notebooks/python/Lab_4_Preprocessing.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/CBravoR/AdvancedAnalyticsLabs/blob/master/notebooks/python/Lab_4_Preprocessing.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "twKpLeRhDeyQ"
      },
      "source": [
        "# Data Preprocessing\n",
        "\n",
        "In this lab, we will set up a **data pipeline** in order to leave the data ready for analysis. Data preprocessing tends to be, by far, the most time-consuming step of the data science process. Errors in this step propagate to the model, so it is really important we do this correctly.\n",
        "\n",
        "The goal of this step is to leave the date ready to apply models to it. \n",
        "\n",
        "Every problem has its own set of data preprocessing functions to apply it to, but we will focus on the ones most common in classification models. In general we want to:\n",
        "\n",
        "1. Eliminate redundant variables.\n",
        "2. Treat null values.\n",
        "3. Treat outliers.\n",
        "4. Remove correlated features.\n",
        "\n",
        "For this goal we will use the excellent [```scikit-learn```](https://scikit-learn.org/stable/) package, which comes with most data-intensive operations. We will also use today (and during the rest of the module) the [```pandas```](https://pandas.pydata.org/) package, which allows for data handling in general."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tTKdVYZLGm-b"
      },
      "source": [
        "## Reading the data\n",
        "\n",
        "Let's download the Bankloan data. I have slightly modified it removing the PD and LGD information (as we will now estimate our own). As a reminder, it has the following variables:\n",
        "\n",
        "- Customer: ID, or unique label, of the borrower (NOT predictive).\n",
        "- Age: Age of the borrower in years.\n",
        "- Education: Maximum education level the borrower reached.\n",
        "1: Complete primary. 2: Completed Secondary. 3: Incomplete Higher Ed. 4: Complete Higher Ed. 5: With postgraduate studies (complete MSc or PhD).\n",
        "- Employ: Years at current job.\n",
        "- Address: Years at current address.\n",
        "- Income: Income in ‘000s USD.\n",
        "- Leverage: Debt/Income Ratio.\n",
        "- CredDebt: Credit card standing debt.\n",
        "- OthDebt: Other debt in ‘000s USD.\n",
        "- MonthlyLoad: Monthly percentage from salary used to repay debts.\n",
        "- Default: 1 If default has occurred, 0 if not (Target variable).\n",
        "\n",
        "We will download it now directly from a link, using the more ubiquitous [```wget```](http://www.gnu.org/software/wget/) command. The command is\n",
        "\n",
        "```\n",
        "wget [-options] path\n",
        "```\n",
        "\n",
        "We need to add the options ```--no-check-certificate``` and ```--output-document=FILENAME``` so it downloads ok."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g3v98YWoGla0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7c5f8bcc-f550-4341-da4e-4104c11947fe"
      },
      "source": [
        "!wget --no-check-certificate --output-document=Bankloan.csv 'https://docs.google.com/spreadsheets/d/1nUJ1fA5f1VeMvulknpsvxpy0GW3CekNnhgeLRK0WlDI/export?gid=1016776666&format=csv'"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2023-03-27 15:41:14--  https://docs.google.com/spreadsheets/d/1nUJ1fA5f1VeMvulknpsvxpy0GW3CekNnhgeLRK0WlDI/export?gid=1016776666&format=csv\n",
            "Resolving docs.google.com (docs.google.com)... 142.251.2.113, 142.251.2.102, 142.251.2.138, ...\n",
            "Connecting to docs.google.com (docs.google.com)|142.251.2.113|:443... connected.\n",
            "HTTP request sent, awaiting response... 307 Temporary Redirect\n",
            "Location: https://doc-0o-8s-sheets.googleusercontent.com/export/mq6he3r7ig44qobar1fsg51390/ej6psm76todsrtmbav85t6tgu8/1679931675000/102123285539774178265/*/1nUJ1fA5f1VeMvulknpsvxpy0GW3CekNnhgeLRK0WlDI?gid=1016776666&format=csv [following]\n",
            "Warning: wildcards not supported in HTTP.\n",
            "--2023-03-27 15:41:15--  https://doc-0o-8s-sheets.googleusercontent.com/export/mq6he3r7ig44qobar1fsg51390/ej6psm76todsrtmbav85t6tgu8/1679931675000/102123285539774178265/*/1nUJ1fA5f1VeMvulknpsvxpy0GW3CekNnhgeLRK0WlDI?gid=1016776666&format=csv\n",
            "Resolving doc-0o-8s-sheets.googleusercontent.com (doc-0o-8s-sheets.googleusercontent.com)... 142.250.141.132, 2607:f8b0:4023:c0b::84\n",
            "Connecting to doc-0o-8s-sheets.googleusercontent.com (doc-0o-8s-sheets.googleusercontent.com)|142.250.141.132|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: unspecified [text/csv]\n",
            "Saving to: ‘Bankloan.csv’\n",
            "\n",
            "Bankloan.csv            [ <=>                ]  65.50K  --.-KB/s    in 0.006s  \n",
            "\n",
            "2023-03-27 15:41:16 (9.88 MB/s) - ‘Bankloan.csv’ saved [67077]\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_ExNpn4rMwfT"
      },
      "source": [
        "To check what we downloaded we can use the ```head``` OS command."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oz4xRtSDM20o",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b6384df6-9104-47a6-9e1f-355346ef8d4d"
      },
      "source": [
        "!head Bankloan.csv"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "customer,Age,Education,Employ,Address,Income,Leverage,Creddebt,OthDebt,MonthlyLoad,Default\r\n",
            "10012,28,Med,7,2,44,17.7,2.99,4.8,0.58,0\r\n",
            "10017,64,Posg,34,17,116,14.7,5.05,12,0.27,0\r\n",
            "10030,40,Bas,20,12,61,4.8,1.04,1.89,0.13,0\r\n",
            "10039,30,Bas,11,3,27,34.5,1.75,7.56,1.62,0\r\n",
            "10069,25,Bas,2,2,30,22.4,0.76,5.96,0.97,1\r\n",
            "10071,35,Bas,2,9,38,10.9,1.46,2.68,0.4,1\r\n",
            "10096,26,SupInc,2,4,38,11.9,0.95,3.57,0.43,1\r\n",
            "10128,25,Bas,4,2,30,14.4,1.05,3.27,0.62,0\r\n",
            "10129,65,SupCom,29,14,189,5,3.36,6.09,0.08,0\r\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "k5sgZk3lMtIK"
      },
      "source": [
        "Now we will use Pandas to read the CSV file. The  function to do so is [```read_csv```](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.read_csv.html). We will store the results in a variable named ```bankloan_data```."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Wj-Yl2t5JW0Y"
      },
      "source": [
        "import pandas as pd\n",
        "\n",
        "bankloan_data = pd.read_csv('Bankloan.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i2A_hACkDcxY"
      },
      "source": [
        "As a reminder, here are the summary statistics of the variables."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XpypXLF6Nm9B",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "outputId": "3969480d-98f8-4cb4-8131-f7adfbf277ac"
      },
      "source": [
        "bankloan_data.describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            customer          Age       Employ      Address       Income  \\\n",
              "count    1500.000000  1500.000000  1500.000000  1499.000000  1499.000000   \n",
              "mean   257714.548000    34.174000     6.952000     6.309540    59.615077   \n",
              "std    139555.147719    13.142306     8.977644     6.047563    67.144768   \n",
              "min     10012.000000    18.000000     0.000000     0.000000    12.000000   \n",
              "25%     98638.750000    24.000000     0.000000     2.000000    27.000000   \n",
              "50%    315991.500000    31.000000     4.000000     5.000000    40.000000   \n",
              "75%    370701.500000    42.000000    10.000000     9.000000    64.000000   \n",
              "max    453777.000000    79.000000    63.000000    34.000000  1079.000000   \n",
              "\n",
              "            Leverage     Creddebt      OthDebt  MonthlyLoad      Default  \n",
              "count    1499.000000  1499.000000  1489.000000  1500.000000  1500.000000  \n",
              "mean      677.046631     1.936197     3.872257     0.365447     0.365333  \n",
              "std     25828.218141     2.974485     5.343150     0.293937     0.481684  \n",
              "min         0.100000     0.000000     0.100000     0.000000     0.000000  \n",
              "25%         4.800000     0.420000     1.120000     0.150000     0.000000  \n",
              "50%         8.500000     0.990000     2.240000     0.290000     0.000000  \n",
              "75%        13.600000     2.200000     4.610000     0.500000     1.000000  \n",
              "max    999999.000000    35.970000    63.470000     2.150000     1.000000  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-e9c4545b-7182-4935-adea-6c26dd8bed5e\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>customer</th>\n",
              "      <th>Age</th>\n",
              "      <th>Employ</th>\n",
              "      <th>Address</th>\n",
              "      <th>Income</th>\n",
              "      <th>Leverage</th>\n",
              "      <th>Creddebt</th>\n",
              "      <th>OthDebt</th>\n",
              "      <th>MonthlyLoad</th>\n",
              "      <th>Default</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1499.000000</td>\n",
              "      <td>1499.000000</td>\n",
              "      <td>1499.000000</td>\n",
              "      <td>1499.000000</td>\n",
              "      <td>1489.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>257714.548000</td>\n",
              "      <td>34.174000</td>\n",
              "      <td>6.952000</td>\n",
              "      <td>6.309540</td>\n",
              "      <td>59.615077</td>\n",
              "      <td>677.046631</td>\n",
              "      <td>1.936197</td>\n",
              "      <td>3.872257</td>\n",
              "      <td>0.365447</td>\n",
              "      <td>0.365333</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>139555.147719</td>\n",
              "      <td>13.142306</td>\n",
              "      <td>8.977644</td>\n",
              "      <td>6.047563</td>\n",
              "      <td>67.144768</td>\n",
              "      <td>25828.218141</td>\n",
              "      <td>2.974485</td>\n",
              "      <td>5.343150</td>\n",
              "      <td>0.293937</td>\n",
              "      <td>0.481684</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>10012.000000</td>\n",
              "      <td>18.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>12.000000</td>\n",
              "      <td>0.100000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.100000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>98638.750000</td>\n",
              "      <td>24.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>27.000000</td>\n",
