judgement_pendency_prediction.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DWrpAu9iHU1Z"
      },
      "source": [
        "<h1> Judgement Pendency Prediction <h1>\n",
        "\n",
        "\n",
        "The data used in this file is an abridged version of the dataset [available here](https://www.devdatalab.org/judicial-data) based on the work of [Bhowmik et. al. (2021)](https://dl.acm.org/doi/abs/10.1145/3460112.3471943). The cleaned up and abrdiged dataset with license terms used in this code can be downloaded at https://drive.google.com/file/d/1aQcfZTi2GFdyod8bFYi0VEGqX6mUUhJ4/view?usp=drive_link \n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ndIpGKhuSQbX"
      },
      "source": [
        "## Import of libraries & dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ax44AOpvAwfu"
      },
      "source": [
        "### Import Libraries"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "styAAPBV2Fzj"
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from datetime import date\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "T6W3V2AG7fs8"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.metrics import classification_report\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.ensemble import BaggingClassifier\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "from sklearn.metrics import roc_curve, auc, roc_auc_score\n",
        "from sklearn.metrics import precision_recall_curve\n",
        "from sklearn.metrics import log_loss\n",
        "from sklearn import tree"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RVUhZZ0uqheR"
      },
      "source": [
        "### Import of Cases 2010"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VG5Ouj_OOG2R",
        "outputId": "07d42b37-a759-4389-e9e3-0988ee0d56fa"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 357
        },
        "id": "CB_Ofwb18gl1",
        "outputId": "4b6c4bf1-778f-490f-c4ff-f98730714cdb"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                ddl_case_id  year  state_code  dist_code  court_no  \\\n",
              "0  01-01-01-201908000152010  2010           1          1         1   \n",
              "1  01-01-01-201908000242010  2010           1          1         1   \n",
              "2  01-01-01-201908000702010  2010           1          1         1   \n",
              "3  01-01-01-201908001322010  2010           1          1         1   \n",
              "4  01-01-01-201908001572010  2010           1          1         1   \n",
              "\n",
              "               cino             judge_position  female_defendant  \\\n",
              "0  MHNB030003482010  chief judicial magistrate                 0   \n",
              "1  MHNB030001272010  chief judicial magistrate                 0   \n",
              "2  MHNB030014152010  chief judicial magistrate                 0   \n",
              "3  MHNB030009002010  chief judicial magistrate             -9998   \n",
              "4  MHNB030010732010  chief judicial magistrate                 0   \n",
              "\n",
              "  female_petitioner  female_adv_def  female_adv_pet  type_name  purpose_name  \\\n",
              "0            0 male           -9999               0       1429        4946.0   \n",
              "1          1 female           -9999           -9998       1429        3627.0   \n",
              "2            0 male               0               1       1429         509.0   \n",
              "3          1 female               0               0       1429        2237.0   \n",
              "4          1 female           -9999               0       1429        3280.0   \n",
              "\n",
              "   disp_name             date_of_filing           date_of_decision  \\\n",
              "0         25  2010-04-20 00:00:00+00:00  2011-03-31 00:00:00+00:00   \n",
              "1         22  2010-02-11 00:00:00+00:00  2013-07-01 00:00:00+00:00   \n",
              "2         25  2010-05-28 00:00:00+00:00  2015-07-27 00:00:00+00:00   \n",
              "3         25  2010-08-31 00:00:00+00:00  2010-09-09 00:00:00+00:00   \n",
              "4         30  2010-11-12 00:00:00+00:00  2013-05-13 00:00:00+00:00   \n",
              "\n",
              "             date_first_list             date_last_list  \\\n",
              "0  2010-06-23 00:00:00+00:00  2011-02-28 00:00:00+00:00   \n",
              "1  2010-03-11 00:00:00+00:00  2013-06-12 00:00:00+00:00   \n",
              "2  2012-12-24 00:00:00+00:00  2015-06-19 00:00:00+00:00   \n",
              "3  2010-09-03 00:00:00+00:00  2010-09-06 00:00:00+00:00   \n",
              "4  2010-12-13 00:00:00+00:00  2013-03-26 00:00:00+00:00   \n",
              "\n",
              "              date_next_list      act    section bailable_ipc  \\\n",
              "0  2011-03-31 00:00:00+00:00  13504.0  1244334.0          NaN   \n",
              "1  2013-07-01 00:00:00+00:00   4759.0   132871.0          NaN   \n",
              "2  2015-07-27 00:00:00+00:00   9533.0  1166313.0          NaN   \n",
              "3  2010-09-09 00:00:00+00:00   4759.0  1123673.0          NaN   \n",
              "4  2013-05-13 00:00:00+00:00   4759.0   132871.0          NaN   \n",
              "\n",
              "   number_sections_ipc  criminal  ddl_filing_judge_id  ddl_decision_judge_id  \\\n",
              "0                  NaN       0.0                  NaN                    NaN   \n",
              "1                  NaN       1.0                  NaN                   51.0   \n",
              "2                  NaN       0.0                  NaN                    NaN   \n",
              "3                  NaN       1.0                  NaN                    NaN   \n",
              "4                  NaN       1.0                  NaN                   92.0   \n",
              "\n",
              "  female_judge_filing start_date_filing end_date_filing female_judge_decision  \\\n",
              "0                 NaN               NaN             NaN                   NaN   \n",
              "1                 NaN               NaN             NaN           0 nonfemale   \n",
              "2                 NaN               NaN             NaN                   NaN   \n",
              "3                 NaN               NaN             NaN                   NaN   \n",
              "4                 NaN               NaN             NaN           0 nonfemale   \n",
              "\n",
              "  start_date_decision end_date_decision  judgediff  pendency_decision_filing  \\\n",
              "0                 NaN               NaN        NaN                        11   \n",
