Bank_loan_classification_analysis.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyPXN8E9xadJcUkKb9kyxWOv",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/AntonioLunardi/Loan-Classification-Prediction-Competition-Case/blob/main/Bank_loan_classification_analysis.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 0.0 Preconfiguration"
      ],
      "metadata": {
        "id": "d8I6EoSPU56R"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Make installations\n",
        "# !pip install graphviz==0.9\n",
        "# !pt-get install graphviz"
      ],
      "metadata": {
        "id": "Ky_cl7FQLr1T"
      },
      "execution_count": 208,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": 209,
      "metadata": {
        "id": "xIs8f3KUr8jn"
      },
      "outputs": [],
      "source": [
        "# Import libraries\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "from scipy.stats import norm, skew \n",
        "from pandas import get_dummies"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# ML libraries\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "from sklearn.dummy import DummyClassifier\n",
        "from sklearn.impute import SimpleImputer\n",
        "from sklearn.linear_model import LogisticRegression, Perceptron\n",
        "from sklearn.svm import SVC\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.model_selection import train_test_split, GridSearchCV, cross_val_score, cross_validate, KFold\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from xgboost import XGBClassifier\n",
        "import lightgbm as lgb"
      ],
      "metadata": {
        "id": "z6-ZnVXk-npG"
      },
      "execution_count": 210,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Avoid generating warning messages\n",
        "import warnings\n",
        "warnings.filterwarnings(\"ignore\") \n",
        "\n",
        "# Styles\n",
        "theme = sns.set_theme(style='whitegrid')\n",
        "colors = sns.color_palette('pastel')\n",
        "\n",
        "# Standardize random parameters for obtaining constant results\n",
        "SEED = 452"
      ],
      "metadata": {
        "id": "KQV2EfagOqrr"
      },
      "execution_count": 211,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 1.0 EDA (Exploratory data analysis)"
      ],
      "metadata": {
        "id": "hL2Ng4V3V5AR"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1.1 Data overview"
      ],
      "metadata": {
        "id": "UsDnbgWZTrau"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Read the dataframe\n",
        "df_train = pd.read_csv('/content/train.csv')\n",
        "df_test = pd.read_csv('/content/test.csv')"
      ],
      "metadata": {
        "id": "pJz9tcygtBwD"
      },
      "execution_count": 212,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df_test.columns"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FNdrWtrha64f",
        "outputId": "d1ec1e54-76d5-4f92-bb74-1d37cb5baf49"
      },
      "execution_count": 213,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['Loan_ID', 'Gender', 'Married', 'Dependents', 'Education',\n",
              "       'Self_Employed', 'ApplicantIncome', 'CoapplicantIncome', 'LoanAmount',\n",
              "       'Loan_Amount_Term', 'Credit_History', 'Property_Area'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 213
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check the number of columns and rows\n",
        "df_train.shape"
      ],
      "metadata": {
        "id": "WLXYpvFhtQfk",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "851cad40-9578-46b3-e05d-39c7189bc7c7"
      },
      "execution_count": 214,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(614, 13)"
            ]
          },
          "metadata": {},
          "execution_count": 214
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df_train"
      ],
      "metadata": {
        "id": "lFXojTZctY9t",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "outputId": "7a30fd3b-5ff6-4b88-dfbe-5621a7007ae4"
      },
      "execution_count": 215,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Loan_ID  Gender Married Dependents     Education Self_Employed  \\\n",
              "0    LP001002    Male      No          0      Graduate            No   \n",
              "1    LP001003    Male     Yes          1      Graduate            No   \n",
              "2    LP001005    Male     Yes          0      Graduate           Yes   \n",
              "3    LP001006    Male     Yes          0  Not Graduate            No   \n",
              "4    LP001008    Male      No          0      Graduate            No   \n",
              "..        ...     ...     ...        ...           ...           ...   \n",
              "609  LP002978  Female      No          0      Graduate            No   \n",
              "610  LP002979    Male     Yes         3+      Graduate            No   \n",
              "611  LP002983    Male     Yes          1      Graduate            No   \n",
              "612  LP002984    Male     Yes          2      Graduate            No   \n",
              "613  LP002990  Female      No          0      Graduate           Yes   \n",
              "\n",
              "     ApplicantIncome  CoapplicantIncome  LoanAmount  Loan_Amount_Term  \\\n",
              "0               5849                0.0         NaN             360.0   \n",
              "1               4583             1508.0       128.0             360.0   \n",
              "2               3000                0.0        66.0             360.0   \n",
              "3               2583             2358.0       120.0             360.0   \n",
              "4               6000                0.0       141.0             360.0   \n",
              "..               ...                ...         ...               ...   \n",
              "609             2900                0.0        71.0             360.0   \n",
              "610             4106                0.0        40.0             180.0   \n",
              "611             8072              240.0       253.0             360.0   \n",
              "612             7583                0.0       187.0             360.0   \n",
              "613             4583                0.0       133.0             360.0   \n",
              "\n",
              "     Credit_History Property_Area Loan_Status  \n",
              "0               1.0         Urban           Y  \n",
              "1               1.0         Rural           N  \n",
              "2               1.0         Urban           Y  \n",
              "3               1.0         Urban           Y  \n",
              "4               1.0         Urban           Y  \n",
              "..              ...           ...         ...  \n",
              "609             1.0         Rural           Y  \n",
              "610             1.0         Rural           Y  \n",
              "611             1.0         Urban           Y  \n",
              "612             1.0         Urban           Y  \n",
              "613             0.0     Semiurban           N  \n",
              "\n",
              "[614 rows x 13 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-efbf2464-f0fb-4821-989a-79eb22189f9c\">\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>Loan_ID</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Married</th>\n",
              "      <th>Dependents</th>\n",
              "      <th>Education</th>\n",
              "      <th>Self_Employed</th>\n",
              "      <th>ApplicantIncome</th>\n",
              "      <th>CoapplicantIncome</th>\n",
              "      <th>LoanAmount</th>\n",
              "      <th>Loan_Amount_Term</th>\n",
              "      <th>Credit_History</th>\n",
              "      <th>Property_Area</th>\n",
              "      <th>Loan_Status</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>LP001002</td>\n",
              "      <td>Male</td>\n",
              "      <td>No</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>5849</td>\n",
              "      <td>0.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>LP001003</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>1</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>4583</td>\n",
              "      <td>1508.0</td>\n",
              "      <td>128.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Rural</td>\n",
              "      <td>N</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>LP001005</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>Yes</td>\n",
              "      <td>3000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>66.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>LP001006</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>0</td>\n",
              "      <td>Not Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>2583</td>\n",
              "      <td>2358.0</td>\n",
              "      <td>120.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>LP001008</td>\n",
              "      <td>Male</td>\n",
              "      <td>No</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>6000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>141.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>609</th>\n",
              "      <td>LP002978</td>\n",
              "      <td>Female</td>\n",
              "      <td>No</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>2900</td>\n",
              "      <td>0.0</td>\n",
              "      <td>71.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Rural</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>610</th>\n",
              "      <td>LP002979</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>3+</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>4106</td>\n",
              "      <td>0.0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>180.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Rural</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>611</th>\n",
              "      <td>LP002983</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>1</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>8072</td>\n",
              "      <td>240.0</td>\n",
              "      <td>253.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>612</th>\n",
              "      <td>LP002984</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>2</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>7583</td>\n",
              "      <td>0.0</td>\n",
              "      <td>187.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>613</th>\n",
              "      <td>LP002990</td>\n",
              "      <td>Female</td>\n",
              "      <td>No</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>Yes</td>\n",
              "      <td>4583</td>\n",
              "      <td>0.0</td>\n",
              "      <td>133.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>Semiurban</td>\n",
              "      <td>N</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>614 rows × 13 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-efbf2464-f0fb-4821-989a-79eb22189f9c')\"\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-efbf2464-f0fb-4821-989a-79eb22189f9c 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-efbf2464-f0fb-4821-989a-79eb22189f9c');\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": 215
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "This is appearently a cleaned and ordered data frame. Although it is necessary to do tests to check if all registers are consistent"
      ],
      "metadata": {
        "id": "qp9O4e9BtzsL"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df_train.columns"
      ],
      "metadata": {
        "id": "POSsYF32tbnW",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "453dcc17-5b59-4cf0-f01f-2128861ca0fb"
      },
      "execution_count": 216,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['Loan_ID', 'Gender', 'Married', 'Dependents', 'Education',\n",
              "       'Self_Employed', 'ApplicantIncome', 'CoapplicantIncome', 'LoanAmount',\n",
              "       'Loan_Amount_Term', 'Credit_History', 'Property_Area', 'Loan_Status'],\n",
              "      dtype='object')"
            ]
          },
          "metadata": {},
          "execution_count": 216
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "The index column is already present"
      ],
      "metadata": {
        "id": "PfUDdHWpusNg"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1.2 Numerical data summary"
      ],
      "metadata": {
        "id": "Wiy4kaHWDTrE"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Describe statistically\n",
        "df_train.describe()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "JfmQ1SurDZ8m",
        "outputId": "27831d93-73e1-4a71-afdd-e8bfa05135c4"
      },
      "execution_count": 217,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       ApplicantIncome  CoapplicantIncome  LoanAmount  Loan_Amount_Term  \\\n",
              "count       614.000000         614.000000  592.000000         600.00000   \n",
              "mean       5403.459283        1621.245798  146.412162         342.00000   \n",
              "std        6109.041673        2926.248369   85.587325          65.12041   \n",
              "min         150.000000           0.000000    9.000000          12.00000   \n",
              "25%        2877.500000           0.000000  100.000000         360.00000   \n",
              "50%        3812.500000        1188.500000  128.000000         360.00000   \n",
              "75%        5795.000000        2297.250000  168.000000         360.00000   \n",
              "max       81000.000000       41667.000000  700.000000         480.00000   \n",
              "\n",
              "       Credit_History  \n",
              "count      564.000000  \n",
              "mean         0.842199  \n",
              "std          0.364878  \n",
              "min          0.000000  \n",
              "25%          1.000000  \n",
              "50%          1.000000  \n",
              "75%          1.000000  \n",
              "max          1.000000  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-7c83d842-ad81-4f3d-8d80-6eb065864e10\">\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>ApplicantIncome</th>\n",
              "      <th>CoapplicantIncome</th>\n",
              "      <th>LoanAmount</th>\n",
              "      <th>Loan_Amount_Term</th>\n",
              "      <th>Credit_History</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>614.000000</td>\n",
              "      <td>614.000000</td>\n",