              "      <td>4.800000</td>\n",
              "      <td>0.420000</td>\n",
              "      <td>1.120000</td>\n",
              "      <td>0.150000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>315991.500000</td>\n",
              "      <td>31.000000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>5.000000</td>\n",
              "      <td>40.000000</td>\n",
              "      <td>8.500000</td>\n",
              "      <td>0.990000</td>\n",
              "      <td>2.240000</td>\n",
              "      <td>0.290000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>370701.500000</td>\n",
              "      <td>42.000000</td>\n",
              "      <td>10.000000</td>\n",
              "      <td>9.000000</td>\n",
              "      <td>64.000000</td>\n",
              "      <td>13.600000</td>\n",
              "      <td>2.200000</td>\n",
              "      <td>4.610000</td>\n",
              "      <td>0.500000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>453777.000000</td>\n",
              "      <td>79.000000</td>\n",
              "      <td>63.000000</td>\n",
              "      <td>34.000000</td>\n",
              "      <td>1079.000000</td>\n",
              "      <td>999999.000000</td>\n",
              "      <td>35.970000</td>\n",
              "      <td>63.470000</td>\n",
              "      <td>2.150000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e9c4545b-7182-4935-adea-6c26dd8bed5e')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-e9c4545b-7182-4935-adea-6c26dd8bed5e button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-e9c4545b-7182-4935-adea-6c26dd8bed5e');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wcp4xeZwpuqi",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b6ce5285-2b39-4d4e-e5e7-9012ab82e858"
      },
      "source": [
        "bankloan_data.columns"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['customer', 'Age', 'Education', 'Employ', 'Address', 'Income',\n",
              "       'Leverage', 'Creddebt', 'OthDebt', 'MonthlyLoad', 'Default'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-m5-Xr8lP5Gm"
      },
      "source": [
        "We can see there are a few null values in different variables and apparently an invalid outlier in Leverage.  Let's visualize the dataset using seaborn to get an idea of the distribution."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dnkTHoUiP4fj"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import numpy as np\n",
        "%matplotlib inline"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Oc-0gLkpPTgu",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "e0d3701a-b3b1-4923-cee2-6d322e97b6f2"
      },
      "source": [
        "sns.set(color_codes=True)\n",
        "\n",
        "for col_id in bankloan_data.columns[np.r_[1,3:9]]:\n",
        "    sns.displot(data = bankloan_data, x = col_id, hue = \"Default\", kind = 'kde')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAFcCAYAAAD1Qqj3AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAAsTAAALEwEAmpwYAAAvcUlEQVR4nO3df1xUdaL/8RczoplgChdwSMra7epsys3yx6UiS0EwwWEtojUrc8Navblpt+TbFkiphdu6pUFtt9J1t70Z7k0SiCWyVqkts/VCj9ByvRAlAxjETU1Fh/P9w5rbhMIYnIPQ+/l4+JCZz+eceZ8eD3x3PmfmTIBhGAYiIiLdzNbTAUREpG9SwYiIiClUMCIiYgoVjIiImEIFIyIiplDBiIiIKfr1dIDeoKnpIG1tXXs399ChZ/PFF191UyLrKLe1lNs63ZU5LCy4G9L0TTqDsUi/fvaejvC9KLe1lNs6vTFzb6OCERERU6hgRETEFCoYERExhQpGRERMoYIRERFTqGBERMQUKhgRETGFZQVTXV1NWloaCQkJpKWlUVNT026Ox+MhOzubuLg44uPjyc/P947l5uYyffp0kpOTmTlzJtu2bfOOHT58mLvvvpv4+HgSExN54403/BoTERHzWPZJ/qysLGbNmoXL5aKgoIDMzEzWr1/vM2fz5s3U1tZSWlpKS0sLKSkpxMTEMHz4cKKjo5k7dy4DBw5k9+7dzJ49m/Lycs466yyee+45goKCeO2116ipqeGmm26itLSUQYMGdTgmIiLmseQMpqmpiaqqKpKSkgBISkqiqqqK5uZmn3nFxcWkpqZis9kICQkhLi6OkpISAGJjYxk4cCAAI0eOxDAMWlpaAHj11VdJS0sDYMSIEYwePZqtW7d2OiYiIuaxpGDcbjcRERHY7SduzWC32wkPD8ftdrebFxkZ6X3scDior69vt79NmzZx3nnnMWzYMADq6uo499xzT7pdR2MiImKeXnezy+3bt/PEE0/w/PPPW/aaoaFBXdre/ads3IBjVlb3BLJYb72Zn3Jbqzfm7o2ZexNLCsbhcNDQ0IDH48Fut+PxeGhsbMThcLSbV1dXR3R0NND+jGbnzp3ce++95OXlceGFF3qfj4yMZN++fYSEhHi3mzhxYqdj/urq3ZQPV1cCsH//ge+9j54SFhas3BZSbut0V2aV1KlZskQWGhqK0+mksLAQgMLCQpxOp/cf/W8kJiaSn59PW1sbzc3NlJWVkZCQAEBlZSWLFi1i9erVXHzxxe2227BhAwA1NTV88MEHxMbGdjomIiLmCTAMo2tfdOKnvXv3kpGRwZdffsngwYPJycnhwgsvJD09nYULFzJmzBg8Hg8PPfQQb731FgDp6eneC/TXXXcd+/btIyIiwrvPlStXMnLkSL766isyMjLYtWsXNpuNe++9l7i4OIAOx/zV1TOYA8/MASB43rrvvY+e0hv/zxSU22q9MbfOYMxnWcH0ZiqY3vUPByi31XpjbhWM+fRJfhERMYUKRkRETKGCERERU6hgRETEFCoYERExhQpGRERMoYIRERFTqGBERMQUKhgRETGFCkZEREyhghEREVOoYERExBQqGBERMYUKRkRETKGCERERU6hgRETEFCoYERExhQpGRERMoYIRERFTqGBERMQUKhgRETGFCkZEREyhghEREVOoYERExBQqGBERMYUKRkRETKGCERERU1hWMNXV1aSlpZGQkEBaWho1NTXt5ng8HrKzs4mLiyM+Pp78/HzvWHl5OTNnzmT06NHk5OT4bHfffffhcrm8f0aNGsXrr78OwJo1a4iJifGOZWdnm3qcIiJyQj+rXigrK4tZs2bhcrkoKCggMzOT9evX+8zZvHkztbW1lJaW0tLSQkpKCjExMQwfPpyoqCiWL19OSUkJra2tPtutXLnS+/Pu3bu59dZbiY2N9T6XkpLCkiVLzD1AERHxYckZTFNTE1VVVSQlJQGQlJREVVUVzc3NPvOKi4tJTU3FZrMREhJCXFwcJSUlAJx//vk4nU769eu4Ezdu3EhycjL9+/c352BERMQvlpzBuN1uIiIisNvtANjtdsLDw3G73YSEhPjMi4yM9D52OBzU19f7/Tqtra1s3ryZdevW+TxfVFREeXk5YWFh3HXXXYwdO/a08oeGBp3W/O868PXfYWHBXdpPT1Fuaym3dXpj5t7EsiUyK5SVlREZGYnT6fQ+d+ONN3LnnXcSGBjIW2+9xfz58ykuLmbo0KF+77ep6SBtbUaX8+3ff6DzSWeYsLBg5baQclunuzKrpE7NkiUyh8NBQ0MDHo8HOHExv7GxEYfD0W5eXV2d97Hb7WbYsGF+v86f//xnrrvuOp/nwsLCCAwMBOCKK67A4XCwZ8+e73soIiLiJ0sKJjQ0FKfTSWFhIQCFhYU4nU6f5TGAxMRE8vPzaWtro7m5mbKyMhISEvx6jfr6et5//32Sk5N9nm9oaPD+vGvXLvbt28cFF1zQxSMSEZHOWLZEtnTpUjIyMsjLy2Pw4MHetxqnp6ezcOFCxowZg8vloqKigqlTpwKwYMECoqKiANixYweLFy/m4MGDGIZBUVERy5cv975b7OWXX+aaa67hnHPO8XndVatW8eGHH2Kz2QgMDGTlypWEhYVZddgiIj9YAYZhdP3iQh/X1WswB56ZA0DwvHXdE8hCvXFtHZTbar0xt67BmE+f5BcREVOoYERExBQqGBERMYUKRkRETKGCERERU6hgRETEFCoYERExhQpGRERMoYIRERFTqGBERMQUKhgRETGFCkZEREyhghEREVOoYERExBQqGBERMYUKRkRETKGCERERU6hgRETEFCoYERExhQpGRERMoYIRERFTqGBERMQUKhgRETGFCkZEREyhghEREVOoYERExBSWFUx1dTVpaWkkJCSQlpZGTU1Nuzkej4fs7Gzi4uKIj48nPz/fO1ZeXs7MmTMZPXo0OTk5PtutWbOGmJgYXC4XLpeL7Oxs79jhw4e5++67iY+PJzExkTfeeMO0YxQRkf/Tz6oXysrKYtasWbhcLgoKCsjMzGT9+vU+czZv3kxtbS2lpaW0tLSQkpJCTEwMw4cPJyoqiuXLl1NSUkJra2u7/aekpLBkyZJ2zz/33HMEBQXx2muvUVNTw0033URpaSmDBg0y7VhFRMSiM5impiaqqqpISkoCICkpiaqqKpqbm33mFRcXk5qais1mIyQkhLi4OEpKSgA4//zzcTqd9Ot3ep346quvkpaWBsCIESMYPXo0W7du7YajEhGRjlhSMG63m4iICOx2OwB2u53w8HDcbne7eZGRkd7HDoeD+vp6v16jqKiI5ORk5s6dy86dO73P19XVce65536vfYqIyPdn2RKZmW688UbuvPNOAgMDeeutt5g/fz7FxcUMHTq0W/YfGhrUpe0PfP13WFhw18P0AOW2lnJbpzdm7k0sKRiHw0FDQwMejwe73Y7H46GxsRGHw9FuXl1dHdHR0UD7M5pTCQsL8/58xRVX4HA42LNnDxMmTCAyMpJ9+/YREhLi3efEiRNPK39T00Ha2ozT2uZk9u8/0PmkM0xYWLByW0i5rdNdmVVSp2bJElloaChOp5PCwkIACgsLcTqd3n/0v5GYmEh+fj5tbW00NzdTVlZGQkJCp/tvaGjw/rxr1y727dvHBRdc4N3nhg0bAKipqeGDDz4gNja2uw5NREROwbIlsqVLl5KRkUFeXh6DBw/2vtU4PT2dhQsXMmbMGFwuFxUVFUydOhWABQsWEBUVBcCOHTtYvHgxBw8exDAMioqKWL58ObGxsaxatYoPP/wQm81GYGAgK1eu9J7V/PznPycjI4P4+HhsNhsPPfQQQUFdW/ISEZHOBRiG0fW1nz6uq0tkB56ZA0DwvHXdE8hCvXHpA5Tbar0xt5bIzKdP8ouIiClUMCIiYgoVjIiImEIFIyIiplDBiIiIKVQwIiJiChWMiIiYQgUjIiKmUMGIiIgpVDAiImIKFYyIiJhCBSMiIqZQwYiIiClUMCIiYgoVjIiImEIFIyIiplDBiIiIKVQwIiJiChWMiIiYQgUjIiKmUMGIiIgpVDAiImIKFYyIiJhCBSMiIqZQwYiIiClUMCIiYgoVjIiImMKygqmuriYtLY2EhATS0tKoqalpN8fj8ZCdnU1cXBzx8fHk5+d7x8rLy5k5cyajR48mJyfHZ7vc3FymT59OcnIyM2fOZNu2bd6xjIwMrrrqKlwuFy6Xi6eeesq0YxQRkf/Tz6oXysrKYtasWbhcLgoKCsjMzGT9+vU+czZv3kxtbS2lpaW0tLSQkpJCTEwMw4cPJyoqiuXLl1NSUkJra6vPdtHR0cydO5eBAweye/duZs+eTXl5OWeddRYA8+bNY/bs2VYdqoiIYNEZTFNTE1VVVSQlJQGQlJREVVUVzc3NPvOKi4tJTU3FZrMREhJCXFwcJSUlAJx//vk4nU769WvfibGxsQwcOBCAkSNHYhgGLS0t5h6UiIh0yJIzGLfbTUREBHa7HQC73U54eDhut5uQkBCfeZGRkd7HDoeD+vr603qtTZs2cd555zFs2DDvc2vXrmXDhg1ERUVxzz338KMf/ei09hkaGnRa87/rwNd/h4UFd2k/PUW5raXc1umNmXsTy5bIrLB9+3aeeOIJnn/+ee9zixYtIiwsDJvNxqZNm7j99tspKyvzlp0/mpoO0tZmdDnf/v0HOp90hgkLC1ZuCym3dbors0rq1CxZInM4HDQ0NODxeIATF/MbGxtxOBzt5tXV1Xkfu91unzORjuzcuZN7772X3NxcLrzwQu/zERER2GwnDjMlJYWvvvrqtM+KRETk9FlSMKGhoTidTgoLCwEoLCzE6XT6LI8BJCYmkp+fT1tbG83NzZSVlZGQkNDp/isrK1m0aBGrV6/m4osv9hlraGjw/rxt2zZsNhsRERHdcFQiItIRy5bIli5dSkZGBnl5eQwePNj7VuP09HQWLlzImDFjcLlcVFRUMHXqVAAWLFhAVFQUADt27GDx4sUcPHgQwzAoKipi+fLlxMbGkp2dzZEjR8jMzPS+3sqVKxk5ciRLliyhqamJgIAAgoKCeOqpp076RgEREeleAYZhdP3iQh/X1WswB56ZA0DwvHXdE8hCvXFtHZTbar0xt67BmE+f5BcREVOoYERExBQqGBERMYUKRkRETOF3wZSVlXH8+HEzs4iISB/id8GsXr2aK6+8koceeoiKigozM4mISB/gd8G88sorrFu3jgEDBnDXXXeRkJBAXl4en332mZn5RESklzqtazCjRo1iyZIl/PWvfyUrK4uSkhLi4+O56aabeOWVV2hrazMrp4iI9DKn/ZH22tpaXnnlFV555RUCAgJYuHAhDoeDF154gdLSUp588kkzcoqISC/jd8G88MILFBQU8MknnzBt2jRWrlzJJZdc4h1PSEjg8ssvNyOjiIj0Qn4XzNatW7ntttuYMmUK/fv3bzc+cOBA1qxZ063hRESk9/L7GsyECROYNm1au3JZu3at9+crr7yy+5KJiEiv5nfB5ObmnvT5p556qtvCiIhI39HpEtnf/vY34MSXhL3zzjt8++bLn332GYMGDTIvnYiI9FqdFsyvfvUrAFpbW7n//vu9zwcEBBAWFsYDDzxgXjoREem1Oi2YLVu2AHDfffexcuVK0wOJiEjf4Pc1GJWLiIicjg7PYKZNm8arr74KwKRJkwgICDjpvDfffLPbg4mISO/WYcE8/PDD3p9//etfmx5GRET6jg4LZty4cd6fJ0yYYHoYERHpO/y+BrN27Vp27doFwH//939z9dVXM3nyZHbu3GlaOBER8d+f/vQnLr/8csaOHcsXX3zRpX1NnjyZt99+u0v78PtWMevWreP6668H4De/+Q1z5sxh0KBBrFixgvz8/C6FEBGRE/+of/7559jtdux2Oz/+8Y9xuVykpaVhs3V8PnDs2DEeffRRXnrpJUaNGtWtudasWcMnn3zCY489dlrb+V0wBw4cIDg4mIMHD/LRRx+xbt067HY7OTk5px1WRERO7umnn+byyy/nwIEDbN++neXLl1NZWckjjzzS4XZNTU0cPXqUH//4xxYl7ZzfS2QOh4O///3vFBcXM27cOOx2OwcPHsRut5uZT0TkByk4OJgpU6bw+OOP8/LLL/Pxxx/T2tpKTk4OV199NZdffjmZmZkcOXKE6upqEhMTARg/fjy33HILAMuWLWPSpElceumlzJw5kx07dnj3n5GRwW9/+1vv43fffZerrrqqXY6tW7fyu9/9jldffZWxY8cyY8YMv4/B7zOY++67j4ULF9K/f39Wr14NwBtvvMGYMWP8fjERETk90dHRDBs2jB07drBx40Zqa2vZtGkT/fr149///d/Jzc3lnnvuobCwkClTpvDee+/Rr9+Jf9rHjBnDggULCA4OZv369fzyl79ky5YtDBgwwO/Xv+qqq7jjjju+1xKZ32cwkyZNory8nC1btjB69GgAEhMTdbNLERGThYeH87//+7+89NJL3H///QwZMoSgoCDuuOMOioqKTrmdy+Vi6NCh9OvXj7lz59La2kp1dbVluU/rGy0PHDhAdXU1hw4d8nk+Jiam022rq6vJyMigpaWFIUOGkJOTw4gRI3zmeDweli1bxrZt2wgICGDevHmkpqYCUF5ezqpVq/j444+5+eabWbJkiV/bdTQmItIbNDQ04PF4OHz4MDNnzvQ+bxhGh19V/9xzz7Fx40YaGxsJCAjg4MGDXX532enwu2D+67/+i4ceeoizzz6bs846y/t8QEAAr7/+eqfbZ2VlMWvWLFwuFwUFBWRmZrJ+/XqfOZs3b6a2tpbS0lJaWlpISUkhJiaG4cOHExUVxfLlyykpKaG1tdXv7ToaExE501VWVtLQ0EBcXBz/8R//QVFREREREZ1ut2PHDp599lnWrVvHRRddhM1mY/z48d474g8cOJAjR45453/++een3Nep7uLSGb+XyH7729/yxBNP8Pbbb7NlyxbvH3/KpampiaqqKpKSkgBISkqiqqqK5uZmn3nFxcWkpqZis9kICQkhLi6OkpISAM4//3ycTqd3bdHf7ToaExE5Ux08eJA33niDxYsXM2PGDEaNGkVqaiorVqygqakJOHFms23btpNuf+jQIex2OyEhIRw/fpwnn3ySgwcPesedTid//etfaWlpYf/+/fz+978/ZZbQ0FD27dvX4dnSyfhdMB6P53t/Y6Xb7SYiIsL7jjO73U54eDhut7vdvMjISO9jh8NBfX29X/s/1Xbfd58iIj3hzjvvZOzYsUyaNImnn36a2267zfsW5XvvvZfzzz+fG264gUsvvZQ5c+ac8prKlVdeSWxsLAkJCUyePJkBAwbgcDi84y6Xi1GjRjF58mTmzp3Ltddee8pM37xDbeLEifz0pz/1+1j8XiJLT0/nqaeeYv78+Z1+4KevCQ0N6tL2B77+OywsuOtheoByW0u5rXOmZf7m61FOZcCAASxevJjFixe3Gxs+fDgfffSR97HdbueRRx7x+fxMenq6z74ef/xxn33MmTPnpFmGDh3Kf/7nf/p7GF6n9Un+zz//nGeffZYhQ4b4jHV2N2WHw+G9SGW32/F4PDQ2Nvq06Tfz6urqiI6OBtqffXS0/1Nt9333+W1NTQdpazM6n9iJ/fsPdD7pDBMWFqzcFlJu63RX5jOtpM4kfhdMV+6mHBoaitPppLCwEJfLRWFhIU6nk5CQEJ95iYmJ5OfnM3XqVFpaWigrK+OFF17odP8dbfd99ykiIl3jd8F09W7KS5cuJSMjg7y8PAYPHuy9xUx6ejoLFy5kzJgxuFwuKioqmDp1KgALFiwgKioKOPGOiMWLF3Pw4EEMw6CoqIjly5cTGxvb4XYdjYmIiHkCjG/es9aJ1tZWcnNzKSwspKWlhffff5/y8nJqamqYPXu22Tl7VFeXyA48MweA4HnruieQhXrj0gcot9V6Y24tkZnP76v1K1as4OOPP+axxx7zvif6oosu+l4XfkREpO/ze4msrKyM0tJSzj77bO+7yCIiImhoaDAtnIiI9F5+n8EEBgbi8Xh8nmtubm73jjIRERE4jYJJTExkyZIlfPrppwA0Njby0EMPMX36dNPCiYhI7+V3wSxatIioqChmzJjBl19+SUJCAmFhYSxYsMDMfCIi0gOqq6tJS0sjISGBtLQ0ampqTnsffl+Dqa2t5YILLuCOO+7A4/EQFxfHyJEjT/sFRUTkzOfPDYo702nBGIbB/fffz6ZNmxg2bBjh4eE0NDSQm5uLy+VixYoV3/tOmyIi4mvLjlpe215ryr7jJ5zH5HHndTrvmxsUr127Fjhxg+KHH36Y5ubmdh+Q70inBbNhwwa2b9/Ohg0bvLdbgRO3kL7nnnt48cUX+dnPfub3C4qIyJmtoxsUd2vBFBQU8MADD/iUC5z4Gs/777+f3/3udyoYEZFuMnmcf2cZvUGnF/n37t3L+PHjTzo2fvx49u7d2+2hRESk53z7BsXAKW9Q3JlOC8bj8RAUdPLb1QcFBZ32F9CIiMiZ7ds3KAZOeYPiznS6RHb8+HHeeecdTnXLsu9++FJERHq/U92g+HR0WjChoaHcf//9pxw/3UYTEZEz349+9CPy8/O7tI9OC6azb1gTERE5mR/Wdx+LiIhlVDAiImIKFYyIiJhCBSMiIqZQwYiIiClUMCIi4iMnJ4fJkyczcuRIPv744++9HxWMiIj4mDJlCi+88ALnnntul/bj9/fBiIjID8O4ceO6ZT8qGBGRM8iByjc5UGHOB9yD/2UywdFXm7Lvk9ESmYiImEJnMCIiZ5Dg6KstPcswk85gRETEFCoYERHxsWzZMq666irq6+u57bbbmD59+vfaj2VLZNXV1WRkZNDS0sKQIUPIyclhxIgRPnM8Hg/Lli1j27ZtBAQEMG/ePFJTUzsdu++++/joo4+8+/noo4/Izc1lypQprFmzhj/96U+Eh4cDcOmll5KVlWXNQYuI9EIPPPAADzzwQJf3Y1nBZGVlMWvWLFwuFwUFBWRmZrJ+/XqfOZs3b6a2tpbS0lJaWlpISUkhJiaG4cOHdzi2cuVK7z52797NrbfeSmxsrPe5lJQUlixZYtWhiogIFi2RNTU1UVVVRVJSEgBJSUlUVVXR3NzsM6+4uJjU1FRsNhshISHExcVRUlLS6di3bdy4keTkZPr372/+gYmIyClZUjBut5uIiAjsdjsAdrud8PBw3G53u3mRkZHexw6Hg/r6+k7HvtHa2srmzZu57rrrfJ4vKioiOTmZuXPnsnPnzm49NhERObk+9TblsrIyIiMjcTqd3uduvPFG7rzzTgIDA3nrrbeYP38+xcXFDB061O/9hoYGdSnXga//DgsL7tJ+eopyW0u5rdMbM/cmlhSMw+GgoaEBj8eD3W7H4/HQ2NiIw+FoN6+uro7o6GjA96ylo7Fv/PnPf2539hIWFub9+YorrsDhcLBnzx4mTJjgd/6mpoO0tRn+H/Ap7N9/oPNJZ5iwsGDltpByW6e7MqukTs2SJbLQ0FCcTieFhYUAFBYW4nQ6CQkJ8ZmXmJhIfn4+bW1tNDc3U1ZWRkJCQqdjAPX19bz//vskJyf77LOhocH7865du9i3bx8XXHCBWYcqIiJfs2yJbOnSpWRkZJCXl8fgwYPJyckBID09nYULFzJmzBhcLhcVFRVMnToVgAULFhAVFQXQ4RjAyy+/zDXXXMM555zj87qrVq3iww8/xGazERgYyMqVK33OakRExBwBhmF0fe2nj+vqEtmBZ+YAEDxvXfcEslBvXPoA5bZab8ytJTLz6ZP8IiJiChWMiIiYQgUjIiKmUMGIiIgpVDAiImIKFYyIiJhCBSMiIqZQwYiIiClUMCIiYgoVjIiImEIFIyIiplDBiIiIKVQwIiJiChWMiIiYQgUjIiKmUMGIiIgpVDAiImIKFYyIiJhCBSMiIqZQwYiIiClUMCIiYgoVjIiImEIFIyIiplDBiIiIKVQwIiJiChWMiIiYQgUjIiKmsKxgqqurSUtLIyEhgbS0NGpqatrN8Xg8ZGdnExcXR3x8PPn5+X6NrVmzhpiYGFwuFy6Xi+zsbO/Y4cOHufvuu4mPjycxMZE33njD1OMUEZET+ln1QllZWcyaNQuXy0VBQQGZmZmsX7/eZ87mzZupra2ltLSUlpYWUlJSiImJYfjw4R2OAaSkpLBkyZJ2r/vcc88RFBTEa6+9Rk1NDTfddBOlpaUMGjTIkuMWEfmhsuQMpqmpiaqqKpKSkgBISkqiqqqK5uZmn3nFxcWkpqZis9kICQkhLi6OkpKSTsc68uqrr5KWlgbAiBEjGD16NFu3bu3mIxQRke+y5AzG7XYTERGB3W4HwG63Ex4ejtvtJiQkxGdeZGSk97HD4aC+vr7TMYCioiLKy8sJCwvjrrvuYuzYsQDU1dVx7rnnnnI7f4SGBp3W/O868PXfYWHBXdpPT1Fuaym3dXpj5t7EsiUyM914443ceeedBAYG8tZbbzF//nyKi4sZOnRot+y/qekgbW1Gl/ezf/+BziedYcLCgpXbQsptne7KrJI6NUuWyBwOBw0NDXg8HuDEBfvGxkYcDke7eXV1dd7HbrebYcOGdToWFhZGYGAgAFdccQUOh4M9e/YAEBkZyb59+066nYiImMeSggkNDcXpdFJYWAhAYWEhTqfTZ3kMIDExkfz8fNra2mhubqasrIyEhIROxxoaGrz72LVrF/v27eOCCy7wbrdhwwYAampq+OCDD4iNjTX9mEVEfugsWyJbunQpGRkZ5OXlMXjwYHJycgBIT09n4cKFjBkzBpfLRUVFBVOnTgVgwYIFREVFAXQ4tmrVKj788ENsNhuBgYGsXLmSsLAwAH7+85+TkZFBfHw8NpuNhx56iKCgrl1TERGRzgUYhtH1iwt9XFevwRx4Zg4AwfPWdU8gC/XGtXVQbqv1xty6BmM+fZJfRERMoYIRERFTqGBERMQUKhgRETGFCkZEREyhghEREVOoYERExBQqGBERMYUKRkRETKGCERERU6hgRETEFCoYERExhQpGRERMoYIRERFTqGBERMQUKhgRETGFCkZEREyhghEREVOoYERExBQqGBERMYUKRkRETKGCERERU6hgRETEFCoYERExhQpGRERMoYIRERFT9LPqhaqrq8nIyKClpYUhQ4aQk5PDiBEjfOZ4PB6WLVvGtm3bCAgIYN68eaSmpnY6lpubS3FxMTabjcDAQBYtWkRsbCwAGRkZvP322wwdOhSAxMREfvGLX1h12CIiP1iWFUxWVhazZs3C5XJRUFBAZmYm69ev95mzefNmamtrKS0tpaWlhZSUFGJiYhg+fHiHY9HR0cydO5eBAweye/duZs+eTXl5OWeddRYA8+bNY/bs2VYdqoiIYNESWVNTE1VVVSQlJQGQlJREVVUVzc3NPvOKi4tJTU3FZrMREhJCXFwcJSUlnY7FxsYycOBAAEaOHIlhGLS0tFhxaCIicgqWFIzb7SYiIgK73Q6A3W4nPDwct9vdbl5kZKT3scPhoL6+vtOxb9u0aRPnnXcew4YN8z63du1akpOTmT9/Pnv37u3WYxMRkZOzbInMCtu3b+eJJ57g+eef9z63aNEiwsLCsNlsbNq0idtvv52ysjJv2fkjNDSoS7kOfP13WFhwl/bTU5TbWsptnd6YuTexpGAcDgcNDQ14PB7sdjsej4fGxkYcDke7eXV1dURHRwO+Zy0djQHs3LmTe++9l7y8PC688ELv8xEREd6fU1JSeOSRR6ivr+fcc8/1O39T00Ha2ozTP/Dv2L//QOeTzjBhYcHKbSHltk53ZVZJnZolS2ShoaE4nU4KCwsBKCwsxOl0EhIS4jMvMTGR/Px82traaG5upqysjISEhE7HKisrWbRoEatXr+biiy/22WdDQ4P3523btmGz2XxKR0REzGHZEtnSpUvJyMggLy+PwYMHk5OTA0B6ejoLFy5kzJgxuFwuKioqmDp1KgALFiwgKioKoMOx7Oxsjhw5QmZmpvf1Vq5cyciRI1myZAlNTU0EBAQQFBTEU089Rb9+fWplUETkjBRgGEbX1376uK4ukR14Zg4AwfPWdU8gC/XGpQ9Qbqv1xtxaIjOfPskvIiKmUMGIiIgpVDAiImIKFYyIiJhCBSMiIqZQwYiIiClUMCIiYgoVjIiImEIFIyIiplDBiIiIKVQwIiJiChWMiIiYQgUjIiKmUMGIiIgpVDAiImIKFYyIiJhCBSMiIqZQwYiIiClUMCIiYgoVjIiImEIFIyIiplDBiIiIKVQwIiJiChWMiIiYQgUjIiKmUMGIiIgpVDAiImIKywqmurqatLQ0EhISSEtLo6ampt0cj8dDdnY2cXFxxMfHk5+fb+qYiIiYp59VL5SVlcWsWbNwuVwUFBSQmZnJ+vXrfeZs3ryZ2tpaSktLaWlpISUlhZiYGIYPH27KmIiImMeSgmlqaqKqqoq1a9cCkJSUxMMPP0xzczMhISHeecXFxaSmpmKz2QgJCSEuLo6SkhJuv/12U8b8ZbMF+DXviwNH+O89n3P0eButx9o4dOQYB79q5X9tP+O4YSO44EOCBvYjaGAggwYGMiDQzkn3fIqXCzjVgIkGDWri0KFWy1+3q5TbWr0x96kyX+AI5ryI4B5I1PdYUjBut5uIiAjsdjsAdrud8PBw3G63T8G43W4iIyO9jx0OB/X19aaN+Wvo0EF+zQsNDeLHI/7pJCMTT+v1RET6Al3kFxERU1hSMA6Hg4aGBjweD3DiwntjYyMOh6PdvLq6Ou9jt9vNsGHDTBsTERHzWFIwoaGhOJ1OCgsLASgsLMTpdPosjwEkJiaSn59PW1sbzc3NlJWVkZCQYNqYiIiYx7J3kS1dupSMjAzy8vIYPHgwOTk5AKSnp7Nw4ULGjBmDy+WioqKCqVOnArBgwQKioqIATBkTERHzBBiGYfR0CBER6Xt0kV9EREyhghEREVOoYERExBQqGBERMYUKpht19YaePcWf3Lm5uUyfPp3k5GRmzpzJtm3brA/6Hf7k/sb//M//8C//8i/edy/2JH9zFxcXk5ycTFJSEsnJyXz++efWBv0Of3I3NTUxb948kpOTmTZtGkuXLuX48ePWh/1aTk4OkydPZuTIkXz88ccnnXMm/k72GYZ0m5tvvtnYtGmTYRiGsWnTJuPmm29uN+fll1825s6da3g8HqOpqcmIjY01Pv30U6uj+vAn99atW42vvvrKMAzD2LVrl3HZZZcZhw8ftjTnd/mT2zAM4/jx48bs2bONxYsXG48++qiVEU/Kn9yVlZXGtGnTjMbGRsMwDOPLL780jhw5YmnO7/In97Jly7z/jVtbW43rr7/eKCoqsjTnt7333ntGXV2dcc011xgfffTRSeecib+TfYXOYLrJNzf0TEpKAk7c0LOqqorm5mafeae6+WZP8Td3bGwsAwcOBGDkyJEYhkFLS4vVcb38zQ3wzDPPcPXVVzNixAiLU7bnb+5169Yxd+5cwsLCAAgODmbAgAGW5/2Gv7kDAgI4dOgQbW1ttLa2cuzYMSIiInoiMgDjxo1rd8eQ7zrTfif7EhVMN+nohp7fndfVm292J39zf9umTZs477zzevSWO/7m3r17N+Xl5cyZM6cHUrbnb+69e/fy6aefctNNN/HTn/6UvLw8jB78yJq/uefPn091dTVXXnml989ll13WE5H9dqb9TvYlKhg5Ldu3b+eJJ57gN7/5TU9H6dSxY8d48MEHyc7O9v7D2Ft4PB4++ugj1q5dyx/+8Ae2bt1KQUFBT8fqVElJCSNHjqS8vJytW7eyY8cOnQ38gKlgukl33NCzJ/ibG2Dnzp3ce++95ObmcuGFF1od1Yc/uffv309tbS3z5s1j8uTJ/P73v+ell17iwQcf7KnYfv/3joyMJDExkf79+xMUFMSUKVOorKzsiciA/7n/+Mc/MmPGDGw2G8HBwUyePJl33323JyL77Uz7nexLVDDdpDtu6NkT/M1dWVnJokWLWL16NRdffHFPRPXhT+7IyEjeffddtmzZwpYtW7j11lu54YYbePjhh3sqtt//vZOSkigvL8cwDI4dO8Y777zDqFGjeiIy4H/u4cOHs3XrVgBaW1v529/+xkUXXWR53tNxpv1O9ik9/CaDPuUf//iHcf311xtTp041rr/+emPv3r2GYRjG7bffblRWVhqGceIdTZmZmcaUKVOMKVOmGC+++GJPRjYMw7/cM2fONCZOnGjMmDHD+2f37t09Gduv3N+2evXqM+JdZP7k9ng8xooVK4zExETj2muvNVasWGF4PJ6ejO1X7k8++cSYM2eOkZSUZEybNs1YunSpcezYsR7L/PDDDxuxsbGG0+k0Lr/8cuPaa69tl/lM/J3sK3SzSxERMYWWyERExBQqGBERMYUKRkRETKGCERERU6hgROQHxZ8bYH5XS0sLixcvJiEhgenTp/Pkk0+anLJv6NfTAURErDRlyhRuueUWbrrpJr+3ycjI4F//9V9ZtWoVcOJDvNI5ncFInzB58mTefvvtno4hvcCpboBZUVHBzTffzMyZM5k5cyZvvvkmADU1NXz88cfceuut3rnf3IRUOqYzGJEu8ng8ve5eZ+Lryy+/JCsri2eeeYbw8HAaGxu5/vrrKSws5B//+AcRERH86le/YteuXfzTP/0T99133xl/h4Izgc5gpM9qa2vjmWeeIS4ujokTJ/LLX/7S+xUDt99+O3/84x995s+YMYPS0lLgxN2Mb7vtNiZMmEBCQgLFxcXeeRkZGWRlZZGens4ll1zCu+++y5tvvklKSgqXXnopkyZNYs2aNT773rRpE9dccw0TJ04kNzfX54yro5xijZ07d/LZZ5+Rnp6Oy+UiPT2dgIAAPvnkE9ra2qioqGDmzJm8/PLLpKam8otf/KKnI/cOPX0rAZHucM011xhvvfWWz3Pr1q0zUlNTDbfbbRw9etR48MEHjUWLFhmGceJLptLS0rxz9+zZY1x22WXG0aNHjUOHDhlXXXWVsXHjRuPYsWPGhx9+aEyYMMHYs2ePYRiGsWTJEuPSSy81duzYYXg8HuPIkSPGO++8Y+zevdvweDzGrl27jJiYGOO1117z7vuSSy4x3nvvPePo0aPGo48+avzkJz/x5u0op5jn219C9sYbbxizZs066bzKykpj8uTJPs9FR0cbTU1Npmfs7XQGI33Wiy++yKJFixg2bBj9+/fn3/7t3/jLX/7C8ePHiYuLY/fu3ezbtw+AzZs3Ex8fT//+/XnzzTc599xzue666+jXrx8/+clPSEhI8Lnt/JQpU7jsssuw2WwMGDCAiRMnMnLkSGw2G6NGjWL69Ols374dOHEL+2uuuYZx48bRv39/Fi5cSEBAgF85xRpjx47lk08+4Z133vE+V1lZiWEYjB49mrPPPps9e/YA8N5773HOOecwdOjQnorba+gajPRZdXV1LFiwAJvt//4/ymaz0dTUREREBJMmTaKoqIh58+ZRWFjIsmXLANi3bx+VlZWMGzfOu53H42HGjBnex9+9SFxRUcFjjz3Gnj17OHbsGK2trSQmJgLQ2Njoc/v3gQMHMmTIEL9zSvdatmwZpaWlfP7559x2220MGTKEoqIi8vLy+PWvf82KFSs4duwYUVFRPP3009hsNlasWMH/+3//j9bWVgYOHMiTTz7p8z8JcnIqGOmzhg0bxooVK075jYpJSUk8+eSTjB8/nqNHjzJx4kTgRHmMHz+etWvX+v1a99xzD7Nnz+bZZ59lwIABLF++nC+++AKA8PBwqqurvXOPHDnic42ls5zSvR544AEeeOCBds9HR0fzhz/84aTbjBkzho0bN5odrc/REpn0GceOHePo0aPeP6mpqTz++OPeZbBvvuvjG5MmTaKuro7Vq1dz7bXXes8grr76ampqati0aRPHjh3j2LFjVFZWsnfv3lO+9qFDhzjnnHMYMGAAlZWV3u9NAUhISGDLli38/e9/p7W1lTVr1vh8/fHPfvazDnOK9FYqGOkz5s2bR3R0tPfPl19+yeTJk5k7dy5jx47lhhtu8PlWyP79+xMfH8/bb79NUlKS9/mgoCCee+45iouLiY2N5corr+Sxxx6jtbX1lK+dlZXF6tWrGTt2LLm5uUybNs07dtFFF/Hggw+yePFiYmNjOfvsswkJCaF///4A3HLLLR3mFOmt9H0wIhY7dOgQ48eP5y9/+QtRUVE9HUfENDqDEbHAli1bOHz4MF999RU5OTn88z//M8OHD+/pWCKmUsGIWOD1118nNjaW2NhYPvnkE1atWqV3IUmfpyUyERExhc5gRETEFCoYERExhQpGRERMoYIRERFTqGBERMQUKhgRETHF/wc+FgU1rfv35wAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DohsQiAl0LJq"
      },
      "source": [
        "Seaborn can do many sophisticated (and aestetically pleasing) \n",
        "graphs. Go to [the Gallery](https://seaborn.pydata.org/examples/index.html) for details and example code. \n",
        "\n",
        "There are severe outliers in OthDebt, Leverage, Creddebt and Income, but we don't know which ones are valid and which invalid. We will treat these now."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HZLgIWHqSvts"
      },
      "source": [
        "## Data Cleaning\n",
        "\n",
        "Now we can focus on cleaning the data. Let's start with the easy part: removing null values. **Remember to check when an outlier is a missing value** (invalid outliers).\n",
        "\n",
        "### Null values\n",
        "\n",
        "The core function here will be Panda's [```fillna```](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.fillna.html). This allows to replace all null values (represented by None or NaN in Python) by a certain value. This also allows to set what the replacement will be with the ```value``` argument. See the help for details.\n",
        "\n",
        "Remember the strategies to deal :\n",
        "\n",
        "1. Keep: If the null values are a category by themselves. In this case, replace by something meaningful.\n",
        "\n",
        "2. Delete: If the null values are too many **either by row or by column** then it is better to just drop the case or the variable.\n",
        "\n",
        "3. Replace: If there are only a few missings for the variable or the row (<1% total), replace by the  replace the null values by the **median** for continous variables, and the **mode** for categorical values.\n",
        "\n",
        "Let's study our dataset's null values. The [```isnull()```](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.isnull.html) function returns which elements in the dataframe are null. The [```any()```](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.any.html) function returns a list with whatever columns (default) or rows (passing ```axis = 1``` to the function) have any element with a boolean of true."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z-s-kkifWlud",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3f4f122d-64c3-4e94-e284-a8a51a9621c3"
