Copy_of_DMA24_Lab_3_Decision_Trees.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.9"
    },
    "colab": {
      "provenance": []
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vAxKKJ84ugjs"
      },
      "source": [
        "# DMA Fall '24\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nFr3Tjz_ugjv"
      },
      "source": [
        "NAME = \"Min Cheah\"\n",
        "COLLABORATORS = \"\""
      ],
      "execution_count": 91,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VtQyVZG7ugj0"
      },
      "source": [
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "0ec61936c13b80ce01fa0818121a29a9",
          "grade": false,
          "grade_id": "cell-c6a6e713740ff888",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "dSxT_unDugj1"
      },
      "source": [
        "# Lab 3: Decision Trees\n",
        "\n",
        "**Please read the following instructions very carefully**\n",
        "\n",
        "## Working on the assignment / FAQs\n",
        "- **Always use the seed/random_state as *42* wherever applicable** (This is to ensure repeatability in answers, across students and coding environments)\n",
        "- Questions can be either autograded and manually graded.\n",
        "- The type of question and the points they carry are indicated in each question cell\n",
        "- An autograded question has 3 cells\n",
        "     - **Question cell** : Read only cell containing the question\n",
        "     - **Code Cell** : This is where you write the code\n",
        "     - **Grading cell** : This is where the grading occurs, and **you are required not to edit this cell**\n",
        "- Manually graded questions only have the question and code cells. **All manually graded questions are explicitly stated**\n",
        "- To avoid any ambiguity, each question also specifies what *value* must be set. Note that these are dummy values and not the answers\n",
        "- If an autograded question has multiple answers (due to differences in handling NaNs, zeros etc.), all answers will be considered.\n",
        "- Most assignments have bonus questions for extra credit, do try them out!\n",
        "- You can delete the `raise NotImplementedError()` for all questions.\n",
        "- **Submitting the assignment** : Download the '.ipynb' and '.pdf' files from Colab and upload them to bcourses. Do not delete any outputs from cells before submitting.\n",
        "- That's about it. Happy coding!\n",
        "\n",
        "\n",
        "## About the dataset\n",
        "This assignment uses a dataset obtained from the JSE Data Archive that contains biological and self-reported activity traits of a sample of college students at a single university uploaded in 2013. The study associated with these data focused on exploring if a correspondence exists between eye color and other traits. You will be using gender as the target/label in this lab.\n",
        "\n",
        "FEATURE DESCRIPTIONS:\n",
        "- Color (Blue, Brown, Green, Hazel, Other)\n",
        "- Age (in years)\n",
        "- YearinSchool (First, Second, Third, Fourth, Other)\n",
        "- Height (in inches)\n",
        "- Miles (distance from home town of student to Ames, IA)\n",
        "- Brothers (number of brothers)\n",
        "- Sisters (number of sisters)\n",
        "- CompTime (number of hours spent on computer per week)\n",
        "- Exercise (whether the student exercises Yes or No)\n",
        "- ExerTime (number of hours spent exercising per week)\n",
        "- MusicCDs (number of music CDs student owns)\n",
        "- PlayGames (number of hours spent playing games per week)\n",
        "- WatchTV (number of hours spent watching TV per week\n",
        "\n",
        "Background Information on the dataset: http://jse.amstat.org/v21n2/froelich/eyecolorgender.txt"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "0fb5897b9d372125635a32c3c25065db",
          "grade": false,
          "grade_id": "cell-0c1c5a5b7543f388",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "xeiDHBc5ugj2",
        "collapsed": true,
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 258
        },
        "outputId": "23e63194-bc3e-41c9-eef3-46a6aca95a9a"
      },
      "source": [
        "from collections import Counter, defaultdict\n",
        "from itertools import combinations\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import operator\n",
        "import math\n",
        "import itertools\n",
        "from sklearn.feature_extraction import DictVectorizer\n",
        "from sklearn import preprocessing, tree\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "\n",
        "!wget -nc http://quadro.ist.berkeley.edu:1331/eye_color.csv\n",
        "!ls\n",
        "df = pd.read_csv('eye_color.csv')\n",
        "# remove NA's and reset the index\n",
        "df = df.dropna(axis=0, how='any', subset=None, inplace=False)\n",
        "df = df.reset_index(drop=True)\n",
        "\n",
        "df.head()"
      ],
      "execution_count": 92,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "File ‘eye_color.csv’ already there; not retrieving.\n",
            "\n",
            "eye_color.csv  sample_data\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   gender  age    year eyecolor  height  miles  brothers  sisters  \\\n",
              "0  female   18   first    hazel    68.0  195.0         0        1   \n",
              "1    male   20   third    brown    70.0  120.0         3        0   \n",
              "2  female   18   first    green    67.0  200.0         0        1   \n",
              "3    male   23  fourth    hazel    74.0  140.0         1        1   \n",
              "4  female   19  second     blue    62.0   60.0         0        1   \n",
              "\n",
              "   computertime exercise  exercisehours  musiccds  playgames  watchtv  \n",
              "0          20.0      Yes            3.0      75.0        6.0     18.0  \n",
              "1          24.0       No            0.0      50.0        0.0      3.0  \n",
              "2          35.0      Yes            3.0      53.0        8.0      1.0  \n",