              "      <td>592.000000</td>\n",
              "      <td>600.00000</td>\n",
              "      <td>564.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>5403.459283</td>\n",
              "      <td>1621.245798</td>\n",
              "      <td>146.412162</td>\n",
              "      <td>342.00000</td>\n",
              "      <td>0.842199</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>6109.041673</td>\n",
              "      <td>2926.248369</td>\n",
              "      <td>85.587325</td>\n",
              "      <td>65.12041</td>\n",
              "      <td>0.364878</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>150.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>9.000000</td>\n",
              "      <td>12.00000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>2877.500000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>100.000000</td>\n",
              "      <td>360.00000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>3812.500000</td>\n",
              "      <td>1188.500000</td>\n",
              "      <td>128.000000</td>\n",
              "      <td>360.00000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>5795.000000</td>\n",
              "      <td>2297.250000</td>\n",
              "      <td>168.000000</td>\n",
              "      <td>360.00000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>81000.000000</td>\n",
              "      <td>41667.000000</td>\n",
              "      <td>700.000000</td>\n",
              "      <td>480.00000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-7c83d842-ad81-4f3d-8d80-6eb065864e10')\"\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-7c83d842-ad81-4f3d-8d80-6eb065864e10 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-7c83d842-ad81-4f3d-8d80-6eb065864e10');\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": 217
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df_train.describe(include = ['object'])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "2Fnw3-UAEKJ2",
        "outputId": "84910778-1465-4fd5-b7c1-7bdf67d71043"
      },
      "execution_count": 218,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "         Loan_ID Gender Married Dependents Education Self_Employed  \\\n",
              "count        614    601     611        599       614           582   \n",
              "unique       614      2       2          4         2             2   \n",
              "top     LP001002   Male     Yes          0  Graduate            No   \n",
              "freq           1    489     398        345       480           500   \n",
              "\n",
              "       Property_Area Loan_Status  \n",
              "count            614         614  \n",
              "unique             3           2  \n",
              "top        Semiurban           Y  \n",
              "freq             233         422  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-e9008dd2-0c06-4b85-a7c0-414d3a8499c5\">\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>Loan_ID</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Married</th>\n",
              "      <th>Dependents</th>\n",
              "      <th>Education</th>\n",
              "      <th>Self_Employed</th>\n",
              "      <th>Property_Area</th>\n",
              "      <th>Loan_Status</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>614</td>\n",
              "      <td>601</td>\n",
              "      <td>611</td>\n",
              "      <td>599</td>\n",
              "      <td>614</td>\n",
              "      <td>582</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>unique</th>\n",
              "      <td>614</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>top</th>\n",
              "      <td>LP001002</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>Semiurban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>freq</th>\n",
              "      <td>1</td>\n",
              "      <td>489</td>\n",
              "      <td>398</td>\n",
              "      <td>345</td>\n",
              "      <td>480</td>\n",
              "      <td>500</td>\n",
              "      <td>233</td>\n",
              "      <td>422</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e9008dd2-0c06-4b85-a7c0-414d3a8499c5')\"\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-e9008dd2-0c06-4b85-a7c0-414d3a8499c5 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-e9008dd2-0c06-4b85-a7c0-414d3a8499c5');\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": 218
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "The data frame doesn't have any duplicated register of loan"
      ],
      "metadata": {
        "id": "UY4lioUjiBW2"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1.3 Categorial and numerical features discrimination"
      ],
      "metadata": {
        "id": "p7mNUJkgdFUy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Separe columns in two types to ease the analysis\n",
        "\n",
        "categorical_cols = df_train.dtypes == 'object'\n",
        "categorical_cols = list(categorical_cols[categorical_cols].index)\n",
        "categorical_cols.remove('Loan_ID') # Not useful as a category\n",
        "categorical_cols.remove('Loan_Status') # Target variable\n",
        "\n",
        "numeric_cols = df_train.dtypes != 'object'\n",
        "numeric_cols = list(numeric_cols[numeric_cols].index) "
      ],
      "metadata": {
        "id": "xStSdlNFdPp-"
      },
      "execution_count": 219,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "numeric_cols"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_QaXESuFeSAU",
        "outputId": "cdb2939e-142f-401a-d11b-4816a71405aa"
      },
      "execution_count": 220,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['ApplicantIncome',\n",
              " 'CoapplicantIncome',\n",
              " 'LoanAmount',\n",
              " 'Loan_Amount_Term',\n",
              " 'Credit_History']"
            ]
          },
          "metadata": {},
          "execution_count": 220
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "categorical_cols"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gCxOWG2leFqN",
        "outputId": "7189688e-9b96-43b9-f6ae-891be4823168"
      },
      "execution_count": 221,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Gender',\n",
              " 'Married',\n",
              " 'Dependents',\n",
              " 'Education',\n",
              " 'Self_Employed',\n",
              " 'Property_Area']"
            ]
          },
          "metadata": {},
          "execution_count": 221
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1.4 Visualization"
      ],
      "metadata": {
        "id": "eFpqn6FfM_cm"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 1.4.1 Loan status (target)"
      ],
      "metadata": {
        "id": "Jkf-P5XCWJ6F"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Pie chart\n",
        "df_train['Loan_Status'].value_counts().plot\\\n",
        "  .pie(autopct = '%1.1f%%',figsize =(6, 6),explode=(0.01,0.01),colors=colors)\\\n",
        "  .set_title('Loan Status Percentage').set_fontsize('16')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "6YrplxdzZtAR",
        "outputId": "9050c245-209f-4bde-b5c0-28b63edc5c2e"
      },
      "execution_count": 222,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x432 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWUAAAFmCAYAAAC1EYsVAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAAsTAAALEwEAmpwYAAA8MklEQVR4nO3deXhU5fn/8fc5Z2ay75CNALIlQEIIOwqogFpQEKHiVlG0lqpAa1Us1qUWa7Xf1qrg+tOitla7uG+tVK2KVDbFsggIGCAs2UhCyDKZ5ZzfH4cMCQRIYCZzZuZ+XReXzmQyc89k8skzz3me+yiGYRgIIYSwBDXYBQghhDhCQlkIISxEQlkIISxEQlkIISxEQlkIISxEQlkIISxEQjkIXn/9dfLy8ti1a1ewS2mXuro6Fi9ezIUXXkhRUREjRoxg6tSp3HvvvRw4cMB3uxdeeIFly5ad8uO8/vrrvPrqq/4ouUMWLlxIXl6e79/o0aP5wQ9+wGeffdbptfhTsF5PcXpswS5AWJvX62X27Nns3buXH/3oRwwYMICGhga2bdvGe++9R3l5OWlpaQD86U9/YujQoVxwwQWn9FhvvPEGHo+HSy+91J9PoV1SU1N56qmnAKisrGTp0qXMmTOH559/njPPPLPT6/GHYL6e4tRJKIsTWr16NRs2bOCJJ57gvPPO810/ceJEbrzxRnRdD2J1/mO32ykqKvJdHj16NOeeey4vvvjiaYWyYRi43W4cDocfqhSRQKYvLMrtdvPII48wYcIECgoKmDBhAo888ghut7vV7RYvXsz06dMZOnQoo0aN4pprruHrr79udZtVq1aRl5fHRx99xKJFixg1ahSjRo3i9ttvp7a29oR1HDx4EICuXbu2+XVVNd9CEyZMYO/evbzzzju+aYCFCxcCsGvXLhYsWMCECRMoLCxk4sSJ/PKXv/TdN8CsWbNYvXo1X331le/7Z82aBcCSJUvIy8s75rEXLlzIhAkTfJc9Hg+PPvoo5513HoMGDWLUqFFceeWVrF279oTPsS3x8fGcccYZ7N6923fdsmXLuOyyyxg8eDDDhw/nJz/5Cfv27Wv1fRMmTOD222/n1VdfZdKkSRQUFPDpp58CsGXLFubOncuoUaMoLCzke9/7Hs8880yr7+/IY7z33ntMnjyZoqIiZsyY0ep5nuj1rKqq4t577+V73/segwcP5pxzzuG2226jrKzsmNfh3XffZdKkSQwaNIipU6fy0UcfMWvWLN99NWu+z3HjxlFQUMCkSZP429/+1uHXXchI2bIWLlzIP//5T3784x8zbNgw1q1bx9NPP82ePXt4+OGHfbcrKyvj2muvJTMzk8bGRt5++22uvvpqXnvttWOC7IEHHmD8+PE8/PDDFBcX87vf/Q5N0/jtb3973Dry8/Ox2Wzce++9zJs3j5EjR5KUlHTM7R5//HHmzJlDXl4e8+fPB8wpAYDy8nKysrL4xS9+QVJSEiUlJTzzzDPMmTPH94v7y1/+kgULFuD1elm0aBFgBmNHPPvss7z44ovccsstDBgwgLq6OjZu3Ngq/NvL4/FQWlpKTk4OAK+88gr33XcfM2bMYO7cudTX17NkyRKuvvpq3n777Va1rlq1ii1btjBv3jzS0tLo1q0b69evZ9asWfTo0YM777yTjIwMdu3axdatW33f15HH+PLLLykuLuanP/0pUVFRPPbYY9x44418/PHHJCYmnvD1rKmpweFwcOutt5Kamkp5eTlLly7lyiuv5J///CdRUVEArFixgttvv50JEyZw5513UlVVxW9+8xuampro1auXr5a6ujquvPJKmpqamD9/Pjk5OSxfvpz77rsPl8t1TICLkzBEp3vttdeM3NxcY+fOnW1+fevWrUZubq6xePHiVtc/8cQTRm5urrF58+Y2v8/j8Rhut9u44IILjPvvv993/cqVK43c3FzjjjvuaHX7X/3qV0ZBQYGh6/oJ6/373/9uFBUVGbm5uUZeXp5x4YUXGg899JBRWlra6nbjx483brvtthPel2EYhtvtNtasWWPk5uYamzZt8l1/9dVXG1dcccUxt1+8eLGRm5t7zPU///nPjfHjx/suz5kzx5g7d+5JH7+t+xk3bpzhdrsNt9tt7N+/37jnnnuM3Nxc4/nnnzfq6uqMoUOHGgsXLmz1fbt37zby8/ON559/3nfd+PHjjcLCQqO8vLzVba+66irj7LPPNhoaGtqsoaOPMXz4cKOmpsZ33fr1643c3Fzj7bff9l13vNfzaB6Px9i3b5+Rm5trLFu2zHf95Zdfblx00UWt3h8bNmwwcnNzjauvvtp33eOPP24UFBQYxcXFre73rrvuMkaOHGm43e6T1iCOkOkLC1qzZg0AF198cavrmy83fx3gv//9L7NmzWLUqFEMHDiQ/Px8du7cSXFx8TH3e84557S6nJubi8vlorKy8oT1zJw5k08++YTf//73XHbZZei6ztKlS5kyZQrbtm076fNxuVw8/fTTTJo0icLCQvLz8/nBD34A0Gadp2rQoEF8+umnPPLII6xduxaXy9Xu7y0rKyM/P5/8/HzOOecc3nnnHX7yk5/4poPq6uq4+OKL8Xg8vn9ZWVn06tXrmOmRwYMHt5ruaWxs5KuvvmLq1KnExMS0+fgdfYyioqJWn1iaPxXt37+/Xc/35Zdf5uKLL2bIkCEMHDiQc889Fzjy8/B6vWzcuJELLrgARVF831dQUOD79NBs+fLlDB48mJycnFa1jx07lpqaGrZv396umoRJpi8s6HjzuM2Xm7++adMm5syZw9ixY3nggQfo2rUrqqpy9913txlIycnJrS43H3xqamo6aU1JSUlMnTqVqVOnAvDhhx8yf/58lixZwuLFi0/4vX/4wx946aWXuPnmmxkyZAhxcXGUlZUxb968dj12e/34xz/G4XDwzjvv8PTTTxMbG8ukSZNYsGCBbyrleNLS0njmmWdQFIXk5GSysrLQNA3At+xv9uzZbX7v0dM5R//camtr0XWdzMzM4z5+Rx/j6Msd+Vn++c9/5te//jXXXXcdY8eOJTExEcMwuOyyy3zfX11djdvt9q2saalLly6tLldVVbFr1y7y8/PbfLyampqT1iSOkFC2oOZfuMrKSnr06OG7vqKiotXXly1bhqZpLFmyBLvd7rtdbW0tiYmJAa3xvPPOo3///u0aBb333ntMmzaNm2++2XfdypUr2/1YzXOcLper1SqGo3/Z7XY7c+bMYc6cOVRUVPDJJ5/w4IMP0tjYyKOPPnrCx7DZbAwaNKjNrzX/MXvooYfo27fvMV+Pi4trdbnlyBIgMTERVVXbPJB2qo9xOt577z3OPPNM34FYgJKSkla3SUlJwW63t1qH3qyyspLs7Gzf5eTkZFJTU7nrrrvafLyW88/i5CSULWjEiBGA+ctz0003+a5/5513ABg5ciRgfixWVbVVCHzxxRfs27fvmI+Yp6q6upr4+PhWoQ/Q0NDA/v37Wx1MtNvtbY7UnE4nNlvrt9rrr79+zO0cDgf19fXHXN8cANu2bfONxmpra1m3bt1xw6pr167MnDmTTz/9tF1TLCcydOhQ4uLi2LVrF9OnT+/w98fExDBs2DDefvtt5s6dS3R0tN8foy3Hez2dTucxB1GP/nlomkZBQQHLli1j/vz5vvfYxo0b2bNnT6tQHjduHC+99BLZ2dltjqxFx0goB9Hy5cvZvHlzq+sSEhIYM2YMU6ZM4fHHH8fr9TJkyBDWrVvHU089xZQpU3xBOG7cOF588UUWLlzI97//fYqLi3nyySfJyMjwW42rVq3iN7/5DVOnTmXo0KEkJiayd+9eXnrpJQ4ePMh1113nu23fvn1Zu3Yt//nPf+jSpQspKSnk5OQwbtw43nzzTXJzc+nZsyfLli1j3bp1xzxWnz59ePnll3n//ffp3r07cXFx9O7dm7PPPpuEhATuuece5s+fj8vl4rnnniM2NrbV9990003079+f/Px8EhMT+eabb1i+fDmXX375ab0G8fHx3HHHHSxatIiqqipfPWVlZaxZs4aRI0f6pnWO54477mDWrFlcfvnlXHfddWRmZlJSUsKWLVu45557/PIYRzve6zlu3DieffZZnn76aQoLC1m5ciUffPDBMd8/f/58rr/+eubOncvll19OdXU1S5YsoWvXrq0GArNnz+b999/nqquuYvbs2fTq1YvGxka+++471q5d69uUI9pHQjmI7r///mOu69evH++++y4PPvggOTk5vPbaazz11FOkp6dzww03MG/ePN9tx40bx913383zzz/PsmXL6NevH//3f//n11+CoqIipk2bxsqVK3n99depra0lISGBQYMGsXTp0lYbK2699VbuuecebrnlFpxOJ9OnT+ehhx7i7rvvxjAM3xTC2WefzcMPP8zMmTNbPdaPfvQjiouLueuuu2hoaGDkyJH8+c9/JjExkaeffpoHH3yQW265hczMTG6++Wa++OILVq9e7fv+ESNG8K9//YuXX36ZxsZGsrKyuOGGG7jxxhtP+3W44ooryMrK4rnnnuPdd9/F6/WSkZHBsGHDGDBgwEm/v7CwkFdeeYXFixfz61//GpfLRXZ2NjNmzPDbYxzteK/n3Llzqa2t5YUXXqCpqYmRI0fy3HPPtdocBDBmzBh+//vf8/jjjzN37lx69uzJwoULeeKJJ0hISPDdLiEhgb/+9a888cQTPPvss5SXl5OQkECvXr1OeXdnJFMMQ04HJYRon9LSUs4//3xuvPFG5s6dG+xywpKMlIUQbXI6nTz44IOcddZZpKSkUFJSwnPPPUdMTMwxn3KE/0goCyHapKoqlZWV3H///dTU1PgOWD722GOkp6cHu7ywJdMXQghhIbKjTwghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLERCWQghLMQW7AKE6AjdMPDqYBigKKAp5n+9Onh08OgGHh28uoHHC17DAEBVFFSFo/4pKAqoKjg0BZtq3o9ugIE5YlFV0FQlqM9ZRBYJZWE5Hq+BboB2+HOc021Q7zI45NQ56DRocBnUu3TqXQZOtxnC/qAAUXaFaBvE2BWi7QrRNoUYh0JitEpStEp8lIIB6LoZ7DZNAlv4l2IYh4cSQnQy3TBHs8rhkeuhJoMDdV4q6nSqGnVqG3Vc3mBXeawYu0JStEJSjEpKrPkvMVr1jbRtmjkKF+JUSCiLTuPVDbyGGcBV9Trlh7wcaNCpbtA55DQI9TditA26xmtkJKhkJWkkRqt4ZUQtOkhCWQRM80hYU6G6Qaekxsv+Wi8H6nUi4V2nKpASq9I1XiU7SSM9XvN9KpB5anE8EsrCr9xeA1WBWqdOSbUZwhV1Orq8ywBIjVXpnqxxRpqN+CgF3QC7jKJFCxLK4rQ1B/GBep3tFR5Kajw0eYJdlfVF2xW6JWn0StNIT9DQZT5aIKFsOXfeeSepqaksWLDAd93s2bO54IILuOqqq4JYWWvNQVxZp7O90kNJtceSB+VChapARoJGv642clI0DBlBRywJZYs5dOgQ06ZN45FHHmHw4MH89a9/5YMPPmDp0qUoQR5BeXXzYFxVvc635R721EgQB4JNhR4pNvpn2EiOVVGQOehIIqFsQStWrOCBBx7giSeeYPbs2bzyyitkZ2cHrR6Xx3yLbKtws7XMQ51L3jKdJc6h0KeLjdx0G3ZNQVNleiPcyeYRCxozZgwjRozg0ksv5c477wxKIHt1A8OAmkadTaVuSqq9crAuCOpdBuv3uVm/z03XeJX8TDvZyZqMnsOYjJQtavfu3Vx66aWsXr26Ux/X7TXfDjsqPWwpc1PrlLeH1cQ5FAZm2unX1YaBzD2HGxkpW5Sqqp06h+z2miPjjfvdbClz+23rsvC/epfBmt0uvt7rol9XGwVZDjRVwjlcSChHOPfhPhPr97r4ttyDVwbGIcPthW9KPWwu83BGqkZhtoNYhyLhHOIklCOU22t2W/t6j4vtlR6ZLw5hhgHFB7wUH2ikW5LGiJ4OYuwSzqFK5pQjjNtr4PEafLXHzXcHPBGx3TkS9UzRGN7DgcMm4RxqJJQjhFc3pyk27HPzTalbRsYRQAH6dLUxrLsDTZoihQwJ5TBnHG4Kv7vaw9rdLpyy/Tni2FTIz7STn2U3TwwgS+ksTUI5jLm9BrVOnS+KXVQ1yHKKSBdtVxjd00FWkiZTGhYmoRyGPF4Dt26waqeL3dWyD1q0lp2kMaaXA7umyJSGBUkohxHDMJvIby1z8/UetyxvE8elqVDUzU5euh1Vtm5bioRymPB4zXPXfbajSaYqRLslxyiM7R1FQrQqUxoWIaEc4poP5G3c72bDfrcscROnJLerjWE9zFUaqhwIDCoJ5RDm9ppnc/5kWxPVjTI6FqcnzqEwPjeKhCgZNQeThHKI8ngNtpa7WbdH1hwL/1GAohw7AzLschAwSCSUQ4xXN/Do8Mk2J2WHZHQsAiM9QeXcvtHYNVnX3NkklEOIx2tw0Knz8bdNNLrlxyYCy6HBuD5RpCfIuubOJKEcIjxeg20VbtaWyME80blyu9oY3sNsDxrsU5JFAglli9MPr674/LsmSmQjiAiS1FiV8/JkOqMzSChbmMdr0OA2+Girk0NN8mMSwRVjVzgvL5qEKNkJGEgSyhbl9hrsr/WyfEcTXjmeJyxCU2BsnyiypX9GwEgoW1Dz/PGa3e5glyJEmwqz7RRkybK5QJBQthiP12DdHheby6THprC27ikaY3tHYZMDgH4loWwhHq/B5981SWc3ETJSYlUu6G8eAJSmRv4hoWwBumFuCPloq5OKOplAFqElPkph8sBoojRF+mb4gYRykHl1A5fH4IMtTmqd8qMQoSnWYQZztE2RJXOnSUI5iHTdoNFt8P43TtmhJ0JetF1h8oBoYh0SzKdDQjlIJJBFOIqywfcGxJAQJcF8qiSUg0DXDZweg/c2SSCL8GNX4YIB0STFqNgkmDtMDXYBkaY5kN+XQBZhyq3DB5udVNfreKSvbIdJKHciXTdo8phTFg0SyCKMeXT491YndU4DrwRzh0god5JWgeySN6kIfx4dPtjSiNNtoEswt5uEcicwDAO3Du9/46ReAllEkCYP/GuzE5fX/D0QJyeh3Ak8Ovx7iwSyiEz1LoMPNjfilo2q7SKhHGAer8FnO5qoapCdeiJyHXQafLjVidsrA5OTkVAOII/X4Ks9LvbWyBBBiMp6nU+2NeGRYD4hCeUAcXsNtld62CLd3oTw2V/rZfUul4yYT0BCOQA8ukFFnZc1u1zBLkUIy9le6WHnAY+MmI9DQtnPdN2grsngk21NyFtOiLat2uXioFOXpXJtkFD2s+aVFh45rifEcekGfPRtk6zIaIOEsh81r7SQ7dNCnJzTbfDRt06ZxjiKhLKfuL0GW8rc7Dsof/qFaK/Kep3Vu1wSzC1IKPuBVzc42Kizbo+c6FSIjtpe6aFYDvz5SCj7gVeH/8iBPSFO2apdLhrdhmzFRkL5tHm8Bp9ulzacQpwO3YD/bHPilQPkEsqnw+012FzmZn+tvJOEOF01jQbr97kjfmOJhPIpMgyD+iaDr2UeWQi/2bTfzcHGyF6/LKF8irw6fLbDKfPIQviRAXy6vYlIHixLKJ8Cj9fgm1I3NY0R/M4RIkDqXUZE98eQUO4gwzDPQr1+n0xbCBEoOyo9lB/yRuSppCSUO8ictmgiAt8rQnSqFd9F5u+ZLdgFhBKP12BbhZsD9eGx2mLTFx+w/M1nqK0sJS65C1Pn3EePvKF8s2oZn73+NIeqyklMzeDcmfPIGz6+zft4ZuGlHKzc77vscbvoU3gWl9/2GM6GQ7zx+M/Zu2MjfQeP4+IbF6GqGgDv/fF++hSeRf8REzvluYrQ4/TAl7tdDOvhwK4pwS6n00god4DLa/BVmKy2+G7DSj7+22PMmPcQ2b0LOFRTCUBtVTlvPXU3M3/2CH0Kz2L7/z7n9SU/Z94f3iUuKfWY+/nxQ6/6/t8wDJ64dSoDRp0PwLqPXyOjZ39m/uxR/vLgj9m69j8MGHkee7b9j7qaCglkcVLfVnjIy7CTHAOKEhnBLNMX7eT2Gny+oylsFrd/9vrTjLtkDt36FqKoKomp6SSmpnOoqozo2AT6Dh6Doij0KxqHIyqa6vKSk97n7i1f0VhXQ//hZtjWVOyj54Dh2OwOuucNoaZ8D7ru5d9/eZgLZt0R6KcowkQ4/d61h4RyO+i6QWWdl9JD4fHO0HUv+4u/oeFQNU/edjGLfzKJf734EG6Xk6zeA+mS3Ytvv/oUXfeyde1/0GwO0rvnnvR+13/+Dv2HT8QRHQNA15w+7Ny0CrfLScnWdXTJ6cOaD16hT+EYUtJzAv00RZiobtT57oAnYg76SSi3g465Nz9c1B+sQvd62LzmQ2bd80dueOAVynZt5fO3nkNVNQaNncKbT/6Ch64bzZtP3cXk6+/yBe3xuJsa2bL6IwrPnuq7ruicS3A21PHCfdfQPW8IGT1y2bDifUZOuor3n3+AP/36h3zyjycC/XRFGPiqxBUxa5dlTvkkvLrBd5Ueap3h846wOaIAGHH+FSQkdwVg1OSr+fyt5zhjwAg+/ttjXP2LZ8k6oz/7d27m73/4GVcsWEJmz7zj3ueWtR8TE59Ij/7DWj3ORT+8x3f5tcV3MP6yuWz87z8xdJ1Zdz3HK7+9mR3rV9CncEyAnq0IBy4vfBUhB/1kpHwSugHr9oTPKBkgJi6RhNQMaOPASdmurXTPG0p274Eoqkp273y69Slg58ZVJ7zPDcvfZdDYKcc9GLNj/QrAoE/hGCpKtpPVayCKopDVayDlu7f542mJMLetwoMzAhp/SSifgNtrsH6fi6YwPCH14LMvZu2yv1J/sIrG+lpW/+sv9C0aR1bvfEq2rqN011YASnduoeTbdaT36Hfc+6qtKmPn5rUUjp3S5tc9riY+/tsSzr/6dgCSu2aze8uXeD1uSrZ9TbLML4t2MIA1u8N/p59MX5yARzfYXBqGiQyMnXYDDYdqeOqOS7DZoxgw8nzGXvxDbI4oxs2Yw+uLF1BfW0VsQgpnTb2e3oPOBGDjivdZ8c7SVkvhNnz+Hjl9B5GS0b3Nx1rxzlIKzppMYmoGAEMmfJ/Xl/ycR26eSN+iscddAy3E0fbUeKl3GSTHhO8UhmJIV+k2ub0G/y1uYleVnN5JCCvJTtI4p29U2M4ty/TFcTS6DQlkISxo30Evh5rCY3lqWySU2+D2GnxVEl4H94QIJ2vDeG5ZQrkNTrfB7moZJQthVaW1OrVOPSzP6SehfBS31wi7JXBChKM1u114wnAWQ0L5KB6vzCULEQrKD+kcbAy/VJZQbsFcl+yWUzwJESLC8USrEsotGMD2ivBclyxEONpb48UTZo2KJJQPMzeKuCOm6YkQ4cDAPAO2J4x+cSWUD1OArWXh0cBeiEiyvcJj/gKHCQllzDNmlNZ6ccrMhRAhx+WFnQc86GGyPE5CGXDrsLlMElmIULWp1I0eJgsxJJQBXYf9B2UZnBCh6mCjQXWYLI+L+FD2eA22lssyOCFC3cb9blxhcMAv4kNZUczm2UKI0La3xhsWx/siPpQr6nQaXKH/11WISKcbUFId+gf8IjqUXR5zbbIQIjxsq/DgCfHDQxEdyooCe+QAnxBho/yQHvLHhyI2lA3DYG+NlxD/pCOEaMEAikN8zXLEhrLba/7whBDhZXuFB28Ir46L2FDWVPO0MkKI8FLVoId057iIDeXSWq80HxIiTJmj5dD8BY/IUHZ7DL6TqQshwlZJjZcQzeTIDGVVhT01MnUhRLg6UB+6k8oRGcoH6nXckslChLXS2tD8JY+4UPZ4DVl1IUQE2FXtDckDfhEXygah+xdUCNF++w96UUOwGUZEhvJBZ+j99RRCdEyj26DRHXq/6xEXyhWHZJQsRKTYXe3BCLHdfR0K5e3bt1NZWQlAfX09ixcv5vHHH6exsTEgxfmbRzfYKxtGhIgYe2u8IXdQv0OhfOutt1JbWwvAb3/7W9asWcPXX3/NvffeG5Di/M0wZD5ZiEhSdkjHFmLzAbaO3Hjv3r307t0bwzD497//zXvvvUd0dDQTJ04MVH1+pQA1jaH1UUYIcep0A+pcBonRoXPEr0OhHBUVRV1dHTt27CArK4vU1FQ8Hg9NTU2Bqs+vKupllCxEpKk45CUxOnSGyx0K5SlTpnDttddSX1/P1VdfDcA333xDTk5OQIrzJ69uyC4+ISJQeZ1Oj1QDuxYao2XF6OChyc8//xybzcbo0aMB2LBhA3V1dZx55pkBKdBfXB6Dj791Ul4XutsvhRAdlxKr8r0B0TjCNZRDlVc3+Me6BlwyWBYioigKXDUsFi1EdpJ0aPriqquuQlHafmJ/+ctf/FJQoHh0JJCFiECGAYeaDJJjwjCUZ86c2epyRUUFr732GlOnTvVrUYFwsFGmLYSIVOWHvCRFK8cdVFrJaU9f7Nq1izvvvJOXX37ZXzX5nWEYbC5zs3a3nLlaiEjUu4uNUT0dIXGw77TXiWRkZLB161Z/1BIwHh2q6iNi6lwI0YbaRj1kTpLcoemLV199tdVlp9PJsmXLKCoq8mdNfmcYUCPTF0JErENNOlqILFXuUCi/9dZbrS7HxsYyZMgQZs+e7c+a/M6myZyyEJGsyWN2iAwFEbEkrsGl8+rXodE0SQgRGNMLY0gIgZ19Hapw5MiRbV5v9Y0jDa6w/7sjhDiJWmdofFruUCi73ceuXnC73ei6tZ9snYSyEBGvplEPid7K7ZpTbt404nK5+MEPftDqa6WlpQwZMiQgxfmDYRgcCpG/kEKIwKl1Gnh0sGvBruTE2hXKM2fOxDAMNmzYwKWXXuq7XlEU0tLSfH0wrMiry/SFEAIOOUNjWVy7Qnn69OkADB48mD59+gS0IH/TDaiXUBYi4tW5DEJgQ1/HlsT16dOHyspK1q9fT3V1dav5mZYjaKuRkbIQosljhMTZrTsUyh9++CELFiygZ8+ebN++nb59+7Jt2zaGDh1q2VBWVWgIwTPaCiH8y+0l/EL50Ucf5Te/+Q2TJ09mxIgRvPnmm7z22mts3749UPWdNlUBp4SyEAJC4kBfh5bE7du3j8mTJ7e6bvr06bz55pv+rMmvPNKuUwhxmMdr/QFah0I5LS2NyspKALp168a6devYvXu3pdcpe3Xr/xCEEJ0jFHqqdyiUZ86cyZdffgnA7Nmzueaaa5g2bRpXXHFFQIrzhxD4wyiE6CQuj/UD4bR6X+zbt4/GxkZLL5M72Kjz1gbpeyGEgAm5UeQkd+hQWqfr0Ej5pptuanU5OzubPn36MG/ePL8W5U8emb4QQhzWGAIH/TsUyqtWrWrz+tWrV/ulmECQA31CiGZNITB90a5x/GOPPQaYzYea/79ZSUkJ2dnZ/q/MTzyhsK9SCNEpQiEO2hXKpaWlgNncp/n/m2VlZTF//nz/V+YnMlIWQjQLhdnMdoXygw8+CMCQIUO47LLLAlqQv7ll+YUQ4jDDMAeXVj6rdbtWXzQ0NADm6Z/AfFL/+Mc/+PbbbxkyZAgXXXRRYKs8Ddsq3HxR7Ap2GSKIVAUmD4wm1mHdX0TROeyqgqZi6VBu10j5Zz/7GZMnT+aSSy4B4Le//S1vvPEGo0aN4te//jVlZWVcf/31gazzlNks/OKLztEjRSMxWg2J08sL0a5Q3rRpE7/73e8AcLlc/P3vf+fJJ59k9OjRrF+/njvuuMOyoaxZfJ+7CLyCLLsZyIYBnqZglyOsQFHB5gh2FW1qVyg3NjaSmJgIwMaNG7HZbL7G9oWFhVRUVASuwtMUKqcVF4GREqOS2HyyTK8Ltn0G9dXBLUoEl6rBkOnBruK42hXK6enpbNmyhf79+7NixQqGDRvm+1ptbS0OhzX/4oBMX0S6gZm2I+0a3U4obnutvYggmnXzCtoZytdffz0//OEPGTJkCJ9//jlLlizxfe3zzz8nLy8vYAWeLhkpRy67Bj3TbKiqAh4X7Pgi2CUJK1AUwLqrstp9jr6ePXuyceNGZs+ezfDhw31fi4qKsvQ2awnlyNWni+3IZgFFgb3rg1qPsAhFtXImt7/J/ciRIxk5cuQx10+cOLHV5alTp/LOO++cfmV+osr0RcTKzzx8gE/3wr5NcpBPmGwOMKzbbtjv48g9e/b4+y5Pi4yUI1NmgorDdvgPsu6VuWRxhC0qskLZaouyZW1qZMrPsmNrfnfXV8Gh8qDWIyzEHh3sCk4o7MeRVj8fl/C/WLtCRqJmDhA8TbDjv8EuSViJTUI5qAwD7GH/LEVLeRk2fJ+PDANKNwezHGE19ujDKzCsye9xdRonMgkI3YBou3V/AMK/VAXy0u1oqgJeD+z60tLzhyII7FHmCgyL8ntlixYt8vddnhbDgBgJ5YjRPUVrMQgyYNfaYJYjrMgWDap1TwnVocpqampYunQpmzdv9nWOa/aXv/wFMJfEWY10B4scrfpcVO0GZ22wSxJWExVn6emLDoXybbfdhsvlYvLkycTExASqJr/SVAnlSJEco5DUss+F7OATbYlNDnYFJ9ShUF63bh0rV660dK+Lo2mqQnyUdeePhP8MzLSjNv+o3U1woDio9QiLik0JdgUn1KG0ysvLO+Z0UKEgMVpGyuHOrsIZaTZzB6fHBcUyShbHERUf7ApOqEMj5dGjR3PDDTcwY8YMunTp0uprl156qV8L8yffR1oRtnof3eei5H9BrUdYlGoLjy5xzdauXUtGRgYrVqxodb2iKJYO5RiHgqKExplsxanJbz7Ap+uwf7P0uRBti0kC3Q1qVLArOa4OhfKf//znQNURUF4dEqMUDjollcNRRoJKlK/PhQeKVwa3IGFdMUmWH52d8ud6wzDQdd33z8oMA5JiZAojXLXqc9FQDbVlQa1HWFhMkqU3jkAHR8plZWUsWrSItWvXUlvbev3n5s3W3cpq0yA5RmV3tTfYpQg/i7ErZEmfC9FecSmg2YNdxQl16E/GL3/5S+x2Oy+88AKxsbG88cYbTJgwgV/96leBqs8vVEWhS7y1/zqKU5Ob3uIAn4H0uRAnlpxj6Y0jcArrlP/zn/8QGxuLoij079+fBx54gCuuuILLLrssUDX6RbJMX4QdRYH+GXZs2uE+F7u/NHsnC3E8CV2DXcFJdSipVFXFZjNzPDExkaqqKmJjYykrs/4cXuzhFRgifHRP1o6cFFX6XIiTsUebZx2xuA6NlAcPHsynn37K+eefz9ixY7nllluIjo6moKAgUPX5jVeHpGiFmkZrH3kV7deqz0X1Hmg8GOyShJUlZpifqCzcjAhAMTrQa7O2thZd10lOTsbpdPLHP/6RhoYGrr32WtLT0wNZ52nzeA3Wlrj4ttwT7FKEHyRFK1xUEINNPXyA78tXofK7YJclrKzXKMibAJq1Q7lD1SUmJvr+Pzo6mrlz5/q9oECxaQrZSZqEcpgYkGk/MnXhcUkgi5NLybF8IEMH55TdbjeLFy9mwoQJDBo0iIkTJ7J48WJcLleg6vOr9Hg5N1Q4sKnQu1WfC9ksItohKSvYFbRLh/5s/O53v2P9+vUsWrSI7Oxs9u3bx5NPPkldXR2/+MUvAlWj39g0iHMo1LtkXjmU9U6z4fsJKgqUfB3EakRIUDWITjz57SygQ6H8r3/9i7feeouUFLP1Xe/evRk4cCDTpk0LiVA2DEhPUCk+IMumQlmrPhelW8HtDHZJwuqSu4HXbYazxXVo+uJ4xwStdl6+47FrCtmJ1v+hiONLj1ePnHNR98B30qJTtEOX3pbfydesQ6E8adIkbrrpJpYvX86OHTv47LPPmDt3LpMnTw5UfX6XKaEc0lr1uWg8CLWh199bBEF6v5AYJUMHpy8WLFjAU089xaJFiygvLycjI4MLL7wwZA70gXlm6ygbNMkijJATbTdX0EifC9EhqgYJXU5+O4vo0DrltjQ1NVFUVGTphkQtub0GK3c2ybxyCCrMtlOQbT+yNvnfD8u2anFyqT1h+GXmjr4QcNoNIRRFCZk5ZTDnlXulWX+tomhNwVybbFOb+1ysk0AW7dOlV8jMJ4MfQhnMYA4lmYkteyaIUJCTclSfi51rglmOCCUhNJ8M7ZxT/uKL4x/hdrvdfiums+i6Gcz7DspIK1T4+lwA1OyFxpqg1iNChOaA+NCZT4Z2hvJdd911wq9nZYXGTplmNg3OSJVQDhWJ0QopsYc/1LmbYIcsgxPtlN7XnOYKge3VzdpV6ccffxzoOjqVqih0T7FBceisGolkAzLs+GYudDdUbA9mOSKUdBsEduueJLUtEdv5XVUgLS5in37IsKnQp4sNTVXMHVnfrQp2SSJUqJp5kC/ERGwqaQr0TA2dyf9I1atlnwuAknXBKkWEmrQzQnKFTsSGsqoq9JalcZbXqs9F2TZwNwa7JBEqsgvAFlpTFxDBoQzmmuV0OaGqZXWJV4mRPhfilCiQkWf5k6S2JaITSVMhLyN0FpVHmvxMO1rzO9RZCwf3BbUeEUJScoJdwSmL6FA2V2FoRxrcCMuItkFOsna4kb30uRAd1GNISO3iayni48gw4IxUmVu2mn7pdo7s3ldg36ZgliNCiWaHrIGghma8hWbVfmTXFAZkSihbiYK5NtmmKebR85KvzTllIdojcwAYerCrOGURH8oACVEqSdGhd0AgXHVL1o7MJRs67Fwd1HpEiOk1KiRXXTSTUMY8QNsvXUbLVtGqz8XB/dBQHdyCROiIS4X4tGBXcVoklAFNVejb1S6d4ywgIUohNa5lnws5wCc6oMfQkFwG15KE8mEK5nZeEVwDMlv2ufBAufS5EO2kKNB9CKih/XssoXyYXVMY3K1FIIhOp7Xsc+FxQ/EqIHROoCCCLCMPwuA3WEK5Bbum0CNF+mEES6+WSxMVpM+F6Jh+54RcR7i2SCi3YNcUinIcwS4jYvn6XBi6OW3hagh2SSJUpORAbHKwq/ALCeWjxDoUMhPkZelsXeJUYh2HP3p6pc+F6KDcc0N2B9/RJH2OYlOR0XIQDGzZ56KpzjzlkxDtEZdmjpRDfNVFMwnloyiKQmqsSmqsvDSdJcoG3VNa9rlYEeySRCjpOw6U8Pl9DZ9n4keqAkNywuOjUCjo2/WoPhd7NwazHBFKouIgq39Ina36ZCSU26CqChkJGl2k13LAKcDATNuRPhd71kufC9F+vc8MdgV+J6lzHJoKo3vK3HKgZSdp2Jq3Uho67JRz8Il2csRBz+Fhc4CvmYTycSiKQkK0Svfk8PlYZEX5Lftc1JZBfVVwCxKhI/dswmGzyNFCez9igNk1hZFnONjzv8YWc57CX+KjlCNTRGHW5+L2lz5n5bZSGlxeuiZGc8P4gcwc3Q+Xx8vtL61gY8kB9lbX86ebz2NU38wO3w/A/up6fvricnZW1DJjZB8WThvm+74b/t/H/HTyYAZ1D+3mPMcVnQg5g0ELvwgLv2fkZw5NIberja3lMs/pb/0zbEfGOYYXyr8NZjl+9ePzCvjNFWfisGnsKDvINU/+mwHdUsnNSmZor65cc3Z/bnnxs1O+n4LuaTzz0SYuGdGbqUPPYPof3ueioWcwqHsa76/bSU5qfPgGMsCAiWG14qKl8HxWfmTXFIbkOOSUUX6mKdCvq93sc+F1mz2Tw+jjSL/MZBw2c+pLUcwP2bsP1OGwacw+ZwDDe6ejtuPMGMe7H4A9VXWM7pdBQoyDQd3TKDlQR53Txf/7eBO3XlQUoGdmAQnpZp+LMFpx0ZKMlNtBVWBQtp11e9zBLiVs9Exr+dZTYNdXQaslUO57dTVvrNmB0+1lYLcUzhmQ7df76ZeZxH+3ltIlIYZNe6q4+YJBPPrP/3Ht2f1JjAnjg9T5k0AJz0AGCeV2sWkKAzLsbK/wcKgpfEZzwVTQss9FxQ5w1Qe7JL+779KR3DNjOOt2VrJ6R5lvxOuv+/nxxALue201f1+5javG5OL26GzdV8O8Cwq57c+fU3qwgcmDe3L1uDx/Pq3g6tILkrJC9vx77RG+z8zPVAXG9gn9DlRWkBanEhchfS40VWV473RKaxp4ZcWpz5m3dT/JcVE8es043l4whWvG9ef+N9Zwz4wR/L+PN9EvK5nnb5zIX7/4lh1lB/31dIJL1aDwYrCF8acAJJTbTVUVkmNU+koj/NM2MNNO8yo4muqhuiSo9XQGr66z+8ChgN3P31Zuo6hnF3Kzkvl2fw0F3VNx2DRys5LZur/mtB/XEvqMBXt0sKsIOAnlDrBrCiN6Ooi2h9/ayM7i0A73uVAV8Ljgu/BZBtfswCEn763bSX2TG6+us3zLPt5bt5Mz+5lL31weL01uLwBur06T24vRxkHOk91Py9u9/Pm3zPteIQA5qfGs2lZGfZObjSVVdE+LD/Az7gSxKdDnzLAfJQMoRlvvBnFcXt1g/0EvH29rCnYpIWlgpo2ibg5zW7XHBR/+wVx9EUaq6pz85IXP2LKvGt2AbilxzBqXx2VnmuuLJ9z/BnurW8+hf3T3JeSkxvP0hxtZ+105z82ZcNL7aXbHyysYPzCHyUU9AXP98k9e/IydFYeYMbI3d04b3jlPPJDOvBaSc8J6LrmZhPIpcHsNlu9oYk+NN9ilhJyZRTHEOFSzz0XJ17Dx/WCXJKwuayAUTo2IUTLI9MUpsWsKY3pFYZdXr0OykzRzhAyH+1ysDm5BwvpsUVBwYcQEMkgonzJNg5FnRM4bxR/ys2xH+lwcKoe6yuAWJKxv4AVhuZX6RCSUT5FNVeiZYpMTrbZTnEOha/zh1yrM+lyIAOnaF7Lyw64L3MlIKJ8Gm6YwpnfUkXPLieNq3edCh7KtwSxHWJ0jFoZMB1tkBTJIKJ82TYFz+0aFYQNB/1EV6Jfeos/FrjVh1edCBEDR9IibtmgmoXyaVFUhKUalSE4fdVw9U7UWf7QU2PVlEKsRltdjqHkiVFVCWZwi++HeGFmJMr/cliN9LgyoLDbPVi1EW+JSYcD5EbXa4mgSyn5i0xTO6RtFjOz2ayU1ViUh6vDbzBueO/iEnygqDLssYkfIzSSU/cimwsTcKFTJZZ+BmbYjr4erEap2B7UeYWH5kyAmKSJ27Z1IZD97P1NVhcRoVbrJHebQoEeqLaz7XAg/yRkM3QZF9LRFMwllP7NpCt2SNAZ3kwN/fbrYjiyyUBTYsyGo9QiLSsqCgskSyIdJKAeAXVPIz7RzRmpkH/jznala12HvRnNOWYiWHHEw8qqI2yByIhLKAWLTFM7qFXXkbM0RJitRPbKl2vBC8argFiSsR1FhxBVmfwvhE5mJ0UlsmsJ5udHER+COv/ws+5GTzdZVQl1FUOsRFlQwGRK6hO0JUE+VhHKA2VS4YEA09gh638U5FDISNBRFAY/0uRBt6DkCsgtAk3nko0koB5iqKsTYFc7Pi0aLkFc7L92GbxO1YUDplmCWI6wmcwAMmCgH9o4jQmIiuLTD5/c7Py867NcwqwrkptuxqYp5UtSda80GREIApJ0BRdPkwN4JSCh3EpumkBqrcl6YB3OPFA2l5fPbtTZotQiLScyE4ZdLIJ+EhHInsmkKXeJUJuRGtQ6uMNKqz8WBndB0+mdwFmEgNgVGz5JAbgcJ5U5m0xTS4zXG9wu/dp/JMeaORuBwn4svgluQsAZHnHniU81B2I5G/EhCOQhsmkJmgsbZYdaHeWCm/cjUjNtpjpRFZHPEwZjrzKb1Ed7Tor3kVQqS5u3Y4/qGx1SGXYMz0lr2uZBRcsSLioexP4ToRFmL3AESykFk0xRykjTOz4s+stEiRB3b52J9UOsRQRadYAZyVLwEcgeFeBSEPpum0CVe5cKBMUSFcBvZ/MzmPhde2P+NuWlERKaYJBhzgzl1IYHcYRLKFmBTFRKiFaYUxBAfFXpzGZkJKg7b4bp1L3y3MrgFieCJTYExPzw8hyyBfCoklC1CUxVibAoX5ceQGhtaP5ZWfS7qq+BQeVDrEUES3wXGXA/2GDmodxrklbMQVVWIsil8b0A0WYmh8aOJtStkJLbocyGN7CNTWk8JZD+RV8+C7JrC+H7R5Ha1/iRzbrrtyLI+w4D9m4NZjgiGboNhxJVmC85wWEoUZBLKFmXTFIb1cDC2twPNou9zVYH+GXa05j4Xu76UPheRJm88FEySnXp+ZP2hWASzawo9UmykxWl8uNVJvcs4+Td1ou7S5yJyqRoUTYeufaTbm5/JSNnibJpCQpTC1EExZCdZ62h2qz4XVbvBWRvskkRnsMfAmbMhva8EcgBIKIcAVVVwaArn9o2yzAlZk2MUklr2uZBG9pEhORvOuQkS0mXKIkBk+iKE2DSFgZl20uM1PtvhpMkTvFoGZNiPHGT3NMGB4uAVIzrHGSOh/wQJ4wCTkXKIsWsK6QkqlxTGkpMcnOkMuwq90myoSnOfC9ksEtZsDrMPct54CeROICPlEKSpCpoKZ/eJoqTGy8qdTbi9nff4vbu0ON2TokDJ15334KJzJWTAyCvBHi2B3EkklEOYTVPonqyRVRjL8h1O9td2znK0/OYDfLoO+7dIn4tw1WMoDLwAVJusP+5EEsohzqYp2DQY3y+a4gMe1ux24QlgNmckqET5+lx4oFhadIadqHgougSSu8noOAgklMOETVPolWajW7LG8h1NlB0KTDLnZ7boc9FQA7VlAXkcESRZA2HQFNBs0lAoSCSUw4g5alaYmBtN6SEvq3a6/LrhJMaukJUkfS7CUlQcFF4MqT1k7XGQSSiHIZumkJWoMW1QDN+Uudmwz43XDwPn3PQWjewNzL7JIvR1KzS3SqsyOrYCCeUwpR0+Wd6ADDu5Xe2s3NnE7upTX6KhHO5zYdMO97ko+crsnSxCV3xXKJxibgSR0bFlSCiHObumYNdgTO8oChp1/lvcRE1jx6c0uidrR06KigE71/i1TtGJ7NHQfyJ0G2SOjBXZrmAlEsoRwq4ppMapXJgfw94aL+v2uKh1tj+cfX0uAKr3QOPBAFUqAkeBnkMhb6LZ81hWVliShHIEURUFVTFHvd2SY9hX4+WrdoRzUrRCcvPZUDxN0uciFKX2gMKp5gE9W1SwqxEnIKEcgVRVQQVykjWyk2PYd9DLuhIXB48TzgMy7UemLjwuqPyu02oVpykxEwacZ645lnnjkCCTSRFMVRVsqkJOksZFBTGM7xdFUnTrnVs2FXq37HNRLH0uQkJiJoy8Gs6abZ6qyY+BPGHCBKZMmYKu662u+/bbb/32GJFMRsrCN3LulqSRlRTDgXqdjfvc7D3opVea9LkIKYkZ0P88SO0e0IN4DQ0NvPXWW0yfPj0g9x/JJJSFT3M4ZyRopPZVcXsNFEU50ueibCu4ncEuU7QlMdNsq5naAxQt4CcvnTdvHo8//jgXXXQRDodMi/iTTF+INtk1hViHSlTzn23dAzukz4WlKCpk5cO4OXDmtZDWy1xR0Qlnky4oKCA/P59XXnkl4I8VaWSkLE5Ibe4O1ngQakuDW4wwOeKg5zCz6byqBm01xS233MI111zDpZdeGpTHD1cSyuLkDANiU2DI92HnKnOdsuh8SVnQ5yxI72deDvI64969e3POOefw/PPPB7WOcCOhLE5OUcyuYZn9zZNluuph15ewb5OcLDXQYpIhpxC6DzF34qm2TpmeaK/58+czY8YMvF7Zcu8vEsqi/VQVVIe5vCr3HPNffRXs/gr2bzbDWpw+ewxkD4Sew81PKBz+o2hBmZmZTJs2jaVLlwa7lLChGIbhv96OIjJ5XOZBp9oys1FR6RZZpdFRjjjo2ge6D4bkHDB02ewRoSSUhX95XOaIur7KPFVUxXao2QfI26w1BZKzID3XbCwfkwSGV7ZACwllEUC612zzqShwYBeUbja3aDsPBbuy4IhOgNSekDUAuvQy/07JGT7EUSSURedpnuZwO6FmLxwoNkfRtaXh15tZUSEp05yK6Nrn8PnubOYmHLuMhsXxSSiL4PG6zTDW7Ob5/qp2m/8OlZvTH15XsCtsH1sUxHcx/yVmQtoZEJ9mPj9VC/rSNRFaJJSFtXiazHXRmt0M5YYaOFRhjqbrq6D+gDn94XV3bl2aw5x+iEmEuDQzfJMyzdURmt2sRznco1hRTn5/QhyHhLIIDV6PudVbUcy1umBOg7gboaneDOrGg9BUZwak4TVH4b5/niP/3zx6VW3mfzVbi/+3m+EbnQhR8RAVay5RU1TzPgxDRr8ioCSURXjxeszlZBiHF3yc7O2tmEGvKGbwKqqMdEVQSSgLIYSFWGe/phBCCAllIYSwEgllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwEAllIYSwkP8PA3tQSKk2oE0AAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 1.4.2 Loan status and numerical features "
      ],
      "metadata": {
        "id": "ZZjOsM65Wbyl"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Separe target from train data frame\n",
        "target_train = df_train['Loan_Status']"
      ],
      "metadata": {
        "id": "btOI1Y-ZJP5M"
      },
      "execution_count": 223,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def boxplot(column, target, train, max_y, title=''):\n",
        "  f, ax = plt.subplots(figsize=(8, 6))\n",
        "  fig = sns.boxplot(x=target, y=column, data=train)\n",
        "  fig.set_title(title).set_fontsize('16')\n",
        "  fig.axis(ymin = 0, ymax = max_y)"
      ],
      "metadata": {
        "id": "bmsu_x4CUiWM"
      },
      "execution_count": 224,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Applicant income\n",
        "\n",
        "boxplot(column=numeric_cols[0], target=target_train,\\\n",
        "        train=df_train, max_y=1.0*max(df_train[numeric_cols[0]]),\\\n",
        "        title='Zoom: 100% maximum income')\n",
        "\n",
        "boxplot(column=numeric_cols[0], target=target_train,\\\n",
        "        train=df_train, max_y=0.3*max(df_train[numeric_cols[0]]),\\\n",
        "        title='Zoom: lowest 30% maximum income')"
      ],
      "metadata": {
        "id": "ce96USe_NDls",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "outputId": "216cbaec-b407-441d-fb0a-eb7e6e346945"
      },
      "execution_count": 225,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "As it can be seen in the figure above, there is a high quantity of outlayers. The categorires of loan status have simililar distributions and medians."