      },
      "source": [
        "bankloan_data.isnull().any()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "customer       False\n",
              "Age            False\n",
              "Education      False\n",
              "Employ         False\n",
              "Address         True\n",
              "Income          True\n",
              "Leverage        True\n",
              "Creddebt        True\n",
              "OthDebt         True\n",
              "MonthlyLoad    False\n",
              "Default        False\n",
              "dtype: bool"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "H2Fp-LCjXv9s"
      },
      "source": [
        "We can see which columns have null values. Let's study them in further detail."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Tgc0zy6UX04R",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b88616ac-12eb-4a55-ed74-359d520d21ba"
      },
      "source": [
        "null_columns = bankloan_data.columns[bankloan_data.isnull().any()]\n",
        "bankloan_data[null_columns].isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Address      1\n",
              "Income       1\n",
              "Leverage     1\n",
              "Creddebt     1\n",
              "OthDebt     11\n",
              "dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OJ7_tgqsYLhj"
      },
      "source": [
        "Given the small number of cases for all variables except OthDebt, we can simply replace those values by the median.\n",
        "\n",
        "Let's study OthDebt cases more in detail."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9uFhZbqGYUqp",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 394
        },
        "outputId": "4c9aee2c-2e04-40ed-ebf6-864318654912"
      },
      "source": [
        "bankloan_data.loc[bankloan_data.isnull().any(axis = 1), :]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      customer  Age Education  Employ  Address  Income  Leverage  Creddebt  \\\n",
              "17       10423   19       Bas       2      0.0    31.0       0.3      0.02   \n",
              "247      71896   20    SupInc       0      0.0    21.0       1.3      0.18   \n",
              "255      72327   19       Med       0      NaN     NaN       NaN       NaN   \n",
              "293      74471   39       Bas       8      9.0    41.0       0.1      0.03   \n",
              "587     244023   31    SupCom       1      7.0    69.0       0.4      0.19   \n",
              "745     315892   22    SupInc       0      1.0    14.0       2.5      0.26   \n",
              "748     315948   41       Med      21      9.0    64.0       0.3      0.12   \n",
              "817     335635   18       Med       0      0.0    24.0       0.9      0.13   \n",
              "1194    372485   20    SupInc       0      0.0    13.0       1.1      0.05   \n",
              "1254    376801   23    SupCom       0      2.0    19.0       1.4      0.18   \n",
              "1395    382157   20    SupInc       0      0.0    13.0  999999.0      0.00   \n",
              "\n",
              "      OthDebt  MonthlyLoad  Default  \n",
              "17        NaN         0.01        1  \n",
              "247       NaN         0.07        1  \n",
              "255       NaN         0.01        1  \n",
              "293       NaN         0.00        0  \n",
              "587       NaN         0.01        0  \n",
              "745       NaN         0.20        0  \n",
              "748       NaN         0.01        0  \n",
              "817       NaN         0.05        1  \n",
              "1194      NaN         0.10        0  \n",
              "1254      NaN         0.09        0  \n",
              "1395      NaN         0.00        0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-3e3beb34-a91a-4640-8591-c5d074467175\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>customer</th>\n",
              "      <th>Age</th>\n",
              "      <th>Education</th>\n",
              "      <th>Employ</th>\n",
              "      <th>Address</th>\n",
              "      <th>Income</th>\n",
              "      <th>Leverage</th>\n",
              "      <th>Creddebt</th>\n",
              "      <th>OthDebt</th>\n",
              "      <th>MonthlyLoad</th>\n",
              "      <th>Default</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>10423</td>\n",
              "      <td>19</td>\n",
              "      <td>Bas</td>\n",
              "      <td>2</td>\n",
              "      <td>0.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>0.02</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.01</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>247</th>\n",
              "      <td>71896</td>\n",
              "      <td>20</td>\n",
              "      <td>SupInc</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>1.3</td>\n",
              "      <td>0.18</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.07</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>255</th>\n",
              "      <td>72327</td>\n",
              "      <td>19</td>\n",
              "      <td>Med</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.01</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>293</th>\n",
              "      <td>74471</td>\n",
              "      <td>39</td>\n",
              "      <td>Bas</td>\n",
              "      <td>8</td>\n",
              "      <td>9.0</td>\n",
              "      <td>41.0</td>\n",
              "      <td>0.1</td>\n",
              "      <td>0.03</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>587</th>\n",
              "      <td>244023</td>\n",
              "      <td>31</td>\n",
              "      <td>SupCom</td>\n",
              "      <td>1</td>\n",
              "      <td>7.0</td>\n",
              "      <td>69.0</td>\n",
              "      <td>0.4</td>\n",
              "      <td>0.19</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>745</th>\n",
              "      <td>315892</td>\n",
              "      <td>22</td>\n",
              "      <td>SupInc</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>2.5</td>\n",
              "      <td>0.26</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.20</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>748</th>\n",
              "      <td>315948</td>\n",
              "      <td>41</td>\n",
              "      <td>Med</td>\n",
              "      <td>21</td>\n",
              "      <td>9.0</td>\n",
              "      <td>64.0</td>\n",
              "      <td>0.3</td>\n",
              "      <td>0.12</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>817</th>\n",
              "      <td>335635</td>\n",
              "      <td>18</td>\n",
              "      <td>Med</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>24.0</td>\n",
              "      <td>0.9</td>\n",
              "      <td>0.13</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.05</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1194</th>\n",
              "      <td>372485</td>\n",
              "      <td>20</td>\n",
              "      <td>SupInc</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>1.1</td>\n",
              "      <td>0.05</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.10</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1254</th>\n",
              "      <td>376801</td>\n",
              "      <td>23</td>\n",
              "      <td>SupCom</td>\n",
              "      <td>0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>19.0</td>\n",
              "      <td>1.4</td>\n",
              "      <td>0.18</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1395</th>\n",
              "      <td>382157</td>\n",
              "      <td>20</td>\n",
              "      <td>SupInc</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>999999.0</td>\n",
              "      <td>0.00</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-3e3beb34-a91a-4640-8591-c5d074467175')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-3e3beb34-a91a-4640-8591-c5d074467175 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-3e3beb34-a91a-4640-8591-c5d074467175');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_ZWCGoKQYfvB"
      },
      "source": [
        "We can see all cases in which OthDebt is null. We can also see the egregious outlier that Leverage has, with a value of 9999999. These values give us some hints about what's happening with OthDebt. Let's study the minimum and maximum of the variables."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nD5BqZ5hYwcY",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d0a1812e-ce38-440d-8f80-16f25a12c33b"
      },
      "source": [
        "bankloan_data.OthDebt.describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "count    1489.000000\n",
              "mean        3.872257\n",
              "std         5.343150\n",
              "min         0.100000\n",
              "25%         1.120000\n",
              "50%         2.240000\n",
              "75%         4.610000\n",
              "max        63.470000\n",
              "Name: OthDebt, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KpgxTyujY0YD"
      },
      "source": [
        "Aha! Note anything? There are no values equal to 0! We can make a good guess that the null values are equal to zero. In real life we would call the sysadmins to confirm this, but in this case we can make a guess.\n",
        "\n",
        "Then, we now should fix these values as follow:\n",
        "\n",
        "1. Replace the leverage equal to 999999 by a null value.\n",
        "2. Replace all OthDebt nulls by 0.\n",
        "3. Replace the remaining null values by the median."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Geu-R8RRa1Tf"
      },
      "source": [
        "# Replace invalid outlier.\n",
        "bankloan_data.Leverage.values[bankloan_data.Leverage.values == 999999] = np.nan"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BFaxJbmNKGcN"
      },
      "source": [
        "# Fills out the null values with zeros. Inplace argument changes dataframe.\n",
        "bankloan_data.OthDebt.fillna(value = 0, inplace=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ceQ6j1xnxHDm"
      },
      "source": [
        "# Remove null case if you want\n",
        "# bankloan_data.drop(index=255, axis = 0, inplace = True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R06ZbW0DcN6f",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ff9fe6e1-644b-4db8-f118-612f41d0625b"
      },
      "source": [
        "# Fill out remaining elements.\n",
        "null_columns = bankloan_data.columns[bankloan_data.isnull().any()]\n",
        "\n",
        "for column in null_columns:\n",
