Breast_Cancer_detection.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Breast_Cancer_detection.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "mount_file_id": "1cHem-TvzG5Uv74MQBaSsqqbr6gOXO8Yh",
      "authorship_tag": "ABX9TyNqa8LwKs1iIZhHX+GcN5RP"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KcdEgTE04d0k"
      },
      "source": [
        "# Breast Cancer Detection\n",
        "___\n",
        "\n",
        "## Motivation:\n",
        "This is one of the many \"Hello World\" machine learning problems in the field. Many data scientists rookies begin with this problem at the beginning of their journey. This means that a simple google search including *breast cancer detection* will lead any of you on many datasets and notebooks dealing with the solutions.\n",
        "\n",
        "So **the motivation** here is not to (once again) follow the Machine learning project process from dataset to the solution but to get deeper at each step of the process to not only understand why each step is necessary but also and especially to understand deeply how it is performed on a mathematical and statistical basis.\n",
        "\n",
        "I believe that is the only why to really master the technics and the concept behind every line of code. But before get to coding, an important step (may be the first one to perform) is to get used to the problem and the data by trying to understand what each features (or column or attribute or statiscal variable) means for the probelm at hand.\n",
        "\n",
        "For juniors or beginners this work can help discovery what Data science work looks like and for more experienced people of teh data world this could be a revision session.\n",
        "\n",
        "## 1. Understand the problem context:\n",
        "You will find on the [Kaggle](https://www.kaggle.com/d3evil3/breast-cancer-detection) page an indepth description of the problem. So for summary's sake, let's just stress some major points here.\n",
        "\n",
        "> *My tips for this first part is to write a little summary like the one below for details to stick into your head.*\n",
        "\n",
        "### 1.1 What breast cancer\n",
        "Breast cancer is a major common malignancy cancer among women. It occurs as a results of **abnormal growth of cells in the breast tissue, commonly referred to as a Tumor**. A tumor does not mean cancer as tumors can be *benign (then not cancerous), pre-malignant (pre-cancerous), or malignant (cancerous)*. So pre-malignant and malignant tumors are cells **of your own body** that behave in a different way and keep dividing and pushes the other cells around it out of the way.\n",
        "These cells start invading into the tissues around it and sometimes entering the bloodstream and becoming what is called **a metastasis**.\n",
        "\n",
        "> The following tests are commonly used to diagnose breast cancer:\n",
        "* MRI\n",
        "* Mammogram\n",
        "* Ultrasound\n",
        "* Biopsy\n",
        "\n",
        "### 1.2 Defining the analytic approach\n",
        "What we are trying to do is to use some data to ease the detection process. Even if the later will always need the advice of an expert de confirm the diagnosis, we can imagine that a ML solution could be used at first hand to help the expert focus on positive patient.\n",
        "\n",
        "Looking at the data in a Excel sheet **(*if possible this step is a must to build a first impression of teh data at hand*)** we can see that:\n",
        "* There are **9 features or predictors** corresponding to cells characteristics\n",
        "* The last column named *\"class\"* corresponds to what we want to predict AKA **the target**\n",
        "\n",
        "Even if our target seems to only have numerical values (**2 - for benign, 4 - for malignant**) the later should be considered as a **nominal (without order) categorical statistical variable**. Thus the analytic approach to implement should be **a classification problem**.\n",
        "\n",
        "### 1.3 Data source and creator\n",
        "> **Data source:**\n",
        "* The data that we will use for this project come from the [UCI data repository](https://archive.ics.uci.edu/ml/datasets/Breast+Cancer)\n",
        "\n",
        "> * Another data source is in the [Kaggle data repository](https://www.kaggle.com/d3evil3/breast-cancer-detection/data). Please pay very close attention to the data range of the later data source when building the ML solution... Fill free to contact me is you don't understand what I mean.\n",
        "\n",
        "> **Creator of the UCI dataset:**\n",
        "* Dr. WIlliam H. Wolberg (physician)\n",
        "* University of Wisconsin Hospitals\n",
        "* Madison, Wisconsin, USA"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WcTD4eQ4NTlF"
      },
      "source": [
        "## 2. ML experimentation and developmeent:\n",
        "The first part of this notebook was about getting a first impression without too deep insights. The ppurpose of this kind of first exploration was to **design the ML solution** (i.e establishing functional and non-functional requirements).\n",
        "\n",
        "In this ML experimention part, we will focus on digging deeper into the data first and secondly experiment with ML algorithms."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n_J1RiUbPb9d"
      },
      "source": [
        "\n",
        "### 2.1 Quick data exploration\n",
        "Now is time to get our hands dirty, but never without thinking first about what we are trying to achieve. Loading the data means import the data into a IDE (Integrated development environment) - in our case a jupyter Notebook on Colab or Anaconda."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rvmtXFVK4VhD"
      },
      "source": [
        "# Let's import the basic necessary librairies\n",
        "# Tensor manipulation\n",
        "import numpy as np\n",
        "\n",
        "# Dataframe manipulation\n",
        "import pandas as pd\n",
        "pd.set_option('display.max_columns', None)\n",
        "pd.set_option('precision', 3)\n",
        "\n",
        "# Dataviz\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "%matplotlib inline"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 383
        },
        "id": "cXbqq47FUj01",
        "outputId": "30c098c8-e55a-4e55-b88e-0f74ad7a56ca"
      },
      "source": [
        "# import the dataset\n",
        "csv_path = '/content/drive/MyDrive/Colab Notebooks/Breast_cancer_detection/breastCancer.csv'\n",
        "#An alternative: pd.read_csv(csv_path, converters={'bare_nucleoli': lambda x: np.nan if x=='?' else int(x)})\n",
        "df = pd.read_csv(csv_path)\n",
        "\n",
        "# display the 5 first rows\n",
        "df.head(10)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>id</th>\n",
              "      <th>clump_thickness</th>\n",
              "      <th>size_uniformity</th>\n",
              "      <th>shape_uniformity</th>\n",
              "      <th>marginal_adhesion</th>\n",
              "      <th>epithelial_size</th>\n",
              "      <th>bare_nucleoli</th>\n",
              "      <th>bland_chromatin</th>\n",
              "      <th>normal_nucleoli</th>\n",
              "      <th>mitoses</th>\n",
              "      <th>class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1000025</td>\n",
              "      <td>5</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1002945</td>\n",
              "      <td>5</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>5</td>\n",
              "      <td>7</td>\n",
              "      <td>10</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1015425</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1016277</td>\n",
              "      <td>6</td>\n",
              "      <td>8</td>\n",
              "      <td>8</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>3</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1017023</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>1017122</td>\n",
              "      <td>8</td>\n",
              "      <td>10</td>\n",
              "      <td>10</td>\n",
              "      <td>8</td>\n",
              "      <td>7</td>\n",
              "      <td>10</td>\n",
              "      <td>9</td>\n",
              "      <td>7</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1018099</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>10</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>1018561</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>1033078</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>1033078</td>\n",
              "      <td>4</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        id  clump_thickness  size_uniformity  shape_uniformity  \\\n",
              "0  1000025                5                1                 1   \n",
              "1  1002945                5                4                 4   \n",
              "2  1015425                3                1                 1   \n",
              "3  1016277                6                8                 8   \n",
              "4  1017023                4                1                 1   \n",
              "5  1017122                8               10                10   \n",
              "6  1018099                1                1                 1   \n",
              "7  1018561                2                1                 2   \n",
              "8  1033078                2                1                 1   \n",
              "9  1033078                4                2                 1   \n",
              "\n",
              "   marginal_adhesion  epithelial_size bare_nucleoli  bland_chromatin  \\\n",
              "0                  1                2             1                3   \n",
              "1                  5                7            10                3   \n",
              "2                  1                2             2                3   \n",
              "3                  1                3             4                3   \n",
              "4                  3                2             1                3   \n",
              "5                  8                7            10                9   \n",
              "6                  1                2            10                3   \n",
              "7                  1                2             1                3   \n",
              "8                  1                2             1                1   \n",
              "9                  1                2             1                2   \n",