              "3           5.0      Yes           25.0      50.0        0.0      7.0  \n",
              "4           5.0      Yes            4.0      30.0        2.0      5.0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1fc8a72d-a54f-44df-8c07-da2484bddd42\" 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>age</th>\n",
              "      <th>year</th>\n",
              "      <th>eyecolor</th>\n",
              "      <th>height</th>\n",
              "      <th>miles</th>\n",
              "      <th>brothers</th>\n",
              "      <th>sisters</th>\n",
              "      <th>computertime</th>\n",
              "      <th>exercise</th>\n",
              "      <th>exercisehours</th>\n",
              "      <th>musiccds</th>\n",
              "      <th>playgames</th>\n",
              "      <th>watchtv</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>female</td>\n",
              "      <td>18</td>\n",
              "      <td>first</td>\n",
              "      <td>hazel</td>\n",
              "      <td>68.0</td>\n",
              "      <td>195.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>20.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>3.0</td>\n",
              "      <td>75.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>18.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>male</td>\n",
              "      <td>20</td>\n",
              "      <td>third</td>\n",
              "      <td>brown</td>\n",
              "      <td>70.0</td>\n",
              "      <td>120.0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>24.0</td>\n",
              "      <td>No</td>\n",
              "      <td>0.0</td>\n",
              "      <td>50.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>female</td>\n",
              "      <td>18</td>\n",
              "      <td>first</td>\n",
              "      <td>green</td>\n",
              "      <td>67.0</td>\n",
              "      <td>200.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>35.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>3.0</td>\n",
              "      <td>53.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>male</td>\n",
              "      <td>23</td>\n",
              "      <td>fourth</td>\n",
              "      <td>hazel</td>\n",
              "      <td>74.0</td>\n",
              "      <td>140.0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>5.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>25.0</td>\n",
              "      <td>50.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>female</td>\n",
              "      <td>19</td>\n",
              "      <td>second</td>\n",
              "      <td>blue</td>\n",
              "      <td>62.0</td>\n",
              "      <td>60.0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>5.0</td>\n",
              "      <td>Yes</td>\n",
              "      <td>4.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1fc8a72d-a54f-44df-8c07-da2484bddd42')\"\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 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\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",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\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-1fc8a72d-a54f-44df-8c07-da2484bddd42 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-1fc8a72d-a54f-44df-8c07-da2484bddd42');\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",
              "\n",
              "\n",
              "<div id=\"df-c01db8c3-361b-4d08-b2fd-28814b95dca0\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-c01db8c3-361b-4d08-b2fd-28814b95dca0')\"\n",
              "            title=\"Suggest charts\"\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",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-c01db8c3-361b-4d08-b2fd-28814b95dca0 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "df",
              "summary": "{\n  \"name\": \"df\",\n  \"rows\": 1988,\n  \"fields\": [\n    {\n      \"column\": \"gender\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"male\",\n          \"female\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"age\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 2,\n        \"min\": 17,\n        \"max\": 56,\n        \"num_unique_values\": 30,\n        \"samples\": [\n          56,\n          41\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"year\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 6,\n        \"samples\": [\n          \"first\",\n          \"third\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"eyecolor\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 5,\n        \"samples\": [\n          \"brown\",\n          \"other\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"height\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 4.207720722595369,\n        \"min\": 44.0,\n        \"max\": 85.0,\n        \"num_unique_values\": 30,\n        \"samples\": [\n          85.0,\n          75.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"miles\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 1078.605752306789,\n        \"min\": 0.0,\n        \"max\": 25000.0,\n        \"num_unique_values\": 328,\n        \"samples\": [\n          347.47,\n          1200.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"brothers\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 1,\n        \"min\": 0,\n        \"max\": 6,\n        \"num_unique_values\": 7,\n        \"samples\": [\n          0,\n          3\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"sisters\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 7,\n        \"num_unique_values\": 8,\n        \"samples\": [\n          0,\n          6\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"computertime\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 11.293750029050782,\n        \"min\": 0.0,\n        \"max\": 110.0,\n        \"num_unique_values\": 66,\n        \"samples\": [\n          110.0,\n          72.