classify_song_genres.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyMsrin2KGni5iARsdVeIEbD",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Parissai/ML-DS-Playground/blob/main/classify_song_genres.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Classify Song Genres from Audio Data\n",
        "\n",
        "Over the past few years, streaming services with huge catalogs have become the primary means through which most people listen to their favorite music. But at the same time, the sheer amount of music on offer can mean users might be a bit overwhelmed when trying to look for newer music that suits their tastes.\n",
        "\n",
        "For this reason, streaming services have looked into means of categorizing music to allow for personalized recommendations. One method involves direct analysis of the raw audio information in a given song, scoring the raw data on a variety of metrics. This is the examination of data compiled by a research group known as The Echo Nest. The goal is to look through this dataset and classify songs as being either 'Hip-Hop' or 'Rock'. In doing so, learn how to clean the data, do some exploratory data visualization, and use feature reduction towards the goal of feeding the data through some simple machine learning algorithms, such as decision trees and logistic regression.\n",
        "\n",
        "There is another dataset that has musical features of each track such as danceability and acousticness on a scale from -1 to 1. These exist in two different files, which are in different formats - CSV and JSON. While CSV is a popular file format for denoting tabular data, JSON is another common file format in which databases often return the results of a given query."
      ],
      "metadata": {
        "id": "X1sptK_bVwMi"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pnon-BvpUJ2z",
        "outputId": "7ea9bf32-6079-4c6a-b26c-60a0a420087d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "\n",
        "# Read in track metadata with genre labels\n",
        "tracks = pd.read_csv('/content/drive/MyDrive/data/fma-rock-vs-hiphop.csv')\n",
        "tracks.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 533
        },
        "id": "TC8VP7vKVo_I",
        "outputId": "e5f83b57-e517-4376-9e3b-a951059740bd"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   track_id  bit_rate  comments        composer         date_created  \\\n",
              "0       135    256000         1             NaN  2008-11-26 01:43:26   \n",
              "1       136    256000         1             NaN  2008-11-26 01:43:35   \n",
              "2       151    192000         0             NaN  2008-11-26 01:44:55   \n",
              "3       152    192000         0             NaN  2008-11-26 01:44:58   \n",
              "4       153    256000         0  Arc and Sender  2008-11-26 01:45:00   \n",
              "\n",
              "         date_recorded  duration  favorites genre_top    genres  ...  \\\n",
              "0  2008-11-26 00:00:00       837          0      Rock  [45, 58]  ...   \n",
              "1  2008-11-26 00:00:00       509          0      Rock  [45, 58]  ...   \n",
              "2                  NaN       192          0      Rock      [25]  ...   \n",
              "3                  NaN       193          0      Rock      [25]  ...   \n",
              "4  2008-11-26 00:00:00       405          5      Rock      [26]  ...   \n",
              "\n",
              "  information interest  language_code  \\\n",
              "0         NaN     2484             en   \n",
              "1         NaN     1948             en   \n",
              "2         NaN      701             en   \n",
              "3         NaN      637             en   \n",
              "4         NaN      354             en   \n",
              "\n",
              "                                             license listens  lyricist number  \\\n",
              "0  Attribution-NonCommercial-ShareAlike 3.0 Inter...    1832       NaN      0   \n",
              "1  Attribution-NonCommercial-ShareAlike 3.0 Inter...    1498       NaN      0   \n",
              "2  Attribution-NonCommercial-ShareAlike 3.0 Inter...     148       NaN      4   \n",
              "3  Attribution-NonCommercial-ShareAlike 3.0 Inter...      98       NaN     11   \n",
              "4  Attribution-NonCommercial-NoDerivatives (aka M...     424       NaN      2   \n",
              "\n",
              "   publisher tags                       title  \n",
              "0        NaN   []                Father's Day  \n",
              "1        NaN   []  Peel Back The Mountain Sky  \n",
              "2        NaN   []                 Untitled 04  \n",
              "3        NaN   []                 Untitled 11  \n",
              "4        NaN   []          Hundred-Year Flood  \n",
              "\n",
              "[5 rows x 21 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1586c061-e1c6-45cf-9f0c-8153cdc58910\" 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>track_id</th>\n",
              "      <th>bit_rate</th>\n",
              "      <th>comments</th>\n",
              "      <th>composer</th>\n",
              "      <th>date_created</th>\n",
              "      <th>date_recorded</th>\n",
              "      <th>duration</th>\n",
              "      <th>favorites</th>\n",
              "      <th>genre_top</th>\n",
              "      <th>genres</th>\n",
              "      <th>...</th>\n",
              "      <th>information</th>\n",
              "      <th>interest</th>\n",
              "      <th>language_code</th>\n",
              "      <th>license</th>\n",
              "      <th>listens</th>\n",
              "      <th>lyricist</th>\n",
              "      <th>number</th>\n",
              "      <th>publisher</th>\n",
              "      <th>tags</th>\n",
              "      <th>title</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>135</td>\n",
              "      <td>256000</td>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2008-11-26 01:43:26</td>\n",
              "      <td>2008-11-26 00:00:00</td>\n",
              "      <td>837</td>\n",
              "      <td>0</td>\n",
              "      <td>Rock</td>\n",
              "      <td>[45, 58]</td>\n",
              "      <td>...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2484</td>\n",
              "      <td>en</td>\n",
              "      <td>Attribution-NonCommercial-ShareAlike 3.0 Inter...</td>\n",
              "      <td>1832</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[]</td>\n",
              "      <td>Father's Day</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>136</td>\n",