              "\n",
              "   normal_nucleoli  mitoses  class  \n",
              "0                1        1      2  \n",
              "1                2        1      2  \n",
              "2                1        1      2  \n",
              "3                7        1      2  \n",
              "4                1        1      2  \n",
              "5                7        1      4  \n",
              "6                1        1      2  \n",
              "7                1        1      2  \n",
              "8                1        5      2  \n",
              "9                1        1      2  "
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yEhfDrNjVQet",
        "outputId": "f77f0975-2594-4cf4-d015-d8c3a29f04b1"
      },
      "source": [
        "# print the shape of the dataset:\n",
        "print('The shape of the dataset: ', df.shape)\n",
        "print('The shape is about knowing exactely how many sample (Nr of rows) and how many features (Nr of columns) do we have in our dataset')"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The shape of the dataset:  (699, 11)\n",
            "The shape is about knowing exactely how many sample (Nr of rows) and how many features (Nr of columns) do we have in our dataset\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3jbrA87JVjA2",
        "outputId": "e8027c44-c43c-41cf-c96c-6a8194d66a8b"
      },
      "source": [
        "# Display the data types\n",
        "df.dtypes"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "id                    int64\n",
              "clump_thickness       int64\n",
              "size_uniformity       int64\n",
              "shape_uniformity      int64\n",
              "marginal_adhesion     int64\n",
              "epithelial_size       int64\n",
              "bare_nucleoli        object\n",
              "bland_chromatin       int64\n",
              "normal_nucleoli       int64\n",
              "mitoses               int64\n",
              "class                 int64\n",
              "dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b2b5jdtmjDrH"
      },
      "source": [
        "Displaying the data type of each columns is about knowing if we are dealing with numérical values or else. Depending on the type, we won't be able to use some python methods or technics. So it is always a good start to look at data type de plan, if any, what we call the **feature engineering** strategies (or the strategy of data transformation."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 320
        },
        "id": "YsNATVb8Wo0U",
        "outputId": "753202ae-5e2e-4c5a-b95f-86c2deba747a"
      },
      "source": [
        "# sumarize the distribution of each attribute (or features)\n",
        "df.describe()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>id</th>\n",
              "      <th>clump_thickness</th>\n",
              "      <th>size_uniformity</th>\n",
              "      <th>shape_uniformity</th>\n",
              "      <th>marginal_adhesion</th>\n",
              "      <th>epithelial_size</th>\n",
              "      <th>bland_chromatin</th>\n",
              "      <th>normal_nucleoli</th>\n",
              "      <th>mitoses</th>\n",
              "      <th>class</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>6.990e+02</td>\n",
              "      <td>699.000</td>\n",
              "      <td>699.000</td>\n",
              "      <td>699.000</td>\n",
              "      <td>699.000</td>\n",
              "      <td>699.000</td>\n",
              "      <td>699.000</td>\n",
              "      <td>699.000</td>\n",
              "      <td>699.000</td>\n",
              "      <td>699.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>1.072e+06</td>\n",
              "      <td>4.418</td>\n",
              "      <td>3.134</td>\n",
              "      <td>3.207</td>\n",
              "      <td>2.807</td>\n",
              "      <td>3.216</td>\n",
              "      <td>3.438</td>\n",
              "      <td>2.867</td>\n",
              "      <td>1.589</td>\n",
              "      <td>2.690</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>6.171e+05</td>\n",
              "      <td>2.816</td>\n",
              "      <td>3.051</td>\n",
              "      <td>2.972</td>\n",
              "      <td>2.855</td>\n",
              "      <td>2.214</td>\n",
              "      <td>2.438</td>\n",
              "      <td>3.054</td>\n",
              "      <td>1.715</td>\n",
              "      <td>0.951</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>6.163e+04</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>2.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>8.707e+05</td>\n",
              "      <td>2.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>2.000</td>\n",
              "      <td>2.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>2.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>1.172e+06</td>\n",
              "      <td>4.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>2.000</td>\n",
              "      <td>3.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>2.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>1.238e+06</td>\n",
              "      <td>6.000</td>\n",
              "      <td>5.000</td>\n",
              "      <td>5.000</td>\n",
              "      <td>4.000</td>\n",
              "      <td>4.000</td>\n",
              "      <td>5.000</td>\n",
              "      <td>4.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>4.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.345e+07</td>\n",
              "      <td>10.000</td>\n",
              "      <td>10.000</td>\n",
              "      <td>10.000</td>\n",
              "      <td>10.000</td>\n",
              "      <td>10.000</td>\n",
              "      <td>10.000</td>\n",
              "      <td>10.000</td>\n",
              "      <td>10.000</td>\n",
              "      <td>4.000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "              id  clump_thickness  size_uniformity  shape_uniformity  \\\n",
              "count  6.990e+02          699.000          699.000           699.000   \n",
              "mean   1.072e+06            4.418            3.134             3.207   \n",
              "std    6.171e+05            2.816            3.051             2.972   \n",
              "min    6.163e+04            1.000            1.000             1.000   \n",
              "25%    8.707e+05            2.000            1.000             1.000   \n",
              "50%    1.172e+06            4.000            1.000             1.000   \n",
              "75%    1.238e+06            6.000            5.000             5.000   \n",
              "max    1.345e+07           10.000           10.000            10.000   \n",
              "\n",
              "       marginal_adhesion  epithelial_size  bland_chromatin  normal_nucleoli  \\\n",
              "count            699.000          699.000          699.000          699.000   \n",
              "mean               2.807            3.216            3.438            2.867   \n",
              "std                2.855            2.214            2.438            3.054   \n",
              "min                1.000            1.000            1.000            1.000   \n",
              "25%                1.000            2.000            2.000            1.000   \n",
              "50%                1.000            2.000            3.000            1.000   \n",
              "75%                4.000            4.000            5.000            4.000   \n",
              "max               10.000           10.000           10.000           10.000   \n",
              "\n",
              "       mitoses    class  \n",
              "count  699.000  699.000  \n",
              "mean     1.589    2.690  \n",
              "std      1.715    0.951  \n",
              "min      1.000    2.000  \n",
              "25%      1.000    2.000  \n",
              "50%      1.000    2.000  \n",
              "75%      1.000    4.000  \n",
              "max     10.000    4.000  "
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U9CP0i4tXVqk"
      },
      "source": [
        "These are what we can observe from this first exploration:\n",
        "* There are 699 samples with 11 columns (the lat column being our target)\n",
        "* Apart from the `id` column and the target the `class` column), the features numerical values range from 1 to 10 (this meaning they are approximetely in the same range)\n",
        "* From the above 'count' row we can see that there are no missing values in any of the samples (i.e. **impute** or filling missing values strategy won't be needed)\n",
        "* Something seems to be wrong with the `bare_nucleoli` column which doesn't appear in the summary table and has an **object data type** despite the fact the all the data seems to be numerical\n",
        "* All the others columns are interger data type\n",
        "\n",
        "These info are the bare minimum you need to ask yourself when looking at a dataset to spot issues like on `bare_nucleoli`. \n",
        "\n",
        "To proceed you need to fix the issue on the later. So let's display a description and the distribution of this columns which seem to be an *object* data type **AKA categorical**."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RxIKrtaYXIHn",
        "outputId": "d7c73c81-a208-479f-d539-934d2fa0c73c"
      },
      "source": [
        "# display the description of bare_nucleoli\n",
        "print(df['bare_nucleoli'].describe(), '\\n')\n",
        "print('unique modalities or bare_nucleoli column: ', df['bare_nucleoli'].unique()) #.value_counts(normalize=True))"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "count     699\n",
            "unique     11\n",
            "top         1\n",
            "freq      402\n",
            "Name: bare_nucleoli, dtype: object \n",
            "\n",
            "unique modalities or bare_nucleoli column:  ['1' '10' '2' '4' '3' '9' '7' '?' '5' '8' '6']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 330
        },
        "id": "rtvBqQoVabat",
        "outputId": "a86bd683-4747-4c87-d128-3372b91985f2"
      },
      "source": [
        "# the class distribution among the 2.3%\n",
        "df.loc[df['bare_nucleoli'] == '?', 'class'].value_counts(normalize=True).plot.bar()\n",
        "plt.title('Target distribution having the question mark in bare_nucleoli column')\n",
        "print(df.loc[df['bare_nucleoli'] == '?', 'class'].value_counts())"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2    14\n",