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"exercise\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"No\",\n          \"Yes\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"exercisehours\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 4.199955905970483,\n        \"min\": 0.0,\n        \"max\": 45.0,\n        \"num_unique_values\": 36,\n        \"samples\": [\n          45.0,\n          2.5\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"musiccds\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 117.78968694297002,\n        \"min\": 0.0,\n        \"max\": 1500.0,\n        \"num_unique_values\": 142,\n        \"samples\": [\n          266.0,\n          10.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"playgames\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 5.190163228429585,\n        \"min\": 0.0,\n        \"max\": 45.0,\n        \"num_unique_values\": 36,\n        \"samples\": [\n          2.5,\n          5.5\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"watchtv\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 8.256714336557256,\n        \"min\": 0.0,\n        \"max\": 100.0,\n        \"num_unique_values\": 51,\n        \"samples\": [\n          100.0,\n          17.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 92
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "f27367e7e9ea794fdff16caaf9834f90",
          "grade": false,
          "grade_id": "cell-80cb77583712aa18",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "S2_bd_sZugj_"
      },
      "source": [
        "---\n",
        "**Question 1 (0.5 points, autograded)**: How many males and females exist in the dataset?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "7bae96e9bc3cf6154faf15c8c953a12d",
          "grade": false,
          "grade_id": "cell-30e88bd16d0691ae",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "Qq43exlvugkA"
      },
      "source": [
        "count_gender = df.groupby('gender').size()\n"
      ],
      "execution_count": 93,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "4b015ffcf2248033a4c736e194051ce0",
          "grade": false,
          "grade_id": "cell-b6afcdf2e16edeaf",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "qowIpB1MugkL"
      },
      "source": [
        "# The value set in the variables must be integers\n",
        "num_males = count_gender['male'] # Replace 0 with the actual value\n",
        "num_females = count_gender['female'] # Replace 0 with the actual value\n"
      ],
      "execution_count": 94,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "b5f6c5a26715775443afa42a101640d3",
          "grade": true,
          "grade_id": "cell-148801df79f2a9ad",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "rxROmUKRugkQ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "701c7534-a3d5-49f8-adb1-fe44415dba74"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(num_males, num_females)"
      ],
      "execution_count": 95,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "910 1078\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "46a9256f2edd92e7d4ab52309a6b4c57",
          "grade": false,
          "grade_id": "cell-c087e3cd6c8885e2",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "LE7EGFnbugkU"
      },
      "source": [
        "---\n",
        "**Question 2 (0.5 points, autograded)**: What is the Gini Index of this dataset, using males and females as the target classes?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "099940e4f38498b7848e39cf16effa72",
          "grade": false,
          "grade_id": "cell-77e48e0f5499c4dc",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "L5-IeBVlugkV"
      },
      "source": [
        "male_pert = num_males/df.shape[0]\n",
        "female_pert = num_females/df.shape[0]"
      ],
      "execution_count": 96,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "9ad8fad77a751f201648404551610cd0",
          "grade": false,
          "grade_id": "cell-48020b6f76b01df8",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "qWJVFWqqugkZ"
      },
      "source": [
        "# The value set in the variable must be float\n",
        "gini_index = 1 - (male_pert**2 + female_pert**2) # Replace 0 with the actual value / formula\n"
      ],
      "execution_count": 97,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "9e64fcb8ce6b4b69fad066fea2ced65a",
          "grade": true,
          "grade_id": "cell-7ae5a16e0caf13e6",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "N6nvXmTHugke",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2e039896-2395-494f-ca8f-3c630b8abb71"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(gini_index)"
      ],
      "execution_count": 98,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.4964292799047807\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "df88f99709afdfc9d92e17de990723e4",
          "grade": false,
          "grade_id": "cell-429dd5b9870bc4f0",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "p4XzX2Jjugkj"
      },
      "source": [
        "---\n",
        "## Best Split of a numeric feature\n",
        "**Question 3 (1.5 points, autograded)**: What is the best split point of the 'height' feature? (Still using males and females as the target classes, assuming a binary split)\n",
        "\n",
        "Recall that, to calculate the best split of this numeric field, you'll need to order your data by **unique** heights, then consider the midpoint between each pair of consecutive heights as a potential split point, then calculate the Gini Index for that partitioning. You'll want to keep track of the best split point and its Gini Index (remember that you are trying to minimize the Gini Index)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "39f67d746e08f968310e08a1e6302254",
          "grade": false,
          "grade_id": "cell-7e178474fbad8e3a",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "C922WXCougkm"
      },
      "source": [
        "df_heights = df.sort_values('height').reset_index(drop=True)\n",
        "unique_heights = df_heights['height'].unique()\n",
        "total_samples = df.shape[0]"
      ],
      "execution_count": 99,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "6f2ddee90f539fef475c3cf581869a7f",