              "1          2013-06-10        2013-09-23        NaN                        41   \n",
              "2                 NaN               NaN        NaN                        62   \n",
              "3                 NaN               NaN        NaN                         1   \n",
              "4          2011-06-06        2014-02-21        NaN                        30   \n",
              "\n",
              "   pendency_first_filing  pendency_decision_first  pendency_next_last  \\\n",
              "0                    2.0                      9.0                 1.0   \n",
              "1                    1.0                     40.0                 1.0   \n",
              "2                   31.0                     31.0                 1.0   \n",
              "3                    1.0                      0.0                 0.0   \n",
              "4                    1.0                     29.0                 2.0   \n",
              "\n",
              "   pendency_decision_last  pendency_decision_next  \\\n",
              "0                     1.0                     0.0   \n",
              "1                     1.0                     0.0   \n",
              "2                     1.0                     0.0   \n",
              "3                     0.0                     0.0   \n",
              "4                     2.0                     0.0   \n",
              "\n",
              "  pendency_decision_filing_categories pendency_first_filing_categories  \\\n",
              "0                         upto 1 year                      upto 1 year   \n",
              "1                        3 to 4 years                      upto 1 year   \n",
              "2                        5 to 6 years                     2 to 3 years   \n",
              "3                         upto 1 year                      upto 1 year   \n",
              "4                        2 to 3 years                      upto 1 year   \n",
              "\n",
              "   filing_judge_tenure  decision_judge_tenure  \n",
              "0                  NaN                    NaN  \n",
              "1                  NaN                    3.0  \n",
              "2                  NaN                    NaN  \n",
              "3                  NaN                    NaN  \n",
              "4                  NaN                   32.0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-efb11d71-e472-4e34-a5f3-6c3bd9ffff04\">\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>ddl_case_id</th>\n",
              "      <th>year</th>\n",
              "      <th>state_code</th>\n",
              "      <th>dist_code</th>\n",
              "      <th>court_no</th>\n",
              "      <th>cino</th>\n",
              "      <th>judge_position</th>\n",
              "      <th>female_defendant</th>\n",
              "      <th>female_petitioner</th>\n",
              "      <th>female_adv_def</th>\n",
              "      <th>female_adv_pet</th>\n",
              "      <th>type_name</th>\n",
              "      <th>purpose_name</th>\n",
              "      <th>disp_name</th>\n",
              "      <th>date_of_filing</th>\n",
              "      <th>date_of_decision</th>\n",
              "      <th>date_first_list</th>\n",
              "      <th>date_last_list</th>\n",
              "      <th>date_next_list</th>\n",
              "      <th>act</th>\n",
              "      <th>section</th>\n",
              "      <th>bailable_ipc</th>\n",
              "      <th>number_sections_ipc</th>\n",
              "      <th>criminal</th>\n",
              "      <th>ddl_filing_judge_id</th>\n",
              "      <th>ddl_decision_judge_id</th>\n",
              "      <th>female_judge_filing</th>\n",
              "      <th>start_date_filing</th>\n",
              "      <th>end_date_filing</th>\n",
              "      <th>female_judge_decision</th>\n",
              "      <th>start_date_decision</th>\n",
              "      <th>end_date_decision</th>\n",
              "      <th>judgediff</th>\n",
              "      <th>pendency_decision_filing</th>\n",
              "      <th>pendency_first_filing</th>\n",
              "      <th>pendency_decision_first</th>\n",
              "      <th>pendency_next_last</th>\n",
              "      <th>pendency_decision_last</th>\n",
              "      <th>pendency_decision_next</th>\n",
              "      <th>pendency_decision_filing_categories</th>\n",
              "      <th>pendency_first_filing_categories</th>\n",
              "      <th>filing_judge_tenure</th>\n",
              "      <th>decision_judge_tenure</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>01-01-01-201908000152010</td>\n",
              "      <td>2010</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MHNB030003482010</td>\n",
              "      <td>chief judicial magistrate</td>\n",
              "      <td>0</td>\n",
              "      <td>0 male</td>\n",
              "      <td>-9999</td>\n",
              "      <td>0</td>\n",
              "      <td>1429</td>\n",
              "      <td>4946.0</td>\n",
              "      <td>25</td>\n",
              "      <td>2010-04-20 00:00:00+00:00</td>\n",
              "      <td>2011-03-31 00:00:00+00:00</td>\n",
              "      <td>2010-06-23 00:00:00+00:00</td>\n",
              "      <td>2011-02-28 00:00:00+00:00</td>\n",
              "      <td>2011-03-31 00:00:00+00:00</td>\n",
              "      <td>13504.0</td>\n",
              "      <td>1244334.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>11</td>\n",
              "      <td>2.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>upto 1 year</td>\n",
              "      <td>upto 1 year</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>01-01-01-201908000242010</td>\n",
              "      <td>2010</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MHNB030001272010</td>\n",
              "      <td>chief judicial magistrate</td>\n",
              "      <td>0</td>\n",
              "      <td>1 female</td>\n",
              "      <td>-9999</td>\n",
              "      <td>-9998</td>\n",
              "      <td>1429</td>\n",
              "      <td>3627.0</td>\n",
              "      <td>22</td>\n",
              "      <td>2010-02-11 00:00:00+00:00</td>\n",
              "      <td>2013-07-01 00:00:00+00:00</td>\n",
              "      <td>2010-03-11 00:00:00+00:00</td>\n",
              "      <td>2013-06-12 00:00:00+00:00</td>\n",
