week-4/MLZoomcamp_wk_4.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "### Import packages"
      ],
      "metadata": {
        "id": "g4iRF0yiGZel"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "cLqoDMqgGPEC"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from sklearn.linear_model import LogisticRegression, Ridge\n",
        "from sklearn.model_selection import train_test_split, KFold\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import mutual_info_score, accuracy_score, mean_squared_error, roc_auc_score, precision_score, recall_score, f1_score\n",
        "from sklearn.feature_extraction import DictVectorizer\n",
        "from math import sqrt"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Loading the dataset"
      ],
      "metadata": {
        "id": "2mmgusw7GwaY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!wget https://raw.githubusercontent.com/alexeygrigorev/mlbookcamp-code/master/chapter-02-car-price/data.csv"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GK2DkfJPG0LK",
        "outputId": "81c3d04e-05a4-4515-abfb-e2f24c1de43a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2023-10-03 21:29:46--  https://raw.githubusercontent.com/alexeygrigorev/mlbookcamp-code/master/chapter-02-car-price/data.csv\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 1475504 (1.4M) [text/plain]\n",
            "Saving to: ‘data.csv.2’\n",
            "\n",
            "data.csv.2          100%[===================>]   1.41M  --.-KB/s    in 0.08s   \n",
            "\n",
            "2023-10-03 21:29:46 (18.6 MB/s) - ‘data.csv.2’ saved [1475504/1475504]\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df = pd.read_csv(\"/content/data.csv\", encoding = \"utf-8\")"
      ],
      "metadata": {
        "id": "Qvy5DwH0G1wq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df.info()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mBoL2yzoHCkd",
        "outputId": "ffb4ec25-8d5b-4d82-bdf4-085527171b6f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 11914 entries, 0 to 11913\n",
            "Data columns (total 16 columns):\n",
            " #   Column             Non-Null Count  Dtype  \n",
            "---  ------             --------------  -----  \n",
            " 0   Make               11914 non-null  object \n",
            " 1   Model              11914 non-null  object \n",
            " 2   Year               11914 non-null  int64  \n",
            " 3   Engine Fuel Type   11911 non-null  object \n",
            " 4   Engine HP          11845 non-null  float64\n",
            " 5   Engine Cylinders   11884 non-null  float64\n",
            " 6   Transmission Type  11914 non-null  object \n",
            " 7   Driven_Wheels      11914 non-null  object \n",
            " 8   Number of Doors    11908 non-null  float64\n",
            " 9   Market Category    8172 non-null   object \n",
            " 10  Vehicle Size       11914 non-null  object \n",
            " 11  Vehicle Style      11914 non-null  object \n",
            " 12  highway MPG        11914 non-null  int64  \n",
            " 13  city mpg           11914 non-null  int64  \n",
            " 14  Popularity         11914 non-null  int64  \n",
            " 15  MSRP               11914 non-null  int64  \n",
            "dtypes: float64(3), int64(5), object(8)\n",
            "memory usage: 1.5+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Data preparation"
      ],
      "metadata": {
        "id": "Cnd9D2XKHAF7"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df.columns = df.columns.str.replace(' ', '_').str.lower()\n",
        "df.info()"
      ],
      "metadata": {
        "id": "Rio4lL36G9Z5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f153a3ca-bd45-4dd5-cf9d-33e0d5d4844b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 11914 entries, 0 to 11913\n",
            "Data columns (total 16 columns):\n",
            " #   Column             Non-Null Count  Dtype  \n",
            "---  ------             --------------  -----  \n",
            " 0   make               11914 non-null  object \n",
            " 1   model              11914 non-null  object \n",
            " 2   year               11914 non-null  int64  \n",
            " 3   engine_fuel_type   11911 non-null  object \n",
            " 4   engine_hp          11845 non-null  float64\n",
            " 5   engine_cylinders   11884 non-null  float64\n",
            " 6   transmission_type  11914 non-null  object \n",
            " 7   driven_wheels      11914 non-null  object \n",
            " 8   number_of_doors    11908 non-null  float64\n",
            " 9   market_category    8172 non-null   object \n",
            " 10  vehicle_size       11914 non-null  object \n",
            " 11  vehicle_style      11914 non-null  object \n",
            " 12  highway_mpg        11914 non-null  int64  \n",
            " 13  city_mpg           11914 non-null  int64  \n",
            " 14  popularity         11914 non-null  int64  \n",
            " 15  msrp               11914 non-null  int64  \n",
            "dtypes: float64(3), int64(5), object(8)\n",
            "memory usage: 1.5+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "features = [\"make\", \"model\", \"year\", \"engine_hp\", \"engine_cylinders\", \"transmission_type\", \"vehicle_style\", \"highway_mpg\", \"city_mpg\", \"msrp\"]\n",
