It can be seen in the table above, the model which has the highest and the most stable AUC and Recall is `XGBoost AUC: 0.839225` and `XGBoost Recall: 0.689061`.\n",
        "\n",
        "In addition, the Recall and AUC scores on the train and test are not much different. It means that we can conclude that this model is 'just right' to classify `target 1` and `target 0`, neither overfitting nor underfitting.\n",
        "\n",
        "If we look back at the features importance by Logistic Regression with Lasso regularization, the selected features seem make sense. Features which affect `loan_status` are:\n",
        "1. Percentage of Income *`('loan_percent_income')`*, \n",
        "1. Loan Amount *`('loan_amnt_WOE')`*,\n",
        "9. Employement Length *`('person_emp_length_WOE')`*,\n",
        "3. Owning Home *`('person_home_ownership_OWN')`*,\n",
        "5. Loan Grade *`('loan_grade')`*,\n",
        "4. Intention for Venture *`('loan_intent_VENTURE')`*, \n",
        "6. Intention for Education *`('loan_intent_EDUCATION')`*, \n",
        "7. Renting home *`('person_home_ownership_RENT')`*,\n",
        "2. Age *`('person_age_WOE')`*,\n",
        "1. Credit History Length *`('cb_person_cred_hist_length_WOE')`*,\n",
        "8. Intention for personal purposes *`('loan_intent_PERSONAL')`*,         \n",
        "10. Intention for home improvement *`('loan_intent_HOMEIMPROVEMENT')`*.\n",
        "\n",
        "After tuning the models and get each metrics, we could predict the holdout sample using our previous models. We see that the XGBoost algorithm shows its best performance among the others. In the holdout sample, XGBoost can reach the `AUC: 0.843304` and the `Recall: 0.696387`. It tells us that XGBoost could be our model for production, because it's not overfitted and it can predicts the holdout sample very well."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x1GDVO32fjoi"
      },
      "source": [
        "# Pipeline Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9qpZOm2tq8Ii"
      },
      "source": [
        "### Splitting Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "96l1HisnjwJi"
      },
      "source": [
        "credit = pd.read_csv('C:\\\\Users\\\\brian\\\\Documents\\\\DataScience\\\\ODP BNI x Mars\\\\trial_deployment\\\\credit_risk_dataset.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tly43nxYjxkC"
      },
      "source": [
        "# Splitting into Train and Test set\n",
        "\n",
        "X = credit.drop('loan_status', axis = 1)\n",
        "y = credit['loan_status']\n",
        "\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, \n",
        "                                                    y, \n",
        "                                                    test_size = 0.2, \n",
        "                                                    random_state = 88, \n",
        "                                                    stratify = y)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "9PF9DLD3qcBK",
        "outputId": "0b9c05e4-02e6-487f-9107-455a172ce266"
      },
      "source": [
        "X_test.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "15817          24         168000              MORTGAGE                0.0   \n",
              "826            23          74800                  RENT                1.0   \n",
              "27171          27         125000              MORTGAGE                2.0   \n",
              "17060          24          62000                  RENT                1.0   \n",
              "15272          26          60000                  RENT                7.0   \n",
              "\n",
              "             loan_intent loan_grade  loan_amnt  loan_int_rate  \\\n",
              "15817           PERSONAL          E      25000          16.45   \n",
              "826              MEDICAL          E      20000          17.04   \n",
              "27171          EDUCATION          A       8000           6.76   \n",
              "17060            MEDICAL          C       1600          13.85   \n",
              "15272  DEBTCONSOLIDATION          A      15000           6.03   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "15817                 0.15                         N   \n",
              "826                   0.23                         Y   \n",
              "27171                 0.06                         N   \n",
              "17060                 0.03                         Y   \n",
              "15272                 0.25                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "15817                           3  \n",
              "826                             2  \n",
              "27171                          10  \n",
              "17060                           4  \n",