      ],
      "metadata": {
        "id": "a6CnU14paf-7"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Coapplicant income\n",
        "\n",
        "boxplot(column=numeric_cols[1],\\\n",
        "        target=target_train,\\\n",
        "        train=df_train,\\\n",
        "        max_y=1.0*max(df_train[numeric_cols[1]]),\\\n",
        "        title='Zoom: 100% maximum income')\n",
        "\n",
        "boxplot(column=numeric_cols[1],\\\n",
        "        target=target_train,\\\n",
        "        train=df_train,\\\n",
        "        max_y=0.25*max(df_train[numeric_cols[1]]),\\\n",
        "        title='Zoom: lowest 25% maximum income')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "Qzz47xeAI4za",
        "outputId": "86f705a9-8291-456a-d2bf-e4a2c3c80ede"
      },
      "execution_count": 226,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgoAAAGLCAYAAACmxbq1AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAAsTAAALEwEAmpwYAAA4c0lEQVR4nO3deVyU9f7//+cACooiYqFoZlmCoLmBgh7NXHLJNY8Fh9RcKnP3e8wUT8rJk0t4zF2zTFuORmblnubWoqWpmR5CW1zS40Y6oALGNtfvD3/Mp0lHh2UYHB/3242bcL2vueY1y+U85329r/dlMgzDEAAAwA14uLoAAABQehEUAACAXQQFAABgF0EBAADYRVAAAAB2ERQAAIBdBAUUWkhIyC1//ve//7m6zCJp27atxo8f7+oyCmTPnj2aN2+eLBbLLddNSUnRzJkz1atXL0VERCgqKkpPP/209u7de92648ePv+FrPGXKFJv1vv76a3Xp0kXh4eEaMWKE0tLSbNrT09PVsmVLffrpp0V6nK7w8ccfu8X7Wrr2PgkJCdGePXtcXQpKOS9XF4Db1wcffHDD5WfOnNHYsWMVEhKiwMDAEq4K3377rebPn68hQ4bIw+Pm3wV++OEHbdy4Ub169VKjRo2Uk5OjFStWqG/fvlq0aJHatGljs35AQIAWLVpks+zuu++2/n7p0iWNHDlSPXv21MMPP6wZM2Zo+vTpmj59unWd2bNnKyQkRJ07dy6GR1uyHnnkEX3wwQdu8b6uV6+ePvjgAz344IOuLgWlHEEBhdaoUaPrlmVnZ2vKlCkqX7685syZo7Jly5Z8YXBYeHi4Nm/eLC+v//uvoGXLlurSpYuWLFlyXVAoU6bMDV/3fAcOHJDFYlFcXJw8PT2Vnp6uV155xdp+5MgRffTRR1q9enVxP5QSERAQoICAAFeXUSwqVKhw09cSyMehBxSr6dOn69ChQ5o2bZpq1qxp0/bll18qOjpaDRo0UHh4uIYOHapjx47ZrGMYht5++2117NhR9evXV8uWLTV58mSlp6fbrBcSEqJZs2Zp6dKlatOmjRo2bKjnnntOFy9e1MWLFzVq1CiFh4erdevWeuONN4r1MR46dEj9+/dX48aN1ahRIz399NM6dOiQtX3z5s0KCQnRuXPnbJ6XkJAQffjhh9Zlu3btUkhIiH7++WfrsiNHjuj5559X06ZN1aBBA8XExGjfvn3X3f+AAQMUGRmpBg0aqF27dvrnP/8pSZo3b57mz58v6do3xvzDA/b4+fnZhARJ8vLyUmhoqM6fP1/g5yYnJ0dlypSRp6enJMnHx0dZWVmSrr22//znPzVw4EDVqlXL4W3md/d/9913GjVqlBo3bqwWLVpo8eLFkq69r3r27KlGjRrpr3/9q5KSkmxuv3PnTj377LNq2bKlGjZsqK5du2rp0qXKy8uzrrNr1y7VrVtXb7/9ts1tx4wZo2bNmuns2bM2tfzx0EPbtm31wgsvaPXq1erYsaMaNGig2NhYnThxQpmZmZo0aZIiIyPVokULTZ8+Xbm5udc9tj8fypg3b951r1txv+dvdOihb9+++tvf/qavv/5ajz/+uPX52rJly3W3P3LkiIYNG2Z9H3bs2NH6mkglty+fOnVKY8aMUVRUlOrXr68ePXrcsF4UHj0KKDbr16/X8uXLNXDgQLVv396m7csvv9TgwYMVFRWlWbNmKTMzU3PnzlVsbKzWrFmjqlWrSpJmzZqlxYsX66mnnlKbNm109OhRzZkzR0eOHNF//vMfm670tWvXqk6dOoqPj9eFCxc0depUvfjii8rIyNDDDz+s6Ohobdq0STNnzlRISIhat25tvW1ISIgef/xxmy5xRxw5ckR9+vTRgw8+qGnTpslkMumNN95Qnz59tHLlStWtW1dNmzaVyWTS7t271bNnT0nS7t275ePjo927d+uJJ56wLrvrrrtUp04dSdcOAzz11FMKDQ3Vv/71L5UrV07vv/+++vfvr8TERNWvX18ZGRl65pln9NBDD2natGny9fXV6dOndeDAAUnSE088oXPnzmnVqlVasWKF9QO7ILKzs/X999/fMGCYzWZFRkbqypUrqlmzpv76179q0KBB1vsJCwtTenq6Pv74Y7Vp00YrVqywfmtdtWqVLl68qOeee67ANUnXxkj06NHD+rq+9tprunz5sr788ks9//zzKl++vGbMmKFhw4Zpy5Yt1t6sU6dOqXnz5urTp4+8vb2VlJSkefPmyWw264UXXpAk/eUvf9HAgQM1c+ZMRUVFqW7duvr444+1fv16zZ07V0FBQTetbd++fTp16pTGjh2r7OxsTZ06VSNGjFDNmjVVq1Ytvfbaa9q7d68WLVqkmjVr6qmnnirUc1CU97yjTp06pSlTpui5555T5cqVtWzZMo0aNUqffvqpNeAdOnRIffv21b333qu4uDhVrVpVv/76q3788UfrdkpiXz579qyefPJJValSRXFxcQoICNDGjRs1YsQILViwQO3atSvU84w/MYBi8MsvvxiNGjUyoqOjjZycnOvaH3/8cePRRx+1aTt58qQRFhZmTJ061TAMw0hNTTXq1atnjBs3zua2q1evNoKDg42tW7dalwUHBxsdOnSw2d7UqVON4OBgY8GCBdZlOTk5RlRUlDF+/HibbYaGhhpxcXG3fFxt2rSxqWfEiBFGeHi4cenSJeuyK1euGE2bNjWGDRtmXdatWzfrfaamphohISHGtGnTjL/85S/WdZ544glj9OjR1r/79etndOrUycjKyrIuy83NNTp16mQMGTLEMAzDOHTokBEcHGwcPnzYbs1z5841goODb/g6OGLmzJlGSEiIsXfvXpvly5YtM959913j66+/Nj7//HPjH//4hxESEmJMmDDhuvVCQ0ON4OBgo02bNsbPP/9spKamGpGRkcYXX3xR4Ho++ugjIzg42Jg3b551Wf7rGhYWZpw8edK6fOvWrUZwcLCxZ8+eG27LYrEYOTk5xsKFC42IiAgjLy/P2padnW306tXL6Ny5s5GcnGw0atTImDhx4g1rOXXqlHVZmzZtjKZNmxqXL1+2LnvnnXeM4ODg656bnj17Gn369Lnp9gzj/17DPyrqe/7Pdu/ebQQHBxu7d++2LuvTp48RFhZmHD9+3LrswoULRt26dY1FixZZl8XGxhoPP/ywkZmZecNtl9S+HBcXZ0RGRhpms9nmfvr372907979po8fjuPQA4osMzNTI0aMkLe3t2bPnn1dV3ZmZqaSk5PVuXNnm7aaNWuqSZMm1hH2Bw8eVE5Ojrp3725z+y5dusjLy+u6kfgtWrSw2V7t2rUlXTvGns/Ly0u1atWydh3nS05O1tSpUwv8WPfu3atHHnlEfn5+1mUVKlRQ27ZtbeqLioqydul+++238vPzU//+/fXbb7/p6NGjSk9P1w8//KDIyEhJ0u+//669e/eqU6dO8vDwUG5urnJzc2UYhlq0aGE9/HDffffJz89P8fHxWrNmzXWPq6jWrVunN954Q0OHDlVERIRNW//+/dW3b181b95crVu31iuvvKJ+/fpp1apVOnHihM163377rTZt2qQtW7bowQcf1MyZM9WsWTM9/PDD+uabb9SzZ081bdpUgwYN0unTpx2q7eGHH7b+nv+63nfffTaHuPLfA398XlJSUjRp0iS1adNG9evXV7169TR79mxdvnxZFy9etK5XpkwZzZw50/otNSgoSHFxcQ7V1qhRI1WsWPG6Ov74Xsxf/sdDUgVVlPe8o/Kf13xVqlRRlSpVdObMGUnS1atX9d1336lbt24qV67cDbdRUvvyV199pdatW6tixYrWfSY3N1ctW7bUkSNHrjvMgcLh0AOKbOLEiTp27JjefPNNVatW7br2y5cvyzCMG44Uv+uuu6wfFPmn0f1xFL107T8If39/Xbp0yWb5Hz+spWv/0UtSpUqVrluef5y8qC5dunRdfdK1x/HH+iIjI/XOO+/o1KlT2rNnj5o2bapq1arp/vvv1+7du1WjRg3l5uYqKirKut28vDwtXLhQCxcuvOF9WywWVaxYUe+++64WLlyol19+WRkZGapTp45GjBihjh07Fumxbd++XXFxcerdu7dGjhzp0G26du2qd955R0lJSTYfLhUqVFCFChUkXfvQ2LhxozZu3Ciz2azhw4frH//4hzp37qxXXnlFY8eO1YoVK255Xzd6Xe29B7KzsyVde86GDBmilJQUjRgxQrVr15a3t7e2bt2q119//br3xX333acmTZpo586devLJJ+1+EP5ZSb0XS+J+/rwtSSpbtqz1Ob18+bIsFssN9/V8JbUvm81mrV692u7g2NTUVOv7EIVHUECRLF++XOvXr9eQIUPUqlWrG67j5+cnk8mk33777bq2CxcuyN/fX5Ks/164cMF63F6ScnNzlZaWdsP/wEpapUqVdOHCheuWX7hwwaa+pk2bytPTU7t379bu3bsVExMj6f96GqpXr66qVataP1wrVqwoDw8PPfXUU+rRo8cN7zv/mG5oaKjmzZun3NxcJSUlafHixRo9erTWrFmj4ODgQj2ub775RqNGjVL79u01efLkAt/eZDLdcLnFYtHLL7+sYcOGqWrVqtq+fbu8vLzUq1cvSVK/fv3UvXt3ZWRkyNfXt1C138zJkyeVlJSkhIQEm+d1x44dN1x/1apV2rlzp+rVq6cFCxaoQ4cOql69erHXlc/b21vStUGgf/TnuSdKEz8/P3l4eNx0sGtJ7cv+/v4KDw/Xs88+e8P2/LFPKBoOPaDQ8s9uiIqKuuk30PLly6tevXratGmTzUjz/EF4zZo1kyQ1bNhQZcqU0YYNG2xuv3HjRuXm5lrXc6WmTZvqyy+/tOnSTE9P144dO2zq8/PzU2hoqDZs2KBffvnF2nMQFRWlb7/9Vt988431sIN07TmKiIjQkSNHVK9ePT300EPX/fyZl5eXGjVqpFGjRslisejo0aOSZB3E9/vvvzv0mA4cOKChQ4eqefPmmjFjxi3nXvijtWvXymQy3bA+SVqxYoVycnLUr18/67Ls7GzryP/MzExJ10bIO0P+c5D/DVW69qG8bt2669Y9fvy4pkyZotjYWL377ruqWLGiXnjhBZv3bHHLDyF/PPMlNzdXO3fudNp9FlW5cuUUHh6utWvX2n2PldS+3KpVK/3444+qU6fODfcZTs8uHvQooFAuXbqk0aNHy8PDQ3379rU5PfCPHnzwQVWoUEGjRo3S4MGDNXjwYMXGxiozM1Pz5s1ThQoVNGDAAEnXvh0MHDhQixcvVrly5dS6dWsdPXpUs2fPVnh4uB555JFiqz8sLEw9e/Ys8DiFoUOH6vPPP1f//v317LPPymQy6c0339TVq1c1bNgwm3UjIyP11ltvqUqVKtZvVc2aNVNaWppSU1NtPjyla6P6+/Tpo0GDBql37966++67lZqaquTkZOXl5emFF17Qjh079MEHH6h9+/a65557dPXqVb333nvy9fVV48aNJUkPPPCAJGnZsmV6+OGH5eHhYfeD/OjRoxo8eLAqV66sQYMG6YcffrBpzz9j4fTp03rxxRf12GOPqVatWsrOztaWLVv0ySefKDo6Wvfee+91275w4YLmzp2rRYsWWY8/529vypQpat++vRYuXKjGjRs7rXu4du3aqlGjhmbNmiUPDw95eXnpnXfeuW697OxsjRkzRvfcc4/Gjx8vb29vzZw5U3369NGiRYs0fPhwp9T30EMP6d5771VCQoIsFovKli1rDVel2Ysvvqi+ffsqOjpaAwYMULVq1XTq1CkdOXJEEydOLLF9eeTIkXriiSf01FNPqU+fPqpRo4YuX76sn376SadOndK0adOK5X7udAQFFMqRI0esYwv+/AH5R++++64iIyP18MMPa/HixVqwYIFGjx6tMmXKqFmzZho7dqxN9+D/+3//TwEBAXr//ff1/vvvy9/fXz179tSYMWMK9E33VvLy8hya4vjP6tatq/fee0+zZs3S+PHjZRiGGjZsqP/85z+qW7euzbr5QeGPPQcBAQEKDg7Wjz/+aLNcujbvwapVqzR//ny98sorunLligICAhQWFqa//e1vkq4NNPPx8dHChQv122+/ydfXVw899JCWLVtmPWbcpk0bxcbGasWKFVqwYIEMw7A5be2PDh48qEuXLunSpUvXBRdJ1tv5+vqqUqVKWrJkiS5cuCAPDw/Vrl1bL730kmJjY2+47YSEBLVr107h4eE2j3/OnDlKSEjQ6tWrrad5OkvZsmW1YMECTZ48WePGjVOlSpX017/+VdWrV9dLL71kXW/WrFn65ZdftGrVKuvhgMaNG2vYsGGaP3++WrRooSZNmhR7fV5eXlq4cKEmT56suLg4VapUSU8//bQaNmxonQ+jNGrQoIHef/99zZ07V6+88oqys7NVvXp16yElqWT25erVq+ujjz7SvHnz9Nprryk1NVX+/v6qU6eO9dRkFJ3JcFafHwAAuO0xRgEAANhFUAAAAHYRFAAAgF0EBQAAYBdnPfyJxWJRRkaGypQpY3cSGQAA3IVhGMrJyZGvr+8Nz0ghKPxJRkaGfvrpJ1eXAQBAiQoODra5Zkk+gsKf5M/gFhwczKxeAAC3l52drZ9++slmBtM/Iij8Sf7hhrJly1onXgEAwN3ZO9zOYEYAAGAXQQEAANhFUAAAAHYRFAAAgF0EBQAAYBdBAQAA2FUiQeHVV19V27ZtFRISYjOZ0fHjxxUdHa2OHTsqOjpaJ06ccGobAAAomBIJCu3atdPy5ctVo0YNm+Xx8fGKjY3V5s2bFRsbq0mTJjm1DQAAFEyJBIWIiAgFBQXZLLt48aKSk5PVtWtXSVLXrl2VnJwss9nslDYAAFBwLpuZ8ezZs6patao8PT0lSZ6engoMDNTZs2dlGEaxtwUEBLjmgQIAcBtjCmc7kpKSXF0CAAAu57KgEBQUpPPnzysvL0+enp7Ky8tTSkqKgoKCZBhGsbcVVP369bnWAwDA7WVlZd30y7HLTo+sUqWKQkNDtX79eknS+vXrFRoaqoCAAKe0AQCAgjMZhmE4+05eeeUVffbZZ7pw4YIqV64sf39/bdiwQUePHtX48eN1+fJl+fn56dVXX1Xt2rUlySltjshPVvQoAADuBLf63CuRoHA7ISgAAO4kt/rcY2ZGAABgF0EBAADYRVAAAAB2ERQAAIBdBAUAAGAXQQEAANhFUAAAAHYRFAAAgF0EBQAAYBdBAQAA2EVQAAAAdhEUAACAXQQFAABgF0EBAADYRVAAAAB2ERQAAIBdBAUAAGAXQQEAANhFUAAAAHYRFAAAgF0EBQAAYBdBAQAA2EVQAAAAdhEU4DbMZrPGjx+v1NRUV5cCAG6DoAC3kZiYqOTkZCUmJrq6FABwGwQFuAWz2axt27bJMAxt3bqVXgUAKCYEBbiFxMREWSwWSZLFYqFXAQCKCUEBbuHzzz9Xbm6uJCk3N1c7duxwcUUA4B4ICnALjzzyiLy8vCRJXl5eatOmjYsrAgD3QFCAW4iJiZGHx7W3s4eHh2JiYlxcEQC4B4IC3EJAQIDatWsnk8mk9u3bq3Llyq4uCQDcgperCwCKS0xMjE6ePElvAgAUI4IC3EZAQICmT5/u6jIAwK1w6AEAANhFUAAAAHYRFAAAgF0EBQAAYBdBAQAA2EVQAAAAdhEUAACAXQQFAABgF0EBAADYRVAAAAB2ERTgNsxms8aPH6/U1FRXlwIAboOgALeRmJio5ORkJSYmuroUAHAbBAW4BbPZrG3btskwDG3dupVeBQAoJgQFuIXExERZLBZJksVioVcBAIoJQQFu4fPPP1dubq4kKTc3Vzt27HBxRQDgHggKcAuPPPKIvLy8JEleXl5q06aNiysCAPdAUIBbiImJkYfHtbezh4eHYmJiXFwRALgHggLcQkBAgFq2bClJatWqlSpXruziigDAPRAU4DYMw3B1CQDgdggKcAtms1m7du2SJH311VecHgkAxYSgALfA6ZEA4BylIijs2LFDPXv2VI8ePdS9e3d99tlnkqTjx48rOjpaHTt2VHR0tE6cOGG9TWHb4J44PRIAnMPlQcEwDL344otKSEjQmjVrlJCQoHHjxslisSg+Pl6xsbHavHmzYmNjNWnSJOvtCtsG98TpkQDgHC4PCtK109muXLkiSbpy5YoCAwOVmpqq5ORkde3aVZLUtWtXJScny2w26+LFi4Vqg/vi9EgAcA4vVxdgMpk0e/ZsDR06VOXLl1dGRobeeOMNnT17VlWrVpWnp6ckydPTU4GBgTp79qwMwyhUW0BAgMseJ5wrICBA7dq106ZNm9S+fXtOjwSAYuLyoJCbm6vFixdr4cKFCg8P1/79+zV69GglJCS4tK6kpCSX3j8K7t5771XZsmV17733av/+/a4uBwDcgsuDwuHDh5WSkqLw8HBJUnh4uMqVKydvb2+dP39eeXl58vT0VF5enlJSUhQUFCTDMArVVhD169eXt7e3Mx4ynGThwoXKzs7WyZMn1aVLF1eXAwC3haysrJt+OXb5GIVq1arp3LlzOnbsmCTp6NGjunjxomrVqqXQ0FCtX79ekrR+/XqFhoYqICBAVapUKVQb3BeXmQYA5zAZpWA6u7Vr1+rNN9+UyWSSJI0cOVLt27fX0aNHNX78eF2+fFl+fn569dVXVbt2bUkqdNut5CcrehRuLwsXLtSWLVuUm5srLy8vdejQQUOGDHF1WQBQ6t3qc69UBIXShKBwe3ryySd19epV69/lypXTypUrXVgRcGcxm83W09sZTHx7udXnnssPPQDFgXkUANdKTExUcnIys6K6IYIC3ALzKACuwxgh90ZQgFvIn0fBZDIxjwJQwrjWinsjKMBtxMTEKCwsjN4EoIRxrRX3RlCA2wgICND06dPpTQBKGGOE3BtBAQBQJIwRcm8EBQBAkTBGyL25fApnAMDtLyYmRidPnqQ3wQ0RFAAARZY/Rgjuh0MPAADALoICAACwi6AAAADsIigAAAC7CAoAAMAuggIAoMjMZrPGjx/PBaHcEEEBAFBkXGbafREU4Db4RgO4BpeZdm8EBbiNt99+Wz/88IPeeecdV5cC3FG4zLR7IyjALZjNZn3xxReSpB07dvCNBihBXGbavREU4Bbefvttm2809CoAJYfLTLs3ggLcwpdffmnz9+eff+6aQoA7EJeZdm8EBQBAkXCZafdGUIBbqFq1qs3f1apVc1ElwJ0pJiZGYWFh9Ca4IS4zDbdgNptt/r548aKLKgHuTFxm2n3RowC30KZNG5lMJkmSyWRS27ZtXVwRALgHggLcQkxMjM2oa7o/AaB4EBTgFgICAtS+fXuZTCY9+uijDKYCgGLicFAwDEMrV65Uv3791K1bN0nS3r17tXHjRqcVBxQEg6kAoPg5HBTmzJmjVatWKTo6WmfPnpV0bWT5kiVLnFYcUBD5g6noTQCA4uNwUPjkk0/0+uuvq0uXLtZBY/fcc49OnTrltOIAAIBrORwU8vLy5OvrK0nWoJCRkaHy5cs7pzIAAOByDgeF1q1ba9q0acrOzpZ0bczCnDlzmNMbAAA35nBQiIuL02+//abw8HBduXJFjRs31pkzZ/TCCy84sz4AAOBCDs/MWKFCBS1YsEAXLlzQmTNnFBQUpLvvvtuZtQEFYjablZCQoHHjxjGgEShh7H/uq8DzKPj4+Khq1aqyWCw6f/68zp8/74y6gAJLTExUcnKyEhMTXV0KcMdh/3NfDvcofP3115o4caLOnDkjwzCsy00mkw4fPuyU4gBHmc1mbdu2TYZhaOvWrYqJieFbDVBC2P/cm8M9Cv/4xz/0/PPPa9++ffrhhx+sP0lJSc6sD3BIYmKiLBaLJMlisfCtBihB7H/uzeGgkJWVpV69esnX11eenp42P4Crff7558rNzZUk5ebmaseOHS6uCLhzsP+5N4eDQv/+/bVkyRKbww5AafHII4/YXBSK03aBksP+594cDgodOnTQypUrFR4ernbt2tn8AK4WExMjD49rb2cPDw+u9wCUIPY/9+bwYMaRI0cqIiJCnTp1ko+PjzNrAgosICBA7dq106ZNm9S+fXsGUgEliP3PvTkcFP73v/9p9erV1tQIlDYxMTE6efIk32YAF2D/c18Of+q3a9dOu3fvdmYtQJFw9UjAddj/3JfDPQrZ2dkaMmSIIiIiVKVKFZu2hISEYi8MAAC4nsNBoU6dOqpTp44zawEAAKWMw0Fh+PDhzqwDAACUQg4HBUnas2ePVq9erZSUFAUGBqpHjx6KiopyVm1AgXBRGgAofg4PZvzwww81evRo3X333Xr00UcVGBioMWPGaOXKlc6sD3AYF6UBgOLncI/CkiVLtGzZMtWtW9e6rHPnzho5cqSefPJJpxQHOIqL0gCAczjco5CWlqYHHnjAZlnt2rV16dKlYi8KKCguSgMAzuFwUGjSpImmT5+uq1evSpIyMzOVkJCgxo0bO604wFFclAYAnMPhoPDyyy/ryJEjioiIUIsWLdS0aVMdOXJEL7/8sjPrAxzCRWkAwDkcHqMQGBio5cuX69y5c9azHqpVq+bM2gCHxcTEaNu2bZK4KA0AFCeHexR27typ48ePq1q1amrQoIGqVaumY8eOadeuXUUuIisrS/Hx8erQoYO6deumiRMnSpKOHz+u6OhodezYUdHR0Tpx4oT1NoVtg3vKvyiNyWTiojQAUIwcDgqTJ0+Wr6+vzTJfX19Nnjy5yEXMmDFD3t7e2rx5s9atW6dRo0ZJkuLj4xUbG6vNmzcrNjZWkyZNst6msG1wXzExMQoLC6M3AQCKkcNB4eLFiwoMDLRZFhgYqN9++61IBWRkZGj16tUaNWqUTCaTJOmuu+7SxYsXlZycrK5du0qSunbtquTkZJnN5kK3wb1xURoAKH4Oj1GoWbOmvvnmGzVv3ty6bM+ePbrnnnuKVMCpU6fk7++v+fPna8+ePfL19dWoUaPk4+OjqlWrytPTU5Lk6empwMBAnT17VoZhFKotICCgSLUCAHCnKdC1HkaMGKHevXurZs2aOnXqlD7++GNNnTq1SAXk5eXp1KlTCgsL07hx43Tw4EE9//zzmjNnTpG2W1RJSUkuvX8AAEoDh4NC+/bttXTpUq1atUpffPGFqlWrpiVLlqhBgwZFKiAoKEheXl7WQwUNGzZU5cqV5ePjo/PnzysvL0+enp7Ky8tTSkqKgoKCZBhGodoKon79+vL29i7SYwMAoLTLysq66ZfjAl0UqkGDBkUOBn8WEBCgyMhI7dq1Sy1bttTx48d18eJF3XfffQoNDdX69evVo0cPrV+/XqGhodbDB4VtAwAAjjMZhmE4smJ2drY++eQTHT58WJmZmTZtCQkJRSri1KlTmjBhgtLS0uTl5aXRo0erdevWOnr0qMaPH6/Lly/Lz89Pr776qmrXri1JhW67lfxkRY8CAOBOcKvPPYeDwt///ncdOXJEbdq0Ubly5Wzahg8fXjzVlgIEBQDAneRWn3sOH3r46quvtG3bNvn5+RVrgQAAoPRyeB6FoKAgZWdnO7MWAABQyjjco9CzZ08NHTpU/fr1U5UqVWza/ji3AgAAcB8OB4X//Oc/kqTXXnvNZrnJZLJejAcAALgXh4PC9u3bnVkHSqHt27dry5Ytri7DYWlpaZIkf39/l9ZRUI8++qjatm3r6jIA4IYKNI8CUJrlX8/jdgsKAFCa3TIoxMbGWi/WZM/y5cuLrSCUHm3btr2tvunGxcVJkqZNm+biSgDAfdwyKDzxxBMlUQcAACiFbhkUHn/8cUnSwYMH1bBhw+vaDx06VPxVAQCAUsHheRQGDBhww+XPPPNMsRUDAABKl1v2KFgsFhmGYfOT7+TJk/L09HRqgQAAwHVuGRTCwsKsgxnDwsJs2jw8PPT88887pzIAAOBytwwK27Ztk2EY6tu3r3XSJenaREsBAQHy8fFxaoEAAMB1bhkUatSoIUnasWOH04sBAACli8MTLqWlpWnp0qU6fPiwMjMzbdqYRwEAAPfkcFAYM2aMsrOz1blzZ5UrV86ZNQEAgFLC4aBw4MAB7d69W2XLlnVmPQBwx7vdrrMi3Z7XWuE6K45xeB6FkJAQnTt3zpm1AABuU2az2Xq9FbgXh3sUoqKi9Mwzz6hXr1666667bNp69+5d7IUBwJ3qdrvOisS1VtyZw0Fh3759qlq1qnbt2mWz3GQyERQAAHBTDgeF9957z5l1AACAUsjhoPBHf57K2cPD4aEOAADgNuJwUDh//rwmT56sffv26fLlyzZthw8fLvbCAACA6zncFRAfH68yZcro7bffVvny5fXJJ5+obdu2evnll51ZHwAAcKECzaOwY8cOlS9fXiaTSXXr1tWUKVMUExOjJ5980pk1AgAAF3G4R8HDw0NeXtdyhZ+fn8xms8qXL6/z5887rTgAAOBaDvcoNGzYUF988YUeffRRtWzZUqNHj5aPj4/q16/vzPoAAIALORwUEhISZLFYJEkTJkzQ0qVLlZGRoaefftppxQEAANdyOCj4+flZf/fx8dHQoUOdUhAAACg9HB6jMHz4cO3bt89m2b59+zRy5MhiLwoAAJQODgeFvXv3qnHjxjbLGjZsqD179hR7UQAAoHRwOCiULVtWV69etVl29epV65kQAADA/TgcFFq2bKlJkyYpPT1dkpSenq7JkyerVatWTisOAAC4lsNBYfz48UpPT1ezZs3UvHlzNWvWTOnp6ZowYYIz6wMAAC7k8HGDSpUq6Y033lBKSorOnTunoKAg3X333c6sDQAAuNhNg4JhGDKZTJJknUPhrrvu0l133WWzjKtHAgDgnm4aFMLDw/Xdd99JksLCwqyhIV9+kODqkQAAuKebBoUNGzZYf9+2bZvTiwEAAKXLTYNCUFCQ9fcaNWo4vRgAAFC63DQojB079rrDDTeSkJBQbAUBAIDS46ZBoVatWiVVBwAAKIVuGhSGDx9eUnUAAIBSqEDzL3/zzTfasGGDUlJSFBgYqC5duqh58+bOqg0AALiYwxMgLF26VH//+99VqVIltW7dWv7+/hozZoyWLl3qzPoAAIALOdyjsGzZMr3zzjsKDg62LuvRo4cGDBiggQMHOqU4AADgWgWaUvHPgxtr1qzp0FkRAADg9uRwUBgxYoQmTJigEydO6Pfff9fx48c1ceJEjRw5UhaLxfoDAADch8OHHiZNmiTp2myNJpNJhmFIktatW6dJkyYxnTMAAG7I4aDAFM4AANx5HA4K+VM4G4ah1NRUVa5cmfEJAAC4OYfHKFy+fFljx47VQw89pBYtWqhBgwYaO3as0tLSnFgeAABwJYeDQlxcnLKysrRmzRodOHBAq1evVnZ2tiZMmODM+gAAgAs5fOhh9+7d2rVrl3x8fCRJDzzwgKZPn65WrVo5rTgAAOBaDvco1K5dW6dPn7ZZdubMGd1///3FXhQAACgdHO5RaN68uQYOHKgePXqoWrVqOnfunNauXasePXpo1apV1vV69+5d6GLmz5+vefPmad26dQoODtb333+vSZMmKSsrSzVq1NCMGTNUpUoVSSp0GwAAcJzDPQoHDhzQvffeqwMHDujTTz/VgQMHVLNmTX333Xdas2aN1qxZo7Vr1xa6kB9++EHff/+99ewKi8WisWPHatKkSdq8ebMiIiL073//u0htAACgYBzuUXjvvfecVkR2drYmT56smTNnql+/fpKkpKQkeXt7KyIiQpIUExOjdu3aadq0aYVuAwAABVOgaz3kMwyjWKdtnjNnjrp376577rnHuuzs2bOqXr269e+AgABZLBalpaUVug0AABSMwz0K58+f1+TJk7Vv3z5dvnzZpq0o0zYfOHBASUlJeuGFFwq9DWdISkpydQkooCtXrkiS9u/f7+JKgDsP+5/7cjgoxMfHy8fHR2+//bb69Omj5cuXa968eWrdunWRCti7d6+OHj2qdu3aSZLOnTunQYMGqW/fvjpz5ox1PbPZLA8PD/n7+ysoKKhQbQVRv359eXt7F+mxoWTlD6oNDw93cSXAnYf97/aVlZV10y/HBRrMOHXqVIWGhspkMqlu3bqaMmWKli5dWqQCn3vuOe3cuVPbt2/X9u3bVa1aNb311lt65pln9Pvvv2vfvn2SpMTERHXq1EnStQ/xwrQBAICCcbhHwcPDQ15e11b38/OT2WxWhQoVdP78eacU5uHhoYSEBMXHx9uc5liUNgAAUDAOB4WGDRvqiy++0KOPPqqWLVtq9OjR8vHxUf369Yu1oO3bt1t/b9KkidatW3fD9QrbBgAAHOdwUEhISLCe4TBhwgS99dZbyszM1NNPP+204gAAgGs5HBT8/Pysv/v4+GjYsGFOKQgAAJQeDg9mzMnJ0dy5c9W2bVs99NBDateunebOnavs7Gxn1gcAAFzI4R6FGTNm6NChQ5o8ebKqV6+uM2fOaOHChUpPT+dS0wAAuCmHg8KmTZu0Zs0aVa5cWdK1q0mGhYWpR48eBAUAANyUw4ceDMMo0HIAAHD7czgodOrUSUOGDNFXX32lo0eP6ssvv9SwYcOYzAgAADfm8KGHsWPHatGiRZo8ebJSUlJUtWpVdenSRUOGDHFmfQAAwIVu2aOwf/9+zZgxQ2XLltWoUaO0ZcsWHTx4UJ999pmys7OVnJxcEnUCAAAXuGVQWLx4sZo2bXrDtsjISL3++uvFXhQAACgdbhkUDh8+rFatWt2wrUWLFlyOGQAAN3bLoJCenq6cnJwbtuXm5iojI6PYiwIAAKXDLYNC7dq1tXPnzhu27dy5U7Vr1y72ogAAQOlwy6DQv39/xcfH67PPPrNeFMpiseizzz7TP//5Tw0YMMDpRQIAANe45emR3bp104ULFzRu3Djl5OTI399faWlpKlOmjEaOHKmuXbuWRJ0AAMAFHJpHYcCAAXriiSd04MABpaWlyd/fX40bN1aFChWcXR8AAHAhhydcqlChgt2zHwAAgHtyeApnAABw5yEoAAAAuwgKAADALoICAACwi6AAAADsIigAAAC7CAoAAMAuggIAALCLoAAAAOwiKAAAALsICgAAwC6CAgAAsIugAAAA7CIoAAAAuwgKAADALoICAACwi6AAAADsIigAAAC7CAoAAMAuggIAALCLoAAAAOwiKAAAALsICgAAwC6CAgAAsIugAAAA7CIoAAAAuwgKAADALoICAACwi6AAAADsIigAAAC7vFxdwJ3izTff1LFjx1xdhlvLf37j4uJcXIn7q127tp599llXlwGgBBAUSsixY8eUlPyjPH38XV2K27LkekqSDh877+JK3Fve72muLqFACOklg6BeMlwR0gkKJcjTx1/la7VzdRlAkWT+us3VJRTIsWPH9PPhH1StAv/dOVM5wyJJunLqRxdX4r7Opee65H7ZcwC4vWoVvDSgQYCrywCKZNkhs0vul8GMAADALoICAACwi6AAAADscnlQSE1N1bPPPquOHTuqW7duGj58uMzma8dhvv/+e3Xv3l0dO3bUwIEDdfHiRevtCtsGAAAc5/KgYDKZ9Mwzz2jz5s1at26datasqX//+9+yWCwaO3asJk2apM2bNysiIkL//ve/JanQbQAAoGBcHhT8/f0VGRlp/btRo0Y6c+aMkpKS5O3trYiICElSTEyMNm3aJEmFbgMAAAXj8qDwRxaLRe+//77atm2rs2fPqnr16ta2gIAAWSwWpaWlFboNAAAUTKmaR+Ff//qXypcvrz59+mjLli0urSUpKalYt3flypVi3R7gSleuXNH+/ftdXYZD2PfgTlyx75WaoPDqq6/q119/1euvvy4PDw8FBQXpzJkz1naz2SwPDw/5+/sXuq0g6tevL29v7yI/rnyrVq2Sfssstu0BrlSxYkWFh4e7ugyHrFq1SlfSXF0FUDycse9lZWXd9MtxqTj08NprrykpKUkLFixQ2bJlJV37oP7999+1b98+SVJiYqI6depUpDYAAFAwLu9R+Pnnn7V48WLdd999iomJkSTdc889WrBggRISEhQfH6+srCzVqFFDM2bMkCR5eHgUqg0AABSMy4NCnTp19OOPN76ISJMmTbRu3bpibQMAAI4rFYceAABA6URQAAAAdhEUAACAXQQFAABgF0EBAADYRVAAAAB2ERQAAIBdBAUAAGAXQQEAANhFUAAAAHYRFAAAgF0EBQAAYBdBAQAA2EVQAAAAdhEUAACAXQQFAABgF0EBAADY5eXqAu4Uqampyvs9TZm/bnN1KUCR5P2eptTUsq4uA0AJoUcBAADYRY9CCalcubLOpWarfK12ri4FKJLMX7epcuXKri4DQAmhRwEAANhFUAAAAHYRFAAAgF0EBQAAYBdBAQAA2EVQAAAAdhEUAACAXQQFAABgF0EBAADYRVAAAAB2ERQAAIBdBAUAAGAXQQEAANhFUAAAAHYRFAAAgF0EBQAAYBdBAQAA2EVQAAAAdhEUAACAXQQFAABgF0EBAADYRVAAAAB2ERQAAIBdBAUAAGAXQQEAANhFUAAAAHYRFAAAgF0EBQAAYJeXqwsAAGdKTU3VhfRcLTtkdnUpQJGcS89Vbmpqid8vPQoAAMAuehQAuLXKlSvLKz1FAxoEuLoUoEiWHTKrYuXKJX6/9CgAAAC7CAoAAMAutw0Kx48fV3R0tDp27Kjo6GidOHHC1SUBAHDbcdugEB8fr9jYWG3evFmxsbGaNGmSq0sCAOC245ZB4eLFi0pOTlbXrl0lSV27dlVycrLMZk6PAgCgINzyrIezZ8+qatWq8vT0lCR5enoqMDBQZ8+eVUDAzUc+G4YhScrOzi7WmsqXL6+K3oY8Ur8t1u06kyUvS8rNcnUZ7s/LWx6e3q6uwmEVvQ2VL19eWVm3x3ujfPnySvH01Ycn8lxdisMysi3KzLG4ugy3V76Mh3zL3j7fl696+qqqE/a9/M+7/M+/P3PLoFAUOTk5kqSffvqpWLf72GOP6bFi3SLgWklJSa4uwSGPPfaYxN4HN+KsfS8nJ0c+Pj7XLXfLoBAUFKTz588rLy9Pnp6eysvLU0pKioKCgm55W19fXwUHB6tMmTIymUwlUC0AAK5jGIZycnLk6+t7w3a3DApVqlRRaGio1q9frx49emj9+vUKDQ295WEHSfLw8FDFihVLoEoAAEqHG/Uk5DMZ9g5K3OaOHj2q8ePH6/Lly/Lz89Orr76q2rVru7osAABuK24bFAAAQNHdPsM9AQBAiSMoAAAAuwgKAADALoICAACwi6AAAADsIijgthcXF6cZM2bYLOvfv79WrFjhooqAO0Pbtm3VtWtXWSwWm2XFPbMtXIuggNvehAkT9Omnn+rgwYOSpMTERJlMJv3tb39zcWWA+8vMzNSaNWtcXQaciKCA217FihX1r3/9S3FxcTp+/LgWLVqkKVOmMAU3UAKGDx+u+fPnF/uF9FB6EBTgFv7yl7+oadOm6t27t0aMGKHq1au7uiTgjlC/fn3Vq1dP77//vqtLgZMQFOA2Bg0aJE9PT/Xu3dvVpQB3lNGjR+vNN99URkaGq0uBExAU4DY8PDw43AC4QO3atdW6dWstW7bM1aXACdzy6pEAgJI1YsQI9erVS3l5ea4uBcWMHgUAQJFVq1ZNPXr0UFpamqtLQTHj6pEAAMAuehQAAIBdBAUAAGAXQQEAANhFUAAAAHYRFAAAgF0EBQAAYBdBAQAA2EVQAO5gbdu21ddff+3qMvThhx+qU6dOaty4sVq0aKFnn31W6enpkqTx48dr1qxZDm/r448/5hLjQDFiCmcALvXtt99q1qxZWrJkicLCwpSWlqYdO3a4uiwA/z96FADYyM7O1pQpU9SyZUu1bNlSU6ZMUXZ2tiTp0qVLGjx4sKKiotS0aVMNHjxY586ds962b9++mj17tmJiYtS4cWMNHDhQZrP5pvf33//+V40aNVJYWJgkyd/fX48//rgqVKigDz74QOvWrdNbb72lxo0b6/nnn5ckvfHGG2rfvr0aN26sxx57TFu2bJEkHT16VPHx8fr+++/VuHFjRUREWOv68MMPrff5x14HwzA0depUNW/eXE2aNFG3bt30008/FdOzCdz+CAoAbCxatEgHDx7UmjVrtHbtWv33v//VwoULJUkWi0W9evXSjh07tGPHDnl7e2vy5Mk2t1+/fr2mTZumb775Rjk5OVq6dOlN769hw4bauXOn5s6dq/3791tDiSRFR0erW7duGjRokA4cOKDXX39dklSzZk0tX75c+/fv1/DhwzV27FilpKTogQce0Msvv6xGjRrpwIED2rdv3y0f786dO7Vv3z5t3rxZ+/fv1+zZs+Xv71/AZw1wXwQFADbWrVunYcOGqUqVKgoICNCwYcO0du1aSVLlypXVsWNHlStXThUqVNCQIUO0d+9em9v36tVL999/v3x8fNSpUycdPnz4pvcXERGhefPmKTk5WYMHD1ZkZKSmTZt206sQdu7cWVWrVpWHh4cee+wx1apVS4cOHSrU4/Xy8lJGRoaOHTsmwzD0wAMPKDAwsFDbAtwRYxQA2EhJSVH16tWtf1evXl0pKSmSpKtXr2ratGn66quvdOnSJUlSRkaG8vLy5OnpKUm6++67rbctV66cMjMzb3mfrVu3VuvWrWWxWLRnzx6NGjVK999/v2JiYm64/urVq7Vs2TKdPn1akpSZmanU1NRCPd7mzZvrqaee0uTJk3X69Gl16NBB48aNU4UKFQq1PcDd0KMAwEZgYKDOnDlj/fvs2bPWb9hLly7V8ePHtXLlSn333Xdavny5pGvH+YuDh4eHmjdvrqioKP3888+SJJPJZLPO6dOn9dJLL2nixInas2eP9u3bpzp16ljb/7y+dC2wXL161fr3hQsXbNr79eunjz/+WBs3btSJEye0ZMmSYnk8gDsgKAB3uJycHGVlZVl/unTpokWLFslsNstsNmvBggXq1q2bpGu9B97e3vLz81NaWprmz59f5PvfunWrNmzYoEuXLskwDB06dEjffvutGjZsKEmqUqWK/ve//1nXv3r1qkwmkwICAiRJH330kTVU5K9//vx5m7EOoaGh2rJli65evapff/1Vq1atsrYdOnRIBw8eVE5OjsqVK6eyZcvKw4P/GoF87A3AHe65555TgwYNrD/Z2dmqX7++unfvru7du6tevXoaOnSoJOnpp59WVlaWoqKiFB0drVatWhX5/itVqqSVK1eqQ4cOatKkicaOHatBgwape/fukqTevXvrl19+UUREhIYOHaoHH3xQAwcOVExMjFq0aKGffvpJTZo0sW4vKipKDz74oFq2bKnIyEhr3WXKlFGLFi00btw4a/CRroWfl156Sc2aNVObNm3k7++vQYMGFflxAe7CZBRXnyEAAHA79CgAAAC7OOsBgNOtXbtW8fHx1y2vXr26NmzY4IKKADiKQw8AAMAuDj0AAAC7CAoAAMAuggIAALCLoAAAAOwiKAAAALv+PyWh2wROHDEgAAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Although the distributions are similar, the refused loans distribuition has a smaller median than the accepted one. It indicates that a half of the denied loans is located in the very narrow space under the line, being this variable zero or almost it."
      ],
      "metadata": {
        "id": "qzC9LqpDKLyt"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Loan amount terms\n",
        "boxplot(column=numeric_cols[3],\\\n",
        "        target=target_train,\\\n",
        "        train=df_train,\\\n",
        "        max_y=max(df_train[numeric_cols[3]]),\\\n",
        "        title='Loan Amount Boxplot')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "mp7d6PrEMowZ",
        "outputId": "398e0b6e-dd1d-4e6f-d5b3-9973e4ba91d3"
      },
      "execution_count": 227,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "With only a few outliers, the vast majority of the amount terms are condensed on the line that represents a bit less than 400 dolars."
      ],
      "metadata": {
        "id": "lduarZcZNVWT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Loan amount terms\n",
        "boxplot(column=numeric_cols[4],\\\n",
        "        target=target_train,\\\n",
        "        train=df_train,\\\n",
        "        max_y=2*max(df_train[numeric_cols[4]]),\\\n",
        "        title='Credit History Boxplot')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "eeqvuiCgOLfJ",
        "outputId": "ca16ce09-fa86-4ce0-947b-6bb416cfd109"
      },
      "execution_count": 228,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x432 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Very different behaviors. Almost all accepted loans in 1.0, while denied ones are very disperse, but also most frequently in 1.0, when it is analysed carefully. The disaproved loans samples have more variaty of credit history."
      ],
      "metadata": {
        "id": "VYqIBMAHPRW4"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Average of numerical features in ACCEPTED loans\n",
        "round(df_train[df_train['Loan_Status'] == 'Y'].mean())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ftrm5HBgi4FM",
        "outputId": "979a916a-769a-48fa-a0ea-4fe9000bfaf3"
      },
      "execution_count": 229,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ApplicantIncome      5384.0\n",
              "CoapplicantIncome    1505.0\n",
              "LoanAmount            144.0\n",
              "Loan_Amount_Term      341.0\n",
              "Credit_History          1.0\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 229
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Average of numerical features in NOT ACCETPTED loans\n",
        "round(df_train[df_train['Loan_Status'] == 'N'].mean())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xtn80EFwkksQ",
        "outputId": "4afd17cb-1304-44b6-dbda-bb81da8caed0"
      },
      "execution_count": 230,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ApplicantIncome      5446.0\n",
              "CoapplicantIncome    1878.0\n",
              "LoanAmount            151.0\n",
              "Loan_Amount_Term      344.0\n",
              "Credit_History          1.0\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 230
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df_train.head(3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "fyeFfSb9zpLM",
        "outputId": "1a43b095-c3d2-4b98-8bff-0cbf799a55c8"
      },
      "execution_count": 231,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "    Loan_ID Gender Married Dependents Education Self_Employed  \\\n",
              "0  LP001002   Male      No          0  Graduate            No   \n",
              "1  LP001003   Male     Yes          1  Graduate            No   \n",
              "2  LP001005   Male     Yes          0  Graduate           Yes   \n",
              "\n",
              "   ApplicantIncome  CoapplicantIncome  LoanAmount  Loan_Amount_Term  \\\n",
              "0             5849                0.0         NaN             360.0   \n",
              "1             4583             1508.0       128.0             360.0   \n",
              "2             3000                0.0        66.0             360.0   \n",
              "\n",
              "   Credit_History Property_Area Loan_Status  \n",
              "0             1.0         Urban           Y  \n",
              "1             1.0         Rural           N  \n",
              "2             1.0         Urban           Y  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1b82b618-d332-4b84-b54d-69b11a3e02b8\">\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>Loan_ID</th>\n",
              "      <th>Gender</th>\n",
              "      <th>Married</th>\n",
              "      <th>Dependents</th>\n",
              "      <th>Education</th>\n",
              "      <th>Self_Employed</th>\n",
