Skip to content

Instantly share code, notes, and snippets.

@alonsosilvaallende

alonsosilvaallende/interpretable_ml_xai_regression_2024-11.ipynb

Created November 15, 2024 06:10
Show Gist options
  • Save alonsosilvaallende/b9618f749916b1a37bda302fe4227ed6 to your computer and use it in GitHub Desktop.
Save alonsosilvaallende/b9618f749916b1a37bda302fe4227ed6 to your computer and use it in GitHub Desktop.
Download ZIP
interpretable_ml_xai_regression_2024-11.ipynb
Display the source blob
Display the rendered blob
Raw
interpretable_ml_xai_regression_2024-11.ipynb
{
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.2"
},
"colab": {
"provenance": [],
"include_colab_link": true
}
},
"cells": [
{
"cell_type": "markdown",
"metadata": {
"id": "view-in-github",
"colab_type": "text"
},
"source": [
"<a href=\"https://colab.research.google.com/gist/alonsosilvaallende/b9618f749916b1a37bda302fe4227ed6/interpretable_ml_xai_regression_2024-11.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"cell_type": "markdown",
"metadata": {
"id": "VCV1UZu-NTgr"
},
"source": [
"# Regression for Iris dataset with 'sepal width' as the target"
]
},
{
"cell_type": "code",
"metadata": {
"id": "vVdIF-vCOYWB"
},
"source": [
"import sklearn\n",
"\n",
"assert sklearn.__version__ >= \"1.0\", \"Please upgrade scikit-learn with %pip install --quiet --upgrade scikit-learn>=1.0\""
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "9ypieMc3NQ-M"
},
"source": [
"import numpy as np\n",
"import pandas as pd\n",
"import matplotlib.pyplot as plt"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "-atdbdkNNQ-i"
},
"source": [
"from sklearn.datasets import load_iris\n",
"\n",
"iris = load_iris()"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"scrolled": false,
"id": "o-LwD0hQNQ-4",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "eda61e46-012e-4743-890b-dee3c06be474"
},
"source": [
"print(iris.DESCR)"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
".. _iris_dataset:\n",
"\n",
"Iris plants dataset\n",
"--------------------\n",
"\n",
"**Data Set Characteristics:**\n",
"\n",
" :Number of Instances: 150 (50 in each of three classes)\n",
" :Number of Attributes: 4 numeric, predictive attributes and the class\n",
" :Attribute Information:\n",
" - sepal length in cm\n",
" - sepal width in cm\n",
" - petal length in cm\n",
" - petal width in cm\n",
" - class:\n",
" - Iris-Setosa\n",
" - Iris-Versicolour\n",
" - Iris-Virginica\n",
" \n",
" :Summary Statistics:\n",
"\n",
" ============== ==== ==== ======= ===== ====================\n",
" Min Max Mean SD Class Correlation\n",
" ============== ==== ==== ======= ===== ====================\n",
" sepal length: 4.3 7.9 5.84 0.83 0.7826\n",
" sepal width: 2.0 4.4 3.05 0.43 -0.4194\n",
" petal length: 1.0 6.9 3.76 1.76 0.9490 (high!)\n",
" petal width: 0.1 2.5 1.20 0.76 0.9565 (high!)\n",
" ============== ==== ==== ======= ===== ====================\n",
"\n",
" :Missing Attribute Values: None\n",
" :Class Distribution: 33.3% for each of 3 classes.\n",
" :Creator: R.A. Fisher\n",
" :Donor: Michael Marshall (MARSHALL%[email protected])\n",
" :Date: July, 1988\n",
"\n",
"The famous Iris database, first used by Sir R.A. Fisher. The dataset is taken\n",
"from Fisher's paper. Note that it's the same as in R, but not as in the UCI\n",
"Machine Learning Repository, which has two wrong data points.\n",
"\n",
"This is perhaps the best known database to be found in the\n",
"pattern recognition literature. Fisher's paper is a classic in the field and\n",
"is referenced frequently to this day. (See Duda & Hart, for example.) The\n",
"data set contains 3 classes of 50 instances each, where each class refers to a\n",
"type of iris plant. One class is linearly separable from the other 2; the\n",
"latter are NOT linearly separable from each other.\n",
"\n",
".. topic:: References\n",
"\n",
" - Fisher, R.A. \"The use of multiple measurements in taxonomic problems\"\n",
" Annual Eugenics, 7, Part II, 179-188 (1936); also in \"Contributions to\n",
" Mathematical Statistics\" (John Wiley, NY, 1950).\n",
" - Duda, R.O., & Hart, P.E. (1973) Pattern Classification and Scene Analysis.\n",
" (Q327.D83) John Wiley & Sons. ISBN 0-471-22361-1. See page 218.\n",
" - Dasarathy, B.V. (1980) \"Nosing Around the Neighborhood: A New System\n",
" Structure and Classification Rule for Recognition in Partially Exposed\n",
" Environments\". IEEE Transactions on Pattern Analysis and Machine\n",
" Intelligence, Vol. PAMI-2, No. 1, 67-71.\n",
" - Gates, G.W. (1972) \"The Reduced Nearest Neighbor Rule\". IEEE Transactions\n",
" on Information Theory, May 1972, 431-433.\n",
" - See also: 1988 MLC Proceedings, 54-64. Cheeseman et al\"s AUTOCLASS II\n",
" conceptual clustering system finds 3 classes in the data.\n",
" - Many, many more ...\n"
]
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "SAFiuZC9NQ_J",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 206
},
"outputId": "b34cf629-f1b5-4d56-ae7a-3b07a38c2238"
},
"source": [
"df_raw = pd.DataFrame(iris.data, columns=iris.feature_names)\n",
"df_raw.head()"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
" sepal length (cm) sepal width (cm) petal length (cm) petal width (cm)\n",
"0 5.1 3.5 1.4 0.2\n",
"1 4.9 3.0 1.4 0.2\n",
"2 4.7 3.2 1.3 0.2\n",
"3 4.6 3.1 1.5 0.2\n",
"4 5.0 3.6 1.4 0.2"
],
"text/html": [
"\n",
" <div id=\"df-1935c552-0530-45e0-a0de-9363dee19b17\" class=\"colab-df-container\">\n",
" <div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>sepal length (cm)</th>\n",
" <th>sepal width (cm)</th>\n",
" <th>petal length (cm)</th>\n",
" <th>petal width (cm)</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>5.1</td>\n",
" <td>3.5</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>4.9</td>\n",
" <td>3.0</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>4.7</td>\n",
" <td>3.2</td>\n",
" <td>1.3</td>\n",
" <td>0.2</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4.6</td>\n",
" <td>3.1</td>\n",
" <td>1.5</td>\n",
" <td>0.2</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5.0</td>\n",
" <td>3.6</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>\n",
" <div class=\"colab-df-buttons\">\n",
"\n",
" <div class=\"colab-df-container\">\n",
" <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1935c552-0530-45e0-a0de-9363dee19b17')\"\n",
" title=\"Convert this dataframe to an interactive table.\"\n",
" style=\"display:none;\">\n",
"\n",
" <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
" <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
" </svg>\n",
" </button>\n",
"\n",
" <style>\n",
" .colab-df-container {\n",
" display:flex;\n",
" gap: 12px;\n",
" }\n",
"\n",
" .colab-df-convert {\n",
" background-color: #E8F0FE;\n",
" border: none;\n",
" border-radius: 50%;\n",
" cursor: pointer;\n",
" display: none;\n",
" fill: #1967D2;\n",
" height: 32px;\n",
" padding: 0 0 0 0;\n",
" width: 32px;\n",
" }\n",
"\n",
" .colab-df-convert:hover {\n",
" background-color: #E2EBFA;\n",
" box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
" fill: #174EA6;\n",
" }\n",
"\n",
" .colab-df-buttons div {\n",
" margin-bottom: 4px;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-convert {\n",
" background-color: #3B4455;\n",