              "15272                           3  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>15817</th>\n",
              "      <td>24</td>\n",
              "      <td>168000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>0.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>E</td>\n",
              "      <td>25000</td>\n",
              "      <td>16.45</td>\n",
              "      <td>0.15</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>826</th>\n",
              "      <td>23</td>\n",
              "      <td>74800</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>E</td>\n",
              "      <td>20000</td>\n",
              "      <td>17.04</td>\n",
              "      <td>0.23</td>\n",
              "      <td>Y</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27171</th>\n",
              "      <td>27</td>\n",
              "      <td>125000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>2.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>A</td>\n",
              "      <td>8000</td>\n",
              "      <td>6.76</td>\n",
              "      <td>0.06</td>\n",
              "      <td>N</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17060</th>\n",
              "      <td>24</td>\n",
              "      <td>62000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>1600</td>\n",
              "      <td>13.85</td>\n",
              "      <td>0.03</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15272</th>\n",
              "      <td>26</td>\n",
              "      <td>60000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>7.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>A</td>\n",
              "      <td>15000</td>\n",
              "      <td>6.03</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 431
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wyuuB7SNq-5K"
      },
      "source": [
        "### Setup The Pipeline"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kad-K_hHa-mQ"
      },
      "source": [
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.preprocessing import OneHotEncoder"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lizkbYVJWLQt"
      },
      "source": [
        "cols_categoric_nominal = ['person_home_ownership', 'loan_intent']\n",
        "cols_categoric_ordinal = ['loan_grade', 'cb_person_default_on_file']\n",
        "cols_categoric         = ['loan_grade', 'cb_person_default_on_file', 'person_home_ownership', 'loan_intent']\n",
        "cols_num               = ['person_age', 'person_income', 'person_emp_length', 'loan_amnt', 'loan_int_rate', 'loan_percent_income', 'cb_person_cred_hist_length']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4VQGX1lHa0bs"
      },
      "source": [
        "# Numeric\n",
        "pipe_num = Pipeline([\n",
        "                     ('scaler', RobustScaler())\n",
        "])\n",
        "\n",
        "# nominal categorical\n",
        "pipe_categoric_nom = Pipeline([\n",
        "                     ('encoder', OneHotEncoder(handle_unknown='ignore'))\n",
        "])\n",
        "\n",
        "pipe_preprocess = ColumnTransformer([\n",
        "                               ('categoric', pipe_categoric_nom, cols_categoric),\n",
        "                               ('numeric', pipe_num, cols_num)\n",
        "])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wQAI4iCUrBK0"
      },
      "source": [
        "### Training the Model\n",
        "using the best algorithm based on the previous training."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HGxm2Gx1VyHi"
      },
      "source": [
        "xg = XGBClassifier()\n",
        "\n",
        "pipe_xg = Pipeline([\n",
        "                    ('preps', pipe_preprocess),\n",
        "                    ('algo', xg)\n",
        "])\n",
        "\n",
        "param_xg = {\n",
        "    \"algo__n_estimators\" : [10, 20, 25, 30, 50, 100, 500],\n",
        "    \"algo__learning_rate\" : np.linspace(0.08, 4, 20),\n",
        "    \"algo__subsample\" : [0.1, 0.5, 1, 4, 7]\n",
        "}\n",
        "\n",
        "gridpipe_xg = RandomizedSearchCV(\n",
        "                                estimator = pipe_xg,\n",
        "                                param_distributions = param_xg,\n",
        "                                cv = 3,\n",
        "                                scoring = 'roc_auc',\n",
        "                                refit = True,\n",
        "                                n_jobs = -1,\n",
        "                                verbose = 3,\n",
        "                                n_iter = 100\n",
        ")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fibaIcNxi5-8",
        "outputId": "91bf0570-b6e3-43cf-bbb3-ce3d9af14337"
      },