        "df = df[features]\n",
        "df.info()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dQ--Mr-RoU0a",
        "outputId": "55e6ee81-6e95-4934-eed1-4fbabce3bc5d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 11914 entries, 0 to 11913\n",
            "Data columns (total 10 columns):\n",
            " #   Column             Non-Null Count  Dtype  \n",
            "---  ------             --------------  -----  \n",
            " 0   make               11914 non-null  object \n",
            " 1   model              11914 non-null  object \n",
            " 2   year               11914 non-null  int64  \n",
            " 3   engine_hp          11845 non-null  float64\n",
            " 4   engine_cylinders   11884 non-null  float64\n",
            " 5   transmission_type  11914 non-null  object \n",
            " 6   vehicle_style      11914 non-null  object \n",
            " 7   highway_mpg        11914 non-null  int64  \n",
            " 8   city_mpg           11914 non-null  int64  \n",
            " 9   msrp               11914 non-null  int64  \n",
            "dtypes: float64(2), int64(4), object(4)\n",
            "memory usage: 930.9+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df = df.fillna(0)\n",
        "df.info()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5U4e1AHIokgF",
        "outputId": "035b122e-db52-468e-946c-ca3ab99b66b8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 11914 entries, 0 to 11913\n",
            "Data columns (total 10 columns):\n",
            " #   Column             Non-Null Count  Dtype  \n",
            "---  ------             --------------  -----  \n",
            " 0   make               11914 non-null  object \n",
            " 1   model              11914 non-null  object \n",
            " 2   year               11914 non-null  int64  \n",
            " 3   engine_hp          11914 non-null  float64\n",
            " 4   engine_cylinders   11914 non-null  float64\n",
            " 5   transmission_type  11914 non-null  object \n",
            " 6   vehicle_style      11914 non-null  object \n",
            " 7   highway_mpg        11914 non-null  int64  \n",
            " 8   city_mpg           11914 non-null  int64  \n",
            " 9   msrp               11914 non-null  int64  \n",
            "dtypes: float64(2), int64(4), object(4)\n",
            "memory usage: 930.9+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df = df.rename({\"msrp\": \"price\"}, axis=1)\n",
        "df.info()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rYYH2ZLxot20",
        "outputId": "ab1669fc-a569-46c5-da8c-7e62cdb9fbb2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 11914 entries, 0 to 11913\n",
            "Data columns (total 10 columns):\n",
            " #   Column             Non-Null Count  Dtype  \n",
            "---  ------             --------------  -----  \n",
            " 0   make               11914 non-null  object \n",
            " 1   model              11914 non-null  object \n",
            " 2   year               11914 non-null  int64  \n",
            " 3   engine_hp          11914 non-null  float64\n",
            " 4   engine_cylinders   11914 non-null  float64\n",
            " 5   transmission_type  11914 non-null  object \n",
            " 6   vehicle_style      11914 non-null  object \n",
            " 7   highway_mpg        11914 non-null  int64  \n",
            " 8   city_mpg           11914 non-null  int64  \n",
            " 9   price              11914 non-null  int64  \n",
            "dtypes: float64(2), int64(4), object(4)\n",
            "memory usage: 930.9+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "avg_price = df.price.mean()\n",
        "avg_price"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "baNv2oSGozJo",
        "outputId": "ba3dd68c-e99d-4799-8523-e07b6a336c06"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "40594.737032063116"
            ]
          },
          "metadata": {},
          "execution_count": 93
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "above_average = (df.price >= avg_price).astype(int)\n",
        "above_average"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3IuoajxSo51l",
        "outputId": "6eb6c156-7e6f-4cfe-9c10-76a801b8b368"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0        1\n",
              "1        1\n",
              "2        0\n",
              "3        0\n",
              "4        0\n",
              "        ..\n",
              "11909    1\n",
              "11910    1\n",
              "11911    1\n",
              "11912    1\n",
              "11913    0\n",
              "Name: price, Length: 11914, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 94
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df[\"above_average\"] = above_average\n",
        "df.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "8N2Aq3dWo8-1",
        "outputId": "dfe51974-ba72-41e0-de34-c28ef3512efb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  make       model  year  engine_hp  engine_cylinders transmission_type  \\\n",
              "0  BMW  1 Series M  2011      335.0               6.0            MANUAL   \n",
              "1  BMW    1 Series  2011      300.0               6.0            MANUAL   \n",
              "2  BMW    1 Series  2011      300.0               6.0            MANUAL   \n",