" fill: #D2E3FC;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-convert:hover {\n",
" background-color: #434B5C;\n",
" box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
" filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
" fill: #FFFFFF;\n",
" }\n",
" </style>\n",
"\n",
" <script>\n",
" const buttonEl =\n",
" document.querySelector('#df-1935c552-0530-45e0-a0de-9363dee19b17 button.colab-df-convert');\n",
" buttonEl.style.display =\n",
" google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
"\n",
" async function convertToInteractive(key) {\n",
" const element = document.querySelector('#df-1935c552-0530-45e0-a0de-9363dee19b17');\n",
" const dataTable =\n",
" await google.colab.kernel.invokeFunction('convertToInteractive',\n",
" [key], {});\n",
" if (!dataTable) return;\n",
"\n",
" const docLinkHtml = 'Like what you see? Visit the ' +\n",
" '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
" + ' to learn more about interactive tables.';\n",
" element.innerHTML = '';\n",
" dataTable['output_type'] = 'display_data';\n",
" await google.colab.output.renderOutput(dataTable, element);\n",
" const docLink = document.createElement('div');\n",
" docLink.innerHTML = docLinkHtml;\n",
" element.appendChild(docLink);\n",
" }\n",
" </script>\n",
" </div>\n",
"\n",
"\n",
"<div id=\"df-ff15ac68-9ba2-4889-8c28-c175417040b9\">\n",
" <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-ff15ac68-9ba2-4889-8c28-c175417040b9')\"\n",
" title=\"Suggest charts\"\n",
" style=\"display:none;\">\n",
"\n",
"<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
" width=\"24px\">\n",
" <g>\n",
" <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
" </g>\n",
"</svg>\n",
" </button>\n",
"\n",
"<style>\n",
" .colab-df-quickchart {\n",
" --bg-color: #E8F0FE;\n",
" --fill-color: #1967D2;\n",
" --hover-bg-color: #E2EBFA;\n",
" --hover-fill-color: #174EA6;\n",
" --disabled-fill-color: #AAA;\n",
" --disabled-bg-color: #DDD;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-quickchart {\n",
" --bg-color: #3B4455;\n",
" --fill-color: #D2E3FC;\n",
" --hover-bg-color: #434B5C;\n",
" --hover-fill-color: #FFFFFF;\n",
" --disabled-bg-color: #3B4455;\n",
" --disabled-fill-color: #666;\n",
" }\n",
"\n",
" .colab-df-quickchart {\n",
" background-color: var(--bg-color);\n",
" border: none;\n",
" border-radius: 50%;\n",
" cursor: pointer;\n",
" display: none;\n",
" fill: var(--fill-color);\n",
" height: 32px;\n",
" padding: 0;\n",
" width: 32px;\n",
" }\n",
"\n",
" .colab-df-quickchart:hover {\n",
" background-color: var(--hover-bg-color);\n",
" box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
" fill: var(--button-hover-fill-color);\n",
" }\n",
"\n",
" .colab-df-quickchart-complete:disabled,\n",
" .colab-df-quickchart-complete:disabled:hover {\n",
" background-color: var(--disabled-bg-color);\n",
" fill: var(--disabled-fill-color);\n",
" box-shadow: none;\n",
" }\n",
"\n",
" .colab-df-spinner {\n",
" border: 2px solid var(--fill-color);\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" animation:\n",
" spin 1s steps(1) infinite;\n",
" }\n",
"\n",
" @keyframes spin {\n",
" 0% {\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" border-left-color: var(--fill-color);\n",
" }\n",
" 20% {\n",
" border-color: transparent;\n",
" border-left-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" }\n",
" 30% {\n",
" border-color: transparent;\n",
" border-left-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" border-right-color: var(--fill-color);\n",
" }\n",
" 40% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" }\n",
" 60% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" }\n",
" 80% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" border-bottom-color: var(--fill-color);\n",
" }\n",
" 90% {\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" }\n",
" }\n",
"</style>\n",
"\n",
" <script>\n",
" async function quickchart(key) {\n",
" const quickchartButtonEl =\n",
" document.querySelector('#' + key + ' button');\n",
" quickchartButtonEl.disabled = true; // To prevent multiple clicks.\n",
" quickchartButtonEl.classList.add('colab-df-spinner');\n",
" try {\n",
" const charts = await google.colab.kernel.invokeFunction(\n",
" 'suggestCharts', [key], {});\n",
" } catch (error) {\n",
" console.error('Error during call to suggestCharts:', error);\n",
" }\n",
" quickchartButtonEl.classList.remove('colab-df-spinner');\n",
" quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
" }\n",
" (() => {\n",
" let quickchartButtonEl =\n",
" document.querySelector('#df-ff15ac68-9ba2-4889-8c28-c175417040b9 button');\n",
" quickchartButtonEl.style.display =\n",
" google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
" })();\n",
" </script>\n",
"</div>\n",
" </div>\n",
" </div>\n"
],
"application/vnd.google.colaboratory.intrinsic+json": {
"type": "dataframe",
"variable_name": "df_raw",
"summary": "{\n \"name\": \"df_raw\",\n \"rows\": 150,\n \"fields\": [\n {\n \"column\": \"sepal length (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.828066127977863,\n \"min\": 4.3,\n \"max\": 7.9,\n \"num_unique_values\": 35,\n \"samples\": [\n 6.2,\n 4.5,\n 5.6\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"sepal width (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.4358662849366982,\n \"min\": 2.0,\n \"max\": 4.4,\n \"num_unique_values\": 23,\n \"samples\": [\n 2.3,\n 4.0,\n 3.5\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"petal length (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 1.7652982332594662,\n \"min\": 1.0,\n \"max\": 6.9,\n \"num_unique_values\": 43,\n \"samples\": [\n 6.7,\n 3.8,\n 3.7\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"petal width (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.7622376689603465,\n \"min\": 0.1,\n \"max\": 2.5,\n \"num_unique_values\": 22,\n \"samples\": [\n 0.2,\n 1.2,\n 1.3\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"
}
},
"metadata": {},
"execution_count": 5
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "e0HkoLYYNQ_f",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 206
},
"outputId": "d7464769-beac-431a-d3e2-1cd5076d9a00"
},
"source": [
"df_raw['class'] = iris.target\n",
"df_raw.head()"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
" sepal length (cm) sepal width (cm) petal length (cm) petal width (cm) \\\n",
"0 5.1 3.5 1.4 0.2 \n",
"1 4.9 3.0 1.4 0.2 \n",
"2 4.7 3.2 1.3 0.2 \n",
"3 4.6 3.1 1.5 0.2 \n",
"4 5.0 3.6 1.4 0.2 \n",
"\n",
" class \n",
"0 0 \n",
"1 0 \n",
"2 0 \n",
"3 0 \n",
"4 0 "
],
"text/html": [
"\n",
" <div id=\"df-129fca03-1c1f-46f1-b946-0be05f117ee2\" class=\"colab-df-container\">\n",
" <div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>sepal length (cm)</th>\n",
" <th>sepal width (cm)</th>\n",
" <th>petal length (cm)</th>\n",
" <th>petal width (cm)</th>\n",
" <th>class</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>5.1</td>\n",
" <td>3.5</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>4.9</td>\n",
" <td>3.0</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>4.7</td>\n",
" <td>3.2</td>\n",
" <td>1.3</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4.6</td>\n",
" <td>3.1</td>\n",