      "source": [
        "gridpipe_xg.fit(X_train, y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fitting 3 folds for each of 100 candidates, totalling 300 fits\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\sklearn\\model_selection\\_search.py:918: UserWarning: One or more of the test scores are non-finite: [0.60472722        nan 0.65378796        nan        nan        nan\n",
            "        nan 0.67186151 0.51968804        nan 0.83774956        nan\n",
            " 0.94019314        nan 0.93943142        nan        nan        nan\n",
            "        nan        nan 0.60509978 0.90209305 0.88085011 0.76827214\n",
            "        nan        nan 0.58973293        nan 0.63513838 0.90429913\n",
            " 0.58388849        nan 0.83455553        nan        nan 0.63382831\n",
            "        nan 0.8041135  0.56303339 0.8630595         nan        nan\n",
            "        nan 0.8423999         nan 0.91452727 0.49846122 0.8822557\n",
            " 0.55033206        nan        nan        nan 0.76893662        nan\n",
            "        nan 0.90375293        nan        nan 0.89855724        nan\n",
            "        nan 0.93266136 0.9341817  0.91099285 0.68569625        nan\n",
            " 0.93621439 0.66192931 0.66117559        nan 0.93507316        nan\n",
            "        nan        nan 0.66117559 0.68370185 0.62136778 0.93543575\n",
            " 0.94159217 0.9433757         nan 0.94056602 0.61843316        nan\n",
            " 0.92715721        nan 0.63642259 0.92482583 0.70031405 0.92206584\n",
            " 0.90551492        nan        nan        nan        nan 0.58973293\n",
            "        nan 0.6326347         nan        nan]\n",
            "  warnings.warn(\n",
            "c:\\users\\brian\\appdata\\local\\programs\\python\\python39\\lib\\site-packages\\xgboost\\sklearn.py:1146: UserWarning: The use of label encoder in XGBClassifier is deprecated and will be removed in a future release. To remove this warning, do the following: 1) Pass option use_label_encoder=False when constructing XGBClassifier object; and 2) Encode your labels (y) as integers starting with 0, i.e. 0, 1, 2, ..., [num_class - 1].\n",
            "  warnings.warn(label_encoder_deprecation_msg, UserWarning)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[03:31:53] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.4.0/src/learner.cc:1095: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior.\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomizedSearchCV(cv=3,\n",
              "                   estimator=Pipeline(steps=[('preps',\n",
              "                                              ColumnTransformer(transformers=[('categoric',\n",
              "                                                                               Pipeline(steps=[('encoder',\n",
              "                                                                                                OneHotEncoder(handle_unknown='ignore'))]),\n",
              "                                                                               ['loan_grade',\n",
              "                                                                                'cb_person_default_on_file',\n",
              "                                                                                'person_home_ownership',\n",
              "                                                                                'loan_intent']),\n",
              "                                                                              ('numeric',\n",
              "                                                                               Pipeline(steps=[('scaler',\n",
              "                                                                                                RobustScaler())]),\n",
              "                                                                               ['person_age',\n",
              "                                                                                'person_income',\n",
              "                                                                                'person_em...\n",
              "                   param_distributions={'algo__learning_rate': array([0.08      , 0.28631579, 0.49263158, 0.69894737, 0.90526316,\n",
              "       1.11157895, 1.31789474, 1.52421053, 1.73052632, 1.93684211,\n",
              "       2.14315789, 2.34947368, 2.55578947, 2.76210526, 2.96842105,\n",
              "       3.17473684, 3.38105263, 3.58736842, 3.79368421, 4.        ]),\n",
              "                                        'algo__n_estimators': [10, 20, 25, 30,\n",
              "                                                               50, 100, 500],\n",