              "      <td>2013-07-01 00:00:00+00:00</td>\n",
              "      <td>4759.0</td>\n",
              "      <td>132871.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>51.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0 nonfemale</td>\n",
              "      <td>2013-06-10</td>\n",
              "      <td>2013-09-23</td>\n",
              "      <td>NaN</td>\n",
              "      <td>41</td>\n",
              "      <td>1.0</td>\n",
              "      <td>40.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3 to 4 years</td>\n",
              "      <td>upto 1 year</td>\n",
              "      <td>NaN</td>\n",
              "      <td>3.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>01-01-01-201908000702010</td>\n",
              "      <td>2010</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MHNB030014152010</td>\n",
              "      <td>chief judicial magistrate</td>\n",
              "      <td>0</td>\n",
              "      <td>0 male</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1429</td>\n",
              "      <td>509.0</td>\n",
              "      <td>25</td>\n",
              "      <td>2010-05-28 00:00:00+00:00</td>\n",
              "      <td>2015-07-27 00:00:00+00:00</td>\n",
              "      <td>2012-12-24 00:00:00+00:00</td>\n",
              "      <td>2015-06-19 00:00:00+00:00</td>\n",
              "      <td>2015-07-27 00:00:00+00:00</td>\n",
              "      <td>9533.0</td>\n",
              "      <td>1166313.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>62</td>\n",
              "      <td>31.0</td>\n",
              "      <td>31.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>5 to 6 years</td>\n",
              "      <td>2 to 3 years</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>01-01-01-201908001322010</td>\n",
              "      <td>2010</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MHNB030009002010</td>\n",
              "      <td>chief judicial magistrate</td>\n",
              "      <td>-9998</td>\n",
              "      <td>1 female</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1429</td>\n",
              "      <td>2237.0</td>\n",
              "      <td>25</td>\n",
              "      <td>2010-08-31 00:00:00+00:00</td>\n",
              "      <td>2010-09-09 00:00:00+00:00</td>\n",
              "      <td>2010-09-03 00:00:00+00:00</td>\n",
              "      <td>2010-09-06 00:00:00+00:00</td>\n",
              "      <td>2010-09-09 00:00:00+00:00</td>\n",
              "      <td>4759.0</td>\n",
              "      <td>1123673.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>upto 1 year</td>\n",
              "      <td>upto 1 year</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>01-01-01-201908001572010</td>\n",
              "      <td>2010</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>MHNB030010732010</td>\n",
              "      <td>chief judicial magistrate</td>\n",
              "      <td>0</td>\n",
              "      <td>1 female</td>\n",
              "      <td>-9999</td>\n",
              "      <td>0</td>\n",
              "      <td>1429</td>\n",
              "      <td>3280.0</td>\n",
              "      <td>30</td>\n",
              "      <td>2010-11-12 00:00:00+00:00</td>\n",
              "      <td>2013-05-13 00:00:00+00:00</td>\n",
              "      <td>2010-12-13 00:00:00+00:00</td>\n",
              "      <td>2013-03-26 00:00:00+00:00</td>\n",
              "      <td>2013-05-13 00:00:00+00:00</td>\n",
              "      <td>4759.0</td>\n",
              "      <td>132871.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>92.0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0 nonfemale</td>\n",
              "      <td>2011-06-06</td>\n",
              "      <td>2014-02-21</td>\n",
              "      <td>NaN</td>\n",
              "      <td>30</td>\n",
              "      <td>1.0</td>\n",
              "      <td>29.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2 to 3 years</td>\n",
              "      <td>upto 1 year</td>\n",
              "      <td>NaN</td>\n",
              "      <td>32.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-efb11d71-e472-4e34-a5f3-6c3bd9ffff04')\"\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-efb11d71-e472-4e34-a5f3-6c3bd9ffff04 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-efb11d71-e472-4e34-a5f3-6c3bd9ffff04');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ],
      "source": [
        "20# Download the dataset from the zip file and keep the file case_2010_final.csv it on your google drive.\n",
        "# You would have to change path below appropriately to load and run the below line of code.\n",
        "df_cases2010_orig = pd.read_csv(\"/content/drive/MyDrive/Law/cases/2010/cases_2010_final.csv\")\n",
        "#print all columns\n",
        "pd.set_option('display.max_columns', None)\n",
        "df_cases2010_orig.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7JD_Ubl6T5St",
        "outputId": "f2832081-c707-44e9-8709-d5bcc7d0c9dc"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "ddl_case_id                             object\n",
              "year                                     int64\n",
              "state_code                               int64\n",
              "dist_code                                int64\n",
              "court_no                                 int64\n",
              "cino                                    object\n",
              "judge_position                          object\n",
              "female_defendant                         int64\n",
              "female_petitioner                       object\n",
              "female_adv_def                           int64\n",
              "female_adv_pet                           int64\n",
              "type_name                                int64\n",
              "purpose_name                           float64\n",
              "disp_name                                int64\n",
              "date_of_filing                          object\n",
              "date_of_decision                        object\n",
              "date_first_list                         object\n",
              "date_last_list                          object\n",
              "date_next_list                          object\n",
              "act                                    float64\n",
              "section                                float64\n",
              "bailable_ipc                            object\n",
              "number_sections_ipc                    float64\n",