              "      <td>256000</td>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2008-11-26 01:43:35</td>\n",
              "      <td>2008-11-26 00:00:00</td>\n",
              "      <td>509</td>\n",
              "      <td>0</td>\n",
              "      <td>Rock</td>\n",
              "      <td>[45, 58]</td>\n",
              "      <td>...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>1948</td>\n",
              "      <td>en</td>\n",
              "      <td>Attribution-NonCommercial-ShareAlike 3.0 Inter...</td>\n",
              "      <td>1498</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[]</td>\n",
              "      <td>Peel Back The Mountain Sky</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>151</td>\n",
              "      <td>192000</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2008-11-26 01:44:55</td>\n",
              "      <td>NaN</td>\n",
              "      <td>192</td>\n",
              "      <td>0</td>\n",
              "      <td>Rock</td>\n",
              "      <td>[25]</td>\n",
              "      <td>...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>701</td>\n",
              "      <td>en</td>\n",
              "      <td>Attribution-NonCommercial-ShareAlike 3.0 Inter...</td>\n",
              "      <td>148</td>\n",
              "      <td>NaN</td>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[]</td>\n",
              "      <td>Untitled 04</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>152</td>\n",
              "      <td>192000</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2008-11-26 01:44:58</td>\n",
              "      <td>NaN</td>\n",
              "      <td>193</td>\n",
              "      <td>0</td>\n",
              "      <td>Rock</td>\n",
              "      <td>[25]</td>\n",
              "      <td>...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>637</td>\n",
              "      <td>en</td>\n",
              "      <td>Attribution-NonCommercial-ShareAlike 3.0 Inter...</td>\n",
              "      <td>98</td>\n",
              "      <td>NaN</td>\n",
              "      <td>11</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[]</td>\n",
              "      <td>Untitled 11</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>153</td>\n",
              "      <td>256000</td>\n",
              "      <td>0</td>\n",
              "      <td>Arc and Sender</td>\n",
              "      <td>2008-11-26 01:45:00</td>\n",
              "      <td>2008-11-26 00:00:00</td>\n",
              "      <td>405</td>\n",
              "      <td>5</td>\n",
              "      <td>Rock</td>\n",
              "      <td>[26]</td>\n",
              "      <td>...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>354</td>\n",
              "      <td>en</td>\n",
              "      <td>Attribution-NonCommercial-NoDerivatives (aka M...</td>\n",
              "      <td>424</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[]</td>\n",
              "      <td>Hundred-Year Flood</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 21 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1586c061-e1c6-45cf-9f0c-8153cdc58910')\"\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-1586c061-e1c6-45cf-9f0c-8153cdc58910 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-1586c061-e1c6-45cf-9f0c-8153cdc58910');\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-a6094473-bfc8-4fb6-9086-000995385a59\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-a6094473-bfc8-4fb6-9086-000995385a59')\"\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-a6094473-bfc8-4fb6-9086-000995385a59 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": "tracks"
            }
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Read in track metrics with the features\n",
        "echonest_metrics = pd.read_json('/content/drive/MyDrive/data/echonest-metrics.json', precise_float=True)\n",
        "echonest_metrics.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "3GlK_B7NWCYL",
        "outputId": "4ef3def7-2391-4030-db42-56903256bfb1"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   track_id  acousticness  danceability    energy  instrumentalness  liveness  \\\n",
              "0         2      0.416675      0.675894  0.634476          0.010628  0.177647   \n",
              "1         3      0.374408      0.528643  0.817461          0.001851  0.105880   \n",
              "2         5      0.043567      0.745566  0.701470          0.000697  0.373143   \n",
              "3        10      0.951670      0.658179  0.924525          0.965427  0.115474   \n",
              "4       134      0.452217      0.513238  0.560410          0.019443  0.096567   \n",
              "\n",
              "   speechiness    tempo   valence  \n",
              "0     0.159310  165.922  0.576661  \n",
              "1     0.461818  126.957  0.269240  \n",
              "2     0.124595  100.260  0.621661  \n",
              "3     0.032985  111.562  0.963590  \n",
              "4     0.525519  114.290  0.894072  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-8780bdac-4821-4f75-80a6-a5b1aaa2ed44\" 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>track_id</th>\n",
              "      <th>acousticness</th>\n",
              "      <th>danceability</th>\n",
              "      <th>energy</th>\n",
              "      <th>instrumentalness</th>\n",
              "      <th>liveness</th>\n",
              "      <th>speechiness</th>\n",
              "      <th>tempo</th>\n",
              "      <th>valence</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>2</td>\n",
              "      <td>0.416675</td>\n",
              "      <td>0.675894</td>\n",
              "      <td>0.634476</td>\n",
              "      <td>0.010628</td>\n",
              "      <td>0.177647</td>\n",
              "      <td>0.159310</td>\n",
              "      <td>165.922</td>\n",
              "      <td>0.576661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>3</td>\n",
              "      <td>0.374408</td>\n",
              "      <td>0.528643</td>\n",
              "      <td>0.817461</td>\n",
              "      <td>0.001851</td>\n",
              "      <td>0.105880</td>\n",
              "      <td>0.461818</td>\n",
              "      <td>126.957</td>\n",
              "      <td>0.269240</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>5</td>\n",
              "      <td>0.043567</td>\n",
              "      <td>0.745566</td>\n",
              "      <td>0.701470</td>\n",
              "      <td>0.000697</td>\n",