            "4     2\n",
            "Name: class, dtype: int64\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s7TSZhzta9fZ"
      },
      "source": [
        "We can see that among the modalities **there is a question mark**. On the associated samples, we can make the following observations:\n",
        "* There are 16 samples **(representing about 2.3% of the sample)**\n",
        "* Among the 2.3% there are 90% tagged as not cancerous and 10% as cancerous.\n",
        "\n",
        "We can then ask ourselves what does the question mark means? (that the data is missing or something else?). The answer to that question require some field knowledge and to know more about the data collection process. **To proceed we will assume that the question mark is equivalent to a missing value. But you should note that you on a real project you shoudn't bother NOT asking the question!**\n",
        "\n",
        "To fix the issue we have the following options:\n",
        "1. Drop the associated samples (i.e dropping the associated rows)\n",
        "2. Impute the missing values with the mean value\n",
        "3. Impute the missing values with the median value of teh distribution\n",
        "4. Use an ML algorithm to predict what the missing value could be.\n",
        "\n",
        "The first strategy is the easiest but could modify the dataset distribution. In our case with the missing value representing only about 2.3% of the entire dataset, this could be a valid solution. The last solution is a more complex one that could be used in a last resort.\n",
        "\n",
        "Apart from the question mark, all the others modalities of the bare_nucleoli column seem to be of type integer, so we can't really use the mean imputing strategy. **We will choose to impute the missing values with the median value.**\n",
        "\n",
        "> **Note the fact that you should always ask yourself what are the possible strattegies before going on. Every thing you choose to do should be volontary and not by default**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 292
        },
        "id": "HKQl0wcDnyos",
        "outputId": "13602b6c-bb61-42dc-c1bd-3055a6206821"
      },
      "source": [
        "# Display the target distribution before transformation\n",
        "df['class'].value_counts(normalize=True).plot(kind='bar')\n",
        "plt.title('class distribution before transformation')\n",
        "plt.xlabel('Target classes')\n",
        "plt.ylabel('class distribution')\n",
        "plt.show()"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yuy9qDTbGUkU"
      },
      "source": [
        "Let's now transform the bare_nucleoli column. For this we will impute the missing values with the median value. Since that is the only operation that we will perform, no need to implement a pipeline *(otherwise a pipeline would be needed to avoid anay data leak)*."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Douf878QCOSP",
        "outputId": "4df5746e-d53c-4766-b100-ae8c831ee725"
      },
      "source": [
        "# Imputing the missing value\n",
        "\n",
        "# Change the question marks to NaN\n",
        "# alternative 1: df['bare_nucleoli'] = df['bare_nucleoli'].replace('?', np.nan)\n",
        "df['bare_nucleoli'] = pd.to_numeric(df['bare_nucleoli'], errors='coerce', downcast='integer')\n",
        "\n",
        "# Fill NaN with median values\n",
        "df['bare_nucleoli'] = df['bare_nucleoli'].fillna(df['bare_nucleoli'].median())\n",
        "\n",
        "# Convert value to integer data type\n",
        "df['bare_nucleoli'] = df['bare_nucleoli'].astype('int64')\n",
        "\n",
        "# display the result\n",
        "print(df.dtypes)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "id                   int64\n",
            "clump_thickness      int64\n",
            "size_uniformity      int64\n",
            "shape_uniformity     int64\n",
            "marginal_adhesion    int64\n",
            "epithelial_size      int64\n",
            "bare_nucleoli        int64\n",
            "bland_chromatin      int64\n",
            "normal_nucleoli      int64\n",
            "mitoses              int64\n",
            "class                int64\n",
            "dtype: object\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LL4VOYX-I50Z",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 292
        },
        "outputId": "cdbca2b8-62b1-44ce-e653-553ec0a52654"
      },
      "source": [
        "# Display the target distribution after transformation\n",
        "df['class'].value_counts(normalize=True).plot(kind='bar');\n",
        "plt.title('class distribution after transformation')\n",
        "plt.xlabel('Target classes')\n",
        "plt.ylabel('class distribution')\n",
        "plt.show()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0ac4fxpSvYvW"
      },
      "source": [
        "We can conclude from this plot that the imputation strategy (as expected) doesn't change the class repartition and that the class **are not imbalanced** (the classes' repartition - that needs special technics - is not an issue)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mwxuNCPNPLJV"
      },
      "source": [
        "### 2.2 Exploratory Data Analysis (EDA)\n",
        "\n",
        "This part intention is to actually build knowledge from series of numbers. Since we are only dealing with numerical value, this means that we need to build a statistical understanding on each features *(i.e. how are their respective distribution, where is located the mean, the median etc.)*.\n",
        "\n",
        "> Start to study each feature alone and then group then together."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aWFvEUABSpce"
      },
      "source": [
        "#### Unimodal data **visualisation**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7VswgI29JG-k",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 607
        },
        "outputId": "4c1d5fe6-e02f-4718-c7e3-76ba926b8626"
      },
      "source": [
        "# features histograms\n",
        "df.hist(figsize=[15, 10]);"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 12 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 484
        },
        "id": "iGUrF7kHVcpD",
        "outputId": "1da30b53-2873-4906-c174-f7f2f438de7e"
      },
      "source": [
        "# Boxplot\n",
        "df.iloc[:,1:-1].boxplot(figsize=(15, 8));"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2cAAAHTCAYAAACjj2AvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3df7xldV0v/tcbZgQlMwHv2A9wLEtRFEvqlnRzyB8k2q8blhUm1tUIL2LeFLRfdksENfNHAokpFKVfpTLFX5jOUYMyQQVBMrs2YSWWQhYkOMjn+8daZ9gzc86ZM+fHPp85PJ+Px3nM2mvvvdZnv9dn/Xittfaeaq0FAACAtbXfWjcAAAAA4QwAAKALwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6sGGaMzv00EPb5s2bpznLFXHLLbfkoIMOWutm3KWo+fSp+fSp+fSp+fSp+fSp+fSp+fTtyzW/8sorv9Bau89cz001nG3evDlXXHHFNGe5ImZmZrJly5a1bsZdippPn5pPn5pPn5pPn5pPn5pPn5pP375c86r6x/mec1sjAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOjAHsNZVb2+qv61qq6ZGHdwVb23qj49/nvv1W0mAADA+rZhEa+5IMnvJvmDiXFnJHlfa+2sqjpjfHz6yjcPYP2pqt3GtdbWoCV3HWrOXYF+Pn1qPn3rveZ7vHLWWvtgkht3Gf3DSS4chy9M8iMr3C6AdWlyp3LUUUfNOZ6VNVnbE088cc7xsK+b7M/Pf/7z5xzPypqs7RFHHDHneFbWbG3322+/vPSlL81+++230/j1YKnfOdvUWvvcOHxDkk0r1B6Au4TWWl7xilesq7N9vWut5ed+7ufUnHWttZbHPe5x+vkUtdZyzjnnqPmU7LfffvnqV7+ao48+Ol/96ld3BLT1YjG3NS6otdaqat7eWFXPSPKMJNm0aVNmZmaWO8sVceyxx67q9Ldu3bqq098Xqfn0qXmfjjrqqMzMzOTmm2/OzMxMjjrqqFx11VXdbB/XoxNPPHGnmp944om56KKL1HyJbFv69PznP3+nfv785z8/L37xi/XzVXTEEUfsVPMjjjgi1113nZqvorPPPnunmp999tl57nOfu25qXotJ+VW1OcklrbUjx8efSrKltfa5qvr6JDOttQfuaTpHH310u+KKK5bX4jWw+Yx3ZNtZT1jrZtylqPn0qfl0zN560VrLzMxMtmzZstM4Vp6ary3blunQz6dPzaevqnZcOZut+f7775877rhjn6p5VV3ZWjt6rueWeh3wbUmeOg4/NcmfL3E6AHdJVZVnP/vZ6+o++d5VVX7/939fzVnXqiqXXnqpfj5FVZVTTjlFzafkjjvuyP77758rrrhiRzBbTxbzU/pvTPJXSR5YVf9UVT+X5Kwkj62qTyd5zPgYgD2YPLN31VVXzTmelTVZ24suumjO8bCvm+zPL37xi+ccz8qarO11110353hW1mxt77jjjjz3uc/dEczWU80X82uNP9la+/rW2sbW2je11n6/tfbF1tqjW2vf2lp7TGtt119zBGAerbW01rJ169Ydw6wuNeeuQD+fPjWfvvVe8/X18yYAAAD7KOEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwBTdvjhh6eqcuyxx6aqcvjhh691k9a9U089NQceeGCOPfbYHHjggTn11FPXuknAOlBVO23Pq2qtm8Q+TjgDmKLDDz88n/3sZ/PIRz4yb3nLW/LIRz4yn/3sZwW0VXTqqafmvPPOy5lnnpl3vetdOfPMM3PeeecJaMCyTAaxF73oRXOOh70lnAFM0Wwwu+yyy3LooYfmsssu2xHQWB3nn39+zj777DznOc/JgQcemOc85zk5++yzc/75569104B1oLWWRz7ykWmtrXVTWAeEM4Apu/jiixd8zMq67bbbcvLJJ+807uSTT85tt922Ri2CvTd7y9xi/iZvsVvsH0tzySWXLPgY9pZwBjBlJ5xwwoKPWVkHHHBAzjvvvJ3GnXfeeTnggAPWqEWw91pri/673+mX7NXrXfFZuic+8YkLPoa9JZwBTNFhhx2Wyy+/PMccc0y+8IUv5Jhjjsnll1+eww47bK2btm49/elPz+mnn56Xv/zlufXWW/Pyl788p59+ep7+9KevddOAdaCqcvnll7sCyYrYsNYNALgruf7663P44Yfn8ssvz+WXX55kCGzXX3/9Grds/Xr1q1+dJHnBC16Q2267LQcccEBOPvnkHeMBlqK1tiOQ/fIv//JO42GpXDkDmLLrr78+rbVs3bo1rTXBbApe/epX59Zbb83WrVtz6623CmbAipi9LXR2ey6YsVzCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdWFY4q6pfrKprq+qaqnpjVR24Ug0DAAC4K9mw1DdW1TcmeVaSB7fWvlxVb07y5CQXrFDbAGBFVNVu41pra9ASAJbjwAMPzG233bbj8QEHHJBbb711DVu0spZ7W+OGJHevqg1J7pHkX5bfJABYOZPB7HnPe96c4wHo32ww27RpU97whjdk06ZNue2223Lggevn5r0lh7PW2j8neVmS65N8LsmXWmuXrlTDAGAltdby+Mc/3hUzgH3UbDC74YYbsnnz5txwww07Atp6sZzbGu+d5IeT3D/Jvyd5S1Wd2Fq7aJfXPSPJM5Jk06ZNmZmZWXpr19C+2u59mZpPn5ov37HHHruq09+6deuqTn+9et7znpeZmZncfPPNmZmZyfOe97y85CUv0eenRJ2nT82Xz/a8T2edddZO2/OzzjorT3va09ZNn6+lnkGsqicl+YHW2s+Nj38myXe31k6Z7z1HH310u+KKK5Y0v7W0+Yx3ZNtZT1jrZtylqPn0qfn0qfl0zN6+2FrLzMxMtmzZstM4Vpd+Pn1qPn1qPh1VtePK2ez2/L73vW8+//nP71Pb86q6srV29FzPLec7Z9cn+e6qukcNe7lHJ7luGdMDgFVTVXnXu97lu2YA+6gDDjggn//853Pf+94327Zt2xHMDjjggLVu2opZ8m2NrbUPV9XFST6a5PYkH0vy2pVqGACshNbajkD2kpe8ZKfxAOw7br311hx44IH5/Oc/n6c97WlJ/FrjTlprv95ae1Br7cjW2lNaa+vn23gArButtbTWsnXr1h3DAOx7br311p225+spmCXL/yl9AAAAVoBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHNqx1AwAAABZj48aNuf3223c83rBhQ7Zv376GLVpZrpwBAADdmw1m9773vXP++efn3ve+d26//fZs3LhxrZu2YoQzAACge7PB7MYbb8wDHvCA3HjjjTsC2nrhtkYA9mlVtarTb62t6vQBWLwPfOADuz1+2MMetkatWXmunAGwT2utLfrvfqdfslevF8wA+vKoRz1qwcf7OuEMAADo3oYNG3LTTTfl4IMPzt///d/n4IMPzk033ZQNG9bPzYDr55MAAADr1vbt27Nx48bcdNNNefrTn57ErzUCAACsie3bt6e1lq1bt6a1tq6CWSKcAQAAdEE4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANCBZYWzqvq6qrq4qv62qq6rqu9ZqYYBAABMOu6447Lffvvl2GOPzX777ZfjjjturZu0opZ75eyVSd7dWntQkqOSXLf8JgEAAOzsuOOOy6WXXpqTTz45b3/723PyySfn0ksvXVcBbcNS31hV90ryfUlOSpLW2leSfGVlmgUAAHCn9773vfmFX/iFnHPOOZmZmck555yTJDnvvPPWuGUrZ8nhLMn9k/xbkjdU1VFJrkxyWmvtlskXVdUzkjwjSTZt2pSZmZllzHJ+z3zfLbll+6pMOkmy+Yx3rNq0D9qYvObRB63a9PdVq9VXmJ+aT5+aT5+aT5+aT5+aT5+ar77WWo4//vjMzMzk5ptvzszMTI4//vice+6566b+ywlnG5J8R5JTW2sfrqpXJjkjya9Ovqi19tokr02So48+um3ZsmUZs5zfLe9+R7ad9YRVmfbMzExWq93JEPxWc/r7pHerydSp+fSp+fSp+fSp+fSp+fSp+VRUVd75znfuuHK2ZcuWnHLKKamqdVP/5YSzf0ryT621D4+PL84QzgAAAFbUYx/72Jx77rlJkuOPPz6nnHJKzj333DzucY9b45atnCWHs9baDVX12ap