ml_intro_pred_maintenance.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "collapsed_sections": [
        "oaoQmknJFdLL"
      ],
      "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/dcfaltesek/One/blob/main/ml_intro_pred_maintenance.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZgqkZPZdi_Pc"
      },
      "source": [
        "#**Introduction to AI & ML**\n",
        "### *Michaela Buchanan - Mark III Systems*"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SP0wDI62k7nm"
      },
      "source": [
        "If you would like to see some more examples using these algorithms, there is are two other versions of this notebook available. One focuses on predicting car prices and the other looks at determining if tumors are malignant or benign. See the links below to check them out.\n",
        "\n",
        "\n",
        "**Car price prediction:**\n",
        "\n",
        "https://colab.research.google.com/drive/1-7g2CRQinW_ytjZvJbOd74vnqu4OAdoL?usp=sharing\n",
        "\n",
        "**Tumor classification:**\n",
        "\n",
        "https://colab.research.google.com/drive/1SOhm2tqj62hb8GmLPJw6LhE5BFnpePXn?usp=sharing\n",
        "\n",
        "---"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Imports for all the code below. Please run first before anything else!"
      ],
      "metadata": {
        "id": "By7Y4eUvaJfW"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "byHwkWWKlCmF",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5c8f0220-71de-4ee0-ca8b-3cc6b44e4b26"
      },
      "source": [
        "!git clone https://github.com/michaelabuchanan/ml_workshop.git\n",
        "!pip install pycm==3.7\n",
        "!pip install shap\n",
        "import sys\n",
        "\n",
        "# data preprocessing\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "# logistic regression\n",
        "from matplotlib import pyplot as plt\n",
        "import numpy as np\n",
        "from sklearn import datasets\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.metrics import f1_score, accuracy_score\n",
        "from pycm import ConfusionMatrix\n",
        "import cv2\n",
        "\n",
        "# random forest\n",
        "from sklearn import tree, metrics\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "# xgboost\n",
        "from xgboost import XGBClassifier\n",
        "import shap\n",
        "\n",
        "print(sys.version)"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "fatal: destination path 'ml_workshop' already exists and is not an empty directory.\n",
            "Requirement already satisfied: pycm==3.7 in /usr/local/lib/python3.10/dist-packages (3.7)\n",
            "Requirement already satisfied: art>=1.8 in /usr/local/lib/python3.10/dist-packages (from pycm==3.7) (6.1)\n",
            "Requirement already satisfied: numpy>=1.9.0 in /usr/local/lib/python3.10/dist-packages (from pycm==3.7) (1.23.5)\n",
            "Requirement already satisfied: shap in /usr/local/lib/python3.10/dist-packages (0.43.0)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from shap) (1.23.5)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from shap) (1.11.3)\n",
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.10/dist-packages (from shap) (1.2.2)\n",
            "Requirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packages (from shap) (1.5.3)\n",
            "Requirement already satisfied: tqdm>=4.27.0 in /usr/local/lib/python3.10/dist-packages (from shap) (4.66.1)\n",
            "Requirement already satisfied: packaging>20.9 in /usr/local/lib/python3.10/dist-packages (from shap) (23.2)\n",
            "Requirement already satisfied: slicer==0.0.7 in /usr/local/lib/python3.10/dist-packages (from shap) (0.0.7)\n",
            "Requirement already satisfied: numba in /usr/local/lib/python3.10/dist-packages (from shap) (0.56.4)\n",
            "Requirement already satisfied: cloudpickle in /usr/local/lib/python3.10/dist-packages (from shap) (2.2.1)\n",
            "Requirement already satisfied: llvmlite<0.40,>=0.39.0dev0 in /usr/local/lib/python3.10/dist-packages (from numba->shap) (0.39.1)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from numba->shap) (67.7.2)\n",
            "Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.10/dist-packages (from pandas->shap) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas->shap) (2023.3.post1)\n",
            "Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->shap) (1.3.2)\n",
            "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn->shap) (3.2.0)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.8.1->pandas->shap) (1.16.0)\n",
            "3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EQQCGfdiHZhJ"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "###Introduction\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8Hs8mq7lyra_"
      },
      "source": [
        "This workshop will be showing both the workflow of a typical ML project as well as demonstrating a few common models. We will be following the ML Project lifecycle mentioned in the previous lecture and covering linear regression, decision trees, random forest, and XGBoost models."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h4v9HLjpyjef"
      },
      "source": [
        "![ml_lifecycle.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "In this notebook we will be looking at a synthetic dataset that \"reflects real predictive maintenance encountered in the industry\". [Click here](https://archive.ics.uci.edu/ml/datasets/AI4I+2020+Predictive+Maintenance+Dataset) to see the source for this dataset. We will be trying to predict the conditions that cause a machine to fail based on factors like tool wear, torque, rpm, and temperature."