              "      <td>0.373143</td>\n",
              "      <td>0.124595</td>\n",
              "      <td>100.260</td>\n",
              "      <td>0.621661</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>10</td>\n",
              "      <td>0.951670</td>\n",
              "      <td>0.658179</td>\n",
              "      <td>0.924525</td>\n",
              "      <td>0.965427</td>\n",
              "      <td>0.115474</td>\n",
              "      <td>0.032985</td>\n",
              "      <td>111.562</td>\n",
              "      <td>0.963590</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>134</td>\n",
              "      <td>0.452217</td>\n",
              "      <td>0.513238</td>\n",
              "      <td>0.560410</td>\n",
              "      <td>0.019443</td>\n",
              "      <td>0.096567</td>\n",
              "      <td>0.525519</td>\n",
              "      <td>114.290</td>\n",
              "      <td>0.894072</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-8780bdac-4821-4f75-80a6-a5b1aaa2ed44')\"\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-8780bdac-4821-4f75-80a6-a5b1aaa2ed44 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-8780bdac-4821-4f75-80a6-a5b1aaa2ed44');\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-1f66e8cb-ddab-47a1-99b7-477adbf35101\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-1f66e8cb-ddab-47a1-99b7-477adbf35101')\"\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-1f66e8cb-ddab-47a1-99b7-477adbf35101 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": "echonest_metrics",
              "summary": "{\n  \"name\": \"echonest_metrics\",\n  \"rows\": 13129,\n  \"fields\": [\n    {\n      \"column\": \"track_id\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 28950,\n        \"min\": 2,\n        \"max\": 124911,\n        \"num_unique_values\": 13129,\n        \"samples\": [\n          51206,\n          15093,\n          14295\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"acousticness\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.3837185743719484,\n        \"min\": 9.035e-07,\n        \"max\": 0.9957964501,\n        \"num_unique_values\": 12905,\n        \"samples\": [\n          0.0996802338,\n          0.349116968,\n          0.6480686921\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"danceability\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.19014800698421913,\n        \"min\": 0.051307487,\n        \"max\": 0.9686446617,\n        \"num_unique_values\": 13129,\n        \"samples\": [\n          0.5986586065,\n          0.939546126,\n          0.3100973491\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"energy\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.2780488871983015,\n        \"min\": 2.01659e-05,\n        \"max\": 0.9999637108,\n        \"num_unique_values\": 13129,\n        \"samples\": [\n          0.766625893,\n          0.5544498465,\n          0.9688740304\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"instrumentalness\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.3614301125491524,\n        \"min\": 0.0,\n        \"max\": 0.9980162209,\n        \"num_unique_values\": 12969,\n        \"samples\": [\n          0.8494615895,\n          0.9713835471,\n          0.8360895325\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"liveness\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.15805059331239463,\n        \"min\": 0.0252973415,\n        \"max\": 0.9803300006,\n        \"num_unique_values\": 13128,\n        \"samples\": [\n          0.2125251147,\n          0.0987544086,\n          0.0942695941\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"speechiness\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.13738050717683195,\n        \"min\": 0.0223236675,\n        \"max\": 0.9661774074,\n        \"num_unique_values\": 13129,\n        \"samples\": [\n          0.0327134194,\n          0.0468909148,\n          0.205923899\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"tempo\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 35.01513746122828,\n        \"min\": 12.753,\n        \"max\": 251.072,\n        \"num_unique_values\": 11902,\n        \"samples\": [\n          84.425,\n          73.914,\n          97.26\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"valence\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.2760283031544492,\n        \"min\": 1e-05,\n        \"max\": 0.99999,\n        \"num_unique_values\": 13127,\n        \"samples\": [\n          0.0522836079,\n          0.9758199984,\n          0.224375033\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Merge the relevant columns of tracks and echonest_metrics\n",
        "echo_tracks = pd.merge(echonest_metrics, tracks[['track_id', 'genre_top']], on='track_id')\n",
        "\n",
        "# Inspect the resultant dataframe\n",
        "echo_tracks.info()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MpqvfQwLWJlX",
        "outputId": "94d8f2de-1501-4f85-8c1a-e9a638d329cb"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 4802 entries, 0 to 4801\n",
            "Data columns (total 10 columns):\n",
            " #   Column            Non-Null Count  Dtype  \n",
            "---  ------            --------------  -----  \n",
            " 0   track_id          4802 non-null   int64  \n",
            " 1   acousticness      4802 non-null   float64\n",
            " 2   danceability      4802 non-null   float64\n",
            " 3   energy            4802 non-null   float64\n",
            " 4   instrumentalness  4802 non-null   float64\n",
            " 5   liveness          4802 non-null   float64\n",
            " 6   speechiness       4802 non-null   float64\n",
            " 7   tempo             4802 non-null   float64\n",
            " 8   valence           4802 non-null   float64\n",
            " 9   genre_top         4802 non-null   object \n",
            "dtypes: float64(8), int64(1), object(1)\n",
            "memory usage: 375.3+ KB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "1. Pairwise relationships between continuous variables\n",
        "\n",
        "  We typically want to avoid using variables that have strong correlations with each other -- hence avoiding feature redundancy -- for a few reasons:\n",
        "\n",
        "  - To keep the model simple and improve interpretability (with many features, we run the risk of overfitting).\n",
        "  - When our datasets are very large, using fewer features can drastically speed up our computation time.\n",
        "  - To get a sense of whether there are any strongly correlated features in our data, we will use built-in functions in the pandas package."