" <td>1.5</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5.0</td>\n",
" <td>3.6</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>\n",
" <div class=\"colab-df-buttons\">\n",
"\n",
" <div class=\"colab-df-container\">\n",
" <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-129fca03-1c1f-46f1-b946-0be05f117ee2')\"\n",
" title=\"Convert this dataframe to an interactive table.\"\n",
" style=\"display:none;\">\n",
"\n",
" <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
" <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
" </svg>\n",
" </button>\n",
"\n",
" <style>\n",
" .colab-df-container {\n",
" display:flex;\n",
" gap: 12px;\n",
" }\n",
"\n",
" .colab-df-convert {\n",
" background-color: #E8F0FE;\n",
" border: none;\n",
" border-radius: 50%;\n",
" cursor: pointer;\n",
" display: none;\n",
" fill: #1967D2;\n",
" height: 32px;\n",
" padding: 0 0 0 0;\n",
" width: 32px;\n",
" }\n",
"\n",
" .colab-df-convert:hover {\n",
" background-color: #E2EBFA;\n",
" box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
" fill: #174EA6;\n",
" }\n",
"\n",
" .colab-df-buttons div {\n",
" margin-bottom: 4px;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-convert {\n",
" background-color: #3B4455;\n",
" fill: #D2E3FC;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-convert:hover {\n",
" background-color: #434B5C;\n",
" box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
" filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
" fill: #FFFFFF;\n",
" }\n",
" </style>\n",
"\n",
" <script>\n",
" const buttonEl =\n",
" document.querySelector('#df-129fca03-1c1f-46f1-b946-0be05f117ee2 button.colab-df-convert');\n",
" buttonEl.style.display =\n",
" google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
"\n",
" async function convertToInteractive(key) {\n",
" const element = document.querySelector('#df-129fca03-1c1f-46f1-b946-0be05f117ee2');\n",
" const dataTable =\n",
" await google.colab.kernel.invokeFunction('convertToInteractive',\n",
" [key], {});\n",
" if (!dataTable) return;\n",
"\n",
" const docLinkHtml = 'Like what you see? Visit the ' +\n",
" '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
" + ' to learn more about interactive tables.';\n",
" element.innerHTML = '';\n",
" dataTable['output_type'] = 'display_data';\n",
" await google.colab.output.renderOutput(dataTable, element);\n",
" const docLink = document.createElement('div');\n",
" docLink.innerHTML = docLinkHtml;\n",
" element.appendChild(docLink);\n",
" }\n",
" </script>\n",
" </div>\n",
"\n",
"\n",
"<div id=\"df-3b12997f-b504-4276-9a3b-0e0b46783dd2\">\n",
" <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-3b12997f-b504-4276-9a3b-0e0b46783dd2')\"\n",
" title=\"Suggest charts\"\n",
" style=\"display:none;\">\n",
"\n",
"<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
" width=\"24px\">\n",
" <g>\n",
" <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
" </g>\n",
"</svg>\n",
" </button>\n",
"\n",
"<style>\n",
" .colab-df-quickchart {\n",
" --bg-color: #E8F0FE;\n",
" --fill-color: #1967D2;\n",
" --hover-bg-color: #E2EBFA;\n",
" --hover-fill-color: #174EA6;\n",
" --disabled-fill-color: #AAA;\n",
" --disabled-bg-color: #DDD;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-quickchart {\n",
" --bg-color: #3B4455;\n",
" --fill-color: #D2E3FC;\n",
" --hover-bg-color: #434B5C;\n",
" --hover-fill-color: #FFFFFF;\n",
" --disabled-bg-color: #3B4455;\n",
" --disabled-fill-color: #666;\n",
" }\n",
"\n",
" .colab-df-quickchart {\n",
" background-color: var(--bg-color);\n",
" border: none;\n",
" border-radius: 50%;\n",
" cursor: pointer;\n",
" display: none;\n",
" fill: var(--fill-color);\n",
" height: 32px;\n",
" padding: 0;\n",
" width: 32px;\n",
" }\n",
"\n",
" .colab-df-quickchart:hover {\n",
" background-color: var(--hover-bg-color);\n",
" box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
" fill: var(--button-hover-fill-color);\n",
" }\n",
"\n",
" .colab-df-quickchart-complete:disabled,\n",
" .colab-df-quickchart-complete:disabled:hover {\n",
" background-color: var(--disabled-bg-color);\n",
" fill: var(--disabled-fill-color);\n",
" box-shadow: none;\n",
" }\n",
"\n",
" .colab-df-spinner {\n",
" border: 2px solid var(--fill-color);\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" animation:\n",
" spin 1s steps(1) infinite;\n",
" }\n",
"\n",
" @keyframes spin {\n",
" 0% {\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" border-left-color: var(--fill-color);\n",
" }\n",
" 20% {\n",
" border-color: transparent;\n",
" border-left-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" }\n",
" 30% {\n",
" border-color: transparent;\n",
" border-left-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" border-right-color: var(--fill-color);\n",
" }\n",
" 40% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" }\n",
" 60% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" }\n",
" 80% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" border-bottom-color: var(--fill-color);\n",
" }\n",
" 90% {\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" }\n",
" }\n",
"</style>\n",
"\n",
" <script>\n",
" async function quickchart(key) {\n",
" const quickchartButtonEl =\n",
" document.querySelector('#' + key + ' button');\n",
" quickchartButtonEl.disabled = true; // To prevent multiple clicks.\n",
" quickchartButtonEl.classList.add('colab-df-spinner');\n",
" try {\n",
" const charts = await google.colab.kernel.invokeFunction(\n",
" 'suggestCharts', [key], {});\n",
" } catch (error) {\n",
" console.error('Error during call to suggestCharts:', error);\n",
" }\n",
" quickchartButtonEl.classList.remove('colab-df-spinner');\n",
" quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
" }\n",
" (() => {\n",
" let quickchartButtonEl =\n",
" document.querySelector('#df-3b12997f-b504-4276-9a3b-0e0b46783dd2 button');\n",
" quickchartButtonEl.style.display =\n",
" google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
" })();\n",
" </script>\n",
"</div>\n",
" </div>\n",
" </div>\n"
],
"application/vnd.google.colaboratory.intrinsic+json": {
"type": "dataframe",
"variable_name": "df_raw",
"summary": "{\n \"name\": \"df_raw\",\n \"rows\": 150,\n \"fields\": [\n {\n \"column\": \"sepal length (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.828066127977863,\n \"min\": 4.3,\n \"max\": 7.9,\n \"num_unique_values\": 35,\n \"samples\": [\n 6.2,\n 4.5,\n 5.6\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"sepal width (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.4358662849366982,\n \"min\": 2.0,\n \"max\": 4.4,\n \"num_unique_values\": 23,\n \"samples\": [\n 2.3,\n 4.0,\n 3.5\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"petal length (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 1.7652982332594662,\n \"min\": 1.0,\n \"max\": 6.9,\n \"num_unique_values\": 43,\n \"samples\": [\n 6.7,\n 3.8,\n 3.7\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"petal width (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.7622376689603465,\n \"min\": 0.1,\n \"max\": 2.5,\n \"num_unique_values\": 22,\n \"samples\": [\n 0.2,\n 1.2,\n 1.3\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"class\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0,\n \"min\": 0,\n \"max\": 2,\n \"num_unique_values\": 3,\n \"samples\": [\n 0,\n 1,\n 2\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"