      ],
      "metadata": {
        "id": "BIMtYfoF8owZ"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oaoQmknJFdLL"
      },
      "source": [
        "\n",
        "---\n",
        "\n",
        "###Problem Definition\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rQVWopn6yxMU"
      },
      "source": [
        "Predict conditions that will produce a failure in the operation of a machine."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ELQD4DppH7GF"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "### Data Analysis/Preprocessing\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kfpLQ2S1y1KY"
      },
      "source": [
        "The first step we need to take is to look at our data. In the first cell you ran, you cloned a GitHub repo which contained the `pred_maintenance.csv` file. A few rows of this file are shown below."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AwgHuD2TIPkr",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 455
        },
        "outputId": "1caad56f-de78-44e4-ab69-314287e13ccf"
      },
      "source": [
        "data = pd.read_csv('./ml_workshop/pred_maintenance.csv', index_col=\"UDI\")\n",
        "data"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      Product ID Type  Air temperature [K]  Process temperature [K]  \\\n",
              "UDI                                                                   \n",
              "1         M14860    M                298.1                    308.6   \n",
              "2         L47181    L                298.2                    308.7   \n",
              "3         L47182    L                298.1                    308.5   \n",
              "4         L47183    L                298.2                    308.6   \n",
              "5         L47184    L                298.2                    308.7   \n",
              "...          ...  ...                  ...                      ...   \n",
              "9996      M24855    M                298.8                    308.4   \n",
              "9997      H39410    H                298.9                    308.4   \n",
              "9998      M24857    M                299.0                    308.6   \n",
              "9999      H39412    H                299.0                    308.7   \n",
              "10000     M24859    M                299.0                    308.7   \n",
              "\n",
              "       Rotational speed [rpm]  Torque [Nm]  Tool wear [min]  Machine failure  \\\n",
              "UDI                                                                            \n",
              "1                        1551         42.8                0                0   \n",
              "2                        1408         46.3                3                0   \n",
              "3                        1498         49.4                5                0   \n",
              "4                        1433         39.5                7                0   \n",
              "5                        1408         40.0                9                0   \n",
              "...                       ...          ...              ...              ...   \n",
              "9996                     1604         29.5               14                0   \n",
              "9997                     1632         31.8               17                0   \n",
              "9998                     1645         33.4               22                0   \n",
              "9999                     1408         48.5               25                0   \n",
              "10000                    1500         40.2               30                0   \n",
              "\n",
              "       TWF  HDF  PWF  OSF  RNF  \n",
              "UDI                             \n",
              "1        0    0    0    0    0  \n",
              "2        0    0    0    0    0  \n",
              "3        0    0    0    0    0  \n",
              "4        0    0    0    0    0  \n",
              "5        0    0    0    0    0  \n",
              "...    ...  ...  ...  ...  ...  \n",
              "9996     0    0    0    0    0  \n",
              "9997     0    0    0    0    0  \n",
              "9998     0    0    0    0    0  \n",
              "9999     0    0    0    0    0  \n",
              "10000    0    0    0    0    0  \n",
              "\n",
              "[10000 rows x 13 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-4dd2a6cd-a9c6-4c7d-92e4-824b62fc2494\" 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>Product ID</th>\n",
              "      <th>Type</th>\n",
              "      <th>Air temperature [K]</th>\n",
              "      <th>Process temperature [K]</th>\n",
              "      <th>Rotational speed [rpm]</th>\n",
              "      <th>Torque [Nm]</th>\n",
              "      <th>Tool wear [min]</th>\n",
              "      <th>Machine failure</th>\n",
              "      <th>TWF</th>\n",
              "      <th>HDF</th>\n",
              "      <th>PWF</th>\n",
              "      <th>OSF</th>\n",
              "      <th>RNF</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>UDI</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>M14860</td>\n",
              "      <td>M</td>\n",
              "      <td>298.1</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1551</td>\n",
              "      <td>42.8</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>L47181</td>\n",
              "      <td>L</td>\n",
              "      <td>298.2</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>46.3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>L47182</td>\n",
              "      <td>L</td>\n",
              "      <td>298.1</td>\n",
              "      <td>308.5</td>\n",
              "      <td>1498</td>\n",
              "      <td>49.4</td>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>L47183</td>\n",
              "      <td>L</td>\n",
              "      <td>298.2</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1433</td>\n",
              "      <td>39.5</td>\n",
              "      <td>7</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>L47184</td>\n",
              "      <td>L</td>\n",
              "      <td>298.2</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>40.0</td>\n",
              "      <td>9</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9996</th>\n",
              "      <td>M24855</td>\n",
              "      <td>M</td>\n",
              "      <td>298.8</td>\n",
              "      <td>308.4</td>\n",
              "      <td>1604</td>\n",
              "      <td>29.5</td>\n",
              "      <td>14</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9997</th>\n",
              "      <td>H39410</td>\n",
              "      <td>H</td>\n",
              "      <td>298.9</td>\n",
              "      <td>308.4</td>\n",
              "      <td>1632</td>\n",
              "      <td>31.8</td>\n",
              "      <td>17</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9998</th>\n",
              "      <td>M24857</td>\n",
              "      <td>M</td>\n",
              "      <td>299.0</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1645</td>\n",
              "      <td>33.4</td>\n",
              "      <td>22</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9999</th>\n",
              "      <td>H39412</td>\n",
              "      <td>H</td>\n",
              "      <td>299.0</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>48.5</td>\n",
              "      <td>25</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10000</th>\n",
              "      <td>M24859</td>\n",
              "      <td>M</td>\n",
              "      <td>299.0</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1500</td>\n",
              "      <td>40.2</td>\n",
              "      <td>30</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10000 rows × 13 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-4dd2a6cd-a9c6-4c7d-92e4-824b62fc2494')\"\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-4dd2a6cd-a9c6-4c7d-92e4-824b62fc2494 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-4dd2a6cd-a9c6-4c7d-92e4-824b62fc2494');\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-1ca9bb5c-a553-4062-85d9-c0b20bc252ae\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-1ca9bb5c-a553-4062-85d9-c0b20bc252ae')\"\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-1ca9bb5c-a553-4062-85d9-c0b20bc252ae 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": 2
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "brxgByKTQOVz"
      },
      "source": [
        "First let's check out what datatypes we have in our dataset. If the data types are not what we expect, we could run into problems with our preprocessing and later on with our training."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "djIagM3nQup0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5fc9ebe8-99af-4b92-bc29-84ebe53f49d9"