              "3  BMW    1 Series  2011      230.0               6.0            MANUAL   \n",
              "4  BMW    1 Series  2011      230.0               6.0            MANUAL   \n",
              "\n",
              "  vehicle_style  highway_mpg  city_mpg  price  above_average  \n",
              "0         Coupe           26        19  46135              1  \n",
              "1   Convertible           28        19  40650              1  \n",
              "2         Coupe           28        20  36350              0  \n",
              "3         Coupe           28        18  29450              0  \n",
              "4   Convertible           28        18  34500              0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1af8d497-6a09-4f06-8f80-e481bcc6151a\" 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>make</th>\n",
              "      <th>model</th>\n",
              "      <th>year</th>\n",
              "      <th>engine_hp</th>\n",
              "      <th>engine_cylinders</th>\n",
              "      <th>transmission_type</th>\n",
              "      <th>vehicle_style</th>\n",
              "      <th>highway_mpg</th>\n",
              "      <th>city_mpg</th>\n",
              "      <th>price</th>\n",
              "      <th>above_average</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>BMW</td>\n",
              "      <td>1 Series M</td>\n",
              "      <td>2011</td>\n",
              "      <td>335.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>MANUAL</td>\n",
              "      <td>Coupe</td>\n",
              "      <td>26</td>\n",
              "      <td>19</td>\n",
              "      <td>46135</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>BMW</td>\n",
              "      <td>1 Series</td>\n",
              "      <td>2011</td>\n",
              "      <td>300.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>MANUAL</td>\n",
              "      <td>Convertible</td>\n",
              "      <td>28</td>\n",
              "      <td>19</td>\n",
              "      <td>40650</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>BMW</td>\n",
              "      <td>1 Series</td>\n",
              "      <td>2011</td>\n",
              "      <td>300.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>MANUAL</td>\n",
              "      <td>Coupe</td>\n",
              "      <td>28</td>\n",
              "      <td>20</td>\n",
              "      <td>36350</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>BMW</td>\n",
              "      <td>1 Series</td>\n",
              "      <td>2011</td>\n",
              "      <td>230.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>MANUAL</td>\n",
              "      <td>Coupe</td>\n",
              "      <td>28</td>\n",
              "      <td>18</td>\n",
              "      <td>29450</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>BMW</td>\n",
              "      <td>1 Series</td>\n",
              "      <td>2011</td>\n",
              "      <td>230.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>MANUAL</td>\n",
              "      <td>Convertible</td>\n",
              "      <td>28</td>\n",
              "      <td>18</td>\n",
              "      <td>34500</td>\n",
              "      <td>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-1af8d497-6a09-4f06-8f80-e481bcc6151a')\"\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-1af8d497-6a09-4f06-8f80-e481bcc6151a 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-1af8d497-6a09-4f06-8f80-e481bcc6151a');\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-a8d12876-a783-4ad3-a1ce-aefec94b2608\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-a8d12876-a783-4ad3-a1ce-aefec94b2608')\"\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-a8d12876-a783-4ad3-a1ce-aefec94b2608 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 95
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Create validation framework"
      ],
      "metadata": {
        "id": "Wih2FXjdpKsx"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def create_valid_framework(df, y_label_name, random_state=1):\n",
        "  df_full_train, df_test = train_test_split(df, test_size = 0.2, random_state = 1)\n",
        "  df_train, df_val = train_test_split(df_full_train, test_size = 0.25, random_state = 1)\n",
        "\n",
        "  print(len(df_train), len(df_val), len(df_test))\n",
        "\n",
        "  df_train = df_train.reset_index(drop=True)\n",
        "  df_test = df_test.reset_index(drop=True)\n",
        "  df_val = df_val.reset_index(drop=True)\n",
        "\n",
        "  y_train = df_train[y_label_name].values\n",
        "  y_test = df_test[y_label_name].values\n",
        "  y_val = df_val[y_label_name].values\n",
        "\n",
        "  return df_train, df_val, df_test, y_train, y_val, y_test\n"
      ],
      "metadata": {
        "id": "WL0LxeX74NTR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df_train, df_val, df_test, y_train, y_val, y_test = create_valid_framework(df=df, y_label_name=\"above_average\", random_state=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2gJk7c94o_nJ",
        "outputId": "33bee15f-397e-4223-e797-b5df500f9c9b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "7148 2383 2383\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df_train.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "JcXklFMdpwV6",