              "criminal                               float64\n",
              "ddl_filing_judge_id                    float64\n",
              "ddl_decision_judge_id                  float64\n",
              "female_judge_filing                     object\n",
              "start_date_filing                       object\n",
              "end_date_filing                         object\n",
              "female_judge_decision                   object\n",
              "start_date_decision                     object\n",
              "end_date_decision                       object\n",
              "judgediff                              float64\n",
              "pendency_decision_filing                 int64\n",
              "pendency_first_filing                  float64\n",
              "pendency_decision_first                float64\n",
              "pendency_next_last                     float64\n",
              "pendency_decision_last                 float64\n",
              "pendency_decision_next                 float64\n",
              "pendency_decision_filing_categories     object\n",
              "pendency_first_filing_categories        object\n",
              "filing_judge_tenure                    float64\n",
              "decision_judge_tenure                  float64\n",
              "dtype: object"
            ]
          },
          "execution_count": 5,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_cases2010_orig.dtypes"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jWhVgsTDOwN2"
      },
      "source": [
        "## Preprocessing & Feature Engineering"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "U0XpwncE8f8c",
        "outputId": "63a2cade-03ec-4307-ba02-81d96da33f08"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Column 'ddl_case_id' has 0 missing values which is 0% of total\n",
            "Column 'year' has 0 missing values which is 0% of total\n",
            "Column 'state_code' has 0 missing values which is 0% of total\n",
            "Column 'dist_code' has 0 missing values which is 0% of total\n",
            "Column 'court_no' has 0 missing values which is 0% of total\n",
            "Column 'cino' has 0 missing values which is 0% of total\n",
            "Column 'judge_position' has 0 missing values which is 0% of total\n",
            "Column 'female_defendant' has 0 missing values which is 0% of total\n",
            "Column 'female_petitioner' has 0 missing values which is 0% of total\n",
            "Column 'female_adv_def' has 0 missing values which is 0% of total\n",
            "Column 'female_adv_pet' has 0 missing values which is 0% of total\n",
            "Column 'type_name' has 0 missing values which is 0% of total\n",
            "Column 'purpose_name' has 233545 missing values which is 6% of total\n",
            "Column 'disp_name' has 0 missing values which is 0% of total\n",
            "Column 'date_of_filing' has 0 missing values which is 0% of total\n",
            "Column 'date_of_decision' has 570401 missing values which is 14% of total\n",
            "Column 'date_first_list' has 279548 missing values which is 7% of total\n",
            "Column 'date_last_list' has 158263 missing values which is 4% of total\n",
            "Column 'date_next_list' has 157923 missing values which is 4% of total\n",
            "Column 'act' has 2054999 missing values which is 49% of total\n",
            "Column 'section' has 2298209 missing values which is 55% of total\n",
            "Column 'bailable_ipc' has 3739286 missing values which is 89% of total\n",
            "Column 'number_sections_ipc' has 3515887 missing values which is 84% of total\n",
            "Column 'criminal' has 2054885 missing values which is 49% of total\n",
            "Column 'ddl_filing_judge_id' has 4101239 missing values which is 98% of total\n",
            "Column 'ddl_decision_judge_id' has 3916281 missing values which is 93% of total\n",
            "Column 'female_judge_filing' has 4101239 missing values which is 98% of total\n",
            "Column 'start_date_filing' has 4101239 missing values which is 98% of total\n",
            "Column 'end_date_filing' has 4101275 missing values which is 98% of total\n",
            "Column 'female_judge_decision' has 3916281 missing values which is 93% of total\n",
            "Column 'start_date_decision' has 3916281 missing values which is 93% of total\n",
            "Column 'end_date_decision' has 3994646 missing values which is 95% of total\n",
            "Column 'judgediff' has 4107182 missing values which is 98% of total\n",
            "Column 'pendency_decision_filing' has 0 missing values which is 0% of total\n",
            "Column 'pendency_first_filing' has 279548 missing values which is 7% of total\n",
            "Column 'pendency_decision_first' has 837324 missing values which is 20% of total\n",
            "Column 'pendency_next_last' has 163977 missing values which is 4% of total\n",
            "Column 'pendency_decision_last' has 719910 missing values which is 17% of total\n",
            "Column 'pendency_decision_next' has 719205 missing values which is 17% of total\n",
            "Column 'pendency_decision_filing_categories' has 0 missing values which is 0% of total\n",
            "Column 'pendency_first_filing_categories' has 279548 missing values which is 7% of total\n",
            "Column 'filing_judge_tenure' has 4101275 missing values which is 98% of total\n",
            "Column 'decision_judge_tenure' has 3994646 missing values which is 95% of total\n"
          ]
        }
      ],
      "source": [
        "missing_values = df_cases2010_orig.isna().sum()\n",
        "total_rows = df_cases2010_orig.shape[0]\n",
        "# print column names and number of missing values\n",
        "for col, num_missing in missing_values.items():\n",
        "    percent_missing = round(num_missing/total_rows*100)\n",
        "    print(f\"Column '{col}' has {num_missing} missing values which is {percent_missing}% of total\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9hGp--fAjETh"
      },
      "source": [
        "We will drop following features as for year 2010 the majority of data about the filling and decision judges is missing\n",
        "\n",
        "* ddl_filing_judge_id\n",
        "* ddl_decision_judge_id\n",
        "* start_date_filing\n",
        "* end_date_filing\n",
        "* start_date_decision\n",
        "* end_date_decision\n",
        "* judgediff\n",
        "* filing_judge_tenure             \n",
        "* decision_judge_tenure    \n",
        "* purpose_name\n",