      },
      "source": [
        "data.dtypes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Product ID                  object\n",
              "Type                        object\n",
              "Air temperature [K]        float64\n",
              "Process temperature [K]    float64\n",
              "Rotational speed [rpm]       int64\n",
              "Torque [Nm]                float64\n",
              "Tool wear [min]              int64\n",
              "Machine failure              int64\n",
              "TWF                          int64\n",
              "HDF                          int64\n",
              "PWF                          int64\n",
              "OSF                          int64\n",
              "RNF                          int64\n",
              "dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vBcVGv6zOCgX"
      },
      "source": [
        "In general, you can't provide text input to a ML model. This is a problem since two of our categories (Product ID and Type) are text based as shown by their `object` data type above. Luckily, it is pretty easy with Numpy to turn all this text into numbers which we will see in a little bit. We are only going to do this for Type as the Product ID column just restates inforation already available in the other columns so we will be removing it using the `pop` command. We are also going to remove the failure type columns as we are only interested in predicting failure in general, not by type, for this problem."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A0Fk59xQUJal",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4003c141-b6cb-49d9-8e20-2bd7ce43604d"
      },
      "source": [
        "# categorical columns\n",
        "data.pop('Product ID')\n",
        "product_type = data.pop('Type')\n",
        "\n",
        "# failure type columns\n",
        "data.pop('TWF')\n",
        "data.pop('HDF')\n",
        "data.pop('PWF')\n",
        "data.pop('OSF')\n",
        "data.pop('RNF')\n",
        "\n",
        "print(\"Done\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Done\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4jOsZ1FWWrMa"
      },
      "source": [
        "Here we check our work and see that the only columns left in data are numerical."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w71cG3LHWgHF",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 455
        },
        "outputId": "d932aab0-4d7b-4bb6-bd60-105156cf9dc8"
      },
      "source": [
        "data"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Air temperature [K]  Process temperature [K]  Rotational speed [rpm]  \\\n",
              "UDI                                                                           \n",
              "1                    298.1                    308.6                    1551   \n",
              "2                    298.2                    308.7                    1408   \n",
              "3                    298.1                    308.5                    1498   \n",
              "4                    298.2                    308.6                    1433   \n",
              "5                    298.2                    308.7                    1408   \n",
              "...                    ...                      ...                     ...   \n",
              "9996                 298.8                    308.4                    1604   \n",
              "9997                 298.9                    308.4                    1632   \n",
              "9998                 299.0                    308.6                    1645   \n",
              "9999                 299.0                    308.7                    1408   \n",
              "10000                299.0                    308.7                    1500   \n",
              "\n",
              "       Torque [Nm]  Tool wear [min]  Machine failure  \n",
              "UDI                                                   \n",
              "1             42.8                0                0  \n",
              "2             46.3                3                0  \n",
              "3             49.4                5                0  \n",
              "4             39.5                7                0  \n",
              "5             40.0                9                0  \n",
              "...            ...              ...              ...  \n",
              "9996          29.5               14                0  \n",
              "9997          31.8               17                0  \n",
              "9998          33.4               22                0  \n",
              "9999          48.5               25                0  \n",
              "10000         40.2               30                0  \n",
              "\n",
              "[10000 rows x 6 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-861e6b51-7bd2-49d3-bdb6-4f5eb345cf43\" 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>Air temperature [K]</th>\n",
              "      <th>Process temperature [K]</th>\n",
              "      <th>Rotational speed [rpm]</th>\n",
              "      <th>Torque [Nm]</th>\n",
              "      <th>Tool wear [min]</th>\n",
              "      <th>Machine failure</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>UDI</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>298.1</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1551</td>\n",
              "      <td>42.8</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>46.3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>298.1</td>\n",
              "      <td>308.5</td>\n",
              "      <td>1498</td>\n",
              "      <td>49.4</td>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1433</td>\n",
              "      <td>39.5</td>\n",
              "      <td>7</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>40.0</td>\n",
              "      <td>9</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9996</th>\n",
              "      <td>298.8</td>\n",
              "      <td>308.4</td>\n",
              "      <td>1604</td>\n",
              "      <td>29.5</td>\n",
              "      <td>14</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9997</th>\n",
              "      <td>298.9</td>\n",
              "      <td>308.4</td>\n",
              "      <td>1632</td>\n",
              "      <td>31.8</td>\n",
              "      <td>17</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9998</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1645</td>\n",
              "      <td>33.4</td>\n",
              "      <td>22</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9999</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>48.5</td>\n",
              "      <td>25</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10000</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1500</td>\n",
              "      <td>40.2</td>\n",
              "      <td>30</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10000 rows × 6 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-861e6b51-7bd2-49d3-bdb6-4f5eb345cf43')\"\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-861e6b51-7bd2-49d3-bdb6-4f5eb345cf43 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-861e6b51-7bd2-49d3-bdb6-4f5eb345cf43');\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-422e1784-0f5a-465f-a21f-c943fcfd6e93\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-422e1784-0f5a-465f-a21f-c943fcfd6e93')\"\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-422e1784-0f5a-465f-a21f-c943fcfd6e93 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": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZfGMLleWWwOk"
      },
      "source": [
        "Now let's perform our one-hot encoding for the one categorical column that we are interested in keeping."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dh6Wy8riW1dr",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f027c7aa-e11a-4c1b-f3d4-c62edd37c94f"
      },
      "source": [
        "oh_type = pd.get_dummies(product_type)\n",
        "\n",
        "print(\"Done\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Done\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ggUI_08xXhRH"
      },
      "source": [
        "And finally we reappend these new one-hot encoded column back onto our original dataframe"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QKBUaWlBXm44",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 455
        },
        "outputId": "3093ff7b-023c-4a00-bd8c-109549a07360"
      },
      "source": [
        "processed_data = pd.concat([data, oh_type], axis=1)\n",
        "processed_data"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Air temperature [K]  Process temperature [K]  Rotational speed [rpm]  \\\n",