              "                                        'algo__subsample': [0.1, 0.5, 1, 4, 7]},\n",
              "                   scoring='roc_auc', verbose=3)"
            ]
          },
          "metadata": {},
          "execution_count": 412
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_Sd-MfVOjEra"
      },
      "source": [
        "pred_pipexg = gridpipe_xg.predict(X_test)\n",
        "train_pipexg = gridpipe_xg.predict(X_train)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YZZ_yBUjldk2"
      },
      "source": [
        "tuned_pipexg = gridpipe_xg.best_estimator_"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OdZ5aNtKo2OS"
      },
      "source": [
        "### Export Model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZUHO1tlynRxh"
      },
      "source": [
        "import pickle"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TIm4RCdOo48F"
      },
      "source": [
        "#### Model with Steps"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vV1JtHdmna6c"
      },
      "source": [
        "with open('xgb_piped', 'wb') as name:\n",
        "  pickle.dump(tuned_pipexg, name)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Qe8vaFMInhs-",
        "outputId": "45bdf1e7-ee36-4128-8a54-8555d5fb5dbb"
      },
      "source": [
        "print(classification_report(y_test, tuned_pipexg.predict(X_test)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.93      0.99      0.96      5095\n",
            "           1       0.96      0.73      0.83      1422\n",
            "\n",
            "    accuracy                           0.93      6517\n",
            "   macro avg       0.95      0.86      0.89      6517\n",
            "weighted avg       0.94      0.93      0.93      6517\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4UQfiS3no-mp"
      },
      "source": [
        "#### Model only and Steps only"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KOH3BPJYnj4_"
      },
      "source": [
        "final_model = gridpipe_xg.best_estimator_.named_steps['algo']\n",
        "final_pipe  = gridpipe_xg.best_estimator_.named_steps['preps']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XIqO_BHJoDP4"
      },
      "source": [
        "with open('xgb_piped_model_only', 'wb') as name: # save model\n",
        "    pickle.dump(final_model, name)\n",
        "    \n",
        "with open('xgb_piped_preps_only', 'wb') as name: # save pipeline\n",
        "    pickle.dump(final_pipe, name)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eeeBy_D3rGSL"
      },
      "source": [
        "### Predict the Single Test Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 222
        },
        "id": "CciemoYcrKae",
        "outputId": "dc3e53a3-281c-4da3-9761-f34b5e909d13"
      },
      "source": [
        "X_test.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       person_age  person_income person_home_ownership  person_emp_length  \\\n",
              "15817          24         168000              MORTGAGE                0.0   \n",
              "826            23          74800                  RENT                1.0   \n",
              "27171          27         125000              MORTGAGE                2.0   \n",
              "17060          24          62000                  RENT                1.0   \n",
              "15272          26          60000                  RENT                7.0   \n",
              "\n",
              "             loan_intent loan_grade  loan_amnt  loan_int_rate  \\\n",
              "15817           PERSONAL          E      25000          16.45   \n",
              "826              MEDICAL          E      20000          17.04   \n",
              "27171          EDUCATION          A       8000           6.76   \n",
              "17060            MEDICAL          C       1600          13.85   \n",
              "15272  DEBTCONSOLIDATION          A      15000           6.03   \n",
              "\n",
              "       loan_percent_income cb_person_default_on_file  \\\n",
              "15817                 0.15                         N   \n",
              "826                   0.23                         Y   \n",
              "27171                 0.06                         N   \n",
              "17060                 0.03                         Y   \n",
              "15272                 0.25                         N   \n",
              "\n",
              "       cb_person_cred_hist_length  \n",
              "15817                           3  \n",
              "826                             2  \n",
              "27171                          10  \n",
              "17060                           4  \n",
              "15272                           3  "
            ],