        "outputId": "7084b3af-053e-4c24-f552-9ca578851944"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "         make              model  year  engine_hp  engine_cylinders  \\\n",
              "0      Toyota  Highlander Hybrid  2016      280.0               6.0   \n",
              "1   Chevrolet     Impala Limited  2014      300.0               6.0   \n",
              "2       Buick              Regal  2016      259.0               4.0   \n",
              "3  Volkswagen                 CC  2015      280.0               6.0   \n",
              "4       Dodge      Grand Caravan  2016      283.0               6.0   \n",
              "\n",
              "  transmission_type      vehicle_style  highway_mpg  city_mpg  price  \\\n",
              "0         AUTOMATIC            4dr SUV           28        27  50485   \n",
              "1         AUTOMATIC              Sedan           30        18  26840   \n",
              "2         AUTOMATIC              Sedan           27        19  36490   \n",
              "3  AUTOMATED_MANUAL              Sedan           25        17  43575   \n",
              "4         AUTOMATIC  Passenger Minivan           25        17  22595   \n",
              "\n",
              "   above_average  \n",
              "0              1  \n",
              "1              0  \n",
              "2              0  \n",
              "3              1  \n",
              "4              0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-63098c6f-8a45-48f5-a5fd-92f682e4827a\" 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>make</th>\n",
              "      <th>model</th>\n",
              "      <th>year</th>\n",
              "      <th>engine_hp</th>\n",
              "      <th>engine_cylinders</th>\n",
              "      <th>transmission_type</th>\n",
              "      <th>vehicle_style</th>\n",
              "      <th>highway_mpg</th>\n",
              "      <th>city_mpg</th>\n",
              "      <th>price</th>\n",
              "      <th>above_average</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Toyota</td>\n",
              "      <td>Highlander Hybrid</td>\n",
              "      <td>2016</td>\n",
              "      <td>280.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>AUTOMATIC</td>\n",
              "      <td>4dr SUV</td>\n",
              "      <td>28</td>\n",
              "      <td>27</td>\n",
              "      <td>50485</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Chevrolet</td>\n",
              "      <td>Impala Limited</td>\n",
              "      <td>2014</td>\n",
              "      <td>300.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>AUTOMATIC</td>\n",
              "      <td>Sedan</td>\n",
              "      <td>30</td>\n",
              "      <td>18</td>\n",
              "      <td>26840</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Buick</td>\n",
              "      <td>Regal</td>\n",
              "      <td>2016</td>\n",
              "      <td>259.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>AUTOMATIC</td>\n",
              "      <td>Sedan</td>\n",
              "      <td>27</td>\n",
              "      <td>19</td>\n",
              "      <td>36490</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Volkswagen</td>\n",
              "      <td>CC</td>\n",
              "      <td>2015</td>\n",
              "      <td>280.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>AUTOMATED_MANUAL</td>\n",
              "      <td>Sedan</td>\n",
              "      <td>25</td>\n",
              "      <td>17</td>\n",
              "      <td>43575</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Dodge</td>\n",
              "      <td>Grand Caravan</td>\n",
              "      <td>2016</td>\n",
              "      <td>283.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>AUTOMATIC</td>\n",
              "      <td>Passenger Minivan</td>\n",
              "      <td>25</td>\n",
              "      <td>17</td>\n",
              "      <td>22595</td>\n",
              "      <td>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-63098c6f-8a45-48f5-a5fd-92f682e4827a')\"\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-63098c6f-8a45-48f5-a5fd-92f682e4827a 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-63098c6f-8a45-48f5-a5fd-92f682e4827a');\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-5c158515-8b40-46fc-bd4d-f9e07e19e46f\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-5c158515-8b40-46fc-bd4d-f9e07e19e46f')\"\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-5c158515-8b40-46fc-bd4d-f9e07e19e46f button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 98
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "del df_train[\"price\"]\n",
        "del df_train[\"above_average\"]\n",
        "\n",
        "del df_test[\"price\"]\n",
        "del df_test[\"above_average\"]\n",
        "\n",
        "del df_val[\"price\"]\n",
        "del df_val[\"above_average\"]"
      ],
      "metadata": {
        "id": "1OGXOr8Ope1I"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df_train.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "GNtuNzIz5PLk",