}
},
"metadata": {},
"execution_count": 6
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "ZAthwy2MNQ_3",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "24aee09d-490b-4c42-f874-4fb8d4df77d1"
},
"source": [
"df_raw['class'].unique()"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"array([0, 1, 2])"
]
},
"metadata": {},
"execution_count": 7
}
]
},
{
"cell_type": "code",
"source": [
"list(iris.target_names)"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "NJfKYx5u-kss",
"outputId": "b2f38ea5-a112-4502-bde2-c4fb4cf9ffcf"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"['setosa', 'versicolor', 'virginica']"
]
},
"metadata": {},
"execution_count": 8
}
]
},
{
"cell_type": "code",
"source": [
"X = df_raw.drop(columns='sepal width (cm)')\n",
"X.head()"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 206
},
"id": "_I57WI_pSh6f",
"outputId": "ec3dbf14-9cff-4599-f7af-b817d7e326d2"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
" sepal length (cm) petal length (cm) petal width (cm) class\n",
"0 5.1 1.4 0.2 0\n",
"1 4.9 1.4 0.2 0\n",
"2 4.7 1.3 0.2 0\n",
"3 4.6 1.5 0.2 0\n",
"4 5.0 1.4 0.2 0"
],
"text/html": [
"\n",
" <div id=\"df-f3e3d8f0-51ba-402e-a806-88ec7a9d962d\" class=\"colab-df-container\">\n",
" <div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>sepal length (cm)</th>\n",
" <th>petal length (cm)</th>\n",
" <th>petal width (cm)</th>\n",
" <th>class</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>5.1</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>4.9</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>4.7</td>\n",
" <td>1.3</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>4.6</td>\n",
" <td>1.5</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>5.0</td>\n",
" <td>1.4</td>\n",
" <td>0.2</td>\n",
" <td>0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>\n",
" <div class=\"colab-df-buttons\">\n",
"\n",
" <div class=\"colab-df-container\">\n",
" <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-f3e3d8f0-51ba-402e-a806-88ec7a9d962d')\"\n",
" title=\"Convert this dataframe to an interactive table.\"\n",
" style=\"display:none;\">\n",
"\n",
" <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
" <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
" </svg>\n",
" </button>\n",
"\n",
" <style>\n",
" .colab-df-container {\n",
" display:flex;\n",
" gap: 12px;\n",
" }\n",
"\n",
" .colab-df-convert {\n",
" background-color: #E8F0FE;\n",
" border: none;\n",
" border-radius: 50%;\n",
" cursor: pointer;\n",
" display: none;\n",
" fill: #1967D2;\n",
" height: 32px;\n",
" padding: 0 0 0 0;\n",
" width: 32px;\n",
" }\n",
"\n",
" .colab-df-convert:hover {\n",
" background-color: #E2EBFA;\n",
" box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
" fill: #174EA6;\n",
" }\n",
"\n",
" .colab-df-buttons div {\n",
" margin-bottom: 4px;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-convert {\n",
" background-color: #3B4455;\n",
" fill: #D2E3FC;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-convert:hover {\n",
" background-color: #434B5C;\n",
" box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
" filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
" fill: #FFFFFF;\n",
" }\n",
" </style>\n",
"\n",
" <script>\n",
" const buttonEl =\n",
" document.querySelector('#df-f3e3d8f0-51ba-402e-a806-88ec7a9d962d button.colab-df-convert');\n",
" buttonEl.style.display =\n",
" google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
"\n",
" async function convertToInteractive(key) {\n",
" const element = document.querySelector('#df-f3e3d8f0-51ba-402e-a806-88ec7a9d962d');\n",
" const dataTable =\n",
" await google.colab.kernel.invokeFunction('convertToInteractive',\n",
" [key], {});\n",
" if (!dataTable) return;\n",
"\n",
" const docLinkHtml = 'Like what you see? Visit the ' +\n",
" '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
" + ' to learn more about interactive tables.';\n",
" element.innerHTML = '';\n",
" dataTable['output_type'] = 'display_data';\n",
" await google.colab.output.renderOutput(dataTable, element);\n",
" const docLink = document.createElement('div');\n",
" docLink.innerHTML = docLinkHtml;\n",
" element.appendChild(docLink);\n",
" }\n",
" </script>\n",
" </div>\n",
"\n",
"\n",
"<div id=\"df-d5fc1737-63dd-4494-bdc5-d79c2cf6929a\">\n",
" <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-d5fc1737-63dd-4494-bdc5-d79c2cf6929a')\"\n",
" title=\"Suggest charts\"\n",
" style=\"display:none;\">\n",
"\n",
"<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
" width=\"24px\">\n",
" <g>\n",
" <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
" </g>\n",
"</svg>\n",
" </button>\n",
"\n",
"<style>\n",
" .colab-df-quickchart {\n",
" --bg-color: #E8F0FE;\n",
" --fill-color: #1967D2;\n",
" --hover-bg-color: #E2EBFA;\n",
" --hover-fill-color: #174EA6;\n",
" --disabled-fill-color: #AAA;\n",
" --disabled-bg-color: #DDD;\n",
" }\n",
"\n",
" [theme=dark] .colab-df-quickchart {\n",
" --bg-color: #3B4455;\n",
" --fill-color: #D2E3FC;\n",
" --hover-bg-color: #434B5C;\n",
" --hover-fill-color: #FFFFFF;\n",
" --disabled-bg-color: #3B4455;\n",
" --disabled-fill-color: #666;\n",
" }\n",
"\n",
" .colab-df-quickchart {\n",
" background-color: var(--bg-color);\n",
" border: none;\n",
" border-radius: 50%;\n",
" cursor: pointer;\n",
" display: none;\n",
" fill: var(--fill-color);\n",
" height: 32px;\n",
" padding: 0;\n",
" width: 32px;\n",
" }\n",
"\n",
" .colab-df-quickchart:hover {\n",
" background-color: var(--hover-bg-color);\n",
" box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
" fill: var(--button-hover-fill-color);\n",
" }\n",
"\n",
" .colab-df-quickchart-complete:disabled,\n",
" .colab-df-quickchart-complete:disabled:hover {\n",
" background-color: var(--disabled-bg-color);\n",
" fill: var(--disabled-fill-color);\n",
" box-shadow: none;\n",
" }\n",
"\n",
" .colab-df-spinner {\n",
" border: 2px solid var(--fill-color);\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" animation:\n",
" spin 1s steps(1) infinite;\n",
" }\n",
"\n",
" @keyframes spin {\n",
" 0% {\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" border-left-color: var(--fill-color);\n",
" }\n",
" 20% {\n",
" border-color: transparent;\n",
" border-left-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" }\n",
" 30% {\n",
" border-color: transparent;\n",
" border-left-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" border-right-color: var(--fill-color);\n",
" }\n",
" 40% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" border-top-color: var(--fill-color);\n",
" }\n",
" 60% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" }\n",
" 80% {\n",
" border-color: transparent;\n",
" border-right-color: var(--fill-color);\n",
" border-bottom-color: var(--fill-color);\n",
" }\n",
" 90% {\n",
" border-color: transparent;\n",
" border-bottom-color: var(--fill-color);\n",
" }\n",
" }\n",
"</style>\n",
"\n",
" <script>\n",
" async function quickchart(key) {\n",
" const quickchartButtonEl =\n",
" document.querySelector('#' + key + ' button');\n",
" quickchartButtonEl.disabled = true; // To prevent multiple clicks.\n",