6YGvtU0keneSTK9c0AACAwXve854cd9xxOe+883LuueemqvK4xz0u73nPe9a6aStmOVfOkuTUJH9UVXdL8pkkT1t+kwAAAHY3G8RW+2tHa2VZ4ay19vEkR69QWwAAAO6ylvv/nAEAALAChDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHdiw1g0AAABYjEMOOSQ33njjjscHH3xwvvjFL65hi1aWK2cAAED3ZoPZQx7ykLzxjW/MQx7ykNx444055JBD1rppK0Y4AwAAujcbzK655prc9773zTXXXLMjoK0XbmtkyY76jUvzpS9vX7Xpbz7jHas27XvdfWOu+vXHrdr0V4uaA6vBtmX61ByW5p3vfOduj+93v/utUWtWnnDGkn3py9uz7awnrMq0Z2ZmsmXLllWZdrK6O63VpObAarBtmT41h6U5/vjjc8011+z0eD1xWyMAANC9gw8+ONdee22OPPLI3HDDDTnyyCNz7bXX5uCDD17rpq0YV84AAIDuffGLX8whhxySa6+9Nj/5kz+ZxK81AgAArIkvfvGLaa1l69ataa2tq2CWCGcAAABdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6MCG5U6gqvZPckWSf26tPXH5TQIAANhdVe02rrW2Bi1ZHStx5ey0JNetwHQAAADmNBvMNm7cmFe+8pXZuHHjTuPXg2WFs6r6piRPSPK6lWkOAADA3DZu3JivfOUredjDHpavfOUrOwLaerHc2xpfkeR5Se453wuq6hlJnpEkmzZtyszMzDJnObd7HnFGHnrhGasy7STJhas36XsekczMHLR6M1hFq7U8b7755lWb9qzVnv5qUfPpeub7bskt21dv+pvPeMeqTfugjclrHr3vbVvUfG3Ytkyfmk+Xbcv68LKXvSwzMzM7+vnLXvaynHbaaftkn5xTa21Jf0memOSccXhLkkv29J5HPOIRbbXc7/RLVm3aW7duXbVpt7a6bV9Naj59aj59aj59aj59aj59aj59ar7vS9I2btzYWruz5hs3bmxDpNl3JLmizZOXlnNb4zFJfqiqtiV5U5Lvr6qLljE9AACAeW3fvj13u9vdcvXVV+dud7tbtm9fxcuha2DJ4ay19vzW2je11jYneXKS97fWTlyxlgEAAIza+KuM27dvz2mnnbYjmM2OXw/8P2cAAMA+Yfb2v61bt05+3WrdWPb/c5YkrbWZJDMrMS0AAIC7IlfOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOjAhrVuAAAAwGJU1W7jWmtr0JLV4coZAADQvclg9uQnP3nO8fs64QwAANhntNby8z//8+vqitkstzUCAEzZPY84Iw+98IzVm8GFqzfpex6RJE9YvRnAAl7wghfs9vjMM89co9asPOEMAGDK/vO6s7LtrNUJODMzM9myZcuqTDtJNp/xjlWbNuzJmWeemRe96EU7PV5P3NYIAADsM6oqv/d7v7euvms2SzgDAAC6N/kdsze96U1zjt/XCWcAAMA+obWW1lq2bt26Y3g9Ec4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6MCSw1lVHVZVW6vqk1V1bVWdtpINAwAAuCvZsIz33p7k/7TWPlpV90xyZVW9t7X2yRVqGwAAwA5Vtdu41toatGR1LPnKWWvtc621j47D/5nkuiTfuFINAwAAmDUZzE444YQ5x+/rVuQ7Z1W1Ocm3J/nwSkwPAABgLq21PPOZz1xXV8xmLee2xiRJVX1Nkj9J8uzW2n/M8fwzkjwjSTZt2pSZmZnlznJem894x6pNO+9evWkftDGrWpfVtFrtvvnmm1e9Jmq+MzWfn5pP1z2POCMPvfCM1ZvBhas36XsekczMHLR6M1hF+vn0OW6ZPv1833fCCSdkZmZmR81POOGEXHzxxeunPq21Jf8l2ZjkPUmes5jXP+IRj2j7ovudfslaN6FLq1mXrVu3rtq0W9t3l6maT5+ary9qMjf9fH1Rk7np5/u+JG2IL3fWfHLcviLJFW2evLScX2usJL+f5LrW2suXnRIBAAD2oKrymte8Zl1912zWcr5zdkySpyT5/qr6+Ph3/Aq1CwAAYIc28R2ziy++eM7x+7olf+estfaXSdZfXAUAALo0G8RmZmayZcuWtW3MKliRX2sEAABgeYQzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANAB4QwAAKADwhkAAEAHhDMAAIAOCGcAAAAdEM4AAAA6IJwBAAB0QDgDAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANCBaq1NbWZHH310u+KKK6Y2v4VU1apOf5p1XSsPvfCha92EZfnEUz+x1k3Ya2o+fWreP9vz5dPP+6efL59+Pn1qPrequrK1dvScz91Vw9nemJmZyZYtW9a6GXcpaj59aj59aj59aj59aj59aj59aj59+3LNFwpnbmsEAADogHAGAADQAeEMAACgA8IZAABAB4QzAACADghnAAAAHRDOAAAAOiCcAQAAdEA4AwAA6IBwBgAA0AHhDAAAoAPCGQAAQAeEMwAAgA4IZwAAAB0QzgAAADognAEAAHRAOAMAAOiAcAYAANCBaq1Nb2ZV/5bkH6c2w5VzaJIvrHUj7mLUfPrUfPrUfPrUfPrUfPrUfPrUfPr25Zrfr7V2n7memGo421dV1RWttaPXuh13JWo+fWo+fWo+fWo+fWo+fWo+fWo+feu15m5rBAAA6IBwBgAA0AHhbHFeu9YNuAtS8+lT8+lT8+lT8+lT8+lT8+lT8+lblzX3nTMAAIAOuHIGAADQAeEMAADoRlX9UFWdMQ7/SFU9eK3bNC1dhrOqemFV/dIazHdLVT1y4vEFVXXCHK/7hqq6eA/T2lZVh65GO6ehql63litCVb2zqr5uHH5WVV1XVX+0CvPZ8Tmr6gUrPf0F5rtP9I9d+3pVvbGqrq6qX1yFeV0+/ru5qn5qpae/Eib75RLeuyLbtaq6efx3S1VdsgLT27ED7EVV/d+qesw4/OyqusfEczfv5bROqqrfHYdPrqqf2cPr93o59VjDSeM6dc1at2NvTC635bx3Mct8CdOfs55VNVNVK/Kz3kvZR8x3zDItux5AT67H68m09t/L6U+T713Ofmsttdbe1lo7a3z4I0nuMuFsw1o3oDNbktyc5PKFXtRa+5cka7YBnIbW2v9a4/kfP/HwlCSPaa3902LeW1UbWmu3L3I+k5/zBUnOXHwr17/Jvl5V903yna21Byz2/Xu5LGZPjGxO8lNJ/njvWrt8e2rvLv1yXWitvS3J29a6HZNaa7828fDZSS5K8l8rMN3zljuNeabbXQ1Xyt6swz1arWW+r1rl5fkjSS5J8slkt/W4C/t6f16KHvdbVbU5ybuT/HWSRyb5SJI3JPmNJP8tyU9nCGNHZzgW+KEkj6qqX0nyY0numeS8JPdI8v+S/Gxr7aaqelaSk5PcnuSTrbUnV9VBSV6d5MgkG5O8sLX251X1kHGed8twserHWmufXv1Pv2ddXDmrqp8Zz8ZfVVV/uMtzk+n/0KraNg6fVFVvrar3jmcx/ndVPaeqPlZVf11VB0+8/5VV9fGquqaqvmueNmzOsEB/cXzt/xif+r6quryqPjN7RmryrFlV7V9VLxunfXVVnbrLdO9eVe+qqqeP77uuqs6vqmur6tKquvv4um+pqndX1ZVV9aGqetA4/knjtK+qqg+O4x5SVX8ztvPqqvrWZdb/oKp6xziPa6rqJ2brPp4R/vj496mq+ofxPY+oqg+M7X1PVX39AtNfaBn+6fi5P11VL5l4z7bxtecl+eYk76qqX6yqg8flfvW4nB82vv6FVfWHVXVZkj8cH1841vIfq+p/VtVLquoT4/w2Tratqs5Kcvfxc/5RDWf8nj3RnhdV1WkrVd/xqVOr6qNjm2aX93dV1V+N/fjyqnrgRK3+fGzvp6vq1yemf+JEf/i9qtp/gbbcPDF8QlVdMA5fUFWvWqivJ7k0yTfOrh9V9fBxGVxdVX9WVfeeqOkrquqKJKeNj3+nqq4Y+/93jsv901X1W3O07awk/2Oczy9W1Qer6uETr/vLqjpql8+1uar+dvwcfzcuw8dU1WXjfL5rD7V9W1W9P8n7quoeVfXmqvrk+Lk+PNF/Z/vlQuvy06vqI+Py/pOauOqzh34y5/uq6v5juz8xWa/R11TVxeNn/6OqqvE9c66fNVyF/uS4zN408flnrzJsrqr3j8+/r6oOX6h/7I25+mlV3Tz2jWvH+d1nYn4n1LCj/YYkW6tq68S0XjTW6a+ratM47j5j3T4y/h0zRxt2XBVbxnLaUw0/PvH35ap6VA3bgNePn/9jVfXDe1u/FbBh7CPXjX3mHlX1a2MNrqmq1070n13X4b3d3p89fta/q3FfWrtcCauqS6pqyzj8AzVsC6+qqvfNMc05l23Nsz/Y5b17fSV0kXar5y7zPbeGbd61VfUbE+O3VdVv1O7b/kNq2I5cW1WvS1ILzbzmP26a65hlSw37wrcl+WRVHVhVbxjn/7GqOnZ83WKPq3Zbd2q46+iHkrx07PvfUhNX8ub73EtR82x/a+X3SW8d+/y1VfWM5bRtoh1zHQsteBw5vuZxNewHPlpVb6mqrxnHP3pcPp+oYRtzwBzv3VZ93uiC0woAAA6VSURBVKnzgCS/neRB499PJfneJL+U4WR5kqS1dnmGk1/Pba09vLX2/5L8QZLTW2sPS/KJJLPHRGck+fZx/MnjuF9O8v7W2nclOTZDHz1ofP6VrbWHZwiBi7oAMBWttTX9S/KQJH+X5NDx8cFJXpjkl8bHM0mOHocPTbJtHD4pyd9nSM/3SfKlJCePz/1OkmdPvP/8cfj7klyzQFt2zHd8fEGSt2QIsQ9O8vfj+M2z00nyC0kuTrJhtv3jv9vG1/1Fkp+ZeN/tSR4+Pn5zkhPH4fcl+dZx+L9n6EjJ0Om+cRz+uvHfVyf56XH4bknuvsxl8GOzNRof32uy7hPj35zkmRnOPFye5D7j+J9I8voFpr/QMvzMOL8Dk/xjksMm6nfoHMOvTvLr4/D3J/n4xLK7crYW4+O/HNt6VIaz7o8fn/uzJD8yR9tunmjz5iQfHYf3y3Bm5pAVrO+2JKeOj09J8rpx+Gsn+tJjkvzJRK0+l+SQJHdPck2GjckRSd6eZOP4unMy9rd52jL5GU9IcsFe9PUdw+Pjq5M8ahz+v0leMVHTc3ZZ/mePw6cl+ZckX5/kgAwbw0Mm25bhCvYlE+9/6sS0vy3JFXN8rs0Z1q2Hjp/hyiSvz3CA88NJ3rqH2v5T7lx3fynJ743DR47Tne0j2zL04dn5zbUuHzLRrt+aWM4vzMT2ZY7PMN/73pY7tyHP3KVOX0ryTeNn/qsMO7Z518+x9gfssj05KcnvjsNvT/LUcfhnk7x1of6xF+vAnP00Scud27Jfm2jHBUlO2HX9Hx+3JD84Dr8kya+Mw3+c5HvH4cOTXDfH59uxDJaxnBas4cTrfjDJh8blceZE//i6DPu8g5ayPVniNmjzWLdjxsevz9DPD554zR9O1HUm4zq8UH+aZ14zSX57HD4+yV/MVaMMV1i2ZNh/fzbJ/cfxB+/6+gWW7Xz7gzmX+RTqOZM7txWzn2P/cfzDJvrzXNv+VyX5tXH4CeP0D51n/rsdNy20no51vmWixv8nd24THpTk+gz74JOyuOOq+dadCzKut/Osx7t97mXUf7ftb1Z+nzRb19l97iETn2W+ZTNn2ybmO9ex0HzHkTMZ9vOHJvlgxm1GktMzbC8PzLDufNs4/g+y87Hv0Xtq71r9jXX69MTjP8id+4JvTvLx7LweT/aleyW5fuK935I7j9fePdbyxCRfM467Ylx+Hx//rs+wT/qpJNeO9fzWta7J5F8PtzV+f5K3tNa+kCSttRurFjxhNGlra+0/k/xnVX0pw84/GQLN5Bm0N47T/mBVfW1VfV1r7d8XOY+3ttbuyHC2adMczz8myXltvEzeWrtx4rk/T/KS1trkd6X+obX28XH4yiSbxzMgj0zylonPPnv247IkF1TVm5P86Tjur5L8clV9U5I/bcu/DPuJJL9dVWdnOCj+0K7LoKqel+TLrbXXVNWRGQ5a3zu+bv8MwWEp3tda+9I4j08muV+Gjc18vjdD2Elr7f3j2cavHZ97W2vtyxOvfVdrbXtVfWJs47snPu/mhRrVWttWVV+sqm9PsinJx1prX9zLzzZrvvrOLs8rk/zPcfheSS6s4Wpoy3BgNOu9s22oqj/NUIvbkzwiyUfGad49yb8usZ176us7VNW9MhyYfmAcdWGGg4JZ/98ub5m95esTSa5trX1unM5nkhyWZKHaviXJr1bVczMEhgvmed0/tNY+MU732gx9q43Lf3P2XNvZdfd7k7wySVpr11TV1QvMb6d1eRw+cjz7+nVJvibJexb4bJPme98xGft8hgPosyfe8zdtvN23qj4+tuHfM//6eXWSP6qqt2YIrLv6ntzZF/8wQ/iZtej+MYdHZ+5+ekfu7CsX5c51YiFfyXBgnwx1f+w4/JgkD57Ydn3t7NnleSx1Oe2phhn72EuTHDtugx6X5Ifqzis4B2YMGYuc50r4bGvtsnH4oiTPSvIP47b9HhlOjF6bO/ejs8vlgdn77f3ktm3zHl773Uk+2Fr7h2S3feis+ZbtQvuD1TZXPSf9+Hi1ZUOGA/8HZ+g7ydzb/u+bHW6tvaOqblpg3rsdN008N996+jezNc5Qt1eP7/3bqvrHDCe+ksUdVy113Znrcy/Vrtvfb8nK75OeVVU/Or7usCTfmoX3VfO1bfMeXr/QcWQyrCMPTnLZuA7cLcNx4APHef3d+LoLM5zAe8Ui2tiD2yaG75h4fEeW/rWrJ2RYl34ww3HyQzOcpP2x1tqndnntdVX14fE976yqn2+tvX+J811RPYSzPbk9d95+eeAuzy12wbZd3rfr44VMzmPRqXF0WZIfqKo/bmN832V6X81wkLJfkn9vw6XVnRva2slV9d8zdJ4rq+oRrbU/XskO1Vr7u6r6jgxnOX+rdrmtpIYv9D4pQ4dPhjpc21r7nkXOYrHL8KtZXp+8Za5pt9buqKrtE8tgsSv+6zKcublvhjOjS7JAfWc/++Tn/s0MO8cfreFW25nJSe066QzL4sLW2vMX25yJ4YWWxd729V3NuSyy83o6+3jBZdFa+6+qem+GK2A/nuEgfy572h4sVNtd27sYc63LyRAef6S1dlVVnZThrPViLPS++bZZc60/C62fc+24Fms5/WPOflpVv7rL6xazbZ5clyfXnf2SfHdr7dZd5jHfdC7I0pbTgjUcQ8Obkzx99oAv8x8cTNNc249zMpxd/2xVvTA7bxNm14m93d4nc2/bJvcDye7bn4Xs7bKdhnmPK6rq/hmupH1nG74Hc0F2/rxz1WelzLeeLnYbt5jjqguytHVnJT/3rtu+Pf3gxV7tk2q45fYxSb5n3AfNZPF9dr59w0LHQgupDCcQf3Knkbvc3r/O/WeGK7pprX2pqm6qqv/RWvtQkqck+UBV7Zfh7qutVfWXSZ6cO08enFpVp44nbL+9tfaxqvrmJJ9prb2qhlv4H5aki3DWw3fO3p/kSVV1SDLcQ77L89ty58HYUn+E4yfGaX9vki/NXqmZw46Fvxfem+Tnq2rDOI/J9v9akpuSvGahCbTW/iPDGcwnjdOo2ZWuqr6ltfbhNnyx9t+SHDbZoTJcndvtPvu9UVXfkOS/WmsXZTjb+x0Tz91vbP+TJq5KfSrJfarqe8bXbKzhi5Xz2ZblL8NZH8rwRdGMG88vjPVbCdtr/C7a6M+S/ECS78zizwzuZqH6zuFeSf55HD5pl+ceW8N3LO6e4YvXl2W4HfaEqvpv47wOHpfZfD5fVUeMG7EfXeB1CxrXoZvqzu9mPiXJBxZ4y96Yaz18XYbbfj7SWlvojPJCFqrtpMsyhMDU8MtjexNgkqHtnxv70k+vwPsuy7CTySKnN+f6ObnjynAbx70y7LgmXb7LvD60F+1fyHz9dL/cuU34qQy3Iu9qsdvlS5Ps+K5GTXxPcR57vZwWWcPXJ3nDeNAwa/bgYPY7Xd++mPmtsMNn+0R2rvUXxkA537Z5b7f389mW5OFVtV9VHZZk9vvff53he1L3H6e/6zFAMv+yXc39wZ7MV89kuIX6liRfGq9ePX4R0/vgOJ1U1eOT3HuB1+7puGlPJuv2bRmu4u7NiYP51p2lHEOtlJXeJ90ryU1jMHtQhqtXy7Utcx8LLXQcmQzryDFV9YDx+YPG5fapDHdfzf5I10ruh3vzpiTPreH7dd+S4esOL63hzpaHZ7iNdf8kF9Vwt8zHkrxqvEvuNzPcKXN1DXfV/OY4zR9Pck0Nd50cmeHWyi6s+ZWz1tq1VfWiDKn3qxkKum3iJS9L8uYabg94xxJnc2tVfSzDwvnZBV739iQX1/Bl7d2+kDmP12W4HeDqqtqe5Pwkkz//e1qS19fwYxfnLDCdn05ybg2/RLMxQ0e8KkPn+9YMZ07eN447PclTxvndkOX/wuBDx/nckWR7hvufXzY+d1KG7zm9dTyu+JfW2vE1fMn3VTXc3rYhw2X0a+eZ/kosw1kvzFDPqzN8j+ypy5zepNdmWI4fba39dGvtKzX8CMG/t9a+uozpzlXf+f4rhpdkuPXuV7J7rf4myZ9k+I7RRa21K5JkfO2l44Hj9gy3NfzjPNM/I8MtYf+W4T7shW772pOnJjmvhi/CfybJ05YxrUlXJ/lqVV2V4Ttxv9Nau7Kq/iPDLyst1UK1nXTO+LpPJvnbDP16vhM6c/nVJB/OUOMPZ/EHK/O977Qkf1xVp2c4GbOgsd/OtX7+XYYd170ybE9e1Vr7912uPpya5A013EL6b1mhZdpa++Q8/fSWJN81PvevGU+k7eK1Sd5dVf/SWjt2gdk8K8lrxm3DhgwHuycv8PqlLKfZnf+cNRwD5wlJvq2qZvc1/yvDwcArMmxf9kvyD0meuIj5raRPJXlmVb0+w6/pnZshAFyTYT/ykbnetEB/mm97P5/LMnzuT2a4nfOj4/T/bdw3/OlYm3/Nnbeqzppv2b4wq7c/2JO56vmDSTJeUfpYhu3HZzN89j35jSRvHA8eL8/wvZg5zXPcdNJetP2cDMcbn8hwNeek1tpttfgrkfOtO29Kcn4NP+SzFr9ovZL7pHcnObmqrsuwrP96Bdo337HQgseR4zpyUob+MfuVl18Z78p5WoavxGzIsA7vE79O2lrbliEQzT4+aZ7nLhjHXZbdf0p/rsD8vXPM68tJfn6O8Wdl+AGy7tSdd4esTzVciv6l2QNZWKzxQOGjGa4arunPq44b5qNba/97LduxVmq4+jiT5EHj9ylWc177Z/jhilvHM3R/keSBrbWvrOZ874qq6ubW2nJOEADAurLmV86gR+PtbJck+bO1DmZ3dTX8B7IvSvKc1Q5mo3tk+On2jRmujpwimAEA07Dur5zNZbwMvOv/WXVZa+2Za9Ge9aKqXpPhl+UmvbK1tpxb0ViCGn4wZtf/7+Qpbfw1Q9aO9WTfYDktTH1W3/idst3+37ckj25L//VgVoBlw2q6S4YzAACA3vTwa40AAAB3ecIZAABAB4QzAACADghnAAAAHRDOAAAAOvD/A2MlCBsWxhyhAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1080x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WswQPEyPS6uR"
      },
      "source": [
        "Despite the fact that all the features range from 2 to 10, their respective distributions are quite different.\n",
        "\n",
        "The distributions are far from normally distributed (the most frequent values are far from the distributions' centers). Some features have **outliers** that skew the distribution mostly to the right."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iFHSLZ8OSd9t"
      },
      "source": [
        "#### Multimodal Data visualizations"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 855
        },
        "id": "APsueA4GSn_2",
        "outputId": "f0be8e45-e8e5-42b5-fac2-49c3412d642e"
      },
      "source": [
        "# Heatmap of the correlation between attributes\n",
        "plt.figure(figsize=(34,14))\n",
        "\n",
        "# create a mask\n",