        "\n",
        "\n",
        "The following are dropped as these dates are converted into categorical variables as periods\n",
        "\n",
        "* date_of_filing                       \n",
        "* date_of_decision                    \n",
        "* date_first_list                    \n",
        "* date_last_list                        \n",
        "* date_next_list\n",
        "\n",
        "The following are dropped as they are just identifiers\n",
        "\n",
        "* ddl_case_id                            \n",
        "* year\n",
        "* cino\n",
        "\n",
        "These variables are not included as features for our modelling as they will not be available when the model will be used for prediction or deemed not important*\n",
        "\n",
        "* disp_name\n",
        "* pendency_first_filing\n",
        "* pendency_decision_first\n",
        "* pendency_decision_last\n",
        "* pendency_decision_filing\n",
        "* pendency_decision_next\n",
        "* pendency_next_last"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "fOJumjHajwIk"
      },
      "outputs": [],
      "source": [
        "df_cases2010_orig[\"court_details\"] = df_cases2010_orig[\"state_code\"].astype(str) + \"_\" + df_cases2010_orig[\"dist_code\"].astype(str) + \"_\" + df_cases2010_orig[\"court_no\"].astype(str)\n",
        "\n",
        "df_cases2010_orig[\"state_district\"] = df_cases2010_orig[\"state_code\"].astype(str) + \"_\" + df_cases2010_orig[\"dist_code\"].astype(str)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "IcAkm9wG_U_O"
      },
      "outputs": [],
      "source": [
        "columns_to_drop = ['ddl_filing_judge_id','ddl_decision_judge_id','start_date_filing','end_date_filing','start_date_decision','end_date_decision',\\\n",
        "                   'judgediff','filing_judge_tenure','decision_judge_tenure','date_of_filing','date_of_decision','date_first_list','date_last_list','date_next_list','disp_name',\\\n",
        "                   'pendency_decision_first','pendency_decision_first','pendency_decision_last', 'pendency_decision_next','pendency_first_filing', \\\n",
        "                   'pendency_next_last','ddl_case_id','year','cino','court_no','dist_code','pendency_first_filing_categories','purpose_name']\n",
        "\n",
        "\n",
        "# Create a copy of the dataframe and start preparing the same for prediction modelling\n",
        "df_cases2010 = df_cases2010_orig.drop(columns=columns_to_drop, axis=1)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cM9gfCrLaddR",
        "outputId": "6d15c23f-0072-4be7-96b5-4501acde2aec"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 4190783 entries, 0 to 4190782\n",
            "Data columns (total 18 columns):\n",
            " #   Column                               Dtype  \n",
            "---  ------                               -----  \n",
            " 0   state_code                           int64  \n",
            " 1   judge_position                       object \n",
            " 2   female_defendant                     int64  \n",
            " 3   female_petitioner                    object \n",
            " 4   female_adv_def                       int64  \n",
            " 5   female_adv_pet                       int64  \n",
            " 6   type_name                            int64  \n",
            " 7   act                                  float64\n",
            " 8   section                              float64\n",
            " 9   bailable_ipc                         object \n",
            " 10  number_sections_ipc                  float64\n",
            " 11  criminal                             float64\n",
            " 12  female_judge_filing                  object \n",
            " 13  female_judge_decision                object \n",
            " 14  pendency_decision_filing             int64  \n",
            " 15  pendency_decision_filing_categories  object \n",
            " 16  court_details                        object \n",
            " 17  state_district                       object \n",
            "dtypes: float64(4), int64(6), object(8)\n",
            "memory usage: 575.5+ MB\n"
          ]
        }
      ],
      "source": [
        "df_cases2010.info()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4_upqghOj4u2"
      },
      "outputs": [],
      "source": [
        "df_cases2010 = df_cases2010.fillna(\"NA\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "dc-HTouChhDX"
      },
      "outputs": [],
      "source": [
        "# Convert all other features to categorical except pendency_first_filing which is numerical\n",
        "# number_sections_ipc is treated as categorical as it has lots of missing values which is better handled as categorical data\n",
        "df_cases2010['state_code'] = df_cases2010['state_code'].astype(str).astype('category')\n",
        "df_cases2010['state_district'] = df_cases2010['state_district'].astype(str).astype('category')\n",
        "df_cases2010['court_details'] = df_cases2010['court_details'].astype(str).astype('category')\n",
        "df_cases2010['judge_position'] = df_cases2010['judge_position'].astype(str).astype('category')\n",
        "df_cases2010['female_defendant'] = df_cases2010['female_defendant'].astype(str).astype('category')\n",
        "df_cases2010['female_petitioner'] = df_cases2010['female_petitioner'].astype(str).astype('category')\n",
        "df_cases2010['female_adv_def'] = df_cases2010['female_adv_def'].astype(str).astype('category')\n",
        "df_cases2010['female_adv_pet'] = df_cases2010['female_adv_pet'].astype(str).astype('category')\n",
        "df_cases2010['type_name'] = df_cases2010['type_name'].astype(str).astype('category')\n",
        "df_cases2010['act'] = df_cases2010['act'].astype(str).astype('category')\n",
        "df_cases2010['section'] = df_cases2010['section'].astype(str).astype('category')\n",
        "df_cases2010['bailable_ipc'] = df_cases2010['bailable_ipc'].astype(str).astype('category')\n",
        "df_cases2010['number_sections_ipc'] = df_cases2010['number_sections_ipc'].astype(str).astype('category')\n",
        "df_cases2010['criminal'] = df_cases2010['criminal'].astype(str).astype('category')\n",
        "df_cases2010['female_judge_filing'] = df_cases2010['female_judge_filing'].astype(str).astype('category')\n",
        "df_cases2010['female_judge_decision'] = df_cases2010['female_judge_decision'].astype(str).astype('category')\n",
        "df_cases2010['pendency_decision_filing_categories'] = df_cases2010['pendency_decision_filing_categories'].astype(str).astype('category')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4kYhkp6_kCPq"