              "      <th>ApplicantIncome</th>\n",
              "      <th>CoapplicantIncome</th>\n",
              "      <th>LoanAmount</th>\n",
              "      <th>Loan_Amount_Term</th>\n",
              "      <th>Credit_History</th>\n",
              "      <th>Property_Area</th>\n",
              "      <th>Loan_Status</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>LP001002</td>\n",
              "      <td>Male</td>\n",
              "      <td>No</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>5849</td>\n",
              "      <td>0.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>LP001003</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>1</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>4583</td>\n",
              "      <td>1508.0</td>\n",
              "      <td>128.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Rural</td>\n",
              "      <td>N</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>LP001005</td>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>Yes</td>\n",
              "      <td>3000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>66.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "      <td>Y</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1b82b618-d332-4b84-b54d-69b11a3e02b8')\"\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-1b82b618-d332-4b84-b54d-69b11a3e02b8 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-1b82b618-d332-4b84-b54d-69b11a3e02b8');\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": 231
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "encoder = LabelEncoder()\n",
        "df_train['Loan_Status'] = encoder.fit_transform(df_train['Loan_Status']) # Convert to numeric for barplotting"
      ],
      "metadata": {
        "id": "H5LuC95S6XOv"
      },
      "execution_count": 232,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### 1.4.3 Loan status and categorical features"
      ],
      "metadata": {
        "id": "mgDf7qb0A1js"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Autoplot categoricals\n",
        "for column in categorical_cols:\n",
        "  sns.barplot(data=df_train, x=column, y='Loan_Status', color='g')\n",
        "  plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "dpCsxD9H2tzc",
        "outputId": "1fb90bda-90e9-4739-b3e0-2e02b04f289d"
      },
      "execution_count": 233,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 2.0 Data cleaning and preparation"
      ],
      "metadata": {
        "id": "20AEia3ZuQNw"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.1 Adjusting data types"
      ],
      "metadata": {
        "id": "1xEezl3gT6Uo"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Treat missing data of the full data frame\n",
        "df_complete = pd.concat([df_train],axis = 0)\n",
        "\n",
        "# Save the id column of the test data frame\n",
        "y =  df_complete['Loan_Status']"
      ],
      "metadata": {
        "id": "8rBiMmaT1d8k"
      },
      "execution_count": 234,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Check df variable types\n",
        "df_complete.dtypes"
      ],
      "metadata": {
        "id": "Ezj3JN6TxmAs",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "60b3de59-2196-455e-ca7e-204625851a41"
      },
      "execution_count": 235,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Loan_ID               object\n",
              "Gender                object\n",
              "Married               object\n",
              "Dependents            object\n",
              "Education             object\n",
              "Self_Employed         object\n",
              "ApplicantIncome        int64\n",
              "CoapplicantIncome    float64\n",
              "LoanAmount           float64\n",
              "Loan_Amount_Term     float64\n",
              "Credit_History       float64\n",
              "Property_Area         object\n",
              "Loan_Status            int64\n",
              "dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 235
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Change 'CoapplicantIncome' type to stablish a standard in data\n",
        "df_complete['ApplicantIncome'] = df_train['ApplicantIncome'].astype(float)\n",
        "df_complete['ApplicantIncome'].dtypes"
      ],
      "metadata": {
        "id": "ikgSXgd0EsZE",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "781d4d5c-2c1a-41a9-e65d-8bbc49dc7129"
      },
      "execution_count": 236,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "dtype('float64')"
            ]
          },
          "metadata": {},
          "execution_count": 236
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.2 Missing values overview"
      ],
      "metadata": {
        "id": "g9ZhLShfUGOZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Drop the target column\n",
        "df_relevant_columns = df_complete.drop(['Loan_Status', 'Loan_ID'], axis = 1)\n",
        "df_relevant_columns.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "dM83hVYUwft9",
        "outputId": "987ac49d-62c2-4826-a6d4-34149cdf1366"
      },
      "execution_count": 237,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  Gender Married Dependents     Education Self_Employed  ApplicantIncome  \\\n",
              "0   Male      No          0      Graduate            No           5849.0   \n",
              "1   Male     Yes          1      Graduate            No           4583.0   \n",
              "2   Male     Yes          0      Graduate           Yes           3000.0   \n",
              "3   Male     Yes          0  Not Graduate            No           2583.0   \n",
              "4   Male      No          0      Graduate            No           6000.0   \n",
              "\n",
              "   CoapplicantIncome  LoanAmount  Loan_Amount_Term  Credit_History  \\\n",
              "0                0.0         NaN             360.0             1.0   \n",
              "1             1508.0       128.0             360.0             1.0   \n",
              "2                0.0        66.0             360.0             1.0   \n",
              "3             2358.0       120.0             360.0             1.0   \n",
              "4                0.0       141.0             360.0             1.0   \n",
              "\n",
              "  Property_Area  \n",
              "0         Urban  \n",
              "1         Rural  \n",
              "2         Urban  \n",
              "3         Urban  \n",
              "4         Urban  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9dccaae4-9083-4761-9fec-b7a0af09247e\">\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>Gender</th>\n",
              "      <th>Married</th>\n",
              "      <th>Dependents</th>\n",
              "      <th>Education</th>\n",
              "      <th>Self_Employed</th>\n",
              "      <th>ApplicantIncome</th>\n",
              "      <th>CoapplicantIncome</th>\n",
              "      <th>LoanAmount</th>\n",
              "      <th>Loan_Amount_Term</th>\n",
              "      <th>Credit_History</th>\n",
              "      <th>Property_Area</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Male</td>\n",
              "      <td>No</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>5849.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>1</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>4583.0</td>\n",
              "      <td>1508.0</td>\n",
              "      <td>128.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Rural</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>Yes</td>\n",
              "      <td>3000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>66.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Male</td>\n",
              "      <td>Yes</td>\n",
              "      <td>0</td>\n",
              "      <td>Not Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>2583.0</td>\n",
              "      <td>2358.0</td>\n",
              "      <td>120.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Male</td>\n",
              "      <td>No</td>\n",
              "      <td>0</td>\n",
              "      <td>Graduate</td>\n",
              "      <td>No</td>\n",
              "      <td>6000.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>141.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>Urban</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9dccaae4-9083-4761-9fec-b7a0af09247e')\"\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-9dccaae4-9083-4761-9fec-b7a0af09247e 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-9dccaae4-9083-4761-9fec-b7a0af09247e');\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": 237
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Function to organize in a table all columns with missing data grouped by columns\n",
        "def missing_values_df(df):\n",
        "    quantity_missing_values_df = df.isnull().sum().sort_values(ascending=False)\n",
        "    percentage_missing_values_df = (df.isnull().sum()/df.isnull().count()*100).sort_values(ascending=False)\n",
        "    percentage_missing_values_df = round(percentage_missing_values_df, 2)\n",
        "    missing_data_df = pd.concat([quantity_missing_values_df, percentage_missing_values_df], axis=1, keys=['Missing values', 'Percentage'])\n",
        "    return missing_data_df"
      ],
      "metadata": {
        "id": "NLJ57Py3zGd0"
      },
      "execution_count": 238,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Columns with missing values\n",
        "missing_data = missing_values_df(df_relevant_columns)\n",
        "missing_data[missing_data['Missing values'] > 0]"
      ],
      "metadata": {
        "id": "0OqeKggKziLG",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "outputId": "aadd6a29-3eea-4bb4-f07d-dbce10a3b22b"
      },
      "execution_count": 239,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                  Missing values  Percentage\n",
              "Credit_History                50        8.14\n",
              "Self_Employed                 32        5.21\n",
              "LoanAmount                    22        3.58\n",
              "Dependents                    15        2.44\n",
              "Loan_Amount_Term              14        2.28\n",
              "Gender                        13        2.12\n",
              "Married                        3        0.49"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-f085aeab-f20b-49bf-bfa2-eee7fcae8b1c\">\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>Missing values</th>\n",
              "      <th>Percentage</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Credit_History</th>\n",
              "      <td>50</td>\n",
              "      <td>8.14</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Self_Employed</th>\n",
              "      <td>32</td>\n",
              "      <td>5.21</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LoanAmount</th>\n",
              "      <td>22</td>\n",
              "      <td>3.58</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Dependents</th>\n",
              "      <td>15</td>\n",
              "      <td>2.44</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Loan_Amount_Term</th>\n",
              "      <td>14</td>\n",
              "      <td>2.28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Gender</th>\n",
              "      <td>13</td>\n",
              "      <td>2.12</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Married</th>\n",
              "      <td>3</td>\n",
              "      <td>0.49</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f085aeab-f20b-49bf-bfa2-eee7fcae8b1c')\"\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-f085aeab-f20b-49bf-bfa2-eee7fcae8b1c 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-f085aeab-f20b-49bf-bfa2-eee7fcae8b1c');\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": 239
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.3 Missing values fulfillment"
      ],
      "metadata": {
        "id": "ey2kQd-j0wEA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Credit history\n",
        "df_relevant_columns['Credit_History'] = df_relevant_columns['Credit_History'].fillna(-1)"
      ],
      "metadata": {
        "id": "El08R5xqTJr7"
      },
      "execution_count": 240,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "The value -1 is a good marker because it clearly differs from this positive classification."