        "mask = np.zeros_like(df.corr())\n",
        "mask[np.triu_indices_from(mask)] = True\n",
        "\n",
        "# plot the heatmap\n",
        "sns.heatmap(df.corr(), vmax=1, square=True,annot=True,mask = mask, cmap='viridis')\n",
        "plt.title('Correlation between different attributes')\n",
        "plt.show()"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 2448x1008 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aSC6kl-nu2H3"
      },
      "source": [
        "Light green show highly correlated features."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 986
        },
        "id": "TMZ6ooujXdS3",
        "outputId": "248422ac-881a-48e6-c20a-3e5c1c1e93a7"
      },
      "source": [
        "# create a pairplot\n",
        "cols = ['shape_uniformity', 'size_uniformity', 'marginal_adhesion', 'epithelial_size', \n",
        "        'bare_nucleoli', 'bland_chromatin', 'normal_nucleoli']\n",
        "\n",
        "sns.pairplot(df.loc[:, cols], corner=True, diag_kind=None, kind='reg');"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1260x1260 with 28 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jVBlncOWZEL3"
      },
      "source": [
        "These are the observations that we can make:\n",
        "* Apart from the target and the dump_thickness column, the others features a not normaly distributed (the mean values are not at the distributions' centers)\n",
        "* Shape_uniformity and size_uniformity seem to be highly correlated with several others attributes\n",
        "* shape_uniformity, size_uniformity and bare_nucleoli have teh same distributions\n",
        "* Some features seem to have outliers\n",
        "* The id column seems to be not correlated with any others features."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RFPXTybF6bda"
      },
      "source": [
        "#### PCA Analysis\n",
        "Finally, let's perform a **PCA analysis** to asses what kind of ML algorithm can we use:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EfSE-onXxD6m"
      },
      "source": [
        "# import of necessary librairies\n",
        "from sklearn.decomposition import PCA"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YQzfTWDYyHsD"
      },
      "source": [
        "# perform the PCA analysis to get only 2 components\n",
        "pca = PCA(n_components=2)\n",
        "data_2D = pca.fit_transform(df.iloc[:, 1:-1])"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "H482wUCxyh6_",
        "outputId": "3b7c4f60-1e5d-40ec-cc7a-d655b04854b6"
      },
      "source": [
        "# plot the class with PCA components\n",
        "scatter = plt.scatter(data_2D[:,0], data_2D[:,1], c=df['class'], cmap=plt.cm.Spectral)\n",
        "\n",
        "# add legend with values\n",
        "plt.legend(*scatter.legend_elements())\n",
        "\n",
        "plt.title('Class repartition on PCA plot')\n",
        "plt.xlabel('PCA_compo_1')\n",
        "plt.ylabel('PCA_compo_2')\n",
        "plt.show()"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "F3JG_VHp11RP",
        "outputId": "ec45dc3d-9305-47c6-9cff-ec2727187942"
      },
      "source": [
        "# Print the explained variance\n",
        "pca.explained_variance_ratio_"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.68900383, 0.07337931])"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZA7nd-cJzpZN"
      },
      "source": [
        "We can clearly identify 2 clusters corresponding to the binary classes. An imaginary line can clearly achieve a reasonable clusters separation. This means that we can use linear algoorithms like **Logistic regression**.\n",
        "\n",
        "Moreover, from the above line of code we can see that the first PCA component explain about 68% of the total variance and the second component allow to explain about 7.3%. This means that the first component achieve a better separation than the second component."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zu8MPlkX6xoS"
      },
      "source": [
        "#### Manifold analysis"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wA1Vm2LZ7Eau"
      },
      "source": [
        "# import the necessary librairies\n",
        "from sklearn.manifold import Isomap"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1o30Qor37YET"
      },
      "source": [
        "# perform the anaysis\n",
        "isomap = Isomap(n_components=2)\n",
        "data_2D_iso = isomap.fit_transform(df.iloc[:, 1:-1])"
      ],
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        },
        "id": "HNCfq-lm73Ly",
        "outputId": "54abc051-508f-44bb-93ed-020f3485a53c"
      },
      "source": [
        "# plot the class with PCA components\n",
        "scatter_iso = plt.scatter(data_2D_iso[:,0], data_2D_iso[:,1], c=df['class'], cmap=plt.cm.Spectral)\n",
        "\n",
        "# add legend with values\n",
        "plt.legend(*scatter_iso.legend_elements())\n",
        "\n",
        "plt.title('Class repartition on Isomap plot')\n",
        "plt.xlabel('Isomap_compo_1')\n",
        "plt.ylabel('Isomap_compo_2')\n",
        "plt.show()"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XAyaVPC-8z8h"
      },
      "source": [
        "#### Feature importances analysis\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gWt99SLh9kBK"
      },
      "source": [
        "# import \n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "# Train de algorithm\n",
        "rdf = RandomForestClassifier(n_jobs=-1).fit(df.iloc[:, 1:-1], df['class'])"
      ],
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 366
        },
        "id": "eOUg06zp-aav",
        "outputId": "f94c3849-39cf-4c17-eaa8-619072a40dba"
      },
      "source": [
        "# plot features importance\n",
        "plt.bar(df.iloc[:, 1:-1].columns, rdf.feature_importances_)\n",
        "plt.title('features importances')\n",
        "plt.xticks(rotation=90)\n",
        "plt.show()"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Hhg_GOR0BeBW"
      },
      "source": [
        "Comparing the above chart with the correlation plot, we can deduct that:\n",
        "* `mitoses` columns has the least contribution for the classification. The feature can then be droppped\n",
        "* As the `size_uniformity` and `shape_uniformity` are highly correlated, we could only consider the first feature (the one with the highest feature importance) for the analysis\n",
        "\n",
        "Of course, those hypotheses have to be checked and confirmed and not just assumed. To be continued..."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1mFoojdHhAkh"
      },
      "source": [
        "### 2.3 Dataset split"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ThN9gLl5Ajzm"
      },
      "source": [
        "Now is time to split the data so that we would be able to properly assess the performance of the final model. To do so, we need to keep some data away from the training process.\n",
        "\n",
        "So first the data are divided into the feature and the target part, and then the features are splitted into a train (*to litterally train or models*) and a test (*to asses the performance on data unseen by the created models*) parts"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mk-1OtKphFfk",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 198
        },
        "outputId": "161cf66a-ba7c-409c-8391-e2783eaab6b5"
      },
      "source": [
        "# import necessary librairies\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "# splitting feature en target\n",
        "target = df[['class']]\n",
        "features = df.iloc[:, 1:-1]\n",
        "\n",
        "# train \\ test split\n",
        "test_fraction = .2\n",
        "seed = 7\n",
        "X_train, X_test, y_train, y_test = train_test_split(features, target, test_size=test_fraction, random_state = seed)\n",
        "\n",
        "# print data shape (number of sample, numbers of columns) just to be sure\n",
        "print('X_train shape:', X_train.shape)\n",
        "print('X_test shape:', X_test.shape)\n",
        "X_train.head(3)"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "X_train shape: (559, 9)\n",
            "X_test shape: (140, 9)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>clump_thickness</th>\n",
              "      <th>size_uniformity</th>\n",
              "      <th>shape_uniformity</th>\n",
              "      <th>marginal_adhesion</th>\n",
              "      <th>epithelial_size</th>\n",
              "      <th>bare_nucleoli</th>\n",
              "      <th>bland_chromatin</th>\n",
              "      <th>normal_nucleoli</th>\n",
              "      <th>mitoses</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>319</th>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>4</td>\n",
              "      <td>6</td>\n",
              "      <td>5</td>\n",
              "      <td>7</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>132</th>\n",
              "      <td>5</td>\n",
              "      <td>10</td>\n",
              "      <td>8</td>\n",
              "      <td>10</td>\n",
              "      <td>8</td>\n",
              "      <td>10</td>\n",
              "      <td>3</td>\n",
              "      <td>6</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>148</th>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>8</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>8</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     clump_thickness  size_uniformity  shape_uniformity  marginal_adhesion  \\\n",
              "319                4                4                 4                  4   \n",
              "132                5               10                 8                 10   \n",
              "148                3                1                 1                  3   \n",
              "\n",
              "     epithelial_size  bare_nucleoli  bland_chromatin  normal_nucleoli  mitoses  \n",
              "319                6              5                7                3        1  \n",
              "132                8             10                3                6        3  \n",
              "148                8              1                5                8        1  "
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 303
        },
        "id": "8-fANPvVh0V9",
        "outputId": "7a1bda74-ce31-430f-a942-4e6798b5a20e"
      },
      "source": [
        "# plot the train \\ test target distribution\n",
        "plt.subplot(121)\n",
        "y_train.value_counts(normalize=True).plot.bar()\n",
        "plt.title('train distribution')\n",
        "plt.subplot(122)\n",
        "y_test.value_counts(normalize=True).plot.bar()\n",