          "grade": false,
          "grade_id": "cell-fdcb9f54f35480c4",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "RiMp3KiOugkv"
      },
      "source": [
        "#The value set in the variable must be float\n",
        "best_split_point = 0\n",
        "best_gini = float('inf')\n",
        "\n",
        "for i in range(len(unique_heights) - 1):\n",
        "    split = (unique_heights[i] + unique_heights[i + 1]) / 2\n",
        "    group1 = df_heights[df_heights['height'] <= split]\n",
        "    group2 = df_heights[df_heights['height'] > split]\n",
        "    male_pert1 = (group1['gender'] == 'male').sum() / len(group1) if len(group1) > 0 else 0\n",
        "    female_pert1 = (group1['gender'] == 'female').sum() / len(group1) if len(group1) > 0 else 0\n",
        "    gini1 = 1 - (male_pert1**2 + female_pert1**2)\n",
        "    male_pert2 = (group2['gender'] == 'male').sum() / len(group2) if len(group2) > 0 else 0\n",
        "    female_pert2 = (group2['gender'] == 'female').sum() / len(group2) if len(group2) > 0 else 0\n",
        "    gini2 = 1 - (male_pert2**2 + female_pert2**2)\n",
        "\n",
        "    # Calculate weighted Gini index\n",
        "    weighted_gini = (len(group1) / total_samples) * gini1 + (len(group2) / total_samples) * gini2\n",
        "\n",
        "    # Update best split point if current split has a lower Gini index\n",
        "    if weighted_gini < best_gini:\n",
        "        best_gini = weighted_gini\n",
        "        best_split_point = split\n"
      ],
      "execution_count": 100,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "2b3dd2ed02ddb601bc36a52f45defd92",
          "grade": true,
          "grade_id": "cell-323f912bd5314e78",
          "locked": true,
          "points": 1.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "61AwBlXhugk0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4510664c-9ea0-4e60-d2e9-c43b254d425d"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(best_split_point)"
      ],
      "execution_count": 101,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "68.5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "166316144c3f9c6f82d2963c96c19ba4",
          "grade": false,
          "grade_id": "cell-aa468156b67c5ab6",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "AawaoNa9ugk2"
      },
      "source": [
        "---\n",
        "**Question 4 (0.5 points, autograded)**: What is the Gini index of the best split point of the 'height' feature? (Still using males and females as the target classes, assuming a binary split)\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "ee6e07cc1c224dbd0852a4105190e3f2",
          "grade": false,
          "grade_id": "cell-11b6c20fa6270892",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "J_8GTuoSugk3"
      },
      "source": [
        "# YOUR CODE HERE\n"
      ],
      "execution_count": 102,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "4264def559ba5fe2e77d3688f04d9cfd",
          "grade": false,
          "grade_id": "cell-bef6191446f0b09e",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "nnY4qayGugk8"
      },
      "source": [
        "# The value set in the variable must be float\n",
        "gini_of_best_split_point = best_gini# Replace 0 with the actual value"
      ],
      "execution_count": 103,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "0a0c63234a8f7fe57d8a19b011609fda",
          "grade": true,
          "grade_id": "cell-a483f7c52ce46218",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "qtHNnrD8ugk_",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c216c6e5-b709-4792-a58b-4581a2af6598"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(gini_of_best_split_point)"
      ],
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.2655288120702919\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "3e4a285a07159ff52a5f4ffe396f74d6",
          "grade": false,
          "grade_id": "cell-3d05b5c3faef353d",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "TILEzh_WuglE"
      },
      "source": [
        "---\n",
        "**Question 5 (0.5 points, autograded)**: How much does this partitioning reduce the Gini Index over the Gini index of the overall dataset?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "58f9b4479c238a525df6aa057e6eaa3b",
          "grade": false,
          "grade_id": "cell-1ab03a8159601e29",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "qOCZ13AAuglG"
      },
      "source": [
        "# YOUR CODE HERE\n"
      ],
      "execution_count": 105,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "0d222b71a5f4dd420fd140eeefb846d4",
          "grade": false,
          "grade_id": "cell-fbcebc71fc8bc842",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "Z_jBrzFLuglN"
      },
      "source": [
        "# The value set in the variable must be float\n",
        "gini_difference = gini_index - best_gini # Replace 0 with the actual value\n",
        "\n"
      ],
      "execution_count": 106,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "86c116a776736ca45f3b04f15f237623",
          "grade": true,
          "grade_id": "cell-f24e621d36c96976",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "XrLS99QPuglQ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "636c3dc0-fbee-458b-8ac6-38aad5fad4bb"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(gini_difference)"
      ],
      "execution_count": 107,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.2309004678344888\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "a3c0d5d5951cda89ddc95d0edd3c006d",
          "grade": false,
          "grade_id": "cell-68fc7f06fbb3bd5c",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "4vI5aG7GuglS"
      },
      "source": [
        "---\n",
        "**Question 6 (0.5 points, autograded)**: How many 'female' and 'male' rows are within 2 inches (+/- 2 inches) of the best height split point?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "3d82c210dd5b8b938066fc8b9a0e70e8",
          "grade": false,
          "grade_id": "cell-cbfbed1feed2c2c6",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "Zmt1qgE5uglT",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7220daf8-3a28-4f2d-9e21-64235d4db7a2"