      ],
      "metadata": {
        "id": "-kkA3IQ5WVZV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a correlation matrix\n",
        "corr_metrics = echo_tracks.select_dtypes(include=['float', 'int']).corr()\n",
        "corr_metrics.style.background_gradient()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 332
        },
        "id": "WePEi9E4WPRd",
        "outputId": "6fdcdaeb-aad7-4b78-9a0d-30e25a343406"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<pandas.io.formats.style.Styler at 0x79c8d995bf40>"
            ],
            "text/html": [
              "<style type=\"text/css\">\n",
              "#T_1b454_row0_col0, #T_1b454_row1_col1, #T_1b454_row2_col2, #T_1b454_row3_col3, #T_1b454_row4_col4, #T_1b454_row5_col5, #T_1b454_row6_col6, #T_1b454_row7_col7, #T_1b454_row8_col8 {\n",
              "  background-color: #023858;\n",
              "  color: #f1f1f1;\n",
              "}\n",
              "#T_1b454_row0_col1, #T_1b454_row1_col0, #T_1b454_row1_col3, #T_1b454_row2_col5, #T_1b454_row2_col7, #T_1b454_row4_col2, #T_1b454_row4_col6, #T_1b454_row4_col8, #T_1b454_row6_col4 {\n",
              "  background-color: #fff7fb;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row0_col2 {\n",
              "  background-color: #d2d2e7;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row0_col3 {\n",
              "  background-color: #b5c4df;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row0_col4 {\n",
              "  background-color: #f5eef6;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row0_col5 {\n",
              "  background-color: #e9e5f1;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row0_col6, #T_1b454_row8_col3 {\n",
              "  background-color: #d1d2e6;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row0_col7, #T_1b454_row1_col7 {\n",
              "  background-color: #e1dfed;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row0_col8, #T_1b454_row3_col6 {\n",
              "  background-color: #dedcec;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row1_col2 {\n",
              "  background-color: #e0dded;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row1_col4, #T_1b454_row4_col1 {\n",
              "  background-color: #97b7d7;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row1_col5, #T_1b454_row2_col4 {\n",
              "  background-color: #f3edf5;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row1_col6, #T_1b454_row6_col1 {\n",
              "  background-color: #b8c6e0;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row1_col8 {\n",
              "  background-color: #e2dfee;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row2_col0, #T_1b454_row5_col0, #T_1b454_row5_col3 {\n",
              "  background-color: #bdc8e1;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row2_col1, #T_1b454_row6_col0, #T_1b454_row7_col0, #T_1b454_row7_col1 {\n",
              "  background-color: #d0d1e6;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row2_col3 {\n",
              "  background-color: #fbf3f9;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row2_col6 {\n",
              "  background-color: #80aed2;\n",
              "  color: #f1f1f1;\n",
              "}\n",
              "#T_1b454_row2_col8 {\n",
              "  background-color: #529bc7;\n",
              "  color: #f1f1f1;\n",
              "}\n",
              "#T_1b454_row3_col0, #T_1b454_row7_col3 {\n",
              "  background-color: #a7bddb;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row3_col1 {\n",
              "  background-color: #f5eff6;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row3_col2, #T_1b454_row7_col2 {\n",
              "  background-color: #fef6fa;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row3_col4 {\n",
              "  background-color: #c4cbe3;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row3_col5, #T_1b454_row5_col7 {\n",
              "  background-color: #dcdaeb;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row3_col7 {\n",
              "  background-color: #adc1dd;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row3_col8, #T_1b454_row4_col7 {\n",
              "  background-color: #d9d8ea;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row4_col0 {\n",
              "  background-color: #f4eef6;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row4_col3 {\n",
              "  background-color: #d2d3e7;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row4_col5 {\n",
              "  background-color: #fdf5fa;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row5_col1 {\n",
              "  background-color: #ced0e6;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row5_col2 {\n",
              "  background-color: #ede8f3;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row5_col4, #T_1b454_row6_col7 {\n",
              "  background-color: #dbdaeb;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row5_col6 {\n",
              "  background-color: #c0c9e2;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row5_col8 {\n",
              "  background-color: #e8e4f0;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row6_col2 {\n",
              "  background-color: #93b5d6;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row6_col3, #T_1b454_row6_col5 {\n",
              "  background-color: #eae6f1;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row6_col8 {\n",
              "  background-color: #bfc9e1;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row7_col4 {\n",
              "  background-color: #c5cce3;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row7_col5 {\n",
              "  background-color: #f0eaf4;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row7_col6 {\n",
              "  background-color: #c8cde4;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row7_col8 {\n",
              "  background-color: #d6d6e9;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row8_col0 {\n",
              "  background-color: #c6cce3;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row8_col1 {\n",
              "  background-color: #cdd0e5;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row8_col2 {\n",