        "  bankloan_data[column].fillna(bankloan_data[column].median(), inplace=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-15-b0af80ee2ce6>:2: FutureWarning: Dropping of nuisance columns in DataFrame reductions (with 'numeric_only=None') is deprecated; in a future version this will raise TypeError.  Select only valid columns before calling the reduction.\n",
            "  bankloan_data.fillna(bankloan_data.median(), inplace=True)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QgFDQXyBiHiU",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "outputId": "0fc50729-d88f-4e29-e5d0-6985793f24ff"
      },
      "source": [
        "bankloan_data.describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            customer          Age       Employ      Address       Income  \\\n",
              "count    1500.000000  1500.000000  1500.000000  1500.000000  1500.000000   \n",
              "mean   257714.548000    34.174000     6.952000     6.308667    59.602000   \n",
              "std    139555.147719    13.142306     8.977644     6.045640    67.124278   \n",
              "min     10012.000000    18.000000     0.000000     0.000000    12.000000   \n",
              "25%     98638.750000    24.000000     0.000000     2.000000    27.000000   \n",
              "50%    315991.500000    31.000000     4.000000     5.000000    40.000000   \n",
              "75%    370701.500000    42.000000    10.000000     9.000000    64.000000   \n",
              "max    453777.000000    79.000000    63.000000    34.000000  1079.000000   \n",
              "\n",
              "          Leverage     Creddebt      OthDebt  MonthlyLoad      Default  \n",
              "count  1500.000000  1500.000000  1500.000000  1500.000000  1500.000000  \n",
              "mean      9.940600     1.935567     3.843860     0.365447     0.365333  \n",
              "std       6.662313     2.973593     5.333758     0.293937     0.481684  \n",
              "min       0.100000     0.000000     0.000000     0.000000     0.000000  \n",
              "25%       4.800000     0.420000     1.107500     0.150000     0.000000  \n",
              "50%       8.500000     0.990000     2.215000     0.290000     0.000000  \n",
              "75%      13.525000     2.200000     4.572500     0.500000     1.000000  \n",
              "max      40.700000    35.970000    63.470000     2.150000     1.000000  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-94947d2f-8dfc-4834-8cd2-3dd57d785575\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>customer</th>\n",
              "      <th>Age</th>\n",
              "      <th>Employ</th>\n",
              "      <th>Address</th>\n",
              "      <th>Income</th>\n",
              "      <th>Leverage</th>\n",
              "      <th>Creddebt</th>\n",
              "      <th>OthDebt</th>\n",
              "      <th>MonthlyLoad</th>\n",
              "      <th>Default</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "      <td>1500.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>257714.548000</td>\n",
              "      <td>34.174000</td>\n",
              "      <td>6.952000</td>\n",
              "      <td>6.308667</td>\n",
              "      <td>59.602000</td>\n",
              "      <td>9.940600</td>\n",
              "      <td>1.935567</td>\n",
              "      <td>3.843860</td>\n",
              "      <td>0.365447</td>\n",
              "      <td>0.365333</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>139555.147719</td>\n",
              "      <td>13.142306</td>\n",
              "      <td>8.977644</td>\n",
              "      <td>6.045640</td>\n",
              "      <td>67.124278</td>\n",
              "      <td>6.662313</td>\n",
              "      <td>2.973593</td>\n",
              "      <td>5.333758</td>\n",
              "      <td>0.293937</td>\n",
              "      <td>0.481684</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>10012.000000</td>\n",
              "      <td>18.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>12.000000</td>\n",
              "      <td>0.100000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>98638.750000</td>\n",
              "      <td>24.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>27.000000</td>\n",
              "      <td>4.800000</td>\n",
              "      <td>0.420000</td>\n",
              "      <td>1.107500</td>\n",
              "      <td>0.150000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>315991.500000</td>\n",
              "      <td>31.000000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>5.000000</td>\n",
              "      <td>40.000000</td>\n",
              "      <td>8.500000</td>\n",
              "      <td>0.990000</td>\n",
              "      <td>2.215000</td>\n",
              "      <td>0.290000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>370701.500000</td>\n",
              "      <td>42.000000</td>\n",
              "      <td>10.000000</td>\n",
              "      <td>9.000000</td>\n",
              "      <td>64.000000</td>\n",
              "      <td>13.525000</td>\n",
              "      <td>2.200000</td>\n",
              "      <td>4.572500</td>\n",
              "      <td>0.500000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>453777.000000</td>\n",
              "      <td>79.000000</td>\n",
              "      <td>63.000000</td>\n",
              "      <td>34.000000</td>\n",
              "      <td>1079.000000</td>\n",
              "      <td>40.700000</td>\n",
              "      <td>35.970000</td>\n",
              "      <td>63.470000</td>\n",
              "      <td>2.150000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-94947d2f-8dfc-4834-8cd2-3dd57d785575')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-94947d2f-8dfc-4834-8cd2-3dd57d785575 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-94947d2f-8dfc-4834-8cd2-3dd57d785575');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rwY4w7DAcwyE"
      },
      "source": [
        "There are no more null values! We are now ready to study the distributions."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1hpzkfzwiPyB"
      },
      "source": [
        "### Outlier cleaning\n",
        "\n",
        "Cleaning outlier, on the other hand, requires a much more manual intervention. We know there are three variables we need to intervene in: Income, Creddebt, and OthDebt. In general, we would like to modify all cases where there is either a discontinuous distribution or a case that is outside of 3 to 6 standard deviations.\n",
        "\n",
        "It is very important to use your judgement in this! Don't just cut in 3 stds, as that is too restrictive in an exponential distribution. If you want to get an idea of the values, then seaborn can help. The function ```displot``` with either the option ```hist``` or ```kde``` allows to check the distribution of values."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dCOmLnVpiOVo",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "outputId": "9a9374e4-b950-4f1b-ca50-43585d38aa3c"
      },
      "source": [
        "fig = sns.displot(bankloan_data['Income'], kind = 'hist')\n",
        "plt.savefig('Income.pdf')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVwAAAFcCAYAAACEFgYsAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAAsTAAALEwEAmpwYAAAdZ0lEQVR4nO3dfXBU5d3/8c/ZPAfQJSHgBhgZHKERrEACdLTYIYigAtZfB0kzYItiKyrGOlCfxkSDDg1hqGUIhnbUf2SInSk3FFAS7x9Yi04tVKNNiUpVkJpISEKQEPKw2ev+I3f2DhjCBpJrH/J+zWTcPdfZ3e83s3xyvM6TY4wxAgD0O1ewCwCAgYLABQBLCFwAsITABQBLCFwAsITABQBLom18yMmTJ/XrX/9aX331lWJjY3X11VcrPz9fSUlJGj9+vMaNGyeXqyP7165dq/Hjx0uS9u7dq7Vr16q9vV0TJkzQmjVrlJCQ0KvPrqtrlM/X85FvQ4cm6uTJpktrLkSEew/hXr9ED6Ei2D2kpAy54JiVLVzHcbRs2TKVlpZq586dGj16tNatW+cfLykp0Y4dO7Rjxw5/2J45c0bPPPOMiouL9dZbb2nQoEF6+eWX+6W+6Oiofnlfm8K9h3CvX6KHUBHKPVgJXLfbrenTp/ufT5o0SVVVVT2+5p133tHEiRM1ZswYSVJWVpbefPPN/iwTAPqVlSmFrnw+n7Zu3arMzEz/siVLlqi9vV0333yzVqxYodjYWFVXVys1NdW/Tmpqqqqrq22XCwB9xnrgrl69WomJiVq8eLEk6e2335bH41FjY6NWrVqloqIi/epXv+qzz0tOHhzQej3Nu4SLcO8h3OuX6CFUhGoPVgO3oKBAR48eVXFxsX8nmcfjkSQNHjxYCxcu1Kuvvupf/v777/tfW1VV5V+3NwLZaZaSMkQnTpzu9XuHknDvIdzrl+ghVAS7h6DvNJOk9evXq6KiQkVFRYqNjZUknTp1Ss3NzZIkr9er0tJSpaWlSZJmzJihf/7znzpy5Iikjh1rt912m61yAaDPWdnCPXz4sDZv3qwxY8YoKytLkjRq1CgtW7ZMubm5chxHXq9XkydPVk5OjqSOLd78/Hz98pe/lM/nU1pamp5++mkb5QJAv3Ai/fKMTCmEh3CvX6KHUBHsHkJiSgEABjoCFwAsIXABwBICFwAsIXB74DhOsEsAEEEI3AtwHEev7ztM6ALoMwRuD5qavcEuAUAEIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXC74TiOHCfYVQCINNHBLiDUOI6jV9+oVEIcvxoAfYtU6caZ5jYZGTls5gLoQ0wpAIAlBC4AWELgAoAlBC4AWGIlcE+ePKn7779fc+bM0fz58/Xwww+rvr5eklReXq4FCxZozpw5uvfee1VXV+d/XU9jABBurASu4zhatmyZSktLtXPnTo0ePVrr1q2Tz+fTqlWrlJubq9LSUmVkZGjdunWS1OMYAIQjK4Hrdrs1ffp0//NJkyapqqpKFRUViouLU0ZGhiQpKytLe/bskaQexwAgHFk/Dtfn82nr1q3KzMxUdXW1UlNT/WNJSUny+XxqaGjoccztdgf8ecnJgwNaLyVliP9xTHSUYqKj5DiOhg0L7PWhoGsP4Sjc65foIVSEag/WA3f16tVKTEzU4sWL9dZbb/X759XVNcrnMz2uk5IyRCdOnJbUMf3R5m1Xm9eR4ziqrW2UMT2/PhR07SEchXv9Ej2EimD30FPYWw3cgoICHT16VMXFxXK5XPJ4PKqqqvKP19fXy+Vyye129zgGAOHI2mFh69evV0VFhYqKihQbGytJmjhxopqbm3Xw4EFJUklJiebOnXvRMQAIR1a2cA8fPqzNmzdrzJgxysrKkiSNGjVKRUVFWrt2rfLy8tTS0qKRI0eqsLBQkuRyuS44BgDhyErgXnvttfr000+7HZsyZYp27tzZ6zEACDecaQYAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgQsAlhC4AGAJgXsRjiM5jhPsMgBEAAK3Bwlx0Xpld6VefaOS0AVw2aKDXUCoO9PcFuwSAEQItnABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwBICFwAsIXABwJJoWx9UUFCg0tJSff3119q5c6fGjRsnScrMzFRsbKzi4uIkSStXrtSMGTMkSeXl5crNzVVLS4tGjhypwsJCJScn2yoZAPqUtS3cWbNmacuWLRo5cuR3xjZs2KAdO3Zox44d/rD1+XxatWqVcnNzVVpaqoyMDK1bt85WuQDQ56wFbkZGhjweT8DrV1RUKC4uThkZGZKkrKws7dmzp7/KA4B+Z21KoScrV66UMUbp6el67LHHdMUVV6i6ulqpqan+dZKSkuTz+dTQ0CC32x3weycnDw5ovZSUIf7HMdFR5/xI0rBhgb1PMHXtIRyFe/0SPYSKUO0h6IG7ZcsWeTwetba26oUXXlB+fn6fTh3U1TXK5zM9rpOSMkQnTpyWJDmOozZvu9q8jtq8LrV52yVJtbWNMqbn9wmmrj2Eo3CvX6KHUBHsHnoK+6AfpdA5zRAbG6vs7Gx98MEH/uVVVVX+9err6+VyuXq1dQsAoSSogdvU1KTTpzv+Ehlj9MYbbygtLU2SNHHiRDU3N+vgwYOSpJKSEs2dOzdotTpOx9YvAFwqa1MKzz//vMrKylRbW6ulS5fK7XaruLhYK1asUHt7u3w+n6655hrl5eVJklwul9auXau8vLxzDgsLhsT4aL2yu1KStPT2tJCeWgAQuhwT4elxKXO4G7d9rMT4aCXGx6i24awS46PV1OyVJD38/74fkoEb7HmryxXu9Uv0ECqC3UNIz+ECwEBB4AKAJQQuAFhC4AKAJQQuAFhC4AKAJQEH7ptvvtntci4oAwCBCThwn3766W6X5+bm9lkxABDJLnqm2bFjxyR1nHrb+bjrWGxsbP9UBgAR5qKBO3v2bDmOI2OMZs+efc7YsGHDtGLFin4rDgAiyUUD95NPPpEkLV68WK+99lq/FwQAkSrgOVzCFgAuT8BXCzt27JhefPFFVVZWqqmp6Zyxt99+u6/rAoCIE3Dgrly5UqNHj9bjjz+uhISE/qwJACJSwIF7+PBhbd26VS4X50oAwKUIOD2nTp2qQ4cO9WctABDRAt7CHTlypJYtW6bZs2dr2LBh54zl5OT0eWEAEGkCDtyzZ89q5syZ8nq9+uabb/qzJgCISAEH7po1a/qzDgCIeL06LOxCRo8e3SfFAEAkCzhwu57i26nztuGVlZV9XxkARJiAA7fzFN9OJ06c0MaNG5WRkdHnRQFAJLrkg2pTUlL09NNPa/369X1ZDwBErMs6i+GLL77Q2bNn+6oWAIhoAU8pZGdn++dspY7DxP7973/roYce6pfCACDSBBy4CxcuPOd5QkKCvve972nMmDF9XRMARKSAA/euu+7qzzoAIOIFPIfb1tamDRs2aNasWbr++us1a9YsbdiwQa2trf1ZHwBEjIC3cAsLC/Xxxx/