      },
      "source": [
        "# YOUR CODE HERE\n",
        "print(best_split_point)"
      ],
      "execution_count": 108,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "68.5\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "71225beec49fc733245cd37612d3712e",
          "grade": false,
          "grade_id": "cell-fed4f19f21abcc31",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "xAB5GTeauglX"
      },
      "source": [
        "# The value set in the variable must be integer\n",
        "height_range_min = 68.5 - 2\n",
        "height_range_max = 68.5 + 2\n",
        "\n",
        "height_filtered_corrected = df[(df['height'] >= height_range_min) & (df['height'] <= height_range_max)]\n",
        "\n",
        "\n",
        "\n",
        "female_rows_within = (height_filtered_corrected['gender'] == 'female').sum() # Replace 0 with the actual value\n",
        "male_rows_within = (height_filtered_corrected['gender'] == 'male').sum() # Replace 0 with the actual value\n"
      ],
      "execution_count": 109,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "12c1b2345e628bec41463cf88a60aeb0",
          "grade": true,
          "grade_id": "cell-8129604534bfec74",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "MiBZAc9vugld",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "dab5f92c-e82e-46ff-b7db-31ff2e2114e4"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(female_rows_within, male_rows_within)"
      ],
      "execution_count": 110,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "377 306\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "9fa8403981b6352444437901acb0d45c",
          "grade": false,
          "grade_id": "cell-95c30c8534082764",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "GmBNDTILuglg"
      },
      "source": [
        "---\n",
        "**Question 7 (0.5 points, autograded)**: How many 'female' and 'male' rows are shorter than the best height split point?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "9754f9fb331109f6c4b2b834211f0021",
          "grade": false,
          "grade_id": "cell-8b16da8f6a50682f",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "dHJLmbwvuglh"
      },
      "source": [
        "# YOUR CODE HERE\n",
        "height_below_split = df[df['height'] < 68.5]\n",
        "\n"
      ],
      "execution_count": 111,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "e2697ca1d9a6feec241ff33460d57600",
          "grade": false,
          "grade_id": "cell-2314033286696a75",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "46G-hbpnugll"
      },
      "source": [
        "#The value set in the variable must be integer\n",
        "female_rows_below = (height_below_split['gender'] == 'female').sum() # Replace 0 with the actual value\n",
        "male_rows_below = (height_below_split['gender'] == 'male').sum() # Replace 0 with the actual value"
      ],
      "execution_count": 112,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "b2c8fbc90d6eb0473ef27e1855483904",
          "grade": true,
          "grade_id": "cell-da98dfdb73b8de96",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "n2seT5oauglo",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "723d2635-fb10-4665-9f62-bd5977712777"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(female_rows_below, male_rows_below)"
      ],
      "execution_count": 113,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "905 142\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "1600c335755c3857d742205f32718e81",
          "grade": false,
          "grade_id": "cell-f0f3bf7078e30689",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "waG6P9fyuglt"
      },
      "source": [
        "---\n",
        "## Best Split of a Categorial Variable\n",
        "\n",
        "**Question 8 (0.5 points, autograded)**: How many possible splits are there of the eyecolor feature? (Assuming binary split)\n",
        "\n",
        "Python tip: the combinations function of the itertools module allows you to enumerate combinations of a list. You might want to Google 'power set'.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "2887cf4ef5a35baf851be351c05f0397",
          "grade": false,
          "grade_id": "cell-e5896ba56062be2d",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "1uYaZLJnuglu"
      },
      "source": [
        "# YOUR CODE HERE\n",
        "unique_eyecolors = df['eyecolor'].unique()"
      ],
      "execution_count": 114,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "c295a12954acf6f936b6a921a548717d",
          "grade": false,
          "grade_id": "cell-7cc929e1f346c97c",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "FozfnlkNuglx"
      },
      "source": [
        "# The value set in the variable must be integer\n",
        "n = len(unique_eyecolors)\n",
        "num_of_splits = 2**n - 2\n",
        "\n"
      ],
      "execution_count": 115,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "d10ef3967a94bfbc24d620dceae732ed",
          "grade": true,
          "grade_id": "cell-e345841556e102c6",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "cLupsW1Sugl0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5ff5e98a-5efb-4ccc-f79c-0e8cf845770b"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(num_of_splits)"
      ],
      "execution_count": 116,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "30\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "3f998cc39260ed85ff0341750544bd86",
          "grade": false,
          "grade_id": "cell-8503b7bec552f92c",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "PuLJsUuvugl3"
      },
      "source": [
        "---\n",
        "**Question 9 (1 points, autograded)**: Which split of eyecolor best splits the female and male rows, as measured by the Gini Index?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "4f117805db47824ae1ef7129672c0d47",
          "grade": false,
          "grade_id": "cell-76f837f4f27e0842",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "HLOlJnsqugl3"
      },
      "source": [