              "UDI                                                                           \n",
              "1                    298.1                    308.6                    1551   \n",
              "2                    298.2                    308.7                    1408   \n",
              "3                    298.1                    308.5                    1498   \n",
              "4                    298.2                    308.6                    1433   \n",
              "5                    298.2                    308.7                    1408   \n",
              "...                    ...                      ...                     ...   \n",
              "9996                 298.8                    308.4                    1604   \n",
              "9997                 298.9                    308.4                    1632   \n",
              "9998                 299.0                    308.6                    1645   \n",
              "9999                 299.0                    308.7                    1408   \n",
              "10000                299.0                    308.7                    1500   \n",
              "\n",
              "       Torque [Nm]  Tool wear [min]  Machine failure  H  L  M  \n",
              "UDI                                                            \n",
              "1             42.8                0                0  0  0  1  \n",
              "2             46.3                3                0  0  1  0  \n",
              "3             49.4                5                0  0  1  0  \n",
              "4             39.5                7                0  0  1  0  \n",
              "5             40.0                9                0  0  1  0  \n",
              "...            ...              ...              ... .. .. ..  \n",
              "9996          29.5               14                0  0  0  1  \n",
              "9997          31.8               17                0  1  0  0  \n",
              "9998          33.4               22                0  0  0  1  \n",
              "9999          48.5               25                0  1  0  0  \n",
              "10000         40.2               30                0  0  0  1  \n",
              "\n",
              "[10000 rows x 9 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-1f512187-6478-49db-9153-6bbae9f87ef6\" 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>Air temperature [K]</th>\n",
              "      <th>Process temperature [K]</th>\n",
              "      <th>Rotational speed [rpm]</th>\n",
              "      <th>Torque [Nm]</th>\n",
              "      <th>Tool wear [min]</th>\n",
              "      <th>Machine failure</th>\n",
              "      <th>H</th>\n",
              "      <th>L</th>\n",
              "      <th>M</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>UDI</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>298.1</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1551</td>\n",
              "      <td>42.8</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>46.3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>298.1</td>\n",
              "      <td>308.5</td>\n",
              "      <td>1498</td>\n",
              "      <td>49.4</td>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1433</td>\n",
              "      <td>39.5</td>\n",
              "      <td>7</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>40.0</td>\n",
              "      <td>9</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9996</th>\n",
              "      <td>298.8</td>\n",
              "      <td>308.4</td>\n",
              "      <td>1604</td>\n",
              "      <td>29.5</td>\n",
              "      <td>14</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9997</th>\n",
              "      <td>298.9</td>\n",
              "      <td>308.4</td>\n",
              "      <td>1632</td>\n",
              "      <td>31.8</td>\n",
              "      <td>17</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9998</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1645</td>\n",
              "      <td>33.4</td>\n",
              "      <td>22</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9999</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>48.5</td>\n",
              "      <td>25</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10000</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1500</td>\n",
              "      <td>40.2</td>\n",
              "      <td>30</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10000 rows × 9 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-1f512187-6478-49db-9153-6bbae9f87ef6')\"\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-1f512187-6478-49db-9153-6bbae9f87ef6 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-1f512187-6478-49db-9153-6bbae9f87ef6');\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-1cc75266-0849-4aa3-8fb9-62fa9401991b\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-1cc75266-0849-4aa3-8fb9-62fa9401991b')\"\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-1cc75266-0849-4aa3-8fb9-62fa9401991b 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": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nh4XMDo-gETV"
      },
      "source": [
        "One last thing to check before we train. If any of the values are missing in our dataset, they will be listed as NaN (not a number) in our dataframe. When we go to train, our algorithm won't know what to do with these values so we are going to go ahead and remove any rows with missing values now. We should also check for and remove any duplicated columns as well."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aDkrtiSegTBg",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 455
        },
        "outputId": "77056fd9-0ccd-4a93-fb5a-aa4738c81b50"
      },
      "source": [
        "processed_data.dropna(inplace=True)\n",
        "processed_data = processed_data.loc[:,~processed_data.columns.duplicated()]\n",
        "processed_data"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       Air temperature [K]  Process temperature [K]  Rotational speed [rpm]  \\\n",
              "UDI                                                                           \n",
              "1                    298.1                    308.6                    1551   \n",
              "2                    298.2                    308.7                    1408   \n",
              "3                    298.1                    308.5                    1498   \n",
              "4                    298.2                    308.6                    1433   \n",
              "5                    298.2                    308.7                    1408   \n",
              "...                    ...                      ...                     ...   \n",
              "9996                 298.8                    308.4                    1604   \n",
              "9997                 298.9                    308.4                    1632   \n",
              "9998                 299.0                    308.6                    1645   \n",
              "9999                 299.0                    308.7                    1408   \n",
              "10000                299.0                    308.7                    1500   \n",
              "\n",
              "       Torque [Nm]  Tool wear [min]  Machine failure  H  L  M  \n",
              "UDI                                                            \n",
              "1             42.8                0                0  0  0  1  \n",
              "2             46.3                3                0  0  1  0  \n",
              "3             49.4                5                0  0  1  0  \n",
              "4             39.5                7                0  0  1  0  \n",
              "5             40.0                9                0  0  1  0  \n",
              "...            ...              ...              ... .. .. ..  \n",
              "9996          29.5               14                0  0  0  1  \n",
              "9997          31.8               17                0  1  0  0  \n",
              "9998          33.4               22                0  0  0  1  \n",
              "9999          48.5               25                0  1  0  0  \n",
              "10000         40.2               30                0  0  0  1  \n",
              "\n",
              "[10000 rows x 9 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-6dfa06b9-c3b7-4359-8006-5fb9a9931bea\" 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>Air temperature [K]</th>\n",
              "      <th>Process temperature [K]</th>\n",
              "      <th>Rotational speed [rpm]</th>\n",
              "      <th>Torque [Nm]</th>\n",
              "      <th>Tool wear [min]</th>\n",
              "      <th>Machine failure</th>\n",