      ],
      "metadata": {
        "id": "HqZPUZEV5PZF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Self employed\n",
        "df_relevant_columns['Self_Employed'] = df_relevant_columns['Self_Employed'].fillna('Other')"
      ],
      "metadata": {
        "id": "d-6gMI7p2DLW"
      },
      "execution_count": 241,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "The lack of information can be due to the fact that there can be job situations that do not fit well in this binary classification. For instance, cases like retirement, unemployment or not being in labor force."
      ],
      "metadata": {
        "id": "73S7l0Yc31BS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Remaining columns\n",
        "numerical_remaining_columns = ['LoanAmount', 'Loan_Amount_Term']\n",
        "categorical_remaining_columns = ['Gender', 'Married','Dependents']"
      ],
      "metadata": {
        "id": "A_oE52BO21uu"
      },
      "execution_count": 242,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a general median filling fuction\n",
        "median_inputer = SimpleImputer(strategy='median')\n",
        "\n",
        "for column in numerical_remaining_columns:\n",
        "    df_relevant_columns[column] = pd.DataFrame(median_inputer.fit_transform(pd.DataFrame(df_relevant_columns[column])))"
      ],
      "metadata": {
        "id": "p4jeoEJ-CXO4"
      },
      "execution_count": 243,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a general median filling fuction\n",
        "most_frequent_inputer = SimpleImputer(strategy='most_frequent')\n",
        "\n",
        "for column in categorical_remaining_columns:\n",
        "    df_relevant_columns[column] = pd.DataFrame(most_frequent_inputer.fit_transform(pd.DataFrame(df_relevant_columns[column])))"
      ],
      "metadata": {
        "id": "lBKJEB4zCvBu"
      },
      "execution_count": 244,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Check number of NaNs\n",
        "df_relevant_columns.apply(lambda x: x.isna().value_counts())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "Sg1xW7LkEVji",
        "outputId": "e5cccf93-a63f-4171-ca80-5ba32829dfbd"
      },
      "execution_count": 245,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Gender  Married  Dependents  Education  Self_Employed  ApplicantIncome  \\\n",
              "False     614      614         614        614            614              614   \n",
              "\n",
              "       CoapplicantIncome  LoanAmount  Loan_Amount_Term  Credit_History  \\\n",
              "False                614         614               614             614   \n",
              "\n",
              "       Property_Area  \n",
              "False            614  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-40c1c67d-73dc-423f-a2bd-99912a8b7a0d\">\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>Gender</th>\n",
              "      <th>Married</th>\n",
              "      <th>Dependents</th>\n",
              "      <th>Education</th>\n",
              "      <th>Self_Employed</th>\n",
              "      <th>ApplicantIncome</th>\n",
              "      <th>CoapplicantIncome</th>\n",
              "      <th>LoanAmount</th>\n",
              "      <th>Loan_Amount_Term</th>\n",
              "      <th>Credit_History</th>\n",
              "      <th>Property_Area</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>False</th>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "      <td>614</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-40c1c67d-73dc-423f-a2bd-99912a8b7a0d')\"\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-40c1c67d-73dc-423f-a2bd-99912a8b7a0d 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-40c1c67d-73dc-423f-a2bd-99912a8b7a0d');\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": 245
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "All nans were successfully fulfilled"
      ],
      "metadata": {
        "id": "dkEs-TrbEj9K"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.4 Correlation analysis"
      ],
      "metadata": {
        "id": "6C7ko0aBZJes"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Graphic plotation\n",
        "# Upper triangle was hidden in order to ease the visualization, because it has the same values from the lower one\n",
        "plt.figure(figsize=(16, 6))\n",
        "mask = np.triu(np.ones_like(df_relevant_columns.corr(), dtype=np.bool)) # Defining the mask to set the values in the upper triangle to True\n",
        "heatmap = sns.heatmap(df_relevant_columns.corr(), mask=mask, vmin=-1, vmax=1, annot=True, cmap='BrBG')\n",
        "heatmap.set_title('Triangle Correlation Heatmap', fontdict={'fontsize':18}, pad=16) "
      ],
      "metadata": {
        "id": "AfXIuOfzatGj",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "outputId": "ccc5c847-4e45-408c-d6b4-f0d155f1dc4e"
      },
      "execution_count": 246,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Triangle Correlation Heatmap')"
            ]
          },
          "metadata": {},
          "execution_count": 246
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1152x432 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "As can be observed in the figure above, there are no variables that present a high correlation with each other."
      ],
      "metadata": {
        "id": "NhM7FgYTdMSO"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.5 Feature engineering"
      ],
      "metadata": {
        "id": "Wab52oc7IVzW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create 'Total_Income' column\n",
        "df_relevant_columns['Total_Income'] = df_relevant_columns['ApplicantIncome'] + df_relevant_columns['CoapplicantIncome']\n",
        "df_relevant_columns = df_relevant_columns.drop(columns=['ApplicantIncome', 'CoapplicantIncome'])"
      ],
      "metadata": {
        "id": "2qZV_HT1IYhb"
      },
      "execution_count": 247,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Creating 'Total_income' by summing 'ApplicantIncome' and 'CoapplicantIncome' and dropping them"
      ],
      "metadata": {
        "id": "psLpPc2H3p2T"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df_relevant_columns = df_relevant_columns.replace({'Dependents': r'3+'}, {'Dependents': 4}, regex=True)"
      ],
      "metadata": {
        "id": "X-9e3ETF255y"
      },
      "execution_count": 248,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Convert 3+ category into 4 (also making it numerical)"
      ],
      "metadata": {
        "id": "kLmHvPnt3fBx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Apply LabelEncoder to categorical features\n",
        "le = LabelEncoder()\n",
        "le.fit(list(df_relevant_columns[\"Dependents\"].values))\n",
        "df_relevant_columns[\"Dependents\"] = le.transform(list(df_relevant_columns[\"Dependents\"].values))\n",
        "print('Shape all_data: {}'.format(df_relevant_columns.shape))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jW4Og_u68i5O",
        "outputId": "dfec80f9-82e5-4785-d712-8cd442e15c50"
      },
      "execution_count": 249,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Shape all_data: (614, 10)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.6 Categorical columns dummyzation "
      ],
      "metadata": {
        "id": "sY_p3THaZ-NJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df_dummies = pd.get_dummies(df_relevant_columns)\n",
        "df_dummies.head(3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "0W6pLT_xPJZy",
        "outputId": "a89a155c-5aaf-4cfc-a364-508d489236ac"
      },
      "execution_count": 250,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   Dependents  LoanAmount  Loan_Amount_Term  Credit_History  Total_Income  \\\n",
              "0           0       128.0             360.0             1.0        5849.0   \n",
              "1           1       128.0             360.0             1.0        6091.0   \n",
              "2           0        66.0             360.0             1.0        3000.0   \n",
              "\n",
              "   Gender_Female  Gender_Male  Married_No  Married_Yes  Education_Graduate  \\\n",
              "0              0            1           1            0                   1   \n",
              "1              0            1           0            1                   1   \n",
              "2              0            1           0            1                   1   \n",
              "\n",
              "   Education_Not Graduate  Self_Employed_No  Self_Employed_Other  \\\n",
              "0                       0                 1                    0   \n",
              "1                       0                 1                    0   \n",
              "2                       0                 0                    0   \n",
              "\n",
              "   Self_Employed_Yes  Property_Area_Rural  Property_Area_Semiurban  \\\n",
              "0                  0                    0                        0   \n",
              "1                  0                    1                        0   \n",
              "2                  1                    0                        0   \n",
              "\n",
              "   Property_Area_Urban  \n",
              "0                    1  \n",
              "1                    0  \n",
              "2                    1  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-426ac7f4-e87c-4262-be67-ce36c3cefdb2\">\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>Dependents</th>\n",
              "      <th>LoanAmount</th>\n",
              "      <th>Loan_Amount_Term</th>\n",
              "      <th>Credit_History</th>\n",
              "      <th>Total_Income</th>\n",
              "      <th>Gender_Female</th>\n",
              "      <th>Gender_Male</th>\n",
              "      <th>Married_No</th>\n",
              "      <th>Married_Yes</th>\n",
              "      <th>Education_Graduate</th>\n",
              "      <th>Education_Not Graduate</th>\n",
              "      <th>Self_Employed_No</th>\n",
              "      <th>Self_Employed_Other</th>\n",
              "      <th>Self_Employed_Yes</th>\n",
              "      <th>Property_Area_Rural</th>\n",
              "      <th>Property_Area_Semiurban</th>\n",
              "      <th>Property_Area_Urban</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>128.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5849.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>128.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>6091.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0</td>\n",
              "      <td>66.0</td>\n",
              "      <td>360.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3000.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-426ac7f4-e87c-4262-be67-ce36c3cefdb2')\"\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-426ac7f4-e87c-4262-be67-ce36c3cefdb2 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-426ac7f4-e87c-4262-be67-ce36c3cefdb2');\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": 250
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2.7 Train and test data frames"
      ],
      "metadata": {
        "id": "nm-AbkJD7ccJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "x_train, x_test, y_train, y_test = train_test_split(df_dummies, y, test_size=0.3, random_state=SEED)\n",
        "x_train.shape, x_test.shape, y_train.shape, y_test.shape"
      ],
      "metadata": {
        "id": "eG1q6sOraEgM",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9718f0d0-2bbf-4476-80eb-1955b23ad6f9"
      },
      "execution_count": 251,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((429, 17), (185, 17), (429,), (185,))"
            ]
          },
          "metadata": {},
          "execution_count": 251
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 3.0 Machine learning models"
      ],
      "metadata": {
        "id": "o_L76RRKFVvM"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.1 Simple models"
      ],
      "metadata": {
        "id": "v03qsku9JhFp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Dummy classifier\n",
        "\n",
        "dummy = DummyClassifier(random_state=SEED)\n",
        "dummy.fit(x_train,y_train)\n",
        "y_pred = dummy.predict(x_test)\n",
        "acc_dummy = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "YthuMf0O63Hr"
      },
      "execution_count": 252,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Logistic regression\n",
        "\n",
        "logreg = LogisticRegression(random_state=SEED)\n",
        "logreg.fit(x_train,y_train)\n",
        "y_pred = logreg.predict(x_test)\n",
        "acc_log = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "rmsuOtDVJry3"
      },
      "execution_count": 253,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Support vector classifier\n",
        "\n",
        "svc = SVC(random_state=SEED)\n",
        "svc.fit(x_train, y_train)\n",
        "y_pred = svc.predict(x_test)\n",
        "acc_svc  = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "JmJkcwGjJuSl"
      },
      "execution_count": 254,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# K-neighbors classifier\n",
        "\n",
        "knn = KNeighborsClassifier(n_neighbors = 3)\n",
        "knn.fit(x_train, y_train)\n",
        "y_pred = knn.predict(x_test)\n",
        "acc_knn = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "CD7pI3E_JuMo"
      },
      "execution_count": 255,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Gaussian Naive Bayes\n",
        "\n",
        "gaussian = GaussianNB()\n",
        "gaussian.fit(x_train, y_train)\n",
        "y_pred = gaussian.predict(x_test)\n",
        "acc_gaussian = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "h7_ul-h3jbr0"
      },
      "execution_count": 256,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Perceptron\n",
        "\n",
        "perceptron = Perceptron(random_state=SEED)\n",
        "perceptron.fit(x_train, y_train)\n",
        "y_pred = perceptron.predict(x_test)\n",
        "acc_perceptron = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "KMUQ5-pNJuGa"
      },
      "execution_count": 257,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Decision Tree classifier\n",
        "\n",
        "decision_tree = DecisionTreeClassifier(random_state=SEED)\n",
        "decision_tree.fit(x_train, y_train)\n",
        "y_pred = decision_tree.predict(x_test)\n",
        "acc_decision_tree = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "FGiojHOJJt_h"
      },
      "execution_count": 258,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Random Forest classifier\n",
        "\n",
        "random_forest = RandomForestClassifier(random_state = SEED)\n",
        "random_forest.fit(x_train, y_train)\n",
        "y_pred = random_forest.predict(x_test)\n",
        "acc_random_forest = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "3zpwFUMJJtnR"
      },
      "execution_count": 259,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Gradient boosting classifier\n",
        "\n",
        "gbc = GradientBoostingClassifier(random_state=SEED)\n",
        "gbc.fit(x_train, y_train)\n",
        "y_pred = gbc.predict(x_test)\n",
        "acc_gbc = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "YoB9vt1JkA9K"
      },
      "execution_count": 260,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Light gradient boosting machine classifier\n",
        "\n",
        "lgbmc = lgb.LGBMClassifier(random_state=SEED)\n",
        "lgbmc.fit(x_train, y_train)\n",
        "y_pred = lgbmc.predict(x_test)\n",
        "acc_lgbmc = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "Siqi7KAqjvci"
      },
      "execution_count": 261,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# X gradient boosting classifier\n",
        "\n",
        "xgbc = XGBClassifier(random_state=SEED)\n",
        "xgbc.fit(x_train, y_train)\n",
        "y_pred = xgbc.predict(x_test)\n",
        "acc_xgbc = round(accuracy_score(y_test,y_pred)*100,2)"
      ],
      "metadata": {
        "id": "HY3N4sCE2Wik"
      },
      "execution_count": 262,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "models = pd.DataFrame({\n",
        "    'Model': ['Dummy classifier','Support vector classifier', 'K-neighbors classifier', 'Logistic regressor', \n",
        "              'Random forest classifier', 'Gaussian Naive-Bayes', 'Perceptron',  \n",
        "              'Decision tree classifier', 'Gradient boosting classifier', 'XGBClassifier', 'LGBMClassifier'],\n",
        "              \n",
        "    'RMSE': [acc_dummy, acc_svc, acc_knn, acc_log, \n",
        "              acc_random_forest, acc_gaussian, acc_perceptron, \n",
        "              acc_decision_tree, acc_gbc, acc_xgbc, acc_lgbmc]})\n",
        "models.sort_values(by='RMSE', ascending=False)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "UqrLSO8Njy9Z",
        "outputId": "63f54b8f-2f5e-49f5-a120-9ccee2805105"
      },
      "execution_count": 263,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                           Model   RMSE\n",
              "4       Random forest classifier  80.00\n",
              "8   Gradient boosting classifier  80.00\n",
              "10                LGBMClassifier  79.46\n",
              "3             Logistic regressor  77.30\n",
              "9                  XGBClassifier  77.30\n",
              "0               Dummy classifier  74.59\n",
              "7       Decision tree classifier  74.59\n",
              "1      Support vector classifier  74.05\n",
              "5           Gaussian Naive-Bayes  74.05\n",
              "6                     Perceptron  56.76\n",
              "2         K-neighbors classifier  56.22"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-cc26855a-d563-4e0a-89c7-5db2b19604e8\">\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>Model</th>\n",
              "      <th>RMSE</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Random forest classifier</td>\n",
              "      <td>80.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Gradient boosting classifier</td>\n",
              "      <td>80.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>LGBMClassifier</td>\n",
              "      <td>79.46</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Logistic regressor</td>\n",
              "      <td>77.30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>XGBClassifier</td>\n",
              "      <td>77.30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Dummy classifier</td>\n",
              "      <td>74.59</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Decision tree classifier</td>\n",
              "      <td>74.59</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Support vector classifier</td>\n",
              "      <td>74.05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Gaussian Naive-Bayes</td>\n",
              "      <td>74.05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Perceptron</td>\n",
              "      <td>56.76</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>K-neighbors classifier</td>\n",
              "      <td>56.22</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-cc26855a-d563-4e0a-89c7-5db2b19604e8')\"\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-cc26855a-d563-4e0a-89c7-5db2b19604e8 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-cc26855a-d563-4e0a-89c7-5db2b19604e8');\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": 263
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.2 Cross-validated models"
      ],
      "metadata": {
        "id": "_f06NhJBUiZo"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculate root mean square error\n",
        "def rmse_cv(model):\n",
        "    cv = KFold(n_splits=5, shuffle=True, random_state=SEED)\n",
        "    rmse = (cross_val_score(model, x_test,\n",
        "                                   y_test, scoring='f1', cv = cv))\n",
        "    return(rmse)"
      ],
      "metadata": {
        "id": "Xy7aSko4WW3M"
      },
      "execution_count": 264,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Dummy classifier\n",
        "dummy_score = rmse_cv(dummy)"
      ],
      "metadata": {
        "id": "1iaem-8N7ryV"
      },
      "execution_count": 265,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Support vector classifier\n",
        "svc_score = rmse_cv(svc)"
      ],
      "metadata": {
        "id": "t1Mmcm_n51kr"
      },
      "execution_count": 266,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Logistic regressor\n",
        "logreg_score = rmse_cv(logreg)"
      ],
      "metadata": {
        "id": "5DdSAtkhmTEZ"
      },
      "execution_count": 267,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# K-neighbors classifier\n",
        "knn_score = rmse_cv(knn)"
      ],
      "metadata": {
        "id": "-hoQCZ16nAYK"
      },
      "execution_count": 268,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Gaussian Naive-Bayes\n",
        "gaussian_score = rmse_cv(gaussian)"
      ],
      "metadata": {
        "id": "jznhP8GUnAb7"
      },