        "# YOUR CODE HERE\n",
        "unique_colors = df['eyecolor'].unique()"
      ],
      "execution_count": 117,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "1904d5cda806153db739b5a397c420ce",
          "grade": false,
          "grade_id": "cell-c6d008b0007ae4d7",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "iTBzhW_Vugl5"
      },
      "source": [
        "# The value set in the variable must be an array\n",
        "def calculate_gini(groups):\n",
        "    total = len(df)\n",
        "    gini = 0\n",
        "    for group in groups:\n",
        "        group_size = len(group)\n",
        "        count_male = group['gender'].value_counts().get('male', 0)\n",
        "        count_female = group['gender'].value_counts().get('female', 0)\n",
        "        prob_male = count_male / group_size if group_size > 0 else 0\n",
        "        prob_female = count_female / group_size if group_size > 0 else 0\n",
        "        gini_group = 1 - (prob_male**2 + prob_female**2)\n",
        "        gini += (group_size / total) * gini_group\n",
        "    return gini\n",
        "\n",
        "# Find the best split\n",
        "min_gini = float('inf')\n",
        "best_split = None\n",
        "\n",
        "for i in range(1, len(unique_colors)):\n",
        "    for combo in combinations(unique_colors, i):\n",
        "        group_1 = df[df['eyecolor'].isin(combo)]\n",
        "        group_2 = df[~df['eyecolor'].isin(combo)]\n",
        "        gini_index = calculate_gini([group_1, group_2])\n",
        "        if gini_index < min_gini:\n",
        "            min_gini = gini_index\n",
        "            best_split = (list(combo), list(set(unique_colors) - set(combo)))\n"
      ],
      "execution_count": 118,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "4e306b120de04cbdc496f0bf5ca4bc75",
          "grade": true,
          "grade_id": "cell-633329070c59b89d",
          "locked": true,
          "points": 1,
          "schema_version": 1,
          "solution": false
        },
        "id": "DBKu0dDvugl-",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b5814e8d-ddb4-40ea-9174-3692dc6be092"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(colour_group_1, colour_group_2)"
      ],
      "execution_count": 119,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "['blue', 'other'] ['hazel', 'brown', 'green']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "a5fa6db03e0b1d03320f702d669e6328",
          "grade": false,
          "grade_id": "cell-54d3a4d63e10789a",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "Q0n3uxWFugmC"
      },
      "source": [
        "---\n",
        "**Question 10 (0.5 points, autograded)**: What is the Gini Index of this best split?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "f1b5c57dc68cb76716cb9dd1e5cb00db",
          "grade": false,
          "grade_id": "cell-f755b4ebebdcbcd3",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "SLWuwwvnugmF"
      },
      "source": [
        "# YOUR CODE HERE\n"
      ],
      "execution_count": 120,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "efce7c9024ee735ae78576a9e6153798",
          "grade": false,
          "grade_id": "cell-b9d691586c5addbd",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "VYJlCiCuugmJ"
      },
      "source": [
        "# The value set in the variable must be float\n",
        "gini_of_best_split_group = chosen_gini# Replace 0 with the actual value\n"
      ],
      "execution_count": 121,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "2d4da39818e70541f2d3a55729adbae0",
          "grade": true,
          "grade_id": "cell-7332ad4509b48564",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "MFn6uVBcugmL",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "26644f20-c44b-4aee-c614-6a892e40e505"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(gini_of_best_split_group)"
      ],
      "execution_count": 122,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.494857288346644\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "473602565cfa2400f2ad7b85fd72b373",
          "grade": false,
          "grade_id": "cell-34f1033ad6a06232",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "_o0PW2rLugmO"
      },
      "source": [
        "---\n",
        "**Question 11 (0.5 points, autograded)**: How much does this partitioning decrease the Gini Index over the Gini index of the overall dataset?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "bf9360740e6f25bb1c156c93c1869845",
          "grade": false,
          "grade_id": "cell-68d41651a95a23df",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "Ja5w5_ucugmP"
      },
      "source": [
        "# YOUR CODE HERE\n"
      ],
      "execution_count": 123,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "7d84ef913eb55fa6ab7d5f327bc8b650",
          "grade": false,
          "grade_id": "cell-7e15f66615ed4361",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "Z1m4Yf-NugmR"
      },
      "source": [
        "#The value set in the variable must be float\n",
        "gini_difference_2 = gini_index - chosen_gini # Replace 0 with the actual value\n",
        "\n"
      ],
      "execution_count": 124,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "6ac2eb45e569ee8da44b7ed0faf33edc",
          "grade": true,
          "grade_id": "cell-a879c9d33c4f102f",
          "locked": true,
          "points": 0.5,
          "schema_version": 1,
          "solution": false
        },
        "id": "-7ij3URDugmT",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e327dbba-526b-46aa-d5e4-b918ce2126ba"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(gini_difference_2)"
      ],
      "execution_count": 125,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.001569052658539738\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "60aae9f1a5f6e166acf954ee3219986a",
          "grade": false,
          "grade_id": "cell-a47e13b3fcc0108a",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "C0UNgzrTugmX"
      },
      "source": [
        "---\n",