              "      <th>H</th>\n",
              "      <th>L</th>\n",
              "      <th>M</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>UDI</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>298.1</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1551</td>\n",
              "      <td>42.8</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>46.3</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>298.1</td>\n",
              "      <td>308.5</td>\n",
              "      <td>1498</td>\n",
              "      <td>49.4</td>\n",
              "      <td>5</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1433</td>\n",
              "      <td>39.5</td>\n",
              "      <td>7</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>298.2</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>40.0</td>\n",
              "      <td>9</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9996</th>\n",
              "      <td>298.8</td>\n",
              "      <td>308.4</td>\n",
              "      <td>1604</td>\n",
              "      <td>29.5</td>\n",
              "      <td>14</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9997</th>\n",
              "      <td>298.9</td>\n",
              "      <td>308.4</td>\n",
              "      <td>1632</td>\n",
              "      <td>31.8</td>\n",
              "      <td>17</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9998</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.6</td>\n",
              "      <td>1645</td>\n",
              "      <td>33.4</td>\n",
              "      <td>22</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9999</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1408</td>\n",
              "      <td>48.5</td>\n",
              "      <td>25</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10000</th>\n",
              "      <td>299.0</td>\n",
              "      <td>308.7</td>\n",
              "      <td>1500</td>\n",
              "      <td>40.2</td>\n",
              "      <td>30</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10000 rows × 9 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-6dfa06b9-c3b7-4359-8006-5fb9a9931bea')\"\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-6dfa06b9-c3b7-4359-8006-5fb9a9931bea 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-6dfa06b9-c3b7-4359-8006-5fb9a9931bea');\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-162b3499-b86f-49a1-8601-786fa3891951\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-162b3499-b86f-49a1-8601-786fa3891951')\"\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-162b3499-b86f-49a1-8601-786fa3891951 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": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ujlEfWORktMc"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "###Single Variable Logistic Regression"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7AxWxRRzceX6"
      },
      "source": [
        "While we could spend much more time going through our data to make it easier to train off of,  I'm going to move on to training using logistic regression due to time constraints for this session. Logistical regression is very similar to linear regression with the main difference being that we get a categorical output rather than a continuous one. This is what we need to be able to classify into our `0` (no failure) and `1` (failure) categories for the `Machine failure` column. To make things simpler we will be starting off with just one variable to make our predictions which means we need to be able to choose the best column from our dataset to use. The code below creates a graph of the distribution of our two categories among data entries in different columns in our dataset. The columns we are testing are listed in the `test_columns` list. Feel free to change which columns are being used to see if you can find one that fits our task even better than the ones used below."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gE59Cz9aitdQ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "096ad697-7d99-4912-c3e5-5a922156afe2"
      },
      "source": [
        "test_columns = [\"Process temperature [K]\", \"Torque [Nm]\", \"Tool wear [min]\", \"Rotational speed [rpm]\"]\n",
        "\n",
        "for feature_name in test_columns:\n",
        "  plt.figure()\n",
        "  processed_data.groupby(\"Machine failure\")[feature_name].plot.hist(alpha=0.5, density=True, legend=True)\n",
        "  plt.xlabel(feature_name)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZRrVLON_dwhv"
      },
      "source": [
        "In these plots we are looking for a clear distinction between where the machine fails and where it doesn't. This would tell us that the feature we are looking at is a good indicator of machine failure. In the graphs shown there are some that look slightly better than others but if you want a better example of what to look for in these graphs check out the tumor classification notebook linked at the top of the page. We can see however that process temperature does not look like a good indicator while torque is the most promising looking option. Now let's run an example of our logistic regression model. We will start by using Process Temperature as our feature to predict Machine failure off of. We will them create the same model using Torque as our feature to see if that increases accuracy."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VpSIeL2Gd7Mg",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d52331b6-b587-4bb2-c929-87dabf816ecc"
      },
      "source": [
        "# feel free to play with different features to use for X, ie tool wear, air temp, etc\n",
        "x_single = processed_data['Process temperature [K]']\n",
        "y_single = processed_data['Machine failure']\n",
        "\n",
        "# some of this code comes from https://pythonspot.com/linear-regression/\n",
        "\n",
        "x_single = x_single.values.reshape(len(x_single),1)\n",
        "y_single = y_single.values.reshape(len(y_single),1)\n",
        "\n",
        "# Split the data into training/testing sets\n",
        "x_single_train = x_single[:-500]\n",
        "x_single_test = x_single[-500:]\n",
        "\n",
        "# Split the targets into training/testing sets\n",
        "y_single_train = y_single[:-500]\n",
        "y_single_test = y_single[-500:]\n",
        "\n",
        "# Create logistic regression object\n",
        "temp_model = LogisticRegression()\n",
        "\n",
        "# Train the model using the training sets\n",
        "temp_model.fit(x_single_train, y_single_train)\n",
        "accuracy = temp_model.score(x_single_test, y_single_test)\n",
        "print(f\"Accuracy on test set: {accuracy:.3f}\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy on test set: 0.974\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JyxUvLnn-SmE"
      },
      "source": [
        "The code cell above fetches our `Torque` values as our X and our `Machine Failure` labels as our Y. Then we split our data into training and testing sets and train a Logistic Regression model off this data using `scikit-learn` and the `fit()` function. You should see a test set accuracy value printed out at the end which gives us an indication of this model's performance. Keep this accuracy in mind as we will use it for comparison next. Let's try doing the same thing but using `Rotational Speed` as our X feature."
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# feel free to play with different features to use for X, ie tool wear, air temp, etc\n",
        "x_single = processed_data['Torque [Nm]']\n",
        "\n",
        "# some of this code comes from https://pythonspot.com/linear-regression/\n",
        "x_single = x_single.values.reshape(len(x_single),1)\n",
        "\n",
        "# Split the data into training/testing sets\n",
        "x_single_train = x_single[:-500]\n",
        "x_single_test = x_single[-500:]\n",
        "\n",
        "# Create logistic regression object\n",
        "torque_model = LogisticRegression()\n",
        "\n",
        "# Train the model using the training sets\n",
        "torque_model.fit(x_single_train, y_single_train)\n",
        "accuracy = torque_model.score(x_single_test, y_single_test)\n",
        "print(f\"Accuracy on test set: {accuracy:.3f}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IaBJqLXFbobh",
        "outputId": "714c9275-d479-4e72-b1c9-cf1a2a513bc7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy on test set: 0.974\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "An important part of data science is being able to tell when things seem a bit \"fishy\" as a lot of times there's another factor at play scewing your model results. The fact that we got the same exact accuracy from both of these models while using different X features is very unusual and should raise some red flags about the approach we are using."