            "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>person_age</th>\n",
              "      <th>person_income</th>\n",
              "      <th>person_home_ownership</th>\n",
              "      <th>person_emp_length</th>\n",
              "      <th>loan_intent</th>\n",
              "      <th>loan_grade</th>\n",
              "      <th>loan_amnt</th>\n",
              "      <th>loan_int_rate</th>\n",
              "      <th>loan_percent_income</th>\n",
              "      <th>cb_person_default_on_file</th>\n",
              "      <th>cb_person_cred_hist_length</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>15817</th>\n",
              "      <td>24</td>\n",
              "      <td>168000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>0.0</td>\n",
              "      <td>PERSONAL</td>\n",
              "      <td>E</td>\n",
              "      <td>25000</td>\n",
              "      <td>16.45</td>\n",
              "      <td>0.15</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>826</th>\n",
              "      <td>23</td>\n",
              "      <td>74800</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>E</td>\n",
              "      <td>20000</td>\n",
              "      <td>17.04</td>\n",
              "      <td>0.23</td>\n",
              "      <td>Y</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>27171</th>\n",
              "      <td>27</td>\n",
              "      <td>125000</td>\n",
              "      <td>MORTGAGE</td>\n",
              "      <td>2.0</td>\n",
              "      <td>EDUCATION</td>\n",
              "      <td>A</td>\n",
              "      <td>8000</td>\n",
              "      <td>6.76</td>\n",
              "      <td>0.06</td>\n",
              "      <td>N</td>\n",
              "      <td>10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17060</th>\n",
              "      <td>24</td>\n",
              "      <td>62000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>1.0</td>\n",
              "      <td>MEDICAL</td>\n",
              "      <td>C</td>\n",
              "      <td>1600</td>\n",
              "      <td>13.85</td>\n",
              "      <td>0.03</td>\n",
              "      <td>Y</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15272</th>\n",
              "      <td>26</td>\n",
              "      <td>60000</td>\n",
              "      <td>RENT</td>\n",
              "      <td>7.0</td>\n",
              "      <td>DEBTCONSOLIDATION</td>\n",
              "      <td>A</td>\n",
              "      <td>15000</td>\n",
              "      <td>6.03</td>\n",
              "      <td>0.25</td>\n",
              "      <td>N</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ]
          },
          "metadata": {},
          "execution_count": 434
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LzVeYAAnqiCi",
        "outputId": "78b95af1-d2a4-474e-ab3f-7389bd464037"
      },
      "source": [
        "X_test.iloc[0].to_json()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'{\"person_age\":24,\"person_income\":168000,\"person_home_ownership\":\"MORTGAGE\",\"person_emp_length\":0.0,\"loan_intent\":\"PERSONAL\",\"loan_grade\":\"E\",\"loan_amnt\":25000,\"loan_int_rate\":16.45,\"loan_percent_income\":0.15,\"cb_person_default_on_file\":\"N\",\"cb_person_cred_hist_length\":3}'"
            ]
          },
          "metadata": {},
          "execution_count": 473
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TNbhTII2pzSs",
        "outputId": "b008875f-a193-48ad-f297-dc4dde4eea8b"
      },
      "source": [
        "# MODEL WITH PIPELINED\n",
        "# predicting single data\n",
        "\n",
        "data_pred = pd.DataFrame(data = [X_test.iloc[0].values],\n",
        "                         columns = ['person_age', 'person_income', 'person_home_ownership',\n",
        "                                    'person_emp_length', 'loan_intent', 'loan_grade', 'loan_amnt',\n",
        "                                    'loan_int_rate', 'loan_percent_income', 'cb_person_default_on_file',\n",
        "                                    'cb_person_cred_hist_length'])\n",
        "\n",
        "pred = tuned_pipexg.predict(data_pred)[0]\n",
        "proba = tuned_pipexg.predict_proba(data_pred)\n",
        "\n",
        "\n",
        "if pred == 0:\n",
        "  pred = \"Non-default\"\n",
        "  proba = f\"{round(proba[0][0]*100, 2)}%\"\n",
        "  print(proba, pred)\n",
        "else:\n",
        "  pred =  \"Default\"\n",
        "  proba = f\"{round(proba[0][1]*100, 2)}%\"\n",
        "  print(proba, pred)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "87.76% Non-default\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7-5i2Kfgr99o"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}