        "**Question 12 (1 points, autograded)** : How many 'female' rows and 'male' rows are in your first partition? How many 'female' rows and 'male' rows are in your second partition?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "83f6848101b44ce8d5d499484d55d3b2",
          "grade": false,
          "grade_id": "cell-da838430a7b931e1",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "JguwfDQmugmY"
      },
      "source": [
        "# YOUR CODE HERE\n",
        "partition1 = df[df['eyecolor'].isin(best_split[0])]\n",
        "partition2 = df[df['eyecolor'].isin(best_split[1])]"
      ],
      "execution_count": 126,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "3c880462e0a6e9e9d64542752b22cbef",
          "grade": false,
          "grade_id": "cell-e4c3bc95376f6df7",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "H-dDxSbDugma"
      },
      "source": [
        "# The value set in the variable must be integer, order doesn't matter\n",
        "partition1_male = (partition1['gender'] == 'male').sum() # Replace 0 with the actual value\n",
        "partition1_female = (partition1['gender'] == 'female').sum() # Replace 0 with the actual value\n",
        "partition2_male = (partition2['gender'] == 'male').sum() # Replace 0 with the actual value\n",
        "partition2_female = (partition2['gender'] == 'female').sum() # Replace 0 with the actual value\n"
      ],
      "execution_count": 127,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "0fdc2e6ab8f5b169f5bfd9566d85fc01",
          "grade": true,
          "grade_id": "cell-1c022d94964cb6cc",
          "locked": true,
          "points": 1,
          "schema_version": 1,
          "solution": false
        },
        "id": "XYCfnyCNugmd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0aa94b61-0ec3-49dc-d536-e326e33e7431"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(partition1_male, partition1_female, partition2_male, partition2_female)"
      ],
      "execution_count": 128,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "107 190 803 888\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "2d2ab95c05c31c5f214bd422e16e450b",
          "grade": false,
          "grade_id": "cell-d81b2ec1e886e1b6",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "JoFggmQGugmi"
      },
      "source": [
        "---\n",
        "## Training a decision tree\n",
        "**Question 13 (1 points, autograded)**: Using all of the features in the original dataframe read in at the top of this notebook, train a decision tree classifier that has a depth of three (not including the root node). What is the accuracy of this classifier on the training data)?\n",
        "\n",
        "Scikit-learn classifiers require class labels and features to be in numeric arrays. As such, you will need to turn your categorical features into numeric arrays using DictVectorizer. This is a helpful notebook for understanding how to do this: http://nbviewer.ipython.org/gist/sarguido/7423289. You can turn a pandas dataframe of features into a dictionary of the form needed by DictVectorizer by using df.to_dict('records'). Make sure you remove the class label first (in this case, gender). If you use the class label as a feature, your classifier will have a training accuracy of 100%! The example notebook link also shows how to turn your class labels into a numeric array using sklearn.preprocessing.LabelEncoder()."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "14960b019250231f745c6db5a38c70f7",
          "grade": false,
          "grade_id": "cell-a77a52f79c83b21a",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "kFIx_xdyugmj"
      },
      "source": [
        "# YOUR CODE HERE\n",
        "X_df = df.drop('gender', axis=1)\n",
        "y = df['gender']"
      ],
      "execution_count": 129,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "f26339bc22d99829fefeb18e1c46cb06",
          "grade": false,
          "grade_id": "cell-55c9f9a2702cf09c",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "CqowI-lrugmn"
      },
      "source": [
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "\n",
        "X_dict = X_df.to_dict('records')\n",
        "vec = DictVectorizer(sparse=False)\n",
        "X = vec.fit_transform(X_dict)\n",
        "encoder = LabelEncoder()\n",
        "Y = encoder.fit_transform(y)\n",
        "clf = tree.DecisionTreeClassifier(max_depth=3, random_state=42)\n",
        "clf.fit(X, Y)\n",
        "y_predicted = clf.predict(X)\n",
        "accuracy = accuracy_score(Y, y_predicted)\n"
      ],
      "execution_count": 130,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "75544e763b81e3e36748ea0f4989c6ae",
          "grade": true,
          "grade_id": "cell-329a69c83b199316",
          "locked": true,
          "points": 1,
          "schema_version": 1,
          "solution": false
        },
        "id": "urP93Al5ugmu",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0da364be-f6d6-4345-eb4b-c41115abc69e"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(accuracy)"
      ],
      "execution_count": 131,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.8646881287726358\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "f23e20cb9c25f2f8d674f98079625ff2",
          "grade": false,
          "grade_id": "cell-e321872a66519e2c",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "MpU-3jviugm1"
      },
      "source": [
        "---\n",
        "**Question 14 (1 points, manually graded)**: Using the following code snippet, visualize your decision tree. In your write-up, **write down the interpretation of the rule at each node** which is used to perform the splitting.\n",
        "\n",
        "We provide **two options** to visualize decision trees. The first option uses `tree.plot_tree` and the other uses an external tool called `GraphViz`. You can **use either of the two options**.  `tree.plot_tree` is the **recommended and easier option** as it is a built-in function in `sklearn` and doesn't require any additional setup.\n",
        "\n",
        "Uncomment the code, **fill in the clf (classifier) and `feature_names` arguments**. Executing the code will display the tree visualization in the output cell.\n",