      ],
      "metadata": {
        "id": "V-bLMFndfYu3"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C8d9hXPyyVRH"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "###Dealing with Class Imbalance\n"
      ],
      "metadata": {
        "id": "NtdpzRfLKsWc"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "If we think a bit more about our dataset itself we start to find a big problem. When we record data points of an industrial machine in operation at set increments, most of these values are going to report no machine failure. Machine failures are (hopefully) rare occurances compared to the overall runtime of the machine in operation. This creates what is called a class imbalance where most of our datapoints go in one category while relatively few go in the other. Let's check for class imbalance in our dataset."
      ],
      "metadata": {
        "id": "MxywvRf8fyxO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "processed_data[\"Machine failure\"].value_counts()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "noGtDg4egfAY",
        "outputId": "61cebacc-a8ab-494b-b04f-dd05460d2656"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    9661\n",
              "1     339\n",
              "Name: Machine failure, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Only 339 samples out of our 10,000 sample dataset report 1, or that a machine failure has occured. This is only about 3% of our dataset. What can happen when an imbalance this drastic occurs is that the model \"gives up\" on trying to predict actual machine failures and instead just predicts 0 for everything. By doing this it actually is able to achieve a higher accuracy value, in our case of around 98%, by doing this rather than trying at all to predict any machine failures. However, this makes the model useless to us."
      ],
      "metadata": {
        "id": "VoxYDbb6hBGV"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "To confirm our theory, let's take a look at the results from our first model we trained using the Torque feature. We are going to create a confusion matrix from our this model's predictions which will tell us what our model is doing."
      ],
      "metadata": {
        "id": "Zc2fh2tNKxEr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "predictions = torque_model.predict(x_single_test)\n",
        "clean_y_single_test = []\n",
        "\n",
        "for i in range(len(y_single_test)):\n",
        "  clean_y_single_test.append(str(y_single_test[i]).replace(\"[\", \"\").replace(\"]\", \"\"))\n",
        "\n",
        "cm1 = ConfusionMatrix(clean_y_single_test, predictions)\n",
        "\n",
        "cm1.plot(cmap=plt.cm.Blues, number_label=True)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "529hlZW2O76O",
        "outputId": "7a3a9675-0e83-42bf-96ab-920e6ef4afcd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: title={'center': 'Confusion Matrix'}, xlabel='Predicted Classes', ylabel='Actual Classes'>"
            ]
          },
          "metadata": {},
          "execution_count": 13
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "If you look at the 1 column of the Predicted label axis you can see two 0s in the squares. This tells us that our model never tried to predict a 1 in all 500 test sample predictions, confirming our theory. This caused us to miss all 13 machine failures that we can see in our confusion matrix. If we divide 13/500 that gives us 0.026, explaining why we got our 97.4% accuracy no matter what we chose for our X feature."
      ],
      "metadata": {
        "id": "GrrO2iVTXYTF"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Luckily this is a common issue and `scikit-learn` already has a setting for our Logistic Regression model that should help. Keep in mind that I could probably do a whole session just on techniques to help with class imbalance and I'm only going to show the simplest on here for now."
      ],
      "metadata": {
        "id": "UvauU6ypjnm9"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's take another stab at our torque model. The setup below is mostly the same as before expect we are going to pass the `class_weight='balanced'` argument to our `LogisticRegression()` function. In a nutshell this will weight the importance of prediction differently for our two classes skewing the weighting to the underrepresented class. [Click here](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) to view the documentation for this if you want to learn more. Essentially this should prevent the model from just predicting 0 all the time. From this we should expect a lower accuracy but we should be able to see some variety in the model's predictions."
      ],
      "metadata": {
        "id": "V7NOfzfFj9L0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# feel free to play with different features to use for X, ie perimeter_mean, compactness_mean, etc\n",
        "x_single = processed_data['Torque [Nm]']\n",
        "\n",
        "# some of this code comes from https://pythonspot.com/linear-regression/\n",
        "\n",
        "x_single = x_single.values.reshape(len(x_single),1)\n",
        "\n",
        "# Split the data into training/testing sets\n",
        "x_single_train = x_single[:-500]\n",
        "x_single_test = x_single[-500:]\n",
        "\n",
        "# Create logistic regression object\n",
        "torque_model_balanced = LogisticRegression(class_weight='balanced')\n",
        "\n",
        "# Train the model using the training sets\n",
        "torque_model_balanced.fit(x_single_train, y_single_train)\n",
        "accuracy = torque_model_balanced.score(x_single_test, y_single_test)\n",
        "print(f\"Accuracy on test set: {accuracy:.3f}\")\n",
        "\n",
        "# format data for CM\n",
        "clean_y_single_test2 = []\n",
        "\n",
        "for i in range(len(y_single_test)):\n",
        "  clean_y_single_test2.append(str(y_single_test[i]).replace(\"[\", \"\").replace(\"]\", \"\"))\n",
        "\n",
        "# Confusion matrix\n",
        "predictions = torque_model_balanced.predict(x_single_test)\n",
        "\n",
        "cm2 = ConfusionMatrix(clean_y_single_test2, predictions)\n",
        "\n",
        "cm2.plot(cmap=plt.cm.Blues, number_label=True)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 524
        },
        "id": "OUXvX5HfXaO4",
        "outputId": "221cbab3-db75-4eb0-afed-cea53b7ffec1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy on test set: 0.704\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: title={'center': 'Confusion Matrix'}, xlabel='Predicted Classes', ylabel='Actual Classes'>"
            ]
          },
          "metadata": {},
          "execution_count": 14
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now we can see that our model is doing quite a bit worse accuracy-wise, but is at least making an attempt at predicting machine failures. This means we are training a model that's actually going to be useful to us rather than just always predicting 0. Now that we have a better starting point, let's see if we can increase our accuracy using some different model types."