" quickchartButtonEl.classList.add('colab-df-spinner');\n",
" try {\n",
" const charts = await google.colab.kernel.invokeFunction(\n",
" 'suggestCharts', [key], {});\n",
" } catch (error) {\n",
" console.error('Error during call to suggestCharts:', error);\n",
" }\n",
" quickchartButtonEl.classList.remove('colab-df-spinner');\n",
" quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
" }\n",
" (() => {\n",
" let quickchartButtonEl =\n",
" document.querySelector('#df-d5fc1737-63dd-4494-bdc5-d79c2cf6929a button');\n",
" quickchartButtonEl.style.display =\n",
" google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
" })();\n",
" </script>\n",
"</div>\n",
" </div>\n",
" </div>\n"
],
"application/vnd.google.colaboratory.intrinsic+json": {
"type": "dataframe",
"variable_name": "X",
"summary": "{\n \"name\": \"X\",\n \"rows\": 150,\n \"fields\": [\n {\n \"column\": \"sepal length (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.828066127977863,\n \"min\": 4.3,\n \"max\": 7.9,\n \"num_unique_values\": 35,\n \"samples\": [\n 6.2,\n 4.5,\n 5.6\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"petal length (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 1.7652982332594662,\n \"min\": 1.0,\n \"max\": 6.9,\n \"num_unique_values\": 43,\n \"samples\": [\n 6.7,\n 3.8,\n 3.7\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"petal width (cm)\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0.7622376689603465,\n \"min\": 0.1,\n \"max\": 2.5,\n \"num_unique_values\": 22,\n \"samples\": [\n 0.2,\n 1.2,\n 1.3\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"class\",\n \"properties\": {\n \"dtype\": \"number\",\n \"std\": 0,\n \"min\": 0,\n \"max\": 2,\n \"num_unique_values\": 3,\n \"samples\": [\n 0,\n 1,\n 2\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}"
}
},
"metadata": {},
"execution_count": 9
}
]
},
{
"cell_type": "code",
"source": [
"y = df_raw['sepal width (cm)']\n",
"y[:5]"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "Y319buLBSnuf",
"outputId": "9ee3e322-7ad9-46ec-ebf8-4b373e70417c"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"0 3.5\n",
"1 3.0\n",
"2 3.2\n",
"3 3.1\n",
"4 3.6\n",
"Name: sepal width (cm), dtype: float64"
]
},
"metadata": {},
"execution_count": 10
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "d0MYo8ixNRAA"
},
"source": [
"from sklearn.model_selection import train_test_split\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(\n",
" X, y, random_state=42)"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "nY28r0Z6NRAK",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "8ed4ba34-46ef-4278-a071-816da4be58ec"
},
"source": [
"X_train.columns"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"Index(['sepal length (cm)', 'petal length (cm)', 'petal width (cm)', 'class'], dtype='object')"
]
},
"metadata": {},
"execution_count": 12
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "TSYfiyqPNRAT",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 36
},
"outputId": "e7a535bf-8c5c-453a-c57b-2310f886c619"
},
"source": [
"y_train.name"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"'sepal width (cm)'"
],
"application/vnd.google.colaboratory.intrinsic+json": {
"type": "string"
}
},
"metadata": {},
"execution_count": 13
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "_Nn9wkJINRAj"
},
"source": [
"categorical_columns = ['class']\n",
"numerical_columns = ['sepal length (cm)', 'petal length (cm)', 'petal width (cm)']"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "SHNFVFSCNRAt"
},
"source": [
"from sklearn.pipeline import Pipeline\n",
"from sklearn.preprocessing import OneHotEncoder\n",
"\n",
"categorical_pipe = Pipeline([\n",
" ('onehot', OneHotEncoder())\n",
"])"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"pd.DataFrame(X_train['class']).head(5).style.hide(axis=\"index\")"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 206
},
"id": "Svs5Vu0IClPD",
"outputId": "085486ac-6d18-45d2-a22e-0ebba0173669"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"<pandas.io.formats.style.Styler at 0x7861598973a0>"
],
"text/html": [
"<style type=\"text/css\">\n",
"</style>\n",
"<table id=\"T_6ffd9\" class=\"dataframe\">\n",
" <thead>\n",
" <tr>\n",
" <th id=\"T_6ffd9_level0_col0\" class=\"col_heading level0 col0\" >class</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td id=\"T_6ffd9_row0_col0\" class=\"data row0 col0\" >0</td>\n",
" </tr>\n",
" <tr>\n",
" <td id=\"T_6ffd9_row1_col0\" class=\"data row1 col0\" >0</td>\n",
" </tr>\n",
" <tr>\n",
" <td id=\"T_6ffd9_row2_col0\" class=\"data row2 col0\" >2</td>\n",
" </tr>\n",
" <tr>\n",
" <td id=\"T_6ffd9_row3_col0\" class=\"data row3 col0\" >1</td>\n",
" </tr>\n",
" <tr>\n",
" <td id=\"T_6ffd9_row4_col0\" class=\"data row4 col0\" >1</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n"
]
},
"metadata": {},
"execution_count": 16
}
]
},
{
"cell_type": "code",
"source": [
"encoder = OneHotEncoder(sparse_output=False)\n",
"X = encoder.fit_transform(pd.DataFrame(X_train['class']))"
],
"metadata": {
"id": "RHVVQuInDJ9H"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"Scikit-Learn loves Pandas: https://colab.research.google.com/gist/ageron/d4e0d3caed542b8a4285ef433f650a8d/scikit-learn-pandas.ipynb#scrollTo=9jZTcN0OC8ZP"
],
"metadata": {
"id": "N70viswyma4H"
}
},
{
"cell_type": "code",
"source": [
"encoder.feature_names_in_"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "qbJssFuqDQ6b",
"outputId": "dc34ea1f-5096-44b8-9269-c6218bb62f04"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"array(['class'], dtype=object)"
]
},
"metadata": {},
"execution_count": 18
}
]
},
{
"cell_type": "code",
"source": [
"encoder.get_feature_names_out()"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "y-MD3HyjDYqW",
"outputId": "ab2aeedc-7520-4e74-936f-66df103de4ea"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"array(['class_0', 'class_1', 'class_2'], dtype=object)"
]
},
"metadata": {},
"execution_count": 19
}
]
},
{
"cell_type": "code",
"source": [
"X_train_df = pd.DataFrame(X, columns=encoder.get_feature_names_out(), index=X_train.index)\n",
"formatdict = {col: '{:.1f}' for col in X_train_df.columns.values}\n",
"X_train_df[:5].style.hide(axis=\"index\").format(formatdict)"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 206
},
"id": "XrVNezgkUavU",
"outputId": "e8f95eff-130a-4e40-f681-b3cff4559994"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"<pandas.io.formats.style.Styler at 0x786159895390>"
],
"text/html": [
"<style type=\"text/css\">\n",
"</style>\n",
"<table id=\"T_196b7\" class=\"dataframe\">\n",
" <thead>\n",
" <tr>\n",
" <th id=\"T_196b7_level0_col0\" class=\"col_heading level0 col0\" >class_0</th>\n",
" <th id=\"T_196b7_level0_col1\" class=\"col_heading level0 col1\" >class_1</th>\n",
" <th id=\"T_196b7_level0_col2\" class=\"col_heading level0 col2\" >class_2</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <td id=\"T_196b7_row0_col0\" class=\"data row0 col0\" >1.0</td>\n",
" <td id=\"T_196b7_row0_col1\" class=\"data row0 col1\" >0.0</td>\n",
" <td id=\"T_196b7_row0_col2\" class=\"data row0 col2\" >0.0</td>\n",
" </tr>\n",
" <tr>\n",
" <td id=\"T_196b7_row1_col0\" class=\"data row1 col0\" >1.0</td>\n",