        "\n",
        "Note for users who want to install graphviz on their local machines (**you don't need to do install graphviz if you're running the notebook Colab**, which is the class' recommended way of doing assignments):\n",
        "\n",
        "\n",
        "\n",
        "> In order to install graphviz, you may need to download the tool from [this website](https://graphviz.gitlab.io), and then pip3/conda install the python libraries you do not have. Mac users can use ```brew install graphviz``` instead of following the link, and linux users can do the same using their favourite package manager (for example, Ubuntu users can use ```sudo apt-get install graphviz```, followed by the necessary pip3/conda installations.\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CcYGDqUIIGZa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "3c3ebd5a-246d-47ea-a1cb-b75ba119ee84"
      },
      "source": [
        "# Option 1 (Recommended Option) - Using `tree.plot_tree`\n",
        "\n",
        "fig, ax = plt.subplots(figsize=(14, 14))\n",
        "tree.plot_tree(clf, fontsize=10, feature_names=vec.feature_names_);"
      ],
      "execution_count": 132,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1400x1400 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "b43dac070b745305841e22382bc87bda",
          "grade": false,
          "grade_id": "cell-5aaab8332072d23e",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "H80MNiT4ugm2"
      },
      "source": [
        "# Option 2 - Using GraphViz. Visualization is prettier, but additional setup may be required if running on your local machine (although no setup required on Colab)\n",
        "\n",
        "from IPython.display import Image\n",
        "import pydotplus\n",
        "import pydot\n",
        "from six import StringIO\n",
        "\n",
        "# clf = your classifier\n",
        "# dotfile = StringIO()\n",
        "# tree.export_graphviz(clf, out_file=dotfile,\n",
        "#                       feature_names=<Names of columns>,\n",
        "#                           class_names=['Female', 'Male'],\n",
        "#                           filled=True, rounded=True,\n",
        "#                           special_characters=True)\n",
        "# graph = pydotplus.graph_from_dot_data(dotfile.getvalue())\n",
        "# Image(graph.create_png())\n",
        "\n",
        "\n",
        "# Ignore the cell below, but do not delete it. It is used to grade the image output of this cell."
      ],
      "execution_count": 133,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this tree, every node visualizes a decision rule based on the features-for example, height <= 68.5, the Gini impurity index that determines the impurity measure of the node, the number of samples at that particular node, and the class distribution for those samples. The root node starts with all the samples and applies the first decision rule, splitting them into two groups based on the threshold of the feature specified. Each subsequent node refines this process by further splitting the groups based on new rules until a maximum specified depth is reached or no significant splits can be made anymore. The leaf nodes of any branch are the final decision nodes of the model, where the majority class of samples determines the predicted class for those paths. This is quite a useful visualization in understanding the logic behind the classifier's predictions-that is, which features are the most relevant, at what thresholds, for the decision-making process."
      ],
      "metadata": {
        "id": "BnYYw4RPE9O2"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "29e84207b6e13109206f7ef78d0ff602",
          "grade": false,
          "grade_id": "cell-06714e6427b2d0d0",
          "locked": true,
          "schema_version": 1,
          "solution": false
        },
        "id": "gIOPTYuiugnF"
      },
      "source": [
        "---\n",
        "## Bonus Question (2 points, auto graded)\n",
        "For each of your leaf nodes, specify the percentage of 'female' rows in that node (out of the total number of rows at that node)\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "nbgrader": {
          "checksum": "88cda91aeb11b6f96d6c2bacf4a4804e",
          "grade": false,
          "grade_id": "cell-650a81bc31e6f608",
          "locked": false,
          "schema_version": 1,
          "solution": true
        },
        "id": "co_iRKrUugnG"
      },
      "source": [
        "# The value set in the variable must be array\n",
        "leaf_indices = np.where(clf.tree_.children_left == -1)[0]\n",
        "\n",
        "ratios = []\n",
        "for leaf_index in leaf_indices:\n",
        "    values = clf.tree_.value[leaf_index][0]\n",
        "    if sum(values) != 0:\n",
        "        percentage_female = (values[0] / sum(values)) * 100\n",
        "        ratios.append(percentage_female)\n",
        "    else:\n",
        "        ratios.append(0)\n",
        "\n"
      ],
      "execution_count": 134,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "deletable": false,
        "editable": false,
        "nbgrader": {
          "checksum": "de18cf01eb62a9398145d3f1879681d5",
          "grade": true,
          "grade_id": "cell-2cc540aa430d101d",
          "locked": true,
          "points": 2,
          "schema_version": 1,
          "solution": false
        },
        "id": "Ugyn_t3lugnH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "83400cef-6b3e-435a-b51c-89ce711b4685"
      },
      "source": [
        "# This is an autograded cell, do not edit\n",
        "print(ratios)"
      ],
      "execution_count": 135,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[93.25301204819277, 68.30985915492957, 81.48148148148148, 25.0, 55.729166666666664, 11.678832116788321, 15.879828326180256, 3.430079155672823]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "*ⓒ Prof. Zachary Pardos, 2022*"
      ],
      "metadata": {
        "id": "TcHv6u9pzZKV"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "yuwA2qNYIppe"
      },
      "execution_count": 135,
      "outputs": []
    }
  ]
}