        "plt.title('test distribution')\n",
        "plt.show()"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AxAemG2PjiSB"
      },
      "source": [
        "### 2.4 Evaluate Algorithms: Establishing the baseline"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GiYRHAK0izRa"
      },
      "source": [
        "# importing the necessary librairies\n",
        "from sklearn.model_selection import KFold\n",
        "from sklearn.model_selection import cross_val_score\n",
        "\n",
        "# Linear models\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
        "\n",
        "# Non-linear models\n",
        "from sklearn.neighbors import KNeighborsClassifier #AKA KNN\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.svm import SVC # Support vector Machine\n",
        "from sklearn.naive_bayes import GaussianNB"
      ],
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3uIBY5Y5lgKS"
      },
      "source": [
        "# Spot-Checking algorithms\n",
        "\n",
        "# models list:\n",
        "models = [\n",
        "          ('LR', LogisticRegression()),\n",
        "          ('LDA', LinearDiscriminantAnalysis()),\n",
        "          ('KNN', KNeighborsClassifier()),\n",
        "          ('DTC', DecisionTreeClassifier()),\n",
        "          ('NB', GaussianNB()),\n",
        "          ('SVM', SVC())\n",
        "]\n",
        "\n",
        "# The scoring metric\n",
        "scoring = 'accuracy'\n",
        "\n",
        "# Number of folds to consider for teh KFold strategy\n",
        "num_folds = 10"
      ],
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A9PDvj9NmoRq"
      },
      "source": [
        "# Defining our spot checking algorithms\n",
        "def spot_check_function(model_list, num_folds, seed, scoring, x, y):\n",
        "  results = []\n",
        "  names = []\n",
        "  for name, model in model_list:\n",
        "    # splitting strategy: dividing the train dataset\n",
        "    kfold = KFold(n_splits=num_folds, random_state=seed, shuffle=True)\n",
        "    # calculate the score for each fold\n",
        "    result = cross_val_score(model, x, y, cv=kfold, scoring=scoring)\n",
        "    # printing results\n",
        "    #print('%s: mean = %f and std = %f)'% (name, result.mean(), result.std()))\n",
        "    # store the results\n",
        "    results.append(result)\n",
        "    # store the model name\n",
        "    names.append(name)\n",
        "  return names, results\n",
        "\n",
        "#\n",
        "def get_kfold_results(alg_list, results_list):\n",
        "    results_df = pd.DataFrame()\n",
        "    for name, elt in zip(alg_list, results_list):\n",
        "      results_df[name] = elt\n",
        "    return results_df\n",
        "#\n",
        "def get_averaged_results(alg_list, pred_list):\n",
        "  name_list = []\n",
        "  mean_list = []\n",
        "  std_list = []\n",
        "  for name, elt in zip (alg_list, pred_list):\n",
        "    name_list.append(name)\n",
        "    mean_list.append(elt.mean())\n",
        "    std_list.append(elt.std())\n",
        "    #print('%s: mean = %f and std = %f)'% (name, elt.mean(), elt.std()))\n",
        "\n",
        "  res_df = pd.DataFrame({'alg_name': name_list,'mean_acc': mean_list, 'std_acc': std_list}).set_index('alg_name')\n",
        "  return res_df"
      ],
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ecpe6FpJosDZ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "61217c57-8fee-4902-b590-6d48579fe249"
      },
      "source": [
        "# apply the function on X_train and y_train\n",
        "alg_names, pred_results = spot_check_function(models, num_folds, seed, scoring, X_train, y_train)\n",
        "\n",
        "# store the kold results into a Dataframe\n",
        "results_df = get_kfold_results(alg_names, pred_results)\n",
        "\n",
        "# display results\n",
        "results_df"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:515: DataConversionWarning: 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",
            "  estimator.fit(X_train, y_train, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>LR</th>\n",
              "      <th>LDA</th>\n",
              "      <th>KNN</th>\n",
              "      <th>DTC</th>\n",
              "      <th>NB</th>\n",
              "      <th>SVM</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.982</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.911</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.929</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.893</td>\n",
              "      <td>0.911</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.982</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.982</td>\n",
              "      <td>1.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.875</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.982</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.982</td>\n",
              "      <td>1.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>0.893</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.929</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.929</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>0.964</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.927</td>\n",
              "      <td>0.945</td>\n",
              "      <td>0.982</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      LR    LDA    KNN    DTC     NB    SVM\n",
              "0  0.982  0.946  0.964  0.911  0.982  0.982\n",
              "1  0.982  0.982  1.000  0.982  0.964  0.964\n",
              "2  0.911  0.911  0.929  0.911  0.893  0.911\n",
              "3  0.982  0.964  0.982  0.911  0.982  1.000\n",
              "4  0.982  0.982  0.982  0.875  0.964  0.964\n",
              "5  0.964  0.964  0.964  0.946  0.964  0.982\n",
              "6  1.000  1.000  1.000  1.000  0.982  1.000\n",
              "7  0.893  0.911  0.929  0.911  0.946  0.929\n",
              "8  0.964  0.946  0.946  0.911  0.964  0.964\n",
              "9  0.982  0.982  0.964  0.927  0.945  0.982"
            ]
          },
          "metadata": {},
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        },
        "id": "Uy6EavaVqWPM",
        "outputId": "9e90a0fa-0c5a-4428-f33e-60be73924235"
      },
      "source": [
        "# display results\n",
        "mean_results_df = get_averaged_results(alg_names, pred_results)\n",
        "mean_results_df"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>mean_acc</th>\n",
              "      <th>std_acc</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>alg_name</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>LR</th>\n",
              "      <td>0.964</td>\n",
              "      <td>0.033</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>LDA</th>\n",
              "      <td>0.959</td>\n",
              "      <td>0.029</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>KNN</th>\n",
              "      <td>0.966</td>\n",
              "      <td>0.025</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>DTC</th>\n",
              "      <td>0.930</td>\n",
              "      <td>0.035</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>NB</th>\n",
              "      <td>0.959</td>\n",
              "      <td>0.025</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>SVM</th>\n",
              "      <td>0.968</td>\n",
              "      <td>0.027</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          mean_acc  std_acc\n",
              "alg_name                   \n",
              "LR           0.964    0.033\n",
              "LDA          0.959    0.029\n",
              "KNN          0.966    0.025\n",
              "DTC          0.930    0.035\n",
              "NB           0.959    0.025\n",
              "SVM          0.968    0.027"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kEKav9d_Dyx4"
      },
      "source": [
        "Looking at averaged results, the **SVM, KNN and Logistic regression** models have the bes accuracy results (with approximately similar standard deviation)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "qrNim_bzsES4",
        "outputId": "e20c632c-9e57-4263-9109-ae71f30c0a83"
      },
      "source": [
        "# compare the algorithms\n",
        "results_df.boxplot()\n",
        "plt.show()"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MYYr6U-EEo24"
      },
      "source": [
        "Looking at the above chart (with median values and inter-quartile ranges) we can clearly see **Logistic regression and SVM** standing out.\n",
        "\n",
        "Now let's see if standardizing the data change anything. We can already guest that the standardizing coud affect the linear and the KNN algorithms' performances"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eLfKzOJRuubM"
      },
      "source": [
        "### 2.5 Evaluate algorithms: with Standardized Data\n",
        "Since we want the standardize our date before applying the ML algorithms, we **must** use a pipeline to avoid any data leak during the steps sequences. So:\n",
        "* Step 1: standardizing the data\n",
        "* Step 2: applying the ML algorithm"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QyGxYKk9sp1h"
      },
      "source": [
        "# import the pipeline librairie\n",
        "from sklearn.pipeline import Pipeline\n",
        "\n",
        "# import the scaling librairies\n",
        "from sklearn.preprocessing import StandardScaler"
      ],
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QovisylKvpil"
      },
      "source": [
        "# model list:\n",
        "models_with_pipe = [\n",
        "          ('scaledLR', Pipeline([('scaler', StandardScaler()), ('LR', LogisticRegression())])),\n",
        "          ('scaledLDA', Pipeline([('scaler', StandardScaler()), ('LDA', LinearDiscriminantAnalysis())])),\n",
        "          ('scaledKNN', Pipeline([('scaler', StandardScaler()), ('KNN', KNeighborsClassifier())])),\n",
        "          ('scaledDTC', Pipeline([('scaler', StandardScaler()), ('DTC', DecisionTreeClassifier())])),\n",
        "          ('scaledNB', Pipeline([('scaler', StandardScaler()), ('NB', GaussianNB())])),\n",
        "          ('scaledSVM', Pipeline([('scaler', StandardScaler()), ('SVM', SVC())]))\n",
        "]"
      ],
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "emKZCgOLxTLp",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "480ad2c4-92ab-4728-e33a-c04faa721832"
      },
      "source": [
        "# apply the function on X_train and y_train\n",