              "  background-color: #4c99c5;\n",
              "  color: #f1f1f1;\n",
              "}\n",
              "#T_1b454_row8_col4 {\n",
              "  background-color: #efe9f3;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row8_col5 {\n",
              "  background-color: #f7f0f7;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row8_col6 {\n",
              "  background-color: #a5bddb;\n",
              "  color: #000000;\n",
              "}\n",
              "#T_1b454_row8_col7 {\n",
              "  background-color: #d3d4e7;\n",
              "  color: #000000;\n",
              "}\n",
              "</style>\n",
              "<table id=\"T_1b454\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr>\n",
              "      <th class=\"blank level0\" >&nbsp;</th>\n",
              "      <th id=\"T_1b454_level0_col0\" class=\"col_heading level0 col0\" >track_id</th>\n",
              "      <th id=\"T_1b454_level0_col1\" class=\"col_heading level0 col1\" >acousticness</th>\n",
              "      <th id=\"T_1b454_level0_col2\" class=\"col_heading level0 col2\" >danceability</th>\n",
              "      <th id=\"T_1b454_level0_col3\" class=\"col_heading level0 col3\" >energy</th>\n",
              "      <th id=\"T_1b454_level0_col4\" class=\"col_heading level0 col4\" >instrumentalness</th>\n",
              "      <th id=\"T_1b454_level0_col5\" class=\"col_heading level0 col5\" >liveness</th>\n",
              "      <th id=\"T_1b454_level0_col6\" class=\"col_heading level0 col6\" >speechiness</th>\n",
              "      <th id=\"T_1b454_level0_col7\" class=\"col_heading level0 col7\" >tempo</th>\n",
              "      <th id=\"T_1b454_level0_col8\" class=\"col_heading level0 col8\" >valence</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row0\" class=\"row_heading level0 row0\" >track_id</th>\n",
              "      <td id=\"T_1b454_row0_col0\" class=\"data row0 col0\" >1.000000</td>\n",
              "      <td id=\"T_1b454_row0_col1\" class=\"data row0 col1\" >-0.372282</td>\n",
              "      <td id=\"T_1b454_row0_col2\" class=\"data row0 col2\" >0.049454</td>\n",
              "      <td id=\"T_1b454_row0_col3\" class=\"data row0 col3\" >0.140703</td>\n",
              "      <td id=\"T_1b454_row0_col4\" class=\"data row0 col4\" >-0.275623</td>\n",
              "      <td id=\"T_1b454_row0_col5\" class=\"data row0 col5\" >0.048231</td>\n",
              "      <td id=\"T_1b454_row0_col6\" class=\"data row0 col6\" >-0.026995</td>\n",
              "      <td id=\"T_1b454_row0_col7\" class=\"data row0 col7\" >-0.025392</td>\n",
              "      <td id=\"T_1b454_row0_col8\" class=\"data row0 col8\" >0.010070</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row1\" class=\"row_heading level0 row1\" >acousticness</th>\n",
              "      <td id=\"T_1b454_row1_col0\" class=\"data row1 col0\" >-0.372282</td>\n",
              "      <td id=\"T_1b454_row1_col1\" class=\"data row1 col1\" >1.000000</td>\n",
              "      <td id=\"T_1b454_row1_col2\" class=\"data row1 col2\" >-0.028954</td>\n",
              "      <td id=\"T_1b454_row1_col3\" class=\"data row1 col3\" >-0.281619</td>\n",
              "      <td id=\"T_1b454_row1_col4\" class=\"data row1 col4\" >0.194780</td>\n",
              "      <td id=\"T_1b454_row1_col5\" class=\"data row1 col5\" >-0.019991</td>\n",
              "      <td id=\"T_1b454_row1_col6\" class=\"data row1 col6\" >0.072204</td>\n",
              "      <td id=\"T_1b454_row1_col7\" class=\"data row1 col7\" >-0.026310</td>\n",
              "      <td id=\"T_1b454_row1_col8\" class=\"data row1 col8\" >-0.013841</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row2\" class=\"row_heading level0 row2\" >danceability</th>\n",
              "      <td id=\"T_1b454_row2_col0\" class=\"data row2 col0\" >0.049454</td>\n",
              "      <td id=\"T_1b454_row2_col1\" class=\"data row2 col1\" >-0.028954</td>\n",
              "      <td id=\"T_1b454_row2_col2\" class=\"data row2 col2\" >1.000000</td>\n",
              "      <td id=\"T_1b454_row2_col3\" class=\"data row2 col3\" >-0.242032</td>\n",
              "      <td id=\"T_1b454_row2_col4\" class=\"data row2 col4\" >-0.255217</td>\n",
              "      <td id=\"T_1b454_row2_col5\" class=\"data row2 col5\" >-0.106584</td>\n",
              "      <td id=\"T_1b454_row2_col6\" class=\"data row2 col6\" >0.276206</td>\n",
              "      <td id=\"T_1b454_row2_col7\" class=\"data row2 col7\" >-0.242089</td>\n",
              "      <td id=\"T_1b454_row2_col8\" class=\"data row2 col8\" >0.473165</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row3\" class=\"row_heading level0 row3\" >energy</th>\n",
              "      <td id=\"T_1b454_row3_col0\" class=\"data row3 col0\" >0.140703</td>\n",
              "      <td id=\"T_1b454_row3_col1\" class=\"data row3 col1\" >-0.281619</td>\n",
              "      <td id=\"T_1b454_row3_col2\" class=\"data row3 col2\" >-0.242032</td>\n",
              "      <td id=\"T_1b454_row3_col3\" class=\"data row3 col3\" >1.000000</td>\n",
              "      <td id=\"T_1b454_row3_col4\" class=\"data row3 col4\" >0.028238</td>\n",
              "      <td id=\"T_1b454_row3_col5\" class=\"data row3 col5\" >0.113331</td>\n",
              "      <td id=\"T_1b454_row3_col6\" class=\"data row3 col6\" >-0.109983</td>\n",
              "      <td id=\"T_1b454_row3_col7\" class=\"data row3 col7\" >0.195227</td>\n",
              "      <td id=\"T_1b454_row3_col8\" class=\"data row3 col8\" >0.038603</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row4\" class=\"row_heading level0 row4\" >instrumentalness</th>\n",
              "      <td id=\"T_1b454_row4_col0\" class=\"data row4 col0\" >-0.275623</td>\n",
              "      <td id=\"T_1b454_row4_col1\" class=\"data row4 col1\" >0.194780</td>\n",
              "      <td id=\"T_1b454_row4_col2\" class=\"data row4 col2\" >-0.255217</td>\n",
              "      <td id=\"T_1b454_row4_col3\" class=\"data row4 col3\" >0.028238</td>\n",
              "      <td id=\"T_1b454_row4_col4\" class=\"data row4 col4\" >1.000000</td>\n",