rueeeU2pqqqqqqrRp0yY1Njbqqaee6s8aASAiBBy4e/bs0Y4dOzR06FBJ0tixY3XdddfpzjvvJHABIAABTyl0PcMskOUAgHMFHLhz587V8uXL9de//lWff/653nnnHT300EOaO3duf9YHABEj4CmFVatW6aWXXlJ+fr5qamo0YsQI3XHHHVq+fHl/1gcAEeOiW7j/+Mc/VFhYqNjYWOXk5Oitt97SRx99pLKyMrW2tnIXCAAI0EUDd/PmzZo6dWq3Y9OnT1dxcXGfFwUAkeiigVtZWakZM2Z0O3bjjTeqoqKiz4sCgEh00cBtbGxUW1tbt2Ner1dnzpzp86IAIBJdNHDHjh2r/fv3dzu2f/9+jR07ts+LAoBIdNHA/fnPf668vDyVlZXJ5/NJknw+n8rKyvTss89q6dKl/V4kAESCix4WNn/+fNXW1urxxx9XW1ub3G63GhoaFBMTo0ceeUTz5s2zUScAhL2AjsNdunSpFi5cqA8//FANDQ1yu92aPHmyBg8e3N/1AUDECPjEh8GDB1/waIWBxHEkzmYGcCku6xY7A01ifLRK9h4+584XABAoAreXmpq9wS4BQJgicAHAEgIXACwhcAHAEgIXACwhcAHAEiuBW1BQoMzMTI0fP16fffaZf/mXX36pRYsWac6cOVq0aJGOHDkS0BgAhCMrgTtr1ixt2bJFI0eOPGd5Xl6esrOzVVpaquzsbOXm5gY0BgDhyErgZmRkyOPxnLOsrq5Ohw4d8l+LYd68eTp06JDq6+t7HAOAcBXwqb19rbq6WiNGjFBUVJQkKSoqSsOHD1d1dbWMMRccS0pKClbJAHBZgha4tiQnB3aBnZSUIf7HMdFR3fwYxURHyXEcDRsWmhft6dpDOAr3+iV6CBWh2kPQAtfj8ej48eNqb29XVFSU2tvbVVNTI4/HI2PMBcd6q66uUT5fz1ebSUkZohMnTkuSHMdRm7ddbV5HbV5Xl8cd/3UcR7W1jTIhdgWbrj2Eo3CvX6KHUBHsHnoK+6AdFpacnKy0tDTt2rVLkrRr1y6lpaUpKSmpxzEACFdWtnCff/55lZWVqba2VkuXLpXb7dbu3bv17LPP6oknntCmTZt0xRVXqKCgwP+ansYAIBw5JtT+37iPXcqUwsZtHysxPlqJ8TGqbTirxPhoNTV7lRgfLcdxtPS2NKYU+li41y/RQ6gIdg8hOaUAAAMNgQsAlhC4AGAJgXsZuNUOgN4gcC+R4zh6fR/3NwMQOAL3MnB/MwC9QeACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCVWbpMeabjmOIBLQeD2UkJctF7ZXamEOH51AHqH1LgEZ5rbZGS4vQ6AXmEO9zxkKID+QuB24TiOtu//MthlAIhQBO55zrZwY0gA/YPABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITABQBLCFwAsITAvUyOIzmOE+wyAISB6GAXIEmZmZmKjY1VXFycJGnlypWaMWOGysvLlZubq5aWFo0cOVKFhYVKTk4OcrX/JyEuWq/srpQkLb09TcaYIFcEIJSFROBK0oYNGzRu3Dj/c5/Pp1WrVmnNmjXKyMjQpk2btG7dOq1ZsyaIVX7Xmea2YJcAIEyE7JRCRUWF4uLilJGRIUnKysrSnj17glwVAFy6kNnCXblypYwxSk9P12OPPabq6mqlpqb6x5OSkuTz+dTQ0CC32x28QgHgEoVE4G7ZskUej0etra164YUXlJ+fr9mzZ/fJeycnDw5ovZSUIZKkmOioC/yYC45J0rBhgX1Of+rsIVyFe/0SPYSKUO0hJALX4/FIkmJjY5Wdna3ly5frnnvuUVVVlX+d+vp6uVyuXm/d1tU1yufreWdWSsoQnThxWo7jqM3b/r8/jtq8ri6Pu1vW8ViSamsbg7rTrLOHcBXu9Uv0ECqC3UNPYR/0OdympiadPt3xyzHG6I033lBaWpomTpyo5uZmHTx4UJJUUlKiuXPnBrNUALgsQd/Craur04oVK9Te3i6fz6drrrlGeXl5crlcWrt2rfLy8s45LAwAwlXQA3f06NHavn17t2NTpkzRzp077RYEAP0k6FMKADBQELgAYAmBCwCWELgAYAmBCwCWELh9hCs0ArgYArcPJMZHq2TvYa6LC6BHBG4faWr2BrsEACGOwAUASwhcALCEwAUASwhcALCEwAUASwhcALAk6JdnjCQdh+H+37G43DYdQFcEbh9JiIvWK7srlRAXrbMtHcfkLr09jdAF4Efg9qEzzW0yMpwEAaBbzOECgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuD2I+64A6ArArefcJ8zAOcjcPsRp/gC6IrABQBLCFwAsITAtYS5XAAEbj9zHMnlcvT6PnagAQMd18PtR+dclLyVHWjAQEfg9rPOi5KzdQuAKYUgIHyBgYnAtcxxmM8FBioCNwg4IQIYmJjDDRJuqQ4MPARuEHBLdWBgInCDhFuqAwMPc7gAYAmBG0Icx+HoBSCCEbghwnGkV9+o1KtvVMrlInSBSETgWuQ4Pd8FonNelwuXA5GJnWaWdD0yQRfJ0q470hzH6fbohc5A5sgGIHywhWvRmeY2NbW0dTvW3Qbthc5KcxzHP/3AljAQPtjCDQGJ8dH6r79+2e1Yd4eNOU5HeAeiayCzNQwEF4EbIjpPgLgYx3G0fX/34dzdujve/VL1p5olcXIFEGwEbhjoehqw4wQezlLHuoFuDQPoXwRuiDv/NODzd7o5jtS50dp1B1vHMb39Wxs77oDeCfmdZl9++aUWLVqkOXPmaNGiRTpy5EiwS7Ki6yFknTvbzt/plhgfrZK9h/3H7XbuYOvcqbb1vw8H+FnOd066uNDOuM7l7LgDei/kAzcvL0/Z2dkqLS1Vdna2cnNzg11Sv+vcqg0kMI2RXtldqeJtH6up2esP6u6OiOgYc875cbk6grPk/x8+J7Bf39cR5OeHcNejJs40d/wROD9vuztjruuy82s4f3lvXOizzh+/0B+TQP7IhALOQrSvP37fIT2lUFdXp0OHDunVV1+VJM2bN0+rV69WfX29kpKSAnqPQM/a6gyXFHeCYqJcio+NVnxclP9xc6v3O8su9Lgv1m043dKrdQcnxmjo4Fjt/ttRxcVEyZM86Dvr7q/4RqcaWxQXE6X4uGj/4ysGxSo+NlqOS4qK6vh9DbsyQbv/dlSSNP/GMTKmI7CHJMb61+n8jL0ffK1bMkb519n53pHvvK5z2YKbxujP7x75Tg2JibE6eersOa+7mAt91n8f/I9uyRglSXrn42qdamzxryOdO97d48uZIYmK6vttmO767E/90YNtl9tD5/doVvqoPp0yc0wIT8BVVFTo8ccf1+7du/3Lbr/9dhUWFmrChAlBrAwAei/8/5QBQJgI6cD1eDw6fvy42tvbJUnt7e2qqamRx+MJcmUA0HshHbjJyclKS0vTrl27JEm7du1SWlpawPO3ABBKQnoOV5I+//xzPfHEE/r22291xRVXqKCgQGPHjg12WQDQayEfuAAQKUJ6SgEAIgmBCwCWELgAYAmBCwCWDOjADYcL45w8eVL333+/5syZo/nz5+vhhx9WfX29JKm8vFwLFizQnDlzdO+996qurs7/up7Ggmnjxo0aP368PvvsM0nh1UNLS4vy8vJ06623av78+XrmmWck9fw9CqXv2L59+/TjH/9Yd955pxYsWKCysrKL1hjs+gsKCpSZmXnOd+Zyag52PzID2JIlS8z27duNMcZs377dLFmyJMgVfdfJkyfN3/72N//z3/zmN+bJJ5807e3t5pZbbjEHDhwwxhhTVFRknnjiCWOM6XEsmCoqKsx9991nZs6caT799NOw62H16tXmhRdeMD6fzxhjzIkTJ4wxPX+PQuU75vP5TEZGhvn000+NMcZUVlaaSZMmmfb29pCu/8CBA6aqqsr/nQmkrlDuZ8AGbm1trUlPTzder9cYY4zX6zXp6emmrq4uyJX1bM+ePeZnP/uZ+eijj8wdd9zhX15XV2cmTZpkjDE9jgVLS0uLufvuu82xY8f8/3jCqYfGxkaTnp5uGhsbz1ne0/colL5jPp/PTJs2zRw8eNAYY8zf//53c+utt4ZN/V0D91JrDoV+QvpqYf2purpaI0aMUFRUlCQpKipKw4cPV3V1dcieyebz+bR161ZlZmaqurpaqamp/rGkpCT5fD41NDT0OOZ2u4NQufS73/1OCxYs0KhRo/zLwqmHY8eOye12a+PGjXr//fc1aNAg5eTkKD4+/oLfI2NMyHzHHMfRiy++qAcffFCJiYk6c+aMfv/73/f47yCU6u/qUmsOhX4G9BxuuFm9erUSExO1ePHiYJfSKx9++KEqKiqUnZ0d7FIuWXt7u44dO6brrrtO27Zt08qVK7VixQo1NTUFu7SAeL1ebd68WZs2bdK+ffv00ksv6dFHHw2b+iPFgN3C7XphnKioqJC/ME5BQYGOHj2q4uJiuVwueTweVVVV+cfr6+vlcrnkdrt7HAuGAwcO6PPPP9esWbMkSd98843uu+8+LVmyJGx68Hg8io6O1rx58yRJN9xwg4YOHar4+PgLfo+MMSHzHausrFRNTY3S09MlSenp6UpISFBcXFxY1N9VT/92e6o5FPoZsFu44XRhnPXr16uiokJFRUWKjY2VJE2cOFHNzc06ePCgJKmkpERz58696Fgw/OIXv9D+/fu1d+9e7d27V1dddZVefvllLVu2LGx6SEpK0vTp0/Xuu+9K6tjbXVdXpzFjxlzwexRK37GrrrpK33zzjb744gtJHdcoqaur09VXXx0W9XfVU12XOmbLgL6WQjhcGOfw4cOaN2+exowZo/j4eEnSqFGjVFRUpA8++EB5eXlqaWnRyJEjVVhYqGHDhklSj2PBlpmZqeLiYo0bNy6sejh27JieeuopNTQ0KDo6Wo8++qh+9KMf9fg9CqXv2J///Gf94Q9/8N865pFHHtEtt9wS0vU///zzKisrU21trYYOHSq3263du3dfcs3B7mdABy4A2DRgpxQAwDYCFwAsIXABwBICFwAsIXABwBICFwAsIXAR9jIzM/Xee+8FuwzgoghcALCEwEXE2LZtm37605+qoKBAU6dOVWZmpv7yl7/4xxsaGvTkk0/qhz/8oaZOnaoHH3zQP/bHP/5Rs2fP1rRp0/TAAw/o+PHj/rHx48dry5YtuvXWWzV58mS9+OKL+uqrr5SVlaUpU6YoJydHra2t/vX37dunO++8UxkZGcrKytInn3xi5xeA0GftQpBAP5k5c6Z59913zZ/+9Cdz3XXXmddff914vV6zZcsWc9NNN/kvGH7//febnJwc09DQYFpbW837779vjDHmvffeM9OmTTMVFRWmpaXF5Ofnm+zsbP/7jxs3zjzwwAPm9OnT5rPPPjMTJkww99xzj/nqq6/Mt99+a2677Tazbds2Y4wx//rXv8wPfvADU15ebrxer9m2bZuZOXOmaWlpsf+LQchhCxcRJTU1VXfffbeioqJ011136cSJE6qtrVVNTY3eeecdPffcc7ryyisVExOjadOmSZJ27typn/zkJ5owYYJiY2P12GOPqby8XP/5z3/877ts2TINHjxY1157rcaNG6ebbrpJo0eP1pAhQ3TzzTfr0KFDkqTXX39dixYt0g033OCvISYmRuXl5cH4dSDEDNjLMyIydb24TUJCgiSpqalJp06d0pVXXqkrr7zyO6+pqanRhAkT/M8HDRokt9ut48eP+y+Y3vV94+LivvO8trZWklRVVaXt27frtdde84+3tbWppqamjzpEOCNwMSBcddVVOnXqlP8qUV0NHz5cX3/9tf95U1OTGhoaNGLEiF5/jsfj0QMPPKDly5dfds2IPEwpYEAYPny4br75Zj333HM6deqU2tradODAAUnSvHnztG3bNlVWVqq1tVXr16/X97///XNuBxSohQsXqqSkRB999JGMMWpqatLbb7+txsbGvm4JYYgtXAwYa9eu1Zo1a3Tbbbepra1N06dP19SpU3XjjTcqJydHK1as0LfffqvJkyfrt7/97SV9xvXXX6/Vq1crPz9fR48eVXx8vKZMmaKMjIw+7gbhiOvhAoAlTCkAgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCUELgBYQuACgCX/A2QhxsL73YmaAAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IPA95piMjhLv",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "outputId": "f167e902-20fa-41f5-d7e2-ef2d9fab5662"
      },
      "source": [
        "sns.displot(data = bankloan_data, x = 'OthDebt', kind = 'kde', hue = 'Default')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AwJ3ND4gZj9S",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 365
        },
        "outputId": "9f40053a-9738-4e59-c141-9aa90e2826ae"
      },
      "source": [
        "fig = sns.displot(data = bankloan_data, x = 'Creddebt', kind = 'kde', hue = 'Default')\n",
        "plt.savefig('Creddebt.pdf')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 412.925x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QtGBJQw6jnmm"
      },
      "source": [
        "Income needs to be trimmed from the right. Potential cutoffs are Income of 450. OthDebt and Creddebt need to be truncated or left as is.\n",
        "\n",
        "To actually cut the values, then we can simply write a proper expression in Pandas:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VQrTABT-lIZa",
        "outputId": "cbdad954-ee2d-4113-9b8f-7c4963480039"
      },
      "source": [
        "bankloan_data['Income'].sort_values(ascending=False).head(n=50)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1198    1079.0\n",
              "468      699.0\n",
              "1461     533.0\n",
              "1435     526.0\n",
              "776      512.0\n",
              "234      497.0\n",
              "1314     493.0\n",
              "489      475.0\n",
              "939      460.0\n",
              "192      410.0\n",
              "469      403.0\n",
              "1431     397.0\n",
              "105      387.0\n",
              "412      384.0\n",
              "667      368.0\n",
              "608      352.0\n",
              "650      327.0\n",
              "617      326.0\n",
              "1475     322.0\n",
              "1361     321.0\n",
              "467      314.0\n",
              "541      308.0\n",
              "204      289.0\n",
              "342      286.0\n",
              "1132     285.0\n",
              "565      274.0\n",
              "620      274.0\n",
              "1101     267.0\n",
              "679      263.0\n",
              "37       259.0\n",
              "392      251.0\n",
              "919      249.0\n",
              "134      238.0\n",
              "359      236.0\n",
              "340      234.0\n",
              "1196     233.0\n",
              "143      232.0\n",
              "678      231.0\n",
              "206      229.0\n",
              "588      229.0\n",
              "417      227.0\n",
              "680      226.0\n",
              "794      225.0\n",
              "93       222.0\n",
              "828      220.0\n",
              "1432     219.0\n",
              "609      215.0\n",
              "297      209.0\n",
              "862      208.0\n",
              "228      206.0\n",
              "Name: Income, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HWhH4fgcji1G"
      },
      "source": [
        "bankloan_data = bankloan_data.loc[bankloan_data['Income'] < 300]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BpbVxiJQmOBs"
      },
      "source": [
        "We can check now how the data looks like."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wgUVYbTKlpQT",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 381
        },
        "outputId": "c04406f4-6e5a-4dad-bd5c-08e9e3e885c7"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(10,5))\n",
        "a = sns.violinplot(x='variable', y='value', data=pd.melt(bankloan_data.iloc[:, np.r_[1,3:9]]), ax=ax)\n",
        "a.set_xticklabels(a.get_xticklabels(), rotation=90);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qFia8JlY0WU1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "outputId": "cc09fe9c-db36-42f1-897d-73c72456558d"
      },
      "source": [
        "bankloan_data.describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            customer          Age       Employ      Address       Income  \\\n",
              "count    1478.000000  1478.000000  1478.000000  1478.000000  1478.000000   \n",
              "mean   257591.192828    33.820704     6.589310     6.173207    53.795670   \n",
              "std    139579.702866    12.838761     8.382677     5.949440    43.104055   \n",
              "min     10012.000000    18.000000     0.000000     0.000000    12.000000   \n",
              "25%     98310.250000    24.000000     0.000000     2.000000    27.000000   \n",