        "alg_names_pipe, pred_results_pipe = spot_check_function(models_with_pipe, num_folds, seed, scoring, X_train, y_train)\n",
        "\n",
        "# store the kold results into a Dataframe\n",
        "results_df_pipe = get_kfold_results(alg_names_pipe, pred_results_pipe)\n",
        "\n",
        "# displaying the results\n",
        "results_df_pipe"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/pipeline.py:354: DataConversionWarning: 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",
            "  self._final_estimator.fit(Xt, y, **fit_params)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/naive_bayes.py:206: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>scaledLR</th>\n",
              "      <th>scaledLDA</th>\n",
              "      <th>scaledKNN</th>\n",
              "      <th>scaledDTC</th>\n",
              "      <th>scaledNB</th>\n",
              "      <th>scaledSVM</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.982</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.929</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.000</td>\n",
              "      <td>0.982</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.911</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.893</td>\n",
              "      <td>0.893</td>\n",
              "      <td>0.893</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.982</td>\n",
              "      <td>1.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.893</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.982</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.982</td>\n",
              "      <td>1.000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>0.911</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.929</td>\n",
              "      <td>0.911</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.946</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>0.964</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.946</td>\n",
              "      <td>0.929</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.964</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>0.982</td>\n",
              "      <td>0.982</td>\n",
              "      <td>0.964</td>\n",
              "      <td>0.927</td>\n",
              "      <td>0.945</td>\n",
              "      <td>0.982</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   scaledLR  scaledLDA  scaledKNN  scaledDTC  scaledNB  scaledSVM\n",
              "0     0.982      0.946      0.982      0.929     0.982      0.982\n",
              "1     1.000      0.982      1.000      0.964     0.964      0.964\n",
              "2     0.911      0.911      0.946      0.893     0.893      0.893\n",
              "3     0.964      0.964      1.000      0.946     0.982      1.000\n",
              "4     0.982      0.982      0.982      0.893     0.964      0.964\n",
              "5     0.964      0.964      0.964      0.946     0.964      0.982\n",
              "6     1.000      1.000      1.000      1.000     0.982      1.000\n",
              "7     0.911      0.911      0.929      0.911     0.946      0.946\n",
              "8     0.964      0.946      0.946      0.929     0.964      0.964\n",
              "9     0.982      0.982      0.964      0.927     0.945      0.982"
            ]
          },
          "metadata": {},
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        },
        "id": "NhIpJNX5ymgR",
        "outputId": "7c04eeaa-67af-4987-e24c-fe2354a97809"
      },
      "source": [
        "# display results\n",
        "mean_results_df_pipe = get_averaged_results(alg_names_pipe, pred_results_pipe)\n",
        "mean_results_df_pipe"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>mean_acc</th>\n",
              "      <th>std_acc</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>alg_name</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>scaledLR</th>\n",
              "      <td>0.966</td>\n",
              "      <td>0.030</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>scaledLDA</th>\n",
              "      <td>0.959</td>\n",
              "      <td>0.029</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>scaledKNN</th>\n",
              "      <td>0.971</td>\n",
              "      <td>0.024</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>scaledDTC</th>\n",
              "      <td>0.934</td>\n",
              "      <td>0.031</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>scaledNB</th>\n",
              "      <td>0.959</td>\n",
              "      <td>0.025</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>scaledSVM</th>\n",
              "      <td>0.968</td>\n",
              "      <td>0.030</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           mean_acc  std_acc\n",
              "alg_name                    \n",
              "scaledLR      0.966    0.030\n",
              "scaledLDA     0.959    0.029\n",
              "scaledKNN     0.971    0.024\n",
              "scaledDTC     0.934    0.031\n",
              "scaledNB      0.959    0.025\n",
              "scaledSVM     0.968    0.030"
            ]
          },
          "metadata": {},
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yOMDdlXHGb2D"
      },
      "source": [
        "Now let's pause a min to explain thing a little bit. If we focus on Logistic regression, KNN and SVM we can see that:\n",
        "* The SVM model accuracy remained the same (-->a non linear model is not affacted by the standardizing)\n",
        "* The logistic regression and the KNN models' accuracies remainded approximetely the same (--> since the data range is the same for all of the features, the standardizing step desn't bring much but allow a litttle improvement)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "LJoIrnY4yzze",
        "outputId": "3ead4e73-460b-4fb1-ac8f-dd06973c2935"
      },
      "source": [
        "# compare the algorithms\n",
        "results_df_pipe.boxplot()\n",
        "plt.show()"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p3fjeZCODQCr"
      },
      "source": [
        "Even after standardizing, the Logistic regression and the SVM are still standing out. So these are the models that we will use next.\n",
        "\n",
        "> **An important tip here is that in the previous steps we didn't only randomly tried some algorithms and hope for the best. We instead select different models typologies and compared them at once using the KFold strategy to get a significant value to compare**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v7xbSjekJDA9"
      },
      "source": [
        "Now that we've selected **Logistic regression and SVM**, the nex step is to try build even tougher models with the **Grid search strategy**. The aim of this strategy is to find the best possible parameters to use with the selected models that will yield the best possible benchmark."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2703beaIJsU5"
      },
      "source": [
        "### 2.6 Models Tuning"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WSa4lMhOKENy"
      },
      "source": [
        "#### Tuning Logistic regression\n",
        "\n",
        "These are the LR parameters than we will try:\n",
        "* solver: {‘newton-cg’, ‘lbfgs’, ‘liblinear’, ‘sag’, ‘saga’}\n",
        "* C (knowing that stronger values mean stronger regularization)\n",
        "\n",
        "For the complete set of parameters, just check the [sklearn documentation](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html?highlight=logistic%20regression#sklearn.linear_model.LogisticRegression)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b0yJo8LpDPmP"
      },
      "source": [
        "# import ncessary libraries\n",
        "from sklearn.model_selection import GridSearchCV"
      ],
      "execution_count": 45,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GAKkz3-6y5_N"
      },
      "source": [
        "# numbers of neighbors to use\n",
        "#neighbors = np.arange(1, 22, 2)\n",
        "\n",
        "# parameters grid\n",
        "param = {'LR__C': [0.0001, 0.001, 0.01, 0.1, 0.3, 0.5, 0.7, 0.9, 1.0],\n",
        "         'LR__solver': ['newton-cg', 'lbfgs', 'liblinear', 'sag', 'saga']}\n",
        "\n",
        "# Data standardizing + modeling: not ncessary but since the accuracy is a litle bit increased we can choose to do it as well\n",
        "LR_pipe_model = Pipeline([('Standardizing', StandardScaler()),('LR', LogisticRegression())])\n",
        "\n",
        "# KFold strategy\n",
        "kfold = KFold(n_splits=num_folds, random_state=seed, shuffle=True)\n",
        "\n",
        "# Grid search strategy\n",
        "LR_grid = GridSearchCV(estimator=LR_pipe_model, param_grid=param, scoring=scoring, cv=kfold)\n",
        "\n",
        "# fit the model\n",
        "LR_grid_results = LR_grid.fit(X_train, y_train)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MeNQbGPAL45t",
        "outputId": "32296828-81a9-409b-de8b-58a14c7ff0eb"
      },
      "source": [
        "# print the best result\n",
        "print(\"Best: %f using %s\" %(LR_grid_results.best_score_, LR_grid_results.best_params_))"
      ],
      "execution_count": 87,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best: 0.971396 using {'LR__C': 0.01, 'LR__solver': 'liblinear'}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3KH304BjQDc5"
      },
      "source": [
        "#### Tuning SVM"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8KuUfyJ7bIPK",
        "outputId": "94d8197d-8ace-484f-93aa-2a002694a25b"
      },
      "source": [
        "np.linspace(0.002, 1, 5)"
      ],
      "execution_count": 106,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.002 , 0.2515, 0.501 , 0.7505, 1.    ])"
            ]
          },
          "metadata": {},
          "execution_count": 106
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9s0X9BopP9bU"
      },
      "source": [
        "# Mmodeling: As we know that standardizing is not necessarily needed\n",
        "svm = SVC()\n",
        "\n",
        "# parameters grid\n",
        "param = {'C': np.arange(0.1, 10, 0.2),\n",
        "         'kernel': ['linear', 'poly', 'rbf', 'sigmoid'], # 'precomputed is only used with square kernel\n",
        "         'gamma': ['scale', 'auto'],\n",
        "         }#'gamma': np.linspace(0.002, 1, 5)} # could be tested as well\n",
        "\n",
        "# KFold strategy\n",
        "kfold = KFold(n_splits=num_folds, random_state=seed, shuffle=True)\n",
        "\n",
        "# Grid search strategy\n",
        "svm_grid = GridSearchCV(estimator=svm, param_grid=param, scoring=scoring, cv=kfold)\n",
        "\n",
        "# fit the model\n",