      },
      "outputs": [],
      "source": [
        "df_cases2010 = df_cases2010[df_cases2010['pendency_decision_filing_categories'] != 'Undecided']"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "U2GiN58skE24",
        "outputId": "8991af09-c97b-4ed4-b9aa-75a22704c71c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mean values:\n",
            " 36.2776519162895\n",
            "\n",
            "Median values:\n",
            " 31.0\n",
            "\n",
            "75th Percentile:\n",
            " 56.0\n"
          ]
        }
      ],
      "source": [
        "mean_vals = df_cases2010['pendency_decision_filing'].mean()\n",
        "median_vals = df_cases2010['pendency_decision_filing'].median()\n",
        "percentile_75th = np.percentile(df_cases2010['pendency_decision_filing'], 75)\n",
        "\n",
        "print(\"Mean values:\\n\", mean_vals)\n",
        "print(\"\\nMedian values:\\n\", median_vals)\n",
        "print(\"\\n75th Percentile:\\n\", percentile_75th)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0VbNIU-4kIiz",
        "outputId": "261773df-fb22-4225-cca2-f4dbb2eb81d6"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<bound method ExtensionArray.tolist of ['upto 1 year', '3 to 4 years', '5 to 6 years', '2 to 3 years', '4 to 5 years', ..., '6 to 7 years', '7 to 8 years', '8 to 9 years', '9 to 10 years', '10 to 11 years']\n",
            "Length: 11\n",
            "Categories (12, object): ['1 to 2 years', '10 to 11 years', '2 to 3 years', '3 to 4 years', ...,\n",
            "                          '8 to 9 years', '9 to 10 years', 'Undecided', 'upto 1 year']>\n"
          ]
        }
      ],
      "source": [
        "print(df_cases2010['pendency_decision_filing_categories'].unique().tolist)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "LHkDH3VH2hGW"
      },
      "outputs": [],
      "source": [
        "# create columns for binary classification problem\n",
        "mapping_three = {'upto 1 year': 'less_three' ,'1 to 2 years': 'less_three', '2 to 3 years': 'less_three','3 to 4 years': 'three_plus','4 to 5 years': 'three_plus', '5 to 6 years': 'three_plus','6 to 7 years': 'three_plus', '7 to 8 years': 'three_plus','8 to 9 years': 'three_plus', '9 to 10 years': 'three_plus',  '10 to 11 years': 'three_plus','upto 1 year': 'less_three' }\n",
        "df_cases2010['pendency_model_threeyear'] =df_cases2010['pendency_decision_filing_categories'].replace(mapping_three)\n",
        "df_cases2010['pendency_model_threeyear'] = df_cases2010['pendency_model_threeyear'].astype(str).astype('category')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KR5-6nKG5pZy",
        "outputId": "7384142e-9589-444e-c521-1a3b758dd31c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "Int64Index: 3620382 entries, 0 to 4190782\n",
            "Data columns (total 19 columns):\n",
            " #   Column                               Dtype   \n",
            "---  ------                               -----   \n",
            " 0   state_code                           category\n",
            " 1   judge_position                       category\n",
            " 2   female_defendant                     category\n",
            " 3   female_petitioner                    category\n",
            " 4   female_adv_def                       category\n",
            " 5   female_adv_pet                       category\n",
            " 6   type_name                            category\n",
            " 7   act                                  category\n",
            " 8   section                              category\n",
            " 9   bailable_ipc                         category\n",
            " 10  number_sections_ipc                  category\n",
            " 11  criminal                             category\n",
            " 12  female_judge_filing                  category\n",
            " 13  female_judge_decision                category\n",
            " 14  pendency_decision_filing             int64   \n",
            " 15  pendency_decision_filing_categories  category\n",
            " 16  court_details                        category\n",
            " 17  state_district                       category\n",
            " 18  pendency_model_threeyear             category\n",
            "dtypes: category(18), int64(1)\n",
            "memory usage: 150.4 MB\n"
          ]
        }
      ],
      "source": [
        "df_cases2010.info()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ptL9G2n7c0F8"
      },
      "outputs": [],
      "source": [
        "y = df_cases2010[\"pendency_model_threeyear\"]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xcuWn5xyd2tg"
      },
      "outputs": [],
      "source": [
        "col = pd.Categorical(y)\n",
        "y = pd.Series(col.codes)\n",
        "y_class = col.categories.tolist()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VbHLc-owUxvK"
      },
      "outputs": [],
      "source": [
        "#We next drop the target variables from the feature matrix\n",
        "df_cases2010.drop(columns='pendency_decision_filing', axis=1, inplace=True)\n",
        "df_cases2010.drop(columns='pendency_decision_filing_categories', axis=1, inplace=True)\n",
        "df_cases2010.drop(columns='pendency_model_threeyear', axis=1, inplace=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "b5RIJA7TrpuJ"
      },
      "outputs": [],
      "source": [
        "# Lable Encode all Categorical Features\n",
        "le = LabelEncoder()\n",
        "cat_cols = df_cases2010.select_dtypes(include=['category']).columns\n",
        "df_cases2010[cat_cols] = df_cases2010[cat_cols].apply(lambda col: le.fit_transform(col))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XFhXMIMO8irw"
      },
      "source": [
        "## Model Building"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U5Rs5nkxxK4Z"
      },
      "source": [
        "### Train Test Split"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Q4PHacVZxKcU"
      },
      "outputs": [],
      "source": [
        "# Random seed\n",
        "seed =100\n",
        "# split-out train/validation and test dataset\n",
        "X_train, X_test, y_train, y_test = train_test_split(df_cases2010,  y, test_size=0.20, random_state=seed, shuffle=True, stratify=y)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rsHsG0-x7f3O",
        "outputId": "553c202d-b4a6-4865-9fe2-91ba47b6cb06"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "4190783\n",
            "2896305\n",
            "724077\n",
            "3620382\n",