      ],
      "metadata": {
        "id": "40PXZYb75bh2"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "blft5H_gjWhz"
      },
      "source": [
        "###Decision Trees\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sJuuUNczy61c"
      },
      "source": [
        "There's only so far we can get in terms of accuracy with our model when all we have to work with is a straight line. Let's try a decision tree instead to see how that performs for our problem. Once again we are going to use the `class_weight='balanced'` argument to try to help with our class imbalance issue."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CNHmc0RIPUXc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 507
        },
        "outputId": "39550367-9cc5-44b1-f918-e4e732276f2d"
      },
      "source": [
        "x_tree = processed_data.drop('Machine failure', axis=1)\n",
        "y_tree = processed_data['Machine failure']\n",
        "\n",
        "x_train_tree, x_test_tree, y_train_tree, y_test_tree = train_test_split(x_tree, y_tree, test_size=0.2)\n",
        "\n",
        "# create a classifier object\n",
        "tree_model = DecisionTreeClassifier(class_weight='balanced')\n",
        "\n",
        "# fit the classifier with X and Y data\n",
        "tree_model.fit(x_train_tree, y_train_tree)\n",
        "\n",
        "accuracy = tree_model.score(x_test_tree, y_test_tree)\n",
        "\n",
        "print(f\"Accuracy on test set: {accuracy:.3f}\")\n",
        "\n",
        "# Confusion matrix\n",
        "predictions = tree_model.predict(x_test_tree)\n",
        "\n",
        "cm3 = ConfusionMatrix(y_test_tree.tolist(), predictions)\n",
        "\n",
        "cm3.plot(cmap=plt.cm.Blues, number_label=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy on test set: 0.977\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: title={'center': 'Confusion Matrix'}, xlabel='Predicted Classes', ylabel='Actual Classes'>"
            ]
          },
          "metadata": {},
          "execution_count": 15
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-XCnMFc9r41Z"
      },
      "source": [
        "Our accuracy is now ~98% which is much better than before. This isn't too surpring as the model now has a lot more information to work with rather than just one feature. We can also see in our confusion matrix that the model isn't just predicting 0.\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hLuas59uyC5Q"
      },
      "source": [
        "###Random Forest\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s4fYxsh3zRXL"
      },
      "source": [
        "Now that we are starting to get some good results, let's see if we can do even better with some more complex techniques. What's better than one decision tree? What about a whole forest of decision trees? Run the code below to see what happens."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PkXM5E5Rbz_6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 507
        },
        "outputId": "df8a9f21-5759-4ff1-a54f-427bc12e7ea7"
      },
      "source": [
        "x_forest = processed_data.drop('Machine failure', axis=1)\n",
        "y_forest = processed_data['Machine failure']\n",
        "\n",
        "x_train_forest, x_test_forest, y_train_forest, y_test_forest = train_test_split(x_forest, y_forest, test_size=0.2)\n",
        "\n",
        "forest_model = RandomForestClassifier(class_weight='balanced')\n",
        "forest_model.fit(x_train_forest, y_train_forest)\n",
        "\n",
        "accuracy = forest_model.score(x_test_forest, y_test_forest)\n",
        "\n",
        "print(f\"Accuracy on test set: {accuracy:.3f}\")\n",
        "\n",
        "# Confusion matrix\n",
        "predictions = forest_model.predict(x_test_tree)\n",
        "\n",
        "cm4 = ConfusionMatrix(y_test_forest.tolist(), predictions)\n",
        "\n",
        "cm4.plot(cmap=plt.cm.Blues, number_label=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy on test set: 0.982\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: title={'center': 'Confusion Matrix'}, xlabel='Predicted Classes', ylabel='Actual Classes'>"
            ]
          },
          "metadata": {},
          "execution_count": 16
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UOr2rH4q1dk7"
      },
      "source": [
        "The accuracy is pretty similar to the Decision Tree model. However, our confusion matrix tells us that we did not make very many/any correct machine failure predictions. This tells us that our Decision Tree model actually performed better despite the accuracies being similar. We could probably improve the Random Forest model by playing around with hyperparamaters and more dataset preprocessing but I'll leave that as an experiment for you.\n",
        "\n",
        "\n",
        "\n",
        "---\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Wq8No4Y31fxd"
      },
      "source": [
        "###XGBoost"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CHvA8OTTGECS"
      },
      "source": [
        "Now let's move away from bagging and try a boost ensemble model, in this case XGBoost. This is one of the most popular ML algorithms you'll come across in Kaggle competitions and other such sites. Below is the code needed to create your XGBoost model and fit it to your training data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rBMdYaWH1kj7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 248
        },
        "outputId": "9f4e1de6-2933-4d02-b5d7-c60226541568"
      },
      "source": [
        "# XGBoost doesn't like [] in the feature names so we need to change the names of our columns\n",
        "processed_data = processed_data.rename(columns={\"Air temperature [K]\": \"Air temperature\",\n",
        "                                                \"Process temperature [K]\": \"Process temperature\",\n",
        "                                                \"Rotational speed [rpm]\": \"Rotational speed\",\n",
        "                                                \"Torque [Nm]\": \"Torque\",\n",
        "                                                \"Tool wear [min]\": \"Tool wear\"})\n",
        "\n",
        "x_boost = processed_data.drop('Machine failure', axis=1)\n",
        "y_boost = processed_data['Machine failure']\n",
        "\n",
        "x_train_boost, x_test_boost, y_train_boost, y_test_boost = train_test_split(x_boost, y_boost, test_size=0.2)\n",
        "\n",
        "boost = XGBClassifier()\n",
        "boost.fit(x_train_boost, y_train_boost)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
              "              colsample_bylevel=None, colsample_bynode=None,\n",
              "              colsample_bytree=None, device=None, early_stopping_rounds=None,\n",
              "              enable_categorical=False, eval_metric=None, feature_types=None,\n",
              "              gamma=None, grow_policy=None, importance_type=None,\n",
              "              interaction_constraints=None, learning_rate=None, max_bin=None,\n",
              "              max_cat_threshold=None, max_cat_to_onehot=None,\n",
              "              max_delta_step=None, max_depth=None, max_leaves=None,\n",
              "              min_child_weight=None, missing=nan, monotone_constraints=None,\n",
              "              multi_strategy=None, n_estimators=None, n_jobs=None,\n",
              "              num_parallel_tree=None, random_state=None, ...)"