              "50%    316030.000000    31.000000     4.000000     5.000000    39.000000   \n",
              "75%    370670.750000    41.000000     9.000000     9.000000    62.000000   \n",
              "max    453777.000000    77.000000    51.000000    34.000000   289.000000   \n",
              "\n",
              "          Leverage     Creddebt      OthDebt  MonthlyLoad      Default  \n",
              "count  1478.000000  1478.000000  1478.000000  1478.000000  1478.000000  \n",
              "mean      9.975237     1.800298     3.590135     0.369465     0.366035  \n",
              "std       6.665023     2.514837     4.467314     0.294106     0.481882  \n",
              "min       0.100000     0.000000     0.000000     0.000000     0.000000  \n",
              "25%       4.900000     0.420000     1.082500     0.160000     0.000000  \n",
              "50%       8.600000     0.950000     2.190000     0.300000     0.000000  \n",
              "75%      13.600000     2.130000     4.450000     0.500000     1.000000  \n",
              "max      40.700000    35.970000    50.220000     2.150000     1.000000  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-0d98c814-aec2-49b6-b1bf-0aa4ffcf725d\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>customer</th>\n",
              "      <th>Age</th>\n",
              "      <th>Employ</th>\n",
              "      <th>Address</th>\n",
              "      <th>Income</th>\n",
              "      <th>Leverage</th>\n",
              "      <th>Creddebt</th>\n",
              "      <th>OthDebt</th>\n",
              "      <th>MonthlyLoad</th>\n",
              "      <th>Default</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>257591.192828</td>\n",
              "      <td>33.820704</td>\n",
              "      <td>6.589310</td>\n",
              "      <td>6.173207</td>\n",
              "      <td>53.795670</td>\n",
              "      <td>9.975237</td>\n",
              "      <td>1.800298</td>\n",
              "      <td>3.590135</td>\n",
              "      <td>0.369465</td>\n",
              "      <td>0.366035</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>139579.702866</td>\n",
              "      <td>12.838761</td>\n",
              "      <td>8.382677</td>\n",
              "      <td>5.949440</td>\n",
              "      <td>43.104055</td>\n",
              "      <td>6.665023</td>\n",
              "      <td>2.514837</td>\n",
              "      <td>4.467314</td>\n",
              "      <td>0.294106</td>\n",
              "      <td>0.481882</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>10012.000000</td>\n",
              "      <td>18.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>12.000000</td>\n",
              "      <td>0.100000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>98310.250000</td>\n",
              "      <td>24.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>27.000000</td>\n",
              "      <td>4.900000</td>\n",
              "      <td>0.420000</td>\n",
              "      <td>1.082500</td>\n",
              "      <td>0.160000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>316030.000000</td>\n",
              "      <td>31.000000</td>\n",
              "      <td>4.000000</td>\n",
              "      <td>5.000000</td>\n",
              "      <td>39.000000</td>\n",
              "      <td>8.600000</td>\n",
              "      <td>0.950000</td>\n",
              "      <td>2.190000</td>\n",
              "      <td>0.300000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>370670.750000</td>\n",
              "      <td>41.000000</td>\n",
              "      <td>9.000000</td>\n",
              "      <td>9.000000</td>\n",
              "      <td>62.000000</td>\n",
              "      <td>13.600000</td>\n",
              "      <td>2.130000</td>\n",
              "      <td>4.450000</td>\n",
              "      <td>0.500000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>453777.000000</td>\n",
              "      <td>77.000000</td>\n",
              "      <td>51.000000</td>\n",
              "      <td>34.000000</td>\n",
              "      <td>289.000000</td>\n",
              "      <td>40.700000</td>\n",
              "      <td>35.970000</td>\n",
              "      <td>50.220000</td>\n",
              "      <td>2.150000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0d98c814-aec2-49b6-b1bf-0aa4ffcf725d')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-0d98c814-aec2-49b6-b1bf-0aa4ffcf725d button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-0d98c814-aec2-49b6-b1bf-0aa4ffcf725d');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mPDWYQFymRX1"
      },
      "source": [
        "Much better, although income still has a large outlier. This plot might be misleading though, as the magnitud of the data is relevant. We might want to recheck this after normalizing the data. We will now save the output to a compressed format which is very efficient to start whole data structures, [pickle](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_pickle.html)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4V1InwIQIM9J"
      },
      "source": [
        "bankloan_data.to_pickle('BankloanClean.pkl')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IwSEaeGlmFzO"
      },
      "source": [
        "## Normalization\n",
        "\n",
        "Most models require the inputs to be in the same scale, this is called **normalization**. It is very important for most models... except for credit scoring as we will use Weight of Evidence (see the lecture!). Still, this is fairly simple to do in Pandas, being smart about what columns we select and **mapping** a function to those columns.\n",
        "\n",
        "The most traditional mapping is the zscore. We can use scipy's version of it. \n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9E_PSudynST3"
      },
      "source": [
        "from scipy.stats import zscore"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oEFMdbTMnW68"
      },
      "source": [
        "We now select all numeric columns, except the customer one. We can do this with ```select_dtypes```."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D0I8jDLlm6vM",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "299c8343-a0e7-42ba-9918-9ae469227197"
      },
      "source": [
        "# Identify the numerical columns\n",
        "numeric_cols = bankloan_data.select_dtypes(include=[np.number]).columns\n",
        "numeric_cols "
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['customer', 'Age', 'Employ', 'Address', 'Income', 'Leverage',\n",
              "       'Creddebt', 'OthDebt', 'MonthlyLoad', 'Default'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QAmoNyjxny0G"
      },
      "source": [
        "# Remove the first and last one\n",
        "numeric_cols = numeric_cols[1:-1]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k1ETL7-Nn11u"
      },
      "source": [
        "# Apply the zscore function to all data\n",
        "bankloan_data[numeric_cols] = bankloan_data[numeric_cols].apply(zscore)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "StvuqRWApixH"
      },
      "source": [
        "Now the data should look much better."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JrahcfLIoN45",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 381
        },
        "outputId": "3fd7206b-10b5-4794-e55b-5cefb4cd3c97"
      },
      "source": [
        "fig, ax = plt.subplots(figsize=(10,5))\n",
        "a = sns.violinplot(x='variable', y='value', data=pd.melt(bankloan_data.iloc[:, np.r_[1,3:9]]), ax=ax)\n",
        "a.set_xticklabels(a.get_xticklabels(), rotation=90);"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h9o6wiTXdlsv"
      },
      "source": [
        "The dataset looks really clean now. We are ready to use further models! To be 100% purist **you should first split between train and test set** before doing all of these analyses (so you should use the median of the train set). Please remember that when solving your coursework."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7F6iCmc-oTb_"
      },
      "source": [
        "## Self-study: Dummy coding\n",
        "\n",
        "The final step will be transform the categorical variables to dummy variables. Again, this is **not necessary for credit scoring** as the WoE transform is much more robust. The best strategy here is to follow the 5% rule:\n",
        "\n",
        "1. Aggregate by expert judgement (i.e. your opinion) until each group has more than 5% of data (1% if very large data) and there are all classes (by target variable) present.\n",
        "\n",
        "2. Use target variable percentage (or classification trees) until you get a proper number of groups.\n",
        "\n",
        "To check the number of cases for each default level, we can calculate a crosstab."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Tnu4zQW0oSMP",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 237
        },
        "outputId": "41d43f5b-9338-4bfc-fabb-ddf231676ca0"
      },
      "source": [
        "pd.crosstab(bankloan_data['Education'], bankloan_data['Default'])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Default      0    1\n",
              "Education          \n",
              "Bas        180   64\n",
              "Med        325  196\n",
              "Posg        54   26\n",
              "SupCom     167  137\n",
              "SupInc     211  118"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-cab5b883-2154-4dc3-a339-44f612ff3504\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Default</th>\n",
              "      <th>0</th>\n",
              "      <th>1</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Education</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Bas</th>\n",
              "      <td>180</td>\n",
              "      <td>64</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Med</th>\n",
              "      <td>325</td>\n",
              "      <td>196</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Posg</th>\n",
              "      <td>54</td>\n",
              "      <td>26</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>SupCom</th>\n",
              "      <td>167</td>\n",
              "      <td>137</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>SupInc</th>\n",
              "      <td>211</td>\n",
              "      <td>118</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-cab5b883-2154-4dc3-a339-44f612ff3504')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-cab5b883-2154-4dc3-a339-44f612ff3504 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-cab5b883-2154-4dc3-a339-44f612ff3504');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MJ_jTjWOrw-o"