        "outputId": "fcde32fa-d84c-4065-e34d-a6442702e6ac"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "         make              model  year  engine_hp  engine_cylinders  \\\n",
              "0      Toyota  Highlander Hybrid  2016      280.0               6.0   \n",
              "1   Chevrolet     Impala Limited  2014      300.0               6.0   \n",
              "2       Buick              Regal  2016      259.0               4.0   \n",
              "3  Volkswagen                 CC  2015      280.0               6.0   \n",
              "4       Dodge      Grand Caravan  2016      283.0               6.0   \n",
              "\n",
              "  transmission_type      vehicle_style  highway_mpg  city_mpg  \n",
              "0         AUTOMATIC            4dr SUV           28        27  \n",
              "1         AUTOMATIC              Sedan           30        18  \n",
              "2         AUTOMATIC              Sedan           27        19  \n",
              "3  AUTOMATED_MANUAL              Sedan           25        17  \n",
              "4         AUTOMATIC  Passenger Minivan           25        17  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-8acd4b10-2874-4577-b197-2e5882052d07\" 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>make</th>\n",
              "      <th>model</th>\n",
              "      <th>year</th>\n",
              "      <th>engine_hp</th>\n",
              "      <th>engine_cylinders</th>\n",
              "      <th>transmission_type</th>\n",
              "      <th>vehicle_style</th>\n",
              "      <th>highway_mpg</th>\n",
              "      <th>city_mpg</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Toyota</td>\n",
              "      <td>Highlander Hybrid</td>\n",
              "      <td>2016</td>\n",
              "      <td>280.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>AUTOMATIC</td>\n",
              "      <td>4dr SUV</td>\n",
              "      <td>28</td>\n",
              "      <td>27</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Chevrolet</td>\n",
              "      <td>Impala Limited</td>\n",
              "      <td>2014</td>\n",
              "      <td>300.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>AUTOMATIC</td>\n",
              "      <td>Sedan</td>\n",
              "      <td>30</td>\n",
              "      <td>18</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Buick</td>\n",
              "      <td>Regal</td>\n",
              "      <td>2016</td>\n",
              "      <td>259.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>AUTOMATIC</td>\n",
              "      <td>Sedan</td>\n",
              "      <td>27</td>\n",
              "      <td>19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Volkswagen</td>\n",
              "      <td>CC</td>\n",
              "      <td>2015</td>\n",
              "      <td>280.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>AUTOMATED_MANUAL</td>\n",
              "      <td>Sedan</td>\n",
              "      <td>25</td>\n",
              "      <td>17</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Dodge</td>\n",
              "      <td>Grand Caravan</td>\n",
              "      <td>2016</td>\n",
              "      <td>283.0</td>\n",
              "      <td>6.0</td>\n",
              "      <td>AUTOMATIC</td>\n",
              "      <td>Passenger Minivan</td>\n",
              "      <td>25</td>\n",
              "      <td>17</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-8acd4b10-2874-4577-b197-2e5882052d07')\"\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-8acd4b10-2874-4577-b197-2e5882052d07 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-8acd4b10-2874-4577-b197-2e5882052d07');\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-11e9e9e6-1f0d-4582-8cd3-a5bca932aa5e\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-11e9e9e6-1f0d-4582-8cd3-a5bca932aa5e')\"\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-11e9e9e6-1f0d-4582-8cd3-a5bca932aa5e button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 100
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Question 1: ROC AUC feature importance\n",
        "\n",
        "ROC AUC could also be used to evaluate feature importance of numerical variables.\n",
        "\n",
        "Let's do that\n",
        "\n",
        "* For each numerical variable, use it as score and compute AUC with the above_average variable\n",
        "* Use the training dataset for that\n",
        "\n",
        "If your AUC is < 0.5, invert this variable by putting \"-\" in front\n",
        "\n",
        "(e.g. -df_train['engine_hp'])\n",
        "\n",
        "AUC can go below 0.5 if the variable is negatively correlated with the target varialble. You can change the direction of the correlation by negating this variable - then negative correlation becomes positive."
      ],
      "metadata": {
        "id": "_P_BRPHvq6DX"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Assuming df_train is your training DataFrame and 'above_average' is the target variable y_train\n",
        "\n",
        "# List to store feature importance results\n",
        "feature_importance = []\n",
        "\n",
        "# Iterate through numerical columns\n",
        "for column in df_train.select_dtypes(include=['number']).columns:\n",
        "    auc_score = roc_auc_score(y_train, df_train[column])\n",