            ],
            "text/html": [
              "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
              "              colsample_bylevel=None, colsample_bynode=None,\n",
              "              colsample_bytree=None, device=None, early_stopping_rounds=None,\n",
              "              enable_categorical=False, eval_metric=None, feature_types=None,\n",
              "              gamma=None, grow_policy=None, importance_type=None,\n",
              "              interaction_constraints=None, learning_rate=None, max_bin=None,\n",
              "              max_cat_threshold=None, max_cat_to_onehot=None,\n",
              "              max_delta_step=None, max_depth=None, max_leaves=None,\n",
              "              min_child_weight=None, missing=nan, monotone_constraints=None,\n",
              "              multi_strategy=None, n_estimators=None, n_jobs=None,\n",
              "              num_parallel_tree=None, random_state=None, ...)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">XGBClassifier</label><div class=\"sk-toggleable__content\"><pre>XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
              "              colsample_bylevel=None, colsample_bynode=None,\n",
              "              colsample_bytree=None, device=None, early_stopping_rounds=None,\n",
              "              enable_categorical=False, eval_metric=None, feature_types=None,\n",
              "              gamma=None, grow_policy=None, importance_type=None,\n",
              "              interaction_constraints=None, learning_rate=None, max_bin=None,\n",
              "              max_cat_threshold=None, max_cat_to_onehot=None,\n",
              "              max_delta_step=None, max_depth=None, max_leaves=None,\n",
              "              min_child_weight=None, missing=nan, monotone_constraints=None,\n",
              "              multi_strategy=None, n_estimators=None, n_jobs=None,\n",
              "              num_parallel_tree=None, random_state=None, ...)</pre></div></div></div></div></div>"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lMp8GauOLHeG"
      },
      "source": [
        "Now that our model is trained, let's see how it is doing on our data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s_j2fYTm8z1I",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 507
        },
        "outputId": "eb1baf81-55f3-42de-d441-f04d2e61cbfe"
      },
      "source": [
        "accuracy = boost.score(x_test_boost, y_test_boost)\n",
        "print(f\"Accuracy on test set: {accuracy:.3f}\")\n",
        "\n",
        "# Confusion matrix\n",
        "predictions = boost.predict(x_test_boost)\n",
        "\n",
        "cm5 = ConfusionMatrix(y_test_boost.tolist(), predictions)\n",
        "\n",
        "cm5.plot(cmap=plt.cm.Blues, number_label=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy on test set: 0.981\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<Axes: title={'center': 'Confusion Matrix'}, xlabel='Predicted Classes', ylabel='Actual Classes'>"
            ]
          },
          "metadata": {},
          "execution_count": 18
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GU0PFdekEnBe"
      },
      "source": [
        "XGBoost seems to work pretty well on this problem. We have one of the highest accuracies we've seen yet and it is doing a good job on actually predicting machine failures.\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now let's finish off by looking at how our features impacted our model's ability to make predictions. We will be using Shapley values to do this. These values tell you how much each feature contributed to the model's predictions. To learn more about how these values are calculated check out the YouTube video linked at the end of this notebook."
      ],
      "metadata": {
        "id": "idtuTad1PWyB"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "shap_values = shap.TreeExplainer(boost).shap_values(x_train_boost)\n",
        "shap.summary_plot(shap_values, x_train_boost, plot_type=\"bar\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        },
        "id": "EddD2Q4XQZpg",
        "outputId": "9c0206c9-72bd-4166-8f6c-1076881a6310"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[15:30:37] WARNING: /workspace/src/c_api/c_api.cc:1240: Saving into deprecated binary model format, please consider using `json` or `ubj`. Model format will default to JSON in XGBoost 2.2 if not specified.\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 800x470 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "A good exercise might be to make another XGBoost model with only the top 3 features and seeing how the accuracy compares to the model we just looked at."
      ],
      "metadata": {
        "id": "1ObiCw4SQ7lm"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "irV2gzF_ElKp"
      },
      "source": [
        "###Conclusion"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fsH_MscJEp_-"
      },
      "source": [
        "This wraps up our whirlwind tour of machine learning. Hopefully this gives you a good starting point to try using some of these algorithms for your own projects. Some last pieces of advice: take some time to get familiar with all they hyperparameters of the algorithm you are using. I didn't have the time to get into that for this session but scikit learn and XGBoost both have a ton of helpful options you can play with for their algorithms. Finally, don't get discouraged if you start a project and end up with a bunch of errors. Google search is your friend and a large part of getting better at ML projects, just like normal programming, is making mistakes and learning from them. Thanks for your time and happy coding!"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Additional Resources\n",
        "\n",
        "\n",
        "*   YouTube video on Shapley values: https://www.youtube.com/watch?v=UJeu29wq7d0&t=444s\n"
      ],
      "metadata": {
        "id": "TWzGuRZ1P2iy"
      }
    }
  ]
}