      },
      "source": [
        "Now we can calculate the mean of each educational level."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4xFFkNvLqz61",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 237
        },
        "outputId": "0a5ccf3c-c7e6-4150-d168-b39ecb5463f3"
      },
      "source": [
        "bankloan_data.groupby('Education', as_index=False).agg({'Default': [\"mean\"]})"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Education   Default\n",
              "                 mean\n",
              "0       Bas  0.262295\n",
              "1       Med  0.376200\n",
              "2      Posg  0.325000\n",
              "3    SupCom  0.450658\n",
              "4    SupInc  0.358663"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-2f2ff9db-b60c-4312-ac23-910009b2de0d\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>Education</th>\n",
              "      <th>Default</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>mean</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Bas</td>\n",
              "      <td>0.262295</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Med</td>\n",
              "      <td>0.376200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Posg</td>\n",
              "      <td>0.325000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>SupCom</td>\n",
              "      <td>0.450658</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>SupInc</td>\n",
              "      <td>0.358663</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2f2ff9db-b60c-4312-ac23-910009b2de0d')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-2f2ff9db-b60c-4312-ac23-910009b2de0d button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-2f2ff9db-b60c-4312-ac23-910009b2de0d');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NJ__rGzgskpG"
      },
      "source": [
        "We could, if we wanted to, group together some educational levels, such as '1upInc', 'Med', and 'Posg'. If not, we can create dummy variables directly with Pandas ```get_dummies```"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RsntGOQ4sjtN",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 433
        },
        "outputId": "d5c71edd-201d-4177-9676-ac9b2003e4b4"
      },
      "source": [
        "bankloan_data = pd.get_dummies(bankloan_data)\n",
        "bankloan_data.describe()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            customer           Age        Employ       Address        Income  \\\n",
              "count    1478.000000  1.478000e+03  1.478000e+03  1.478000e+03  1.478000e+03   \n",
              "mean   257591.192828 -2.427768e-16  1.201865e-17 -4.326715e-17  4.807461e-17   \n",
              "std    139579.702866  1.000338e+00  1.000338e+00  1.000338e+00  1.000338e+00   \n",
              "min     10012.000000 -1.232678e+00 -7.863288e-01 -1.037963e+00 -9.699741e-01   \n",
              "25%     98310.250000 -7.651850e-01 -7.863288e-01 -7.016827e-01 -6.218612e-01   \n",
              "50%    316030.000000 -2.197765e-01 -3.089927e-01 -1.972629e-01 -3.433709e-01   \n",
              "75%    370670.750000  5.593785e-01  2.876773e-01  4.752968e-01  1.904022e-01   \n",
              "max    453777.000000  3.364336e+00  5.299706e+00  4.678795e+00  5.458511e+00   \n",
              "\n",
              "           Leverage      Creddebt       OthDebt   MonthlyLoad      Default  \\\n",
              "count  1.478000e+03  1.478000e+03  1.478000e+03  1.478000e+03  1478.000000   \n",
              "mean  -2.764290e-17 -2.644104e-17  1.490313e-16  2.331619e-16     0.366035   \n",
              "std    1.000338e+00  1.000338e+00  1.000338e+00  1.000338e+00     0.481882   \n",
              "min   -1.482152e+00 -7.161127e-01 -8.039171e-01 -1.256656e+00     0.000000   \n",
              "25%   -7.617310e-01 -5.490474e-01 -5.615195e-01 -7.124507e-01     0.000000   \n",
              "50%   -2.064062e-01 -3.382268e-01 -3.135238e-01 -2.362716e-01     0.000000   \n",
              "75%    5.440327e-01  1.311472e-01  1.925443e-01  4.439844e-01     1.000000   \n",
              "max    4.611412e+00  1.359184e+01  1.044154e+01  6.056096e+00     1.000000   \n",
              "\n",
              "       Education_Bas  Education_Med  Education_Posg  Education_SupCom  \\\n",
              "count    1478.000000    1478.000000     1478.000000       1478.000000   \n",
              "mean        0.165088       0.352503        0.054127          0.205683   \n",
              "std         0.371386       0.477911        0.226345          0.404337   \n",
              "min         0.000000       0.000000        0.000000          0.000000   \n",
              "25%         0.000000       0.000000        0.000000          0.000000   \n",
              "50%         0.000000       0.000000        0.000000          0.000000   \n",
              "75%         0.000000       1.000000        0.000000          0.000000   \n",
              "max         1.000000       1.000000        1.000000          1.000000   \n",
              "\n",
              "       Education_SupInc  \n",
              "count       1478.000000  \n",
              "mean           0.222598  \n",
              "std            0.416131  \n",
              "min            0.000000  \n",
              "25%            0.000000  \n",
              "50%            0.000000  \n",
              "75%            0.000000  \n",
              "max            1.000000  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-353ea684-a812-43d7-b67f-8cad5aa56fb9\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>customer</th>\n",
              "      <th>Age</th>\n",
              "      <th>Employ</th>\n",
              "      <th>Address</th>\n",
              "      <th>Income</th>\n",
              "      <th>Leverage</th>\n",
              "      <th>Creddebt</th>\n",
              "      <th>OthDebt</th>\n",
              "      <th>MonthlyLoad</th>\n",
              "      <th>Default</th>\n",
              "      <th>Education_Bas</th>\n",
              "      <th>Education_Med</th>\n",
              "      <th>Education_Posg</th>\n",
              "      <th>Education_SupCom</th>\n",
              "      <th>Education_SupInc</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1.478000e+03</td>\n",
              "      <td>1.478000e+03</td>\n",
              "      <td>1.478000e+03</td>\n",
              "      <td>1.478000e+03</td>\n",
              "      <td>1.478000e+03</td>\n",
              "      <td>1.478000e+03</td>\n",
              "      <td>1.478000e+03</td>\n",
              "      <td>1.478000e+03</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "      <td>1478.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>257591.192828</td>\n",
              "      <td>-2.427768e-16</td>\n",
              "      <td>1.201865e-17</td>\n",
              "      <td>-4.326715e-17</td>\n",
              "      <td>4.807461e-17</td>\n",
              "      <td>-2.764290e-17</td>\n",
              "      <td>-2.644104e-17</td>\n",
              "      <td>1.490313e-16</td>\n",
              "      <td>2.331619e-16</td>\n",
              "      <td>0.366035</td>\n",
              "      <td>0.165088</td>\n",
              "      <td>0.352503</td>\n",
              "      <td>0.054127</td>\n",
              "      <td>0.205683</td>\n",
              "      <td>0.222598</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>139579.702866</td>\n",
              "      <td>1.000338e+00</td>\n",
              "      <td>1.000338e+00</td>\n",
              "      <td>1.000338e+00</td>\n",
              "      <td>1.000338e+00</td>\n",
              "      <td>1.000338e+00</td>\n",
              "      <td>1.000338e+00</td>\n",
              "      <td>1.000338e+00</td>\n",
              "      <td>1.000338e+00</td>\n",
              "      <td>0.481882</td>\n",
              "      <td>0.371386</td>\n",
              "      <td>0.477911</td>\n",
              "      <td>0.226345</td>\n",
              "      <td>0.404337</td>\n",
              "      <td>0.416131</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>10012.000000</td>\n",
              "      <td>-1.232678e+00</td>\n",
              "      <td>-7.863288e-01</td>\n",
              "      <td>-1.037963e+00</td>\n",
              "      <td>-9.699741e-01</td>\n",
              "      <td>-1.482152e+00</td>\n",
              "      <td>-7.161127e-01</td>\n",
              "      <td>-8.039171e-01</td>\n",
              "      <td>-1.256656e+00</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>98310.250000</td>\n",
              "      <td>-7.651850e-01</td>\n",
              "      <td>-7.863288e-01</td>\n",
              "      <td>-7.016827e-01</td>\n",
              "      <td>-6.218612e-01</td>\n",
              "      <td>-7.617310e-01</td>\n",
              "      <td>-5.490474e-01</td>\n",
              "      <td>-5.615195e-01</td>\n",
              "      <td>-7.124507e-01</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>316030.000000</td>\n",
              "      <td>-2.197765e-01</td>\n",
              "      <td>-3.089927e-01</td>\n",
              "      <td>-1.972629e-01</td>\n",
              "      <td>-3.433709e-01</td>\n",
              "      <td>-2.064062e-01</td>\n",
              "      <td>-3.382268e-01</td>\n",
              "      <td>-3.135238e-01</td>\n",
              "      <td>-2.362716e-01</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>370670.750000</td>\n",
              "      <td>5.593785e-01</td>\n",
              "      <td>2.876773e-01</td>\n",
              "      <td>4.752968e-01</td>\n",
              "      <td>1.904022e-01</td>\n",
              "      <td>5.440327e-01</td>\n",
              "      <td>1.311472e-01</td>\n",
              "      <td>1.925443e-01</td>\n",
              "      <td>4.439844e-01</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>453777.000000</td>\n",
              "      <td>3.364336e+00</td>\n",
              "      <td>5.299706e+00</td>\n",
              "      <td>4.678795e+00</td>\n",
              "      <td>5.458511e+00</td>\n",
              "      <td>4.611412e+00</td>\n",
              "      <td>1.359184e+01</td>\n",
              "      <td>1.044154e+01</td>\n",
              "      <td>6.056096e+00</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-353ea684-a812-43d7-b67f-8cad5aa56fb9')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-353ea684-a812-43d7-b67f-8cad5aa56fb9 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-353ea684-a812-43d7-b67f-8cad5aa56fb9');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VCaA3YQStIc2"
      },
      "source": [
        "Now we are ready to apply general models!"
      ]
    }
  ]
}