              "      <td id=\"T_1b454_row4_col5\" class=\"data row4 col5\" >-0.091022</td>\n",
              "      <td id=\"T_1b454_row4_col6\" class=\"data row4 col6\" >-0.366762</td>\n",
              "      <td id=\"T_1b454_row4_col7\" class=\"data row4 col7\" >0.022215</td>\n",
              "      <td id=\"T_1b454_row4_col8\" class=\"data row4 col8\" >-0.219967</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row5\" class=\"row_heading level0 row5\" >liveness</th>\n",
              "      <td id=\"T_1b454_row5_col0\" class=\"data row5 col0\" >0.048231</td>\n",
              "      <td id=\"T_1b454_row5_col1\" class=\"data row5 col1\" >-0.019991</td>\n",
              "      <td id=\"T_1b454_row5_col2\" class=\"data row5 col2\" >-0.106584</td>\n",
              "      <td id=\"T_1b454_row5_col3\" class=\"data row5 col3\" >0.113331</td>\n",
              "      <td id=\"T_1b454_row5_col4\" class=\"data row5 col4\" >-0.091022</td>\n",
              "      <td id=\"T_1b454_row5_col5\" class=\"data row5 col5\" >1.000000</td>\n",
              "      <td id=\"T_1b454_row5_col6\" class=\"data row5 col6\" >0.041173</td>\n",
              "      <td id=\"T_1b454_row5_col7\" class=\"data row5 col7\" >0.002732</td>\n",
              "      <td id=\"T_1b454_row5_col8\" class=\"data row5 col8\" >-0.045093</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row6\" class=\"row_heading level0 row6\" >speechiness</th>\n",
              "      <td id=\"T_1b454_row6_col0\" class=\"data row6 col0\" >-0.026995</td>\n",
              "      <td id=\"T_1b454_row6_col1\" class=\"data row6 col1\" >0.072204</td>\n",
              "      <td id=\"T_1b454_row6_col2\" class=\"data row6 col2\" >0.276206</td>\n",
              "      <td id=\"T_1b454_row6_col3\" class=\"data row6 col3\" >-0.109983</td>\n",
              "      <td id=\"T_1b454_row6_col4\" class=\"data row6 col4\" >-0.366762</td>\n",
              "      <td id=\"T_1b454_row6_col5\" class=\"data row6 col5\" >0.041173</td>\n",
              "      <td id=\"T_1b454_row6_col6\" class=\"data row6 col6\" >1.000000</td>\n",
              "      <td id=\"T_1b454_row6_col7\" class=\"data row6 col7\" >0.008241</td>\n",
              "      <td id=\"T_1b454_row6_col8\" class=\"data row6 col8\" >0.149894</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row7\" class=\"row_heading level0 row7\" >tempo</th>\n",
              "      <td id=\"T_1b454_row7_col0\" class=\"data row7 col0\" >-0.025392</td>\n",
              "      <td id=\"T_1b454_row7_col1\" class=\"data row7 col1\" >-0.026310</td>\n",
              "      <td id=\"T_1b454_row7_col2\" class=\"data row7 col2\" >-0.242089</td>\n",
              "      <td id=\"T_1b454_row7_col3\" class=\"data row7 col3\" >0.195227</td>\n",
              "      <td id=\"T_1b454_row7_col4\" class=\"data row7 col4\" >0.022215</td>\n",
              "      <td id=\"T_1b454_row7_col5\" class=\"data row7 col5\" >0.002732</td>\n",
              "      <td id=\"T_1b454_row7_col6\" class=\"data row7 col6\" >0.008241</td>\n",
              "      <td id=\"T_1b454_row7_col7\" class=\"data row7 col7\" >1.000000</td>\n",
              "      <td id=\"T_1b454_row7_col8\" class=\"data row7 col8\" >0.052221</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th id=\"T_1b454_level0_row8\" class=\"row_heading level0 row8\" >valence</th>\n",
              "      <td id=\"T_1b454_row8_col0\" class=\"data row8 col0\" >0.010070</td>\n",
              "      <td id=\"T_1b454_row8_col1\" class=\"data row8 col1\" >-0.013841</td>\n",
              "      <td id=\"T_1b454_row8_col2\" class=\"data row8 col2\" >0.473165</td>\n",
              "      <td id=\"T_1b454_row8_col3\" class=\"data row8 col3\" >0.038603</td>\n",
              "      <td id=\"T_1b454_row8_col4\" class=\"data row8 col4\" >-0.219967</td>\n",
              "      <td id=\"T_1b454_row8_col5\" class=\"data row8 col5\" >-0.045093</td>\n",
              "      <td id=\"T_1b454_row8_col6\" class=\"data row8 col6\" >0.149894</td>\n",
              "      <td id=\"T_1b454_row8_col7\" class=\"data row8 col7\" >0.052221</td>\n",
              "      <td id=\"T_1b454_row8_col8\" class=\"data row8 col8\" >1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "2. Splitting our data\n",
        "\n",
        "  It can be particularly useful to simplify our models and use as few features as necessary to achieve the best result. Since we didn't find any particularly strong correlations between our features, we can now split our data into an array containing our features, and another containing the labels - the genre of the track.\n"
      ],
      "metadata": {
        "id": "s_0eClcHWU0l"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "# Create features\n",
        "features = echo_tracks.drop([\"genre_top\", \"track_id\"], axis=1).values\n",
        "\n",
        "# Create labels\n",
        "labels = echo_tracks[\"genre_top\"].values\n",
        "\n",
        "# Split our data\n",
        "train_features, test_features, train_labels, test_labels = train_test_split(features, labels, random_state=13)"
      ],
      "metadata": {
        "id": "Z1j0d1vuW3tE"
      },
      "execution_count": 31,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "3. Normalizing the feature data\n",
        "\n",
        "  Using a common approach to reduce the number of features: principal component analysis (PCA).\n",
        "\n",
        "  It is possible that the variance between genres can be explained by just a few features in the dataset. PCA rotates the data along the axis of highest variance, thus allowing us to determine the relative contribution of each feature of our data towards the variance between classes.\n",
        "\n",
        "  However, since PCA uses the absolute variance of a feature to rotate the data, a feature with a broader range of values will overpower and bias the algorithm relative to the other features. To avoid this, we must first normalize our train and test features. A common way is through standardization, such that all features have a mean = 0 and standard deviation = 1 (the resultant is a z-score)."