        "\n",
        "    # Check if AUC is less than 0.5, invert the variable if needed\n",
        "    if auc_score < 0.5:\n",
        "        inverted_column = '-' + column\n",
        "        auc_score = roc_auc_score(y_train, -df_train[column])\n",
        "        feature_importance.append((inverted_column, auc_score))\n",
        "    else:\n",
        "        feature_importance.append((column, auc_score))\n",
        "\n",
        "# Sort feature importance by AUC score in descending order\n",
        "feature_importance.sort(key=lambda x: x[1], reverse=True)\n",
        "\n",
        "# Print feature importance results\n",
        "for feature, auc_score in feature_importance:\n",
        "    print(f\"Variable: {feature}, ROC AUC Score: {auc_score}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Tqp8Cmcfpxzt",
        "outputId": "1ad54d2b-078a-4d30-bc31-2bb146627782"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Variable: engine_hp, ROC AUC Score: 0.9171031265539011\n",
            "Variable: engine_cylinders, ROC AUC Score: 0.766116490165669\n",
            "Variable: year, ROC AUC Score: 0.687551492804683\n",
            "Variable: -city_mpg, ROC AUC Score: 0.6734244643245233\n",
            "Variable: -highway_mpg, ROC AUC Score: 0.6330587871772013\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Which numerical variable (among the following 4) has the highest AUC?\n",
        "\n",
        "* engine_hp\n",
        "* engine_cylinders\n",
        "* highway_mpg\n",
        "* city_mpg\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "Answer Q.1\n",
        "\n",
        "---\n",
        "**engine_hp**\n",
        "\n"
      ],
      "metadata": {
        "id": "Ub8CZWeh281V"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Question 2: Training the model\n",
        "\n",
        "Apply one-hot-encoding using DictVectorizer and train the logistic regression with these parameters:\n",
        "\n",
        "*LogisticRegression(solver='liblinear', C=1.0, max_iter=1000)*"
      ],
      "metadata": {
        "id": "EPsJu05j36oe"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def one_hot_encoding(train_df, val_df, test_df=\"\"):\n",
        "  train_dicts = train_df.to_dict(orient=\"records\")\n",
        "  val_dicts = val_df.to_dict(orient=\"records\")\n",
        "  if len(test_df)>0:\n",
        "    test_dicts = test_df.to_dict(orient=\"records\")\n",
        "\n",
        "  dv = DictVectorizer(sparse=False)\n",
        "  X_train = dv.fit_transform(train_dicts)\n",
        "  X_val = dv.transform(val_dicts)\n",
        "  if len(test_df)>0:\n",
        "    X_test = dv.transform(test_dicts)\n",
        "    return X_train, X_val, X_test\n",
        "  else:\n",
        "    return X_train, X_val\n"
      ],
      "metadata": {
        "id": "JcuJySB122wB"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def train_logistic_model(X_train, X_val, X_test, y_train, y_val, y_test, solver=\"liblinear\", C=1.0, max_iter=1000, random_state=1, evaluation_on=\"valid\"):\n",
        "  # Initialize and train the logistic regression model\n",
        "  model = LogisticRegression(solver=solver, C=C, max_iter=max_iter, random_state=random_state)\n",
        "  model.fit(X_train, y_train)\n",
        "\n",
        "  def get_accuracy_score(X, y):\n",
        "    # Predict on the validation set\n",
        "    predictions = model.predict(X)\n",
        "    predict_probabilities = model.predict_proba(X)\n",
        "\n",
        "    # Calculate accuracy\n",
        "    accuracy = accuracy_score(y, predictions)\n",
        "    return accuracy, predictions, predict_probabilities\n",
        "  if evaluation_on == \"valid\":\n",
        "    return get_accuracy_score(X_val, y_val)\n",
        "  elif evaluation_on == \"test\":\n",
        "    return get_accuracy_score(X_test, y_test)"
      ],
      "metadata": {
        "id": "mAg6T_kP5iW2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_train, X_val, X_test = one_hot_encoding(df_train, df_val, df_test)"
      ],
      "metadata": {
        "id": "T9O1tnEl6RFA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "accuracy, predictions, predict_probabilities = train_logistic_model(X_train, X_val, X_test, y_train, y_val, y_test, solver=\"liblinear\", C=1.0, max_iter=1000, evaluation_on=\"valid\")"
      ],
      "metadata": {
        "id": "tawkvDeQ9kpy"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"Accuracy score: \", round(accuracy, 3))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mVDMfW5J6D6M",
        "outputId": "ba06da78-99af-47ee-efaf-08caaa560947"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy score:  0.935\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "What's the AUC of this model on the validation dataset? (round to 3 digits)\n",
        "\n",
        "* 0.678\n",
        "* 0.779\n",
        "* 0.878\n",
        "* 0.979\n",
        "\n",
        "\n",
        "---\n",
        "Answer Q.2\n",
        "\n",
        "---\n",
        "**0.935 (~0.979)**\n",
        "\n"
      ],
      "metadata": {
        "id": "nt6O_AKc7VTC"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Question 3: Precision and Recall\n",
        "\n",
        "Now let's compute precision and recall for our model.\n",
        "\n",