" <td id=\"T_196b7_row1_col1\" class=\"data row1 col1\" >0.0</td>\n",
" <td id=\"T_196b7_row1_col2\" class=\"data row1 col2\" >0.0</td>\n",
" </tr>\n",
" <tr>\n",
" <td id=\"T_196b7_row2_col0\" class=\"data row2 col0\" >0.0</td>\n",
" <td id=\"T_196b7_row2_col1\" class=\"data row2 col1\" >0.0</td>\n",
" <td id=\"T_196b7_row2_col2\" class=\"data row2 col2\" >1.0</td>\n",
" </tr>\n",
" <tr>\n",
" <td id=\"T_196b7_row3_col0\" class=\"data row3 col0\" >0.0</td>\n",
" <td id=\"T_196b7_row3_col1\" class=\"data row3 col1\" >1.0</td>\n",
" <td id=\"T_196b7_row3_col2\" class=\"data row3 col2\" >0.0</td>\n",
" </tr>\n",
" <tr>\n",
" <td id=\"T_196b7_row4_col0\" class=\"data row4 col0\" >0.0</td>\n",
" <td id=\"T_196b7_row4_col1\" class=\"data row4 col1\" >1.0</td>\n",
" <td id=\"T_196b7_row4_col2\" class=\"data row4 col2\" >0.0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n"
]
},
"metadata": {},
"execution_count": 20
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "DZIhlDdKNRA8"
},
"source": [
"from sklearn.preprocessing import StandardScaler\n",
"\n",
"numerical_pipe = Pipeline([\n",
" ('scaler', StandardScaler())\n",
"])"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "RwfiM8TKNRBD"
},
"source": [
"from sklearn.compose import ColumnTransformer\n",
"\n",
"preprocessing = ColumnTransformer(\n",
" [('cat', categorical_pipe, categorical_columns),\n",
" ('num', numerical_pipe, numerical_columns)]\n",
")"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"## KNeighborsRegressor"
],
"metadata": {
"id": "DV5JV17b3jQz"
}
},
{
"cell_type": "code",
"source": [
"from sklearn.neighbors import KNeighborsRegressor\n",
"\n",
"knn = Pipeline([\n",
" ('preprocess', preprocessing),\n",
" ('knn', KNeighborsRegressor(n_neighbors=1))\n",
"])"
],
"metadata": {
"id": "q90veasnBAGS"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"from sklearn import set_config\n",
"\n",
"set_config(display='diagram')\n",
"\n",
"knn.fit(X_train, y_train)"
],
"metadata": {
"id": "adLj3uqmBT1X",
"outputId": "c685fe34-b05e-406d-c59c-f0d5de0fa7d2",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 257
}
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"Pipeline(steps=[('preprocess',\n",
" ColumnTransformer(transformers=[('cat',\n",
" Pipeline(steps=[('onehot',\n",
" OneHotEncoder())]),\n",
" ['class']),\n",
" ('num',\n",
" Pipeline(steps=[('scaler',\n",
" StandardScaler())]),\n",
" ['sepal length (cm)',\n",
" 'petal length (cm)',\n",
" 'petal width (cm)'])])),\n",
" ('knn', KNeighborsRegressor(n_neighbors=1))])"
],
"text/html": [
"<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;,\n",
" &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])),\n",
" (&#x27;knn&#x27;, KNeighborsRegressor(n_neighbors=1))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;,\n",
" &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])),\n",
" (&#x27;knn&#x27;, KNeighborsRegressor(n_neighbors=1))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">preprocess: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;, OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">cat</label><div class=\"sk-toggleable__content\"><pre>[&#x27;class&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">OneHotEncoder</label><div class=\"sk-toggleable__content\"><pre>OneHotEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num</label><div class=\"sk-toggleable__content\"><pre>[&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;, &#x27;petal width (cm)&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" ><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" ><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">KNeighborsRegressor</label><div class=\"sk-toggleable__content\"><pre>KNeighborsRegressor(n_neighbors=1)</pre></div></div></div></div></div></div></div>"
]
},
"metadata": {},
"execution_count": 24
}
]
},
{
"cell_type": "code",
"source": [
"from sklearn.metrics import median_absolute_error\n",
"\n",
"print(\"train error: %0.3f, test error: %0.3f\" %\n",
" (median_absolute_error(y_train, knn.predict(X_train)),\n",
" median_absolute_error(y_test, knn.predict(X_test))))"
],
"metadata": {
"id": "CmbjfLmABb1F",
"outputId": "2e3a7fb2-a121-4220-cbc8-dd742abf82ea",
"colab": {
"base_uri": "https://localhost:8080/"
}
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"train error: 0.000, test error: 0.300\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"def scatter_predictions(y_pred, y_true):\n",
" plt.figure(figsize=(6, 6))\n",
" plt.xlabel('true target')\n",
" plt.ylabel('prediction')\n",
" plt.scatter(y_true, y_pred, color=\"C0\")\n",
" plt.plot(y_true,y_true, color=\"C1\")\n",
"\n",
"scatter_predictions(knn.predict(X_test), y_test)"
],
"metadata": {
"id": "_qVERwk2BsR-",
"outputId": "4961dfdd-183c-4ea6-dcda-6d3e5d2aa42a",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 545
}
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 600x600 with 1 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"from sklearn.inspection import permutation_importance\n",
"\n",
"def boxplot_pi(model, X_test, y_test):\n",
" result = permutation_importance(model, X_test, y_test, n_repeats=10,\n",
" random_state=42, n_jobs=-1)\n",
" sorted_idx = result.importances_mean.argsort()\n",
"\n",
" fig, ax = plt.subplots(figsize=(8,6))\n",
" ax.boxplot(result.importances[sorted_idx].T,\n",
" vert=False, labels=X_test.columns[sorted_idx])\n",
" ax.set_title(\"Permutation Importances (test set)\")\n",
" fig.tight_layout()\n",
" plt.show()\n",
"\n",
"boxplot_pi(knn, X_test, y_test)"
],
"metadata": {
"id": "0eCWxLhCEMuj",
"outputId": "b49bfa8d-75c7-40c3-89a1-2969c680e916",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 607
}
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 800x600 with 1 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"import seaborn as sns\n",
"\n",
"sorted_idx = diff.median().sort_values(ascending=False).index\n",
"sns.boxplot(diff[sorted_idx], orient=\"h\");"
],
"metadata": {
"id": "ndQy-OQTwuNa",
"outputId": "07d6c887-fb9d-4fd6-8b2e-8545e4b61b2b",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 430
}
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"y_pred = knn.predict(X_test)"
],
"metadata": {
"id": "-CVcFyCSwIbb"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"from sklearn.metrics import r2_score"
],
"metadata": {
"id": "wOkdo_DgwbZA"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"diff = pd.DataFrame()\n",
"for col in X_test.columns:\n",
" diff_col = []\n",
" for n in range(100):\n",
" X_aux = X_test.copy()\n",
" X_aux[col] = np.random.permutation(X_aux[col])\n",
" y_pred_aux = knn.predict(X_aux)\n",
" diff_col.append(r2_score(y_test, y_pred) - r2_score(y_test, y_pred_aux))\n",
" # print(r2_score(y_test, y_pred) - r2_score(y_test, y_pred_aux))\n",
" diff[col] = diff_col"
],
"metadata": {
"id": "D9R4ZZu1vtx6"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"diff.median().sort_values(ascending=False)"
],
"metadata": {
"id": "EAbyIN4wwQSh",
"outputId": "8995c0f5-f0b1-40cd-b1f9-96f3131a098d",
"colab": {
"base_uri": "https://localhost:8080/"
}
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"sepal length (cm) 0.439174\n",
"class 0.360237\n",