      ],
      "metadata": {
        "id": "Sw7oiwA2XBfT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "# Scale the features and set the values to a new variable\n",
        "scaler = StandardScaler()\n",
        "\n",
        "# Scale train_features and test_features\n",
        "scaled_train_features = scaler.fit_transform(train_features)\n",
        "scaled_test_features = scaler.transform(test_features)"
      ],
      "metadata": {
        "id": "FE-z9zQbW9h2"
      },
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "4. Principal Component Analysis(PCA)\n",
        "\n",
        "  Using scree-plots and cumulative explained ratio plots to find the number of components to use in further analyses.\n",
        "\n",
        "  Scree-plots display the number of components against the variance explained by each component, sorted in descending order of variance. Scree-plots help us get a better sense of which components explain a sufficient amount of variance in our data. When using scree plots, an 'elbow' (a steep drop from one data point to the next) in the plot is typically used to decide on an appropriate cutoff."
      ],
      "metadata": {
        "id": "vRn2NPnMXIqm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "from sklearn.decomposition import PCA\n",
        "\n",
        "# Get our explained variance ratios from PCA using all features\n",
        "pca = PCA()\n",
        "pca.fit(scaled_train_features)\n",
        "\n",
        "exp_variance = pca.explained_variance_\n",
        "\n",
        "# plot the explained variance using a barplot\n",
        "fig, ax = plt.subplots()\n",
        "ax.bar(range(pca.n_components_) , exp_variance)\n",
        "ax.set_xlabel('Principal Component #')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "id": "xjVFuEskXGIP",
        "outputId": "ac592a21-0038-4821-f439-cab0b1fbdfac"
      },
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'Principal Component #')"
            ]
          },
          "metadata": {},
          "execution_count": 33
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "5. Further visualization of PCA\n",
        "\n",
        "  Unfortunately, there does not appear to be a clear elbow in this scree plot, which means it is not straightforward to find the number of intrinsic dimensions using this method.\n",
        "\n",
        "  So instead, we can also look at the cumulative explained variance plot to determine how many features are required to explain about 85% of the variance."
      ],
      "metadata": {
        "id": "lt5_3Xt6XSrB"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "\n",
        "cumulative_explained_variance = np.cumsum(pca.explained_variance_ratio_)\n",
        "\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.plot(np.arange(1, len(cumulative_explained_variance) + 1), cumulative_explained_variance, 'o-', linewidth=2, color='green')\n",
        "plt.title('Cumulative Explained Variance')\n",
        "plt.xlabel('Principal Component')\n",
        "plt.ylabel('Cumulative Explained Variance Ratio')\n",
        "plt.axhline(y=0.85, color='r', linestyle='--')  # Add a horizontal line for 85% threshold\n",
        "plt.grid(True)\n",
        "plt.show()\n",
        "\n",
        "n_components = np.argmax(cumulative_explained_variance >= 0.85) + 1\n",
        "print(f\"Number of components required to explain 85% of the variance: {n_components}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 581
        },
        "id": "Nz9XFDAGXPE6",
        "outputId": "65c0d894-ba03-47b4-8d1f-a4abf5e2a944"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x600 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of components required to explain 85% of the variance: 6\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "6. Projecting on to our features\n",
        "\n",
        "  From the plot, 6 features can explain 85% of the variance. Therefore, performing PCA and reduce the dimensionality of train and test features with 6 features."
      ],
      "metadata": {
        "id": "AssSSJiVXZdA"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Perform PCA with the chosen number of components and project data onto components\n",
        "pca = PCA(n_components)\n",
        "\n",
        "# Fit and transform the scaled training features using pca\n",
        "train_pca = pca.fit_transform(scaled_train_features)\n",
        "\n",
        "# Fit and transform the scaled test features using pca\n",
        "test_pca = pca.transform(scaled_test_features)"
      ],
      "metadata": {
        "id": "_alIIcT5XWfM"
      },
      "execution_count": 35,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "7. Train a decision tree to classify genre\n",
        "\n",
        "  Using a simple algorithm such as a decision tree. Decision trees are rule-based classifiers that take in features and follow a 'tree structure' of binary decisions to ultimately classify a data point into one of two or more categories. In addition to being easy to both use and interpret, decision trees allow us to visualize the *logic flowchart* that the model generates from the training data."
      ],
      "metadata": {
        "id": "-Uh2Was7XhAK"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.tree import DecisionTreeClassifier\n",
        "\n",
        "# Train our decision tree\n",
        "tree = DecisionTreeClassifier(random_state=13)\n",
        "tree.fit(train_pca, train_labels)\n",
        "\n",
        "# Predict the labels for the test data\n",
        "pred_labels_tree = tree.predict(test_pca)"
      ],
      "metadata": {
        "id": "HUjto4ynXdbn"
      },
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "8. Compare decision tree to a logistic regression"
      ],
      "metadata": {
        "id": "jqiVH5uQXmoW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "# Train our logistic regression and predict labels for the test set\n",
        "logreg = LogisticRegression(random_state=13)\n",
        "logreg.fit(train_pca, train_labels)\n",
        "pred_labels_logit = logreg.predict(test_pca)\n",
        "\n",
        "# Create the classification report for both models\n",
        "from sklearn.metrics import classification_report\n",
        "class_rep_tree = classification_report(test_labels, pred_labels_tree)\n",
        "class_rep_log = classification_report(test_labels, pred_labels_logit)\n",
        "\n",
        "print(\"Decision Tree: \\n\", class_rep_tree)\n",
        "print(\"Logistic Regression: \\n\", class_rep_log)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Pncro50KXj-6",
        "outputId": "9a26dfd7-364b-459a-8b45-9441e7b69316"
      },
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Decision Tree: \n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "     Hip-Hop       0.65      0.63      0.64       218\n",
            "        Rock       0.92      0.92      0.92       983\n",
            "\n",
            "    accuracy                           0.87      1201\n",
            "   macro avg       0.78      0.78      0.78      1201\n",
            "weighted avg       0.87      0.87      0.87      1201\n",
            "\n",
            "Logistic Regression: \n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "     Hip-Hop       0.81      0.54      0.64       218\n",
            "        Rock       0.90      0.97      0.94       983\n",
            "\n",
            "    accuracy                           0.89      1201\n",
            "   macro avg       0.86      0.75      0.79      1201\n",
            "weighted avg       0.89      0.89      0.88      1201\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "9. Balance our data for greater performance\n",
        "\n",
        "  Both our models do similarly well, boasting an average precision of 87% each. However, looking at our classification report, we can see that rock songs are fairly well classified, but hip-hop songs are disproportionately misclassified as rock songs.\n",
        "\n",
        "  By looking at the number of data points we have for each class, we see that we have far more data points for the rock classification than for hip-hop, potentially skewing our model's ability to distinguish between classes. This also tells us that most of our model's accuracy is driven by its ability to classify just rock songs, which is less than ideal.\n",
        "\n",
        "  To account for this, we can weight the value of a correct classification in each class inversely to the occurrence of data points for each class. Since a correct classification for \"Rock\" is not more important than a correct classification for \"Hip-Hop\" (and vice versa), we only need to account for differences in sample size of our data points when weighting our classes here, and not relative importance of each class."