        "* Evaluate the model on all thresholds from 0.0 to 1.0 with step 0.01\n",
        "* For each threshold, compute precision and recall\n",
        "* Plot them"
      ],
      "metadata": {
        "id": "dnyb9jYu8OrK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Initialize lists to store precision and recall values\n",
        "precision_scores = []\n",
        "recall_scores = []\n",
        "thresholds = np.arange(0, 1.01, 0.01)\n",
        "\n",
        "# Compute precision and recall at different thresholds\n",
        "for threshold in thresholds:\n",
        "    predictions = (predict_probabilities >= threshold).astype(int)[:, 1]  # Convert probabilities to binary predictions\n",
        "    precision = precision_score(y_val, predictions)\n",
        "    recall = recall_score(y_val, predictions)\n",
        "    precision_scores.append(precision)\n",
        "    recall_scores.append(recall)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZETXoHV97Gzy",
        "outputId": "e8255fc3-f70b-48a8-f4b3-b2c0236c56d7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1344: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot precision-recall curve\n",
        "plt.figure(figsize=(8, 6))\n",
        "plt.plot(thresholds, precision_scores, label='Precision')\n",
        "plt.plot(thresholds, recall_scores, label='Recall')\n",
        "plt.xlabel('Threshold')\n",
        "plt.ylabel('Score')\n",
        "plt.title('Precision-Recall Curve')\n",
        "plt.legend()\n",
        "plt.grid(True)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "id": "htvEWVrv-gv-",
        "outputId": "491f14ea-b753-43f2-e37e-cc984d11881a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "At which threshold precision and recall curves intersect?\n",
        "\n",
        "* 0.28\n",
        "* 0.48\n",
        "* 0.68\n",
        "* 0.88\n",
        "\n",
        "\n",
        "---\n",
        "Answer Q.3\n",
        "\n",
        "---\n",
        "**0.48**\n",
        "\n"
      ],
      "metadata": {
        "id": "mOA8UOW88XD3"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Question 4: F1 score\n",
        "\n",
        "Precision and recall are conflicting - when one grows, the other goes down. That's why they are often combined into the F1 score - a metrics that takes into account both\n",
        "\n",
        "This is the formula for computing F1:\n",
        "\n",
        "$$F_1 = 2 \\cdot \\cfrac{P \\cdot R}{P + R}$$\n",
        "\n",
        "Where $P$ is precision and $R$ is recall.\n",
        "\n",
        "Let's compute F1 for all thresholds from 0.0 to 1.0 with increment 0.01"
      ],
      "metadata": {
        "id": "qSJVvgtB_V1v"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "accuracy, predictions, predict_probabilities = train_logistic_model(X_train, X_val, X_test, y_train, y_val, y_test, solver=\"liblinear\", C=1.0, max_iter=1000, evaluation_on=\"valid\")"
      ],
      "metadata": {
        "id": "5BMlx3zBAEaS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Initialize a list to store F1 scores\n",
        "f1_scores = []\n",
        "thresholds = np.arange(0, 1.01, 0.01)\n",
        "\n",
        "# Compute F1 score at different thresholds\n",
        "for threshold in thresholds:\n",
        "    predictions = (predict_probabilities >= threshold).astype(int)[:, 1]  # Convert probabilities to binary predictions\n",
        "    f1 = f1_score(y_val, predictions)\n",
        "    f1_scores.append(f1)\n",
        "\n",
        "# Find the threshold with the highest F1 score\n",
        "best_f1_threshold = thresholds[np.argmax(f1_scores)]\n",
        "best_f1_score = max(f1_scores)\n",
        "\n",
        "print(\"Threshold with the highest F1 score:\", best_f1_threshold)\n",
        "print(\"Best F1 score:\", best_f1_score)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k4Q1h2pl8Zq7",
        "outputId": "9bedceb2-3056-419c-e597-bb989de4d58a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Threshold with the highest F1 score: 0.49\n",
            "Best F1 score: 0.8852963818321786\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "At which threshold F1 is maximal?\n",
        "\n",
        "* 0.12\n",
        "* 0.32\n",
        "* 0.52\n",
        "* 0.72\n",
        "\n",
        "\n",
        "---\n",
        "Answer Q.4\n",
        "\n",
        "---\n",
        "\n",
        "**0.49 (~0.52)**\n"
      ],
      "metadata": {
        "id": "mLlHF9MoAZn6"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Question 5: 5-Fold CV\n",
        "\n",
        "Use the KFold class from Scikit-Learn to evaluate our model on 5 different folds:\n",
        "\n",
        "***KFold(n_splits=5, shuffle=True, random_state=1)***\n",
        "\n",
        "* Iterate over different folds of df_full_train\n",
        "* Split the data into train and validation\n",
        "* Train the model on train with these parameters: ***LogisticRegression(solver='liblinear', C=1.0, max_iter=1000)***\n",
        "* Use AUC to evaluate the model on validation"
      ],
      "metadata": {
        "id": "pyaOAz3vAsVR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df_full_train, df_test = train_test_split(df, test_size = 0.2, random_state = 1)\n",
        "df_train = df_full_train.reset_index(drop=True)\n",
        "df_test = df_test.reset_index(drop=True)\n",
        "\n",
        "\n",