"petal length (cm) 0.167202\n",
"petal width (cm) 0.102617\n",
"dtype: float64"
]
},
"metadata": {},
"execution_count": 38
}
]
},
{
"cell_type": "markdown",
"source": [
"## LinearRegression"
],
"metadata": {
"id": "AMh5g6oW34-x"
}
},
{
"cell_type": "code",
"metadata": {
"id": "dzk8LAsaNRBR"
},
"source": [
"from sklearn.linear_model import LinearRegression\n",
"\n",
"lm = Pipeline([\n",
" ('preprocess', preprocessing),\n",
" ('regressor', LinearRegression())\n",
"])"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"scrolled": true,
"id": "UMRyr788NRBc",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 257
},
"outputId": "a9690606-fdd0-4c19-f591-f77ae28be702"
},
"source": [
"from sklearn import set_config\n",
"\n",
"set_config(display='diagram')\n",
"\n",
"lm.fit(X_train, y_train)"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"Pipeline(steps=[('preprocess',\n",
" ColumnTransformer(transformers=[('cat',\n",
" Pipeline(steps=[('onehot',\n",
" OneHotEncoder())]),\n",
" ['class']),\n",
" ('num',\n",
" Pipeline(steps=[('scaler',\n",
" StandardScaler())]),\n",
" ['sepal length (cm)',\n",
" 'petal length (cm)',\n",
" 'petal width (cm)'])])),\n",
" ('regressor', LinearRegression())])"
],
"text/html": [
"<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;,\n",
" &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])),\n",
" (&#x27;regressor&#x27;, LinearRegression())])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-8\" type=\"checkbox\" ><label for=\"sk-estimator-id-8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;,\n",
" &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])),\n",
" (&#x27;regressor&#x27;, LinearRegression())])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-9\" type=\"checkbox\" ><label for=\"sk-estimator-id-9\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">preprocess: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;, OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-10\" type=\"checkbox\" ><label for=\"sk-estimator-id-10\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">cat</label><div class=\"sk-toggleable__content\"><pre>[&#x27;class&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-11\" type=\"checkbox\" ><label for=\"sk-estimator-id-11\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">OneHotEncoder</label><div class=\"sk-toggleable__content\"><pre>OneHotEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-12\" type=\"checkbox\" ><label for=\"sk-estimator-id-12\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num</label><div class=\"sk-toggleable__content\"><pre>[&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;, &#x27;petal width (cm)&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-13\" type=\"checkbox\" ><label for=\"sk-estimator-id-13\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-14\" type=\"checkbox\" ><label for=\"sk-estimator-id-14\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LinearRegression</label><div class=\"sk-toggleable__content\"><pre>LinearRegression()</pre></div></div></div></div></div></div></div>"
]
},
"metadata": {},
"execution_count": 41
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "o9rVdLtCNRBv",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "10044a88-02a8-4af4-d4b2-3f4bb1e6f66d"
},
"source": [
"from sklearn.metrics import median_absolute_error\n",
"\n",
"print(\"train error: %0.3f, test error: %0.3f\" %\n",
" (median_absolute_error(y_train, lm.predict(X_train)),\n",
" median_absolute_error(y_test, lm.predict(X_test))))"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"train error: 0.163, test error: 0.193\n"
]
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "53e0p0D8NRBz",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 545
},
"outputId": "c96c388f-b216-444c-8dd6-ae3dfc652abd"
},
"source": [
"scatter_predictions(lm.predict(X_test), y_test)"
],
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 600x600 with 1 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"boxplot_pi(lm, X_test, y_test)"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 607
},
"id": "X8bwlzZDcu2G",
"outputId": "b5f55506-4f4e-4b4f-8e43-46c2638a2e15"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 800x600 with 1 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"%pip install --quiet shap"
],
"metadata": {
"id": "aEnZLzSgxgMe",
"outputId": "194c51d0-ce44-4308-c572-0c971ccda239",
"colab": {
"base_uri": "https://localhost:8080/"
}
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"\u001b[?25l \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/540.5 kB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r\u001b[2K \u001b[91m━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m112.6/540.5 kB\u001b[0m \u001b[31m3.4 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r\u001b[2K \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━━━━━━━━\u001b[0m \u001b[32m378.9/540.5 kB\u001b[0m \u001b[31m5.5 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m540.5/540.5 kB\u001b[0m \u001b[31m5.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
"\u001b[?25h"
]
}
]
},
{
"cell_type": "code",
"source": [
"import shap\n",
"\n",
"explainer = shap.Explainer(lm.predict, X_train)"
],
"metadata": {
"id": "eM9xA30DxaU8"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"shap_values = explainer(X_test)"
],
"metadata": {
"id": "vTVBK1MVyH0T",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "6dce23c7-986f-4967-c32e-53dd10f50c2c"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stderr",
"text": [
"ExactExplainer explainer: 39it [00:18, 2.06it/s] \n"
]
}
]
},
{
"cell_type": "code",
"source": [
"shap.plots.waterfall(shap_values[5])"
],
"metadata": {
"id": "yKkqztX-ydYx",
"outputId": "a13ca535-33c2-4ead-a965-61495a4a59d5",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 386
}
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 800x350 with 3 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"X_test.iloc[5]"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "LnqyOi0TptRT",
"outputId": "5145ada6-eb00-4190-9bd6-faa8d8d94acb"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"sepal length (cm) 5.4\n",
"petal length (cm) 1.5\n",
"petal width (cm) 0.4\n",
"class 0.0\n",
"Name: 31, dtype: float64"
]
},
"metadata": {},
"execution_count": 51
}
]
},
{
"cell_type": "code",
"source": [
"y_test[:10]"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "bRdknk72qviM",
"outputId": "2e096337-f7b8-4258-fcf2-633f16106636"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"73 2.8\n",
"18 3.8\n",
"118 2.6\n",
"78 2.9\n",
"76 2.8\n",
"31 3.4\n",
"64 2.9\n",
"141 3.1\n",
"68 2.2\n",
"82 2.7\n",
"Name: sepal width (cm), dtype: float64"
]
},
"metadata": {},
"execution_count": 53
}
]
},
{
"cell_type": "code",
"source": [
"lm.predict(X_test)[:10]"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "G47VzGtk5Tzw",
"outputId": "d392bf15-75d0-4855-88b0-f0a08a5bbf76"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"array([2.68882271, 3.67002562, 3.38976912, 2.87491372, 3.0414682 ,\n",
" 3.65914168, 2.72435311, 3.34795138, 2.94223694, 2.68893794])"
]
},
"metadata": {},
"execution_count": 52
}
]
},
{
"cell_type": "code",
"source": [