      ],
      "metadata": {
        "id": "RJQ4sNoaXsai"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Subset only the hip-hop tracks, and then only the rock tracks\n",
        "hop_only = echo_tracks.loc[echo_tracks['genre_top'] == 'Hip-Hop']\n",
        "rock_only = echo_tracks.loc[echo_tracks['genre_top'] == 'Rock']\n",
        "\n",
        "# sample the rocks songs to be the same number as there are hip-hop songs\n",
        "rock_only = rock_only.sample(hop_only.shape[0], random_state=13)\n",
        "\n",
        "# concatenate the dataframes rock_only and hop_only\n",
        "rock_hop_bal = pd.concat([rock_only, hop_only])\n",
        "\n",
        "# The features, labels, and pca projection are created for the balanced dataframe\n",
        "features = rock_hop_bal.drop(['genre_top', 'track_id'], axis=1)\n",
        "labels = rock_hop_bal['genre_top']\n",
        "\n",
        "# Redefine the train and test set with the pca_projection from the balanced data\n",
        "train_features, test_features, train_labels, test_labels = train_test_split(features, labels, random_state=13)\n",
        "\n",
        "train_pca = pca.fit_transform(scaler.fit_transform(train_features))\n",
        "test_pca = pca.transform(scaler.transform(test_features))"
      ],
      "metadata": {
        "id": "iDle7L3KXpRY"
      },
      "execution_count": 38,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "10. Check if balancing dataset improved model bias:\n",
        "  - Check if model bias towards the \"Rock\" classification has been improved.  \n",
        "  - Check for overall classification performance, as a lot of data points has been removed in this process"
      ],
      "metadata": {
        "id": "mcnvHPILX6Mp"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Train decision tree on the balanced data\n",
        "tree = DecisionTreeClassifier(random_state=13)\n",
        "tree.fit(train_pca, train_labels)\n",
        "pred_labels_tree = tree.predict(test_pca)\n",
        "\n",
        "# Train logistic regression on the balanced data\n",
        "logreg = LogisticRegression(random_state=13)\n",
        "logreg.fit(train_pca, train_labels)\n",
        "pred_labels_logit = logreg.predict(test_pca)\n",
        "\n",
        "# Compare the models\n",
        "print(\"Decision Tree: \\n\", classification_report(test_labels, pred_labels_tree))\n",
        "print(\"Logistic Regression: \\n\", classification_report(test_labels, pred_labels_logit))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gSt-0dOiX3zg",
        "outputId": "65fb54c9-950a-436e-867a-9750a20480ba"
      },
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Decision Tree: \n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "     Hip-Hop       0.80      0.81      0.81       233\n",
            "        Rock       0.80      0.79      0.79       222\n",
            "\n",
            "    accuracy                           0.80       455\n",
            "   macro avg       0.80      0.80      0.80       455\n",
            "weighted avg       0.80      0.80      0.80       455\n",
            "\n",
            "Logistic Regression: \n",
            "               precision    recall  f1-score   support\n",
            "\n",
            "     Hip-Hop       0.83      0.75      0.79       233\n",
            "        Rock       0.76      0.84      0.80       222\n",
            "\n",
            "    accuracy                           0.79       455\n",
            "   macro avg       0.80      0.79      0.79       455\n",
            "weighted avg       0.80      0.79      0.79       455\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "11. Using cross-validation to evaluate our models"
      ],
      "metadata": {
        "id": "AbHOjewxX-MN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import KFold, cross_val_score\n",
        "from sklearn.pipeline import Pipeline\n",
        "tree_pipe = Pipeline([(\"scaler\", StandardScaler()), (\"pca\", PCA(n_components=6)),\n",
        "                      (\"tree\", DecisionTreeClassifier(random_state=13))])\n",
        "logreg_pipe = Pipeline([(\"scaler\", StandardScaler()), (\"pca\", PCA(n_components=6)),\n",
        "                        (\"logreg\", LogisticRegression(random_state=13))])\n",
        "\n",
        "# Set up our K-fold cross-validation\n",
        "kf = KFold(n_splits=10)\n",
        "\n",
        "# Train our models using KFold cv\n",
        "tree_score = cross_val_score(tree_pipe, features, labels, cv=kf)\n",
        "logit_score = cross_val_score(logreg_pipe, features, labels, cv=kf)\n",
        "\n",
        "# Print the mean of each array of scores\n",
        "print(\"Decision Tree:\", np.mean(tree_score), \"Logistic Regression:\", np.mean(logit_score))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tsJfhavKX9DD",
        "outputId": "54ebe43b-8498-41ec-d4c7-d5f6f695eb89"
      },
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Decision Tree: 0.743956043956044 Logistic Regression: 0.7862637362637361\n"
          ]
        }
      ]
    }
  ]
}