        "svm_grid_results = svm_grid.fit(X_train, y_train)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iIc2y6tzURKV",
        "outputId": "a9ae7d03-5f4a-4428-e462-1311fbda2e81"
      },
      "source": [
        "# print the best result\n",
        "print(\"Best: %f using %s\" %(svm_grid_results.best_score_, svm_grid_results.best_params_))"
      ],
      "execution_count": 112,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best: 0.969610 using {'C': 1.1000000000000003, 'gamma': 'scale', 'kernel': 'rbf'}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "1CLp6g1_ZAFX",
        "outputId": "203da9d0-92e3-477d-ecac-4301b0af2298"
      },
      "source": [
        "# display the 5 first gris search results\n",
        "pd.DataFrame({'means':svm_grid_results.cv_results_['mean_test_score'], \n",
        "              'stds':svm_grid_results.cv_results_['std_test_score'],\n",
        "              'params':svm_grid_results.cv_results_['params']}).sort_values(by='means', ascending=False).head()"
      ],
      "execution_count": 105,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>means</th>\n",
              "      <th>stds</th>\n",
              "      <th>params</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>143</th>\n",
              "      <td>0.97</td>\n",
              "      <td>0.028</td>\n",
              "      <td>{'C': 1.5000000000000004, 'gamma': 0.002, 'ker...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>682</th>\n",
              "      <td>0.97</td>\n",
              "      <td>0.028</td>\n",
              "      <td>{'C': 6.900000000000001, 'gamma': 0.002, 'kern...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>902</th>\n",
              "      <td>0.97</td>\n",
              "      <td>0.028</td>\n",
              "      <td>{'C': 9.100000000000001, 'gamma': 0.002, 'kern...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>542</th>\n",
              "      <td>0.97</td>\n",
              "      <td>0.031</td>\n",
              "      <td>{'C': 5.500000000000001, 'gamma': 0.002, 'kern...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>922</th>\n",
              "      <td>0.97</td>\n",
              "      <td>0.028</td>\n",
              "      <td>{'C': 9.3, 'gamma': 0.002, 'kernel': 'rbf'}</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     means   stds                                             params\n",
              "143   0.97  0.028  {'C': 1.5000000000000004, 'gamma': 0.002, 'ker...\n",
              "682   0.97  0.028  {'C': 6.900000000000001, 'gamma': 0.002, 'kern...\n",
              "902   0.97  0.028  {'C': 9.100000000000001, 'gamma': 0.002, 'kern...\n",
              "542   0.97  0.031  {'C': 5.500000000000001, 'gamma': 0.002, 'kern...\n",
              "922   0.97  0.028        {'C': 9.3, 'gamma': 0.002, 'kernel': 'rbf'}"
            ]
          },
          "metadata": {},
          "execution_count": 105
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fMMEnUESeiHk"
      },
      "source": [
        "#### Tuning conclusion\n",
        "\n",
        "Tuning the best models brougth a little change for both models. We ended up with:\n",
        "* 97.1% accuracy with Logistic regression (after data standardizing)\n",
        "* 97% accuracy with SVM (without standardizing)\n",
        "\n",
        "So we are sure now that both models are equally qualified and everything will depends on the performances on the test set. But before the last check, let's try to do better with boosting and bagging methods to be sure we've got the best possible models."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xMt-UFwSgZAH"
      },
      "source": [
        "### 2.7 Ensemble methods (boosting an dbagging)\n",
        "These methods use decision tree based algorithms to find the best possible prediction split with the best possible features."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aTVFzqsAaPVI"
      },
      "source": [
        "# import the necessary libraries\n",
        "from sklearn.ensemble import AdaBoostClassifier, GradientBoostingClassifier, ExtraTreesClassifier"
      ],
      "execution_count": 89,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dJLrFrWvg9H2"
      },
      "source": [
        "# Ensemble methods' list\n",
        "ensemble_model_list = [\n",
        "                 ('adaboost', AdaBoostClassifier()),\n",
        "                 ('gdboost', GradientBoostingClassifier()),\n",
        "                 ('RF', RandomForestClassifier()),\n",
        "                 ('ET', ExtraTreesClassifier())\n",
        "]\n",
        "\n",
        "# apply the function on X_train and y_train\n",
        "alg_names, pred_results = spot_check_function(ensemble_model_list, num_folds, seed, scoring, X_train, y_train)\n",
        "\n",
        "# store the kold results into a Dataframe\n",
        "results_df = get_kfold_results(alg_names, pred_results)\n",
        "\n",
        "# display results\n",
        "#results_df\n",
        "\n",
        "# display averaged results\n",
        "mean_results_df = get_averaged_results(alg_names, pred_results)\n",
        "#mean_results_df"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "p3u2GXZdiOLG",
        "outputId": "80e2729d-73dc-4672-dc26-668726efb131"
      },
      "source": [
        "# plot boxplotq\n",
        "results_df.boxplot();"
      ],
      "execution_count": 91,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "1-Wzj0F3inox",
        "outputId": "7c2d1d4b-26d2-4da2-b250-6421a6d5a392"
      },
      "source": [
        "# Averaged results\n",
        "mean_results_df.sort_values(by='mean_acc', ascending=False)"
      ],
      "execution_count": 93,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>mean_acc</th>\n",
              "      <th>std_acc</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>alg_name</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>ET</th>\n",
              "      <td>0.968</td>\n",
              "      <td>0.022</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>RF</th>\n",
              "      <td>0.964</td>\n",
              "      <td>0.031</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>adaboost</th>\n",
              "      <td>0.961</td>\n",
              "      <td>0.022</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>gdboost</th>\n",
              "      <td>0.957</td>\n",
              "      <td>0.029</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          mean_acc  std_acc\n",
              "alg_name                   \n",
              "ET           0.968    0.022\n",
              "RF           0.964    0.031\n",
              "adaboost     0.961    0.022\n",
              "gdboost      0.957    0.029"
            ]
          },
          "metadata": {},
          "execution_count": 93
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nYvpivC6jEi1"
      },
      "source": [
        "WE can see that even though out of the box, the ensemble methods accuracies are equivalent to tuned previous models. Let's keep it 'simple' and continue with the SVM model with tuned parameters."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Lz9-t_qCj_a1"
      },
      "source": [
        "### 2.8 Final model and final test"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ymJvRm7VjEJP",
        "outputId": "8b6e129d-2383-414e-f014-6824fa09c98c"
      },
      "source": [
        "# initialize the model:\n",
        "svm_final = SVC(C=1.1, \n",
        "                gamma='auto', \n",
        "                kernel='rbf')\n",
        "svm_final"
      ],
      "execution_count": 113,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SVC(C=1.1, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
              "    decision_function_shape='ovr', degree=3, gamma='auto', kernel='rbf',\n",
              "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
              "    tol=0.001, verbose=False)"
            ]
          },
          "metadata": {},
          "execution_count": 113
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JxoOmHUNiwIu",
        "outputId": "a99a917f-a1cf-4d27-873e-b03a5be86854"
      },
      "source": [
        "# import necessary librairie\n",
        "from sklearn.metrics import classification_report\n",
        "\n",
        "#  train the model\n",
        "svm_final.fit(X_train, y_train)\n",
        "\n",
        "# print the accuracy\n",
        "print('train accuracy', svm_final.score(X_train, y_train))\n",
        "print('test accuracy', svm_final.score(X_test, y_test))\n",
        "\n",
        "# perform prediction\n",
        "svm_final_pred = svm_final.predict(X_test)\n",
        "\n",
        "# print the classification report\n",
        "print(classification_report(svm_final_pred, y_test))"
      ],
      "execution_count": 114,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "train accuracy 1.0\n",
            "test accuracy 0.9642857142857143\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           2       0.96      0.99      0.97        88\n",
            "           4       0.98      0.92      0.95        52\n",
            "\n",
            "    accuracy                           0.96       140\n",
            "   macro avg       0.97      0.96      0.96       140\n",
            "weighted avg       0.96      0.96      0.96       140\n",
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py:760: DataConversionWarning: 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",
            "  y = column_or_1d(y, warn=True)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "3Q4fKE7nmHRJ",
        "outputId": "af853d1d-2703-42cf-b68a-763961153aa3"
      },
      "source": [
        "# display the confusion matrix\n",
        "pd.crosstab(svm_final_pred, np.array(y_test['class'].tolist()), colnames=['predicted'], rownames=['true'])"
      ],
      "execution_count": 133,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<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>predicted</th>\n",
              "      <th>2</th>\n",
              "      <th>4</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>true</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>87</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>48</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "predicted   2   4\n",
              "true             \n",
              "2          87   1\n",
              "4           4  48"
            ]
          },
          "metadata": {},
          "execution_count": 133
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AwS9XRr9udK-"
      },
      "source": [
        ""
      ]
    }
  ]
}