      "execution_count": 269,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Decision tree classifier\n",
        "decision_tree_score = rmse_cv(decision_tree)"
      ],
      "metadata": {
        "id": "M7fwy38EnAT5"
      },
      "execution_count": 270,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Random forest classifier\n",
        "random_forest_score = rmse_cv(random_forest)"
      ],
      "metadata": {
        "id": "WVdv1llLnAPs"
      },
      "execution_count": 271,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Gradient boosting classifier\n",
        "gbc_score = rmse_cv(gbc)"
      ],
      "metadata": {
        "id": "oc3xol4ZnAKt"
      },
      "execution_count": 272,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# X boosting gradient classifier\n",
        "xgbc_score = rmse_cv(xgbc)"
      ],
      "metadata": {
        "id": "mtI110fZnAFn"
      },
      "execution_count": 273,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Light gradient boosting machine classifier\n",
        "lgbmc_score = rmse_cv(lgbmc)"
      ],
      "metadata": {
        "id": "Qy54ye31nAAL"
      },
      "execution_count": 274,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Perceptron\n",
        "perceptron_score = rmse_cv(perceptron)"
      ],
      "metadata": {
        "id": "naCxqap4m_3W"
      },
      "execution_count": 275,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "models = pd.DataFrame({\n",
        "    'Model': ['Dummy classifier', 'Support vector classifier', 'K-neighbors classifier', 'Logistic regressor', \n",
        "              'Random forest classifier', 'Gaussian Naive-Bayes', 'Perceptron',  \n",
        "              'Decision tree classifier', 'Gradient boosting classifier', 'XGBClassifier', 'LGBMCClassifier'],\n",
        "\n",
        "    'Accuracy': [dummy_score.mean(), svc_score.mean(), knn_score.mean(), logreg_score.mean(), \n",
        "              random_forest_score.mean(), gaussian_score.mean(), perceptron_score.mean(), \n",
        "              decision_tree_score.mean(), gbc_score.mean(), xgbc_score.mean(), lgbmc_score.mean()],\n",
        "    \n",
        "    'Standard_deviation': [dummy_score.std(), svc_score.std(), knn_score.std(), logreg_score.std(), \n",
        "              random_forest_score.std(), gaussian_score.std(), perceptron_score.std(), \n",
        "              decision_tree_score.std(), gbc_score.std(), xgbc_score.std(), lgbmc_score.std()]})\n",
        "\n",
        "models.sort_values(by='Accuracy', ascending=False)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        },
        "id": "8NBPdO6dnzIB",
        "outputId": "00c4733b-14c6-40a4-e4d1-b6d55553f819"
      },
      "execution_count": 276,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                           Model  Accuracy  Standard_deviation\n",
              "8   Gradient boosting classifier  0.880498            0.053142\n",
              "7       Decision tree classifier  0.870118            0.028097\n",
              "4       Random forest classifier  0.865857            0.049547\n",
              "0               Dummy classifier  0.853297            0.037423\n",
              "1      Support vector classifier  0.849847            0.035668\n",
              "9                  XGBClassifier  0.841327            0.046854\n",
              "10               LGBMCClassifier  0.821934            0.038596\n",
              "3             Logistic regressor  0.815888            0.045711\n",
              "5           Gaussian Naive-Bayes  0.793908            0.026536\n",
              "2         K-neighbors classifier  0.789943            0.042254\n",
              "6                     Perceptron  0.677539            0.340592"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-e2be8817-165d-4e0d-86ef-b53fb2ecb788\">\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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Standard_deviation</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Gradient boosting classifier</td>\n",
              "      <td>0.880498</td>\n",
              "      <td>0.053142</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Decision tree classifier</td>\n",
              "      <td>0.870118</td>\n",
              "      <td>0.028097</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Random forest classifier</td>\n",
              "      <td>0.865857</td>\n",
              "      <td>0.049547</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Dummy classifier</td>\n",
              "      <td>0.853297</td>\n",
              "      <td>0.037423</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Support vector classifier</td>\n",
              "      <td>0.849847</td>\n",
              "      <td>0.035668</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>XGBClassifier</td>\n",
              "      <td>0.841327</td>\n",
              "      <td>0.046854</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>LGBMCClassifier</td>\n",
              "      <td>0.821934</td>\n",
              "      <td>0.038596</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Logistic regressor</td>\n",
              "      <td>0.815888</td>\n",
              "      <td>0.045711</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Gaussian Naive-Bayes</td>\n",
              "      <td>0.793908</td>\n",
              "      <td>0.026536</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>K-neighbors classifier</td>\n",
              "      <td>0.789943</td>\n",
              "      <td>0.042254</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Perceptron</td>\n",
              "      <td>0.677539</td>\n",
              "      <td>0.340592</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e2be8817-165d-4e0d-86ef-b53fb2ecb788')\"\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-e2be8817-165d-4e0d-86ef-b53fb2ecb788 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-e2be8817-165d-4e0d-86ef-b53fb2ecb788');\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": 276
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3.3 Cross-validated models with hiperparameters tuning"
      ],
      "metadata": {
        "id": "YlI2mjhUzUIJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Mean square error with grid search parameters and cross-validation\n",
        "def rmse_optimised_cv(model, parameters_space):\n",
        "    cv = KFold(n_splits=5, shuffle=True, random_state=SEED)\n",
        "    search = GridSearchCV(model, parameters_space, cv = cv)\n",
        "    search.fit(x_test, y_test)\n",
        "    rmse = (cross_val_score(search, x_test,\n",
        "                                   y_test, scoring='f1', cv = cv))\n",
        "    best_hiper_parameters = search.best_estimator_\n",
        "    print(\"Best hiperparameters:\\n\", best_hiper_parameters)\n",
        "    return(rmse)"
      ],
      "metadata": {
        "id": "Y3wdxz6wzfaF"
      },
      "execution_count": 277,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "In order not to make executions too slow, in most cases on the following cells, only the best found hyperparameters are shown."
      ],
      "metadata": {
        "id": "sI8JJIwgSrLW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Dummy classifier\n",
        "\n",
        "# Hiperparameters grid\n",
        "parameters_space = {\n",
        "    'strategy' : ['most_frequent', 'prior', 'stratified', 'uniform']}\n",
        "\n",
        "dummy_score = rmse_optimised_cv(dummy, parameters_space)"
      ],
      "metadata": {
        "id": "kH0C_L6Z7CcK",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "94bd5d5a-6bc7-4904-c4f3-d8e5dbeb9dfe"
      },
      "execution_count": 278,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " DummyClassifier(random_state=452, strategy='most_frequent')\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# K-neighbors classifier\n",
        "\n",
        "parameters_space = {\n",
        "    'n_neighbors' : [13],\n",
        "    'weights': ['uniform', 'distance'],\n",
        "    'algorithm': ['kd_tree']}\n",
        "\n",
        "knn_score = rmse_optimised_cv(knn, parameters_space)"
      ],
      "metadata": {
        "id": "KyDTxXNa_5r7",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0aba2545-a52c-487d-9ece-f645f071a279"
      },
      "execution_count": 279,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " KNeighborsClassifier(algorithm='kd_tree', n_neighbors=13)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Gradient boosting classifier\n",
        "\n",
        "parameters_space = {\n",
        "    'learning_rate' : [1],\n",
        "    'n_estimators': [1],\n",
        "    'subsample': [0.5, 1],\n",
        "    'min_samples_leaf': [1, 5, 10]\n",
        "    }\n",
        "\n",
        "gbc_score = rmse_optimised_cv(gbc, parameters_space)"
      ],
      "metadata": {
        "id": "_hD9lMAOGW0D",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "742e1bb9-2abe-4c8b-a7f0-c253311e4b28"
      },
      "execution_count": 280,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " GradientBoostingClassifier(learning_rate=1, n_estimators=1, random_state=452,\n",
            "                           subsample=1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Support vector classifier\n",
        "\n",
        "parameters_space = {\n",
        "    'C' : [1.0],\n",
        "    'gamma': ['scale', 'auto'],\n",
        "    'decision_function_shape': ['ovo', 'ovr'],\n",
        "    'random_state': [SEED]\n",
        "    }\n",
        "\n",
        "\n",
        "svc_score = rmse_optimised_cv(svc, parameters_space)"
      ],
      "metadata": {
        "id": "uI-evSeXG5N3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "45a475d2-23a1-44b3-dbd8-c41f69e25e4c"
      },
      "execution_count": 281,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " SVC(decision_function_shape='ovo', gamma='auto', random_state=452)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Logistic regressor\n",
        "\n",
        "parameters_space = {\n",
        "    'solver' : ['lbfgs', 'liblinear', 'newton-cg', 'newton-cholesky', 'sag', 'saga'],\n",
        "    'max_iter' : [100, 200, 300]\n",
        "    }\n",
        "\n",
        "logreg_score = rmse_optimised_cv(logreg, parameters_space)"
      ],
      "metadata": {
        "id": "msFnAdPxG5N4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "540229c9-4254-4c76-b6e2-5f2730d8ff3e"
      },
      "execution_count": 282,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " LogisticRegression(random_state=452, solver='sag')\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Gaussian Naive-Bayes\n",
        "\n",
        "parameters_space = {\n",
        "    \"var_smoothing\" : [5e-7, 1e-7, 5e-6]}\n",
        "\n",
        "gaussian_score = rmse_optimised_cv(gaussian, parameters_space)"
      ],
      "metadata": {
        "id": "dwYQxyUpG5N4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8b3df707-02bd-4a70-f4d9-6ab36ebbf9ac"
      },
      "execution_count": 283,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " GaussianNB(var_smoothing=1e-07)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Decision tree classifier\n",
        "\n",
        "parameters_space = {\n",
        "    'max_depth' : [1,2,3],\n",
        "    'min_samples_split' : [1,2,3],\n",
        "    'min_samples_leaf' : [1,2,3,4],\n",
        "    'criterion' : ['gini', 'entropy']}\n",
        "\n",
        "decision_tree_score = rmse_optimised_cv(decision_tree, parameters_space)"
      ],
      "metadata": {
        "id": "Y0QU4y4CG5N5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "aefb8a2a-f326-4242-dbf3-a1ffb7893ce6"
      },
      "execution_count": 292,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " DecisionTreeClassifier(max_depth=2, min_samples_leaf=3, random_state=452)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Random forest classifier\n",
        "\n",
        "parameters_space = {\n",
        "    'n_estimators': [10],\n",
        "    'max_depth' : [2],\n",
        "    'min_samples_split' : [2],\n",
        "    'min_samples_leaf' : [1,2,3],\n",
        "    'criterion' : ['gini', 'entropy']}\n",
        "\n",
        "random_forest_score = rmse_optimised_cv(random_forest, parameters_space)"
      ],
      "metadata": {
        "id": "TVlDipKJG5N5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d1f2fa93-22f5-43dd-e192-b33ff818c5b4"
      },
      "execution_count": 285,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " RandomForestClassifier(max_depth=2, n_estimators=10, random_state=452)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# X boosting gradient classifier\n",
        "\n",
        "parameters_space = {\n",
        "    'max_depth': [2],\n",
        "    'learning_rate' : [1],\n",
        "    'n_estimators' : [32]}\n",
        "\n",
        "xgbc_score = rmse_optimised_cv(xgbc, parameters_space)"
      ],
      "metadata": {
        "id": "JlnM4Z-MG5N5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7ffc9083-c11c-4f1d-b72b-d930f1215f63"
      },
      "execution_count": 286,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
            "              colsample_bylevel=None, colsample_bynode=None,\n",
            "              colsample_bytree=None, early_stopping_rounds=None,\n",
            "              enable_categorical=False, eval_metric=None, feature_types=None,\n",
            "              gamma=None, gpu_id=None, grow_policy=None, importance_type=None,\n",
            "              interaction_constraints=None, learning_rate=1, max_bin=None,\n",
            "              max_cat_threshold=None, max_cat_to_onehot=None,\n",
            "              max_delta_step=None, max_depth=2, max_leaves=None,\n",
            "              min_child_weight=None, missing=nan, monotone_constraints=None,\n",
            "              n_estimators=32, n_jobs=None, num_parallel_tree=None,\n",
            "              predictor=None, random_state=452, ...)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Light gradient boosting machine classifier\n",
        "\n",
        "parameters_space = {\n",
        "    'learning_rate' : [1],\n",
        "    'num_leaves' : [10],\n",
        "    'n_estimators ' : [5],\n",
        "    'subsample': [1],\n",
        "    'min_samples_leaf': [1],\n",
        "    'num_iterations': [1]\n",
        "    }\n",
        "\n",
        "lgbmc_score = rmse_optimised_cv(lgbmc, parameters_space)"
      ],
      "metadata": {
        "id": "ssIw1oLyG5N6",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2068caac-bced-4d84-eb94-61f042d04bc1"
      },
      "execution_count": 287,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " LGBMClassifier(learning_rate=1, min_samples_leaf=1, n_estimators =5,\n",
            "               num_iterations=1, num_leaves=10, random_state=452, subsample=1)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Perceptron\n",
        "\n",
        "parameters_space = {\n",
        "    'alpha' : [0.0001],\n",
        "    'max_iter': [1],\n",
        "    'eta0': [0.0001],\n",
        "    'n_iter_no_change': [1],\n",
        "    'random_state': [SEED]\n",
        "    }\n",
        "\n",
        "perceptron_score = rmse_optimised_cv(perceptron, parameters_space)"
      ],
      "metadata": {
        "id": "XV8PDqD-G5N6",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "56d3267f-8871-43da-a920-21ef5e8d0ecd"
      },
      "execution_count": 288,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best hiperparameters:\n",
            " Perceptron(eta0=0.0001, max_iter=1, n_iter_no_change=1, random_state=452)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "models = pd.DataFrame({\n",
        "    'Model': ['Dummy classifier', 'Support vector classifier', 'K-neighbors classifier', 'Logistic regressor', \n",
        "              'Random forest classifier', 'Gaussian Naive-Bayes', 'Perceptron',  \n",
        "              'Decision tree classifier', 'Gradient boosting classifier', 'XGBClassifier', 'LGBMClassifier'],\n",
        "\n",
        "    'Accuracy': [dummy_score.mean(), svc_score.mean(), knn_score.mean(), logreg_score.mean(), \n",
        "              random_forest_score.mean(), gaussian_score.mean(), perceptron_score.mean(), \n",
        "              decision_tree_score.mean(), gbc_score.mean(), xgbc_score.mean(), lgbmc_score.mean()],\n",
        "    \n",
        "    'Standard_deviation': [dummy_score.std(), svc_score.std(), knn_score.std(), logreg_score.std(), \n",
        "              random_forest_score.std(), gaussian_score.std(), perceptron_score.std(), \n",
        "              decision_tree_score.std(), gbc_score.std(), xgbc_score.std(), lgbmc_score.std()]})\n",
        "\n",
        "models.sort_values(by='Accuracy', ascending=False)"
      ],
      "metadata": {
        "id": "1-7yVH_2GWiO",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 394
        },
        "outputId": "6e195574-4bd1-4e6c-dffa-3d78ca4de1cf"
      },
      "execution_count": 289,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                           Model  Accuracy  Standard_deviation\n",
              "8   Gradient boosting classifier  0.908527            0.032712\n",
              "7       Decision tree classifier  0.908271            0.025814\n",
              "4       Random forest classifier  0.876466            0.048833\n",
              "9                  XGBClassifier  0.873582            0.050462\n",
              "10                LGBMClassifier  0.867413            0.046710\n",
              "5           Gaussian Naive-Bayes  0.853766            0.047531\n",
              "0               Dummy classifier  0.853297            0.037423\n",
              "3             Logistic regressor  0.853297            0.037423\n",
              "6                     Perceptron  0.853297            0.037423\n",
              "2         K-neighbors classifier  0.851852            0.040009\n",
              "1      Support vector classifier  0.849847            0.035668"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1a9fd499-0006-4471-b551-b0b7b3f9640b\">\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>Model</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Standard_deviation</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>Gradient boosting classifier</td>\n",
              "      <td>0.908527</td>\n",
              "      <td>0.032712</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Decision tree classifier</td>\n",
              "      <td>0.908271</td>\n",
              "      <td>0.025814</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Random forest classifier</td>\n",
              "      <td>0.876466</td>\n",
              "      <td>0.048833</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>XGBClassifier</td>\n",
              "      <td>0.873582</td>\n",
              "      <td>0.050462</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>LGBMClassifier</td>\n",
              "      <td>0.867413</td>\n",
              "      <td>0.046710</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Gaussian Naive-Bayes</td>\n",
              "      <td>0.853766</td>\n",
              "      <td>0.047531</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Dummy classifier</td>\n",
              "      <td>0.853297</td>\n",
              "      <td>0.037423</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Logistic regressor</td>\n",
              "      <td>0.853297</td>\n",
              "      <td>0.037423</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Perceptron</td>\n",
              "      <td>0.853297</td>\n",
              "      <td>0.037423</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>K-neighbors classifier</td>\n",
              "      <td>0.851852</td>\n",
              "      <td>0.040009</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Support vector classifier</td>\n",
              "      <td>0.849847</td>\n",
              "      <td>0.035668</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1a9fd499-0006-4471-b551-b0b7b3f9640b')\"\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-1a9fd499-0006-4471-b551-b0b7b3f9640b 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-1a9fd499-0006-4471-b551-b0b7b3f9640b');\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": 289
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "As it could be noticed, the models with best accuracy are decision tree classifier and gradient boosting classifier, with a difference of accuracy's RMSE lower than 0.1%. Therefore, the used criterion to choose algorithm is standard deviation. Due to the fact of Decision tree classifier has a stardard deviation not only better than gradient boosting classifier, but the whole set. \n",
        "\n",
        "This study came to the conclusion that the best model for being used to predict the customers's loan status is the decision tree classifier. Considering, of course, the best tuning hyperparameters found in the section number 3.3 (training and test section)."
      ],
      "metadata": {
        "id": "0GQws7u4IZSS"
      }
    }
  ]
}