            "86.38915448497333\n"
          ]
        }
      ],
      "source": [
        "# Training and Test Data Size\n",
        "print(df_cases2010_orig.shape[0])\n",
        "print(X_train.shape[0])\n",
        "print(X_test.shape[0])\n",
        "print(X_test.shape[0]+X_train.shape[0])\n",
        "print((X_test.shape[0]+X_train.shape[0])/df_cases2010_orig.shape[0]*100)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vlZyq9SZC2oI"
      },
      "source": [
        "### DecisionTree Classifiers  ( Baseline Model )\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aaPDssrUL1qT",
        "outputId": "af19d733-567e-41f2-8e38-48e3d3ca718b"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Accuracy : 0.7379104708477137\n",
            "Classification report :\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "  less_three       0.74      0.83      0.78    407627\n",
            "  three_plus       0.74      0.62      0.67    316450\n",
            "\n",
            "    accuracy                           0.74    724077\n",
            "   macro avg       0.74      0.72      0.73    724077\n",
            "weighted avg       0.74      0.74      0.73    724077\n",
            "\n",
            "Log Los: 0.5243180576195146\n",
            "ROC AUC: 0.7244867201888228\n",
            "PR AUC: 0.32518873037341334\n"
          ]
        }
      ],
      "source": [
        "# Train a Decision Tree Classifier for binary classification of six year or more\n",
        "# max_depth is set to avoid overfitting and ensure probalistic output from predict_proba which is used to calculate log_loss\n",
        "dt = DecisionTreeClassifier(random_state=seed, max_depth = 10)\n",
        "dt.fit(X_train, y_train)\n",
        "\n",
        "# Predict the target values for the test set\n",
        "y_pred = dt.predict(X_test)\n",
        "\n",
        "# Evaluate the classifier's performance\n",
        "accuracy = accuracy_score(y_test, y_pred)\n",
        "print(\"Accuracy :\", accuracy)\n",
        "print(\"Classification report :\\n\", classification_report(y_test, y_pred, target_names = y_class))\n",
        "# Calculate Log Loss\n",
        "logloss = log_loss(y_test, dt.predict_proba(X_test))\n",
        "print(\"Log Los:\", logloss)\n",
        "#Calculate AUC and plot ROC curve\n",
        "print(\"ROC AUC:\",roc_auc_score(y_test, y_pred))\n",
        "fpr, tpr, treshold = precision_recall_curve(y_test, y_pred)\n",
        "pr_roc_auc = auc(fpr, tpr)\n",
        "print(\"PR AUC:\",pr_roc_auc)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bTTMKnvApFKg"
      },
      "source": [
        "### Bagging Tree Classifiers"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gtihe6RIwate",
        "outputId": "b3ad6e1b-f19f-4ee0-8bf1-0a6e328e143a"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Accuracy : 0.8117396354255141\n",
            "Classification report :\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "  less_three       0.83      0.84      0.83    407627\n",
            "  three_plus       0.79      0.77      0.78    316450\n",
            "\n",
            "    accuracy                           0.81    724077\n",
            "   macro avg       0.81      0.81      0.81    724077\n",
            "weighted avg       0.81      0.81      0.81    724077\n",
            "\n",
            "Log Los: 0.8900483890428973\n",
            "ROC AUC: 0.807602181585274\n",
            "PR AUC: 0.394731954598174\n"
          ]
        }
      ],
      "source": [
        "# Train a Bagging Tree Classifier for binary classification of six year or more\n",
        "dtbag = BaggingClassifier(random_state=seed)\n",
        "dtbag.fit(X_train, y_train)\n",
        "\n",
        "# Predict the target values for the test set\n",
        "y_pred = dtbag.predict(X_test)\n",
        "\n",
        "# Evaluate the classifier's performance\n",
        "accuracy = accuracy_score(y_test, y_pred)\n",
        "print(\"Accuracy :\", accuracy)\n",
        "print(\"Classification report :\\n\", classification_report(y_test, y_pred, target_names = y_class))\n",
        "# Calculate Log Loss\n",
        "logloss = log_loss(y_test, dtbag.predict_proba(X_test))\n",
        "print(\"Log Los:\", logloss)\n",
        "#Calculate AUC and plot ROC curve\n",
        "print(\"ROC AUC:\",roc_auc_score(y_test, y_pred))\n",
        "fpr, tpr, treshold = precision_recall_curve(y_test, y_pred)\n",
        "pr_roc_auc = auc(fpr, tpr)\n",
        "print(\"PR AUC:\",pr_roc_auc)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ILUd90Iuqhzj"
      },
      "source": [
        "### XGBoost Classifier"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "background_save": true,
          "base_uri": "https://localhost:8080/"
        },
        "id": "ChWmjMLpxe_M",
        "outputId": "77bc5d2c-0708-4143-b4de-a357d8ce848e"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.9/dist-packages/xgboost/sklearn.py:835: UserWarning: `eval_metric` in `fit` method is deprecated for better compatibility with scikit-learn, use `eval_metric` in constructor or`set_params` instead.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "CPU times: user 1h 37min 22s, sys: 4.33 s, total: 1h 37min 26s\n",
            "Wall time: 28min 5s\n",
            "Train Accuracy:  0.8110126523277073\n",
            "Test Accuraccy:  0.8086557092684894\n",
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "  less_three       0.82      0.85      0.83    407627\n",
            "  three_plus       0.79      0.76      0.78    316450\n",
            "\n",
            "    accuracy                           0.81    724077\n",
            "   macro avg       0.81      0.80      0.80    724077\n",
            "weighted avg       0.81      0.81      0.81    724077\n",
            "\n",
            "Log Los: 0.4067858171658429\n",
            "ROC AUC: 0.8029479919715824\n",
            "PR AUC: 0.39220289626723087\n"
          ]
        }
      ],
      "source": [
        "import time\n",
        "from xgboost import XGBClassifier\n",
        "# create a default XGBoost classifier\n",
        "xg = XGBClassifier(n_estimators=500, random_state=seed)\n",
        "# define the eval set and metric\n",
        "eval_set = [(X_train, y_train), (X_test, y_test)]\n",
        "eval_metric = [\"auc\",\"error\"]\n",
        "# fit the model\n",
        "%time xg.fit(X_train, y_train, eval_metric=eval_metric, eval_set=eval_set, verbose=False)\n",
        "\n",
        "# final model assessment\n",
        "pred_test = xg.predict(X_test)\n",
        "pred_train = xg.predict(X_train)\n",