        "y_full_train = df_train[\"above_average\"].values\n",
        "y_full_test = df_test[\"above_average\"].values"
      ],
      "metadata": {
        "id": "4dzWhpjWGpJ6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "del df_train[\"price\"]\n",
        "del df_train[\"above_average\"]\n",
        "\n",
        "del df_test[\"price\"]\n",
        "del df_test[\"above_average\"]"
      ],
      "metadata": {
        "id": "-WLeZ0ffHxbl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_full_train, X_full_val = one_hot_encoding(df_train, df_test)"
      ],
      "metadata": {
        "id": "_BgRNbDrJENE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def perform_kfold(n_splits, shuffle, random_state,solver, C, max_iter, df_full_train):\n",
        "  # Initialize KFold with 5 folds, shuffling the data and setting random_state for reproducibility\n",
        "  kf = KFold(n_splits=n_splits, shuffle=shuffle, random_state=random_state)\n",
        "\n",
        "  # Initialize a list to store AUC scores for each fold\n",
        "  auc_scores = []\n",
        "  auc_ls = []\n",
        "  # Iterate over different folds\n",
        "  for train_idx, val_idx in kf.split(df_full_train):\n",
        "      # Split the data into train and validation sets\n",
        "      X_train, X_val = X_full_train[train_idx], X_full_train[val_idx]\n",
        "      y_train, y_val = y_full_train[train_idx], y_full_train[val_idx]\n",
        "\n",
        "      # Initialize and train the Logistic Regression model\n",
        "      model = LogisticRegression(solver=solver, C=C, max_iter=max_iter)\n",
        "      model.fit(X_train, y_train)\n",
        "\n",
        "      # Make predictions on the validation set\n",
        "      val_predictions = model.predict_proba(X_val)[:, 1]  # Predict probabilities of class 1\n",
        "\n",
        "      # Calculate AUC score for this fold\n",
        "      auc = roc_auc_score(y_val, val_predictions)\n",
        "      auc_scores.append(auc)\n",
        "      # auc_ls.append(accuracy_score(y_val, val_predictions))\n",
        "  # Calculate and print the standard deviation of AUC scores\n",
        "  std_dev = np.std(auc_scores)\n",
        "  mean_auc = np.average(auc_scores)\n",
        "\n",
        "  print(f\"C value: {C}\")\n",
        "  print(f\"Mean of AUC Scores: {round(mean_auc,3)}\")\n",
        "  print(f\"Standard Deviation of AUC Scores: {round(std_dev,3)}\")\n",
        "  print()"
      ],
      "metadata": {
        "id": "FemaFxPV_vPy"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "perform_kfold(5, True, 1, \"liblinear\", 1.0, 1000, df_full_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pzDmbhw1dTgz",
        "outputId": "d43c9050-968f-4739-addb-df3759c3fb4a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "C value: 1.0\n",
            "Mean of AUC Scores: 0.979\n",
            "Standard Deviation of AUC Scores: 0.003\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "How large is standard devidation of the scores across different folds?\n",
        "\n",
        "* 0.003\n",
        "* 0.030\n",
        "* 0.090\n",
        "* 0.140\n",
        "\n",
        "---\n",
        "Answer Q.5\n",
        "\n",
        "---\n",
        "**0.003**\n",
        "\n"
      ],
      "metadata": {
        "id": "FbOpGzItKqgG"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Question 6: Hyperparemeter Tuning\n",
        "\n",
        "Now let's use 5-Fold cross-validation to find the best parameter C\n",
        "\n",
        "* Iterate over the following C values: **[0.01, 0.1, 0.5, 10]**\n",
        "* Initialize KFold with the same parameters as previously\n",
        "* Use these parametes for the model: **LogisticRegression(solver='liblinear', C=C, max_iter=1000)**\n",
        "* Compute the mean score as well as the std (round the mean and std to 3 decimal digits)"
      ],
      "metadata": {
        "id": "kAnKKnG9cpz5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "for C in [0.01, 0.1, 0.5, 10]:\n",
        "  perform_kfold(5, True, 1, \"liblinear\", C, 1000, df_full_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gEu13WSHJvAV",
        "outputId": "e880d17b-d4d0-438e-cd4b-a52113def0a0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "C value: 0.01\n",
            "Mean of AUC Scores: 0.952\n",
            "Standard Deviation of AUC Scores: 0.002\n",
            "\n",
            "C value: 0.1\n",
            "Mean of AUC Scores: 0.972\n",
            "Standard Deviation of AUC Scores: 0.002\n",
            "\n",
            "C value: 0.5\n",
            "Mean of AUC Scores: 0.981\n",
            "Standard Deviation of AUC Scores: 0.002\n",
            "\n",
            "C value: 10\n",
            "Mean of AUC Scores: 0.982\n",
            "Standard Deviation of AUC Scores: 0.003\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Which C leads to the best mean score?\n",
        "\n",
        "* 0.01\n",
        "* 0.1\n",
        "* 0.5\n",
        "* 10\n",
        "\n",
        "If you have ties, select the score with the lowest std. If you still have ties, select the smallest C.\n",
        "\n",
        "\n",
        "---\n",
        "\n",
        "Answer Q.6\n",
        "\n",
        "---\n",
        "**10**\n"
      ],
      "metadata": {
        "id": "eanHcNXjfjvz"
      }
    }
  ]
}