        "print('Train Accuracy: ', accuracy_score(y_train, pred_train))\n",
        "print('Test Accuraccy: ', accuracy_score(y_test, pred_test))\n",
        "print('Classification Report:')\n",
        "print(classification_report(y_test,pred_test, target_names = y_class))\n",
        "# Calculate Log Loss\n",
        "logloss = log_loss(y_test, xg.predict_proba(X_test))\n",
        "print(\"Log Los:\", logloss)\n",
        "#Calculate AUC and plot ROC curve\n",
        "print(\"ROC AUC:\",roc_auc_score(y_test, pred_test))\n",
        "fpr, tpr, treshold = precision_recall_curve(y_test, pred_test)\n",
        "pr_roc_auc = auc(fpr, tpr)\n",
        "print(\"PR AUC:\",pr_roc_auc)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "42dwUofRp5M1"
      },
      "source": [
        "### Random Forest Classifier"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Q4aCqALEwcmR",
        "outputId": "875f69ed-d4b8-4b86-e3e9-0afa3cb44eba"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy: 0.8152558360505858\n",
            "Classification report :\n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "  less_three       0.83      0.84      0.84    407627\n",
            "  three_plus       0.79      0.78      0.79    316450\n",
            "\n",
            "    accuracy                           0.82    724077\n",
            "   macro avg       0.81      0.81      0.81    724077\n",
            "weighted avg       0.82      0.82      0.82    724077\n",
            "\n"
          ]
        }
      ],
      "source": [
        "# Train a Random Forest Classifier for binary classification of six year or more\n",
        "dtrf = RandomForestClassifier(random_state=seed)\n",
        "dtrf.fit(X_train, y_train)\n",
        "\n",
        "# Predict the target values for the test set\n",
        "y_pred = dtrf.predict(X_test)\n",
        "\n",
        "# Evaluate the classifier's performance\n",
        "accuracy = accuracy_score(y_test, y_pred)\n",
        "print(\"Accuracy:\", accuracy)\n",
        "print(\"Classification report :\\n\", classification_report(y_test, y_pred, target_names = y_class))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 709
        },
        "id": "cG3sTI2-oHkg",
        "outputId": "48f92f0d-4131-43fc-f16f-572e15ae972b"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 750x750 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "conf = confusion_matrix(y_test, y_pred)\n",
        "classes = [\"<3 years\", \"> 3 years\"]\n",
        "# Print the confusion matrix using Matplotlib\n",
        "\n",
        "fig, ax = plt.subplots(figsize=(7.5, 7.5))\n",
        "ax.matshow(conf, cmap=plt.cm.Blues, alpha=0.3)\n",
        "totalless3 = conf[0,0]+conf[0,1]\n",
        "totalmore3 = conf[1,0]+conf[1,1]\n",
        "for i in range(conf.shape[0]):\n",
        "    for j in range(conf.shape[1]):\n",
        "        if (i == 0 ):\n",
        "          ax.text(x=j, y=i,s=format((conf[i, j]/(totalless3)), '.2%'), va='center', ha='center', size='xx-large')\n",
        "        else:\n",
        "          ax.text(x=j, y=i,s=format((conf[i, j]/(totalmore3)), '.2%'), va='center', ha='center', size='xx-large')\n",
        "\n",
        "plt.xticks(np.arange(len(classes)), classes, fontsize=18)\n",
        "plt.yticks(np.arange(len(classes)), classes, fontsize=18)\n",
        "plt.xlabel('Predictions', fontsize=18)\n",
        "plt.ylabel('Actuals', fontsize=18)\n",
        "plt.title('Confusion Matrix', fontsize=18)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(totalless3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VCWORtlKT4Qr",
        "outputId": "b6b51841-1572-45e5-bee1-5bc2891f58f9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "407627\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "FlSeq6X62vzm",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 484
        },
        "outputId": "0a4707fc-1544-411e-8935-dafb0a5a00bb"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Log Los: 0.5663113659156319\n",
            "ROC AUC: 0.8114425708966659\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Calculate Log Loss\n",
        "logloss = log_loss(y_test, dtrf.predict_proba(X_test))\n",
        "print(\"Log Los:\", logloss)\n",
        "#Calculate AUC and plot ROC curve\n",
        "print(\"ROC AUC:\",roc_auc_score(y_test, y_pred))\n",
        "fpr, tpr, treshold = roc_curve(y_test, y_pred)\n",
        "roc_auc = auc(fpr, tpr)\n",
        "# Plot the AUC ROC curve\n",
        "plt.plot(fpr, tpr, color='red',\n",
        "         lw=2, label='ROC curve')\n",
        "plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--', label=\"Mean\")\n",
        "plt.xlim([-0.05, 1.0])\n",
        "plt.ylim([0.0, 1.05])\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.legend(loc=\"lower right\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "id": "znmqB5JItdAh",
        "outputId": "d2f27fbb-6e4e-467e-8fbc-bc4906befb15"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "PR AUC: 0.39787443694525865\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "#Calculate PR AUC and plot ROC curve\n",
        "fpr, tpr, treshold = precision_recall_curve(y_test, y_pred)\n",
        "pr_roc_auc = auc(fpr, tpr)\n",
        "print(\"PR AUC:\",pr_roc_auc)\n",
        "# Plot the PR ROC curve\n",
        "plt.plot(fpr, tpr, color='red',\n",
        "         lw=2, label='PR ROC')\n",
        "plt.xlim([-0.05, 1.0])\n",
        "plt.ylim([0.0, 1.05])\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.legend(loc=\"lower right\")\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "d-9e23E547fs",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "outputId": "68eca184-33eb-40df-afc3-f098fe9f284d"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Features')"
            ]
          },
          "metadata": {},
          "execution_count": 22
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# get the feature importances\n",
        "features = np.array(X_train.columns)\n",
        "importances = dtrf.feature_importances_\n",
        "\n",
        "# sort the indexes\n",
        "sorted_index = np.argsort(importances)\n",
        "sorted_importances = importances[sorted_index]\n",
        "sorted_features = features[sorted_index]\n",
        "# plot the explained variance using a barplot\n",
        "fig, ax = plt.subplots()\n",
        "ax.barh(sorted_features , sorted_importances)\n",
        "ax.set_xlabel('Importances')\n",
        "ax.set_ylabel('Features')"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "machine_shape": "hm",
      "provenance": []
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}