"X_test.mean()"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "cX70r-wm3er7",
"outputId": "f5468823-62e9-4f20-8dbd-218832cdd0ba"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"sepal length (cm) 5.881579\n",
"petal length (cm) 3.613158\n",
"petal width (cm) 1.155263\n",
"class 0.921053\n",
"dtype: float64"
]
},
"metadata": {},
"execution_count": 54
}
]
},
{
"cell_type": "code",
"source": [
"y_train.mean()"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "aUVr_KHG31bi",
"outputId": "94e43c75-ed73-4996-c96a-04a4e45fe5d2"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"3.0401785714285716"
]
},
"metadata": {},
"execution_count": 55
}
]
},
{
"cell_type": "code",
"source": [
"X_test.iloc[0]"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "aGwZbxlI5MBM",
"outputId": "69564210-972b-4863-ef4d-46147df43c40"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"sepal length (cm) 6.1\n",
"petal length (cm) 4.7\n",
"petal width (cm) 1.2\n",
"class 1.0\n",
"Name: 73, dtype: float64"
]
},
"metadata": {},
"execution_count": 56
}
]
},
{
"cell_type": "code",
"source": [
"X_test.shape"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/"
},
"id": "DgbxkBa6tNSR",
"outputId": "7cd8d7d8-82ad-4d06-f83b-48b4270e18c3"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"(38, 4)"
]
},
"metadata": {},
"execution_count": 62
}
]
},
{
"cell_type": "code",
"source": [
"shap.plots.waterfall(shap_values[9])"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 386
},
"id": "2TTNLtkDH8rt",
"outputId": "327add0e-1174-473d-a0e4-1db789846de8"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 800x350 with 3 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "markdown",
"source": [
"- The plot is dubbed “waterfall” because each step resembles flowing\n",
"water. Water can flow in either direction, just as SHAP values can be\n",
"positive or negative. Positive SHAP values point to the right.\n",
"- The y-axis exhibits the individual features, along with the values for\n",
"the selected data instance.\n",
"- The feature values are ordered by the magnitudes of their SHAP values.\n",
"- The x-axis is on the scale of SHAP values.\n",
"- Each bar signifies the SHAP value for that specific feature value.\n",
"- The x-axis also shows the estimated expected prediction 𝔼(𝑓(𝑋)) and\n",
"the actual prediction of the instance 𝑓(𝑥(𝑖)).\n",
"- The bars start at the bottom from the expected prediction and add up\n",
"to the actual prediction."
],
"metadata": {
"id": "nx0JDAUVQn9O"
}
},
{
"cell_type": "markdown",
"source": [
"- The x-axis represents the SHAP values, while the y-axis shows the\n",
"features, and the color indicates the feature’s value.\n",
"- Each row corresponds to a feature.\n",
"- The feature order is determined by importance, defined as the average\n",
"of absolute SHAP values: $𝐼_𝑗 =\\frac1n\\sum_{i=1}^n\\phi_j^{(i)}$\n",
"- Each dot represents the SHAP value of a feature for a data point,\n",
"resulting in a total of 𝑝 ⋅ 𝑛 dots."
],
"metadata": {
"id": "Dx2ZAkoiRKBI"
}
},
{
"cell_type": "code",
"source": [
"shap.plots.beeswarm(shap_values)"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 329
},
"id": "uy26MdaIHqVk",
"outputId": "0f1e46c0-2c78-4e79-dad4-ee8c25e6d511"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 800x310 with 2 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"shap.plots.bar(shap_values)"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 354
},
"id": "OV2zeVLzIp2X",
"outputId": "070a438b-3b74-4eeb-e139-a9aac680da08"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 800x350 with 1 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"shap.plots.partial_dependence(\n",
" \"sepal length (cm)\", lm.predict, X_test, ice=False,\n",
" model_expected_value=True, feature_expected_value=True\n",
")"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 474
},
"id": "F-MSZ5XUHRbr",
"outputId": "9f0f3a14-9f9b-4acd-a101-bc762b7163b3"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 640x480 with 4 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"shap.plots.scatter(shap_values[:,\"sepal length (cm)\"], color=shap_values)"
],
"metadata": {
"id": "cyPGqs1CgSeO",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 470
},
"outputId": "89218855-0e9a-4750-8091-d4a9f8939906"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 750x500 with 3 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "U-WxMmdxNRB_",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "6906c530-80f6-4b73-f143-2b808a388b4d"
},
"source": [
"print(\"The hyper-parameters for a linear model are:\")\n",
"for param_name in LinearRegression().get_params().keys():\n",
" print(param_name)"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"The hyper-parameters for a linear model are:\n",
"copy_X\n",
"fit_intercept\n",
"n_jobs\n",
"positive\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"%pip install --quiet interpret"
],
"metadata": {
"id": "Z2_s7mrGO3Ro"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"from interpret.glassbox import ExplainableBoostingRegressor\n",
"\n",
"model = ExplainableBoostingRegressor(interactions=0)\n",
"model.fit(X_train, y_train)"
],
"metadata": {
"id": "_rPpM48ZO6K7",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 74
},
"outputId": "6bd9545e-68c9-4b39-c212-54cf6911015a"
},
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"ExplainableBoostingRegressor(interactions=0)"
],
"text/html": [
"<style>#sk-container-id-3 {color: black;background-color: white;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>ExplainableBoostingRegressor(interactions=0)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-15\" type=\"checkbox\" checked><label for=\"sk-estimator-id-15\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">ExplainableBoostingRegressor</label><div class=\"sk-toggleable__content\"><pre>ExplainableBoostingRegressor(interactions=0)</pre></div></div></div></div></div>"
]
},
"metadata": {},
"execution_count": 68
}
]
},
{
"cell_type": "code",
"source": [
"print(\"train error: %0.3f, test error: %0.3f\" %\n",
" (median_absolute_error(y_train, model.predict(X_train)),\n",
" median_absolute_error(y_test, model.predict(X_test))))"
],
"metadata": {
"id": "X4UJo6OQO-qY",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "b5240ddf-d0f1-436d-dc2c-a5b5302d3e4f"
},
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"train error: 0.149, test error: 0.220\n"
]
}
]
},
{
"cell_type": "code",
"source": [
"explainer = shap.Explainer(model.predict, X_train)\n",
"shap_values = explainer(X_test)"
],
"metadata": {
"id": "bLLMyhyBPLdu"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"source": [
"shap.plots.waterfall(shap_values[0])"
],
"metadata": {
"id": "kSZVYLjxPPIq",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 386
},
"outputId": "5c7f55d3-f8a3-4ea1-b028-630f59c2f12b"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 800x350 with 3 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"shap.plots.partial_dependence(\n",
" \"sepal length (cm)\", model.predict, X_test, ice=False,\n",
" model_expected_value=True, feature_expected_value=True\n",
")"
],
"metadata": {
"colab": {
"base_uri": "https://localhost:8080/",
"height": 474
},
"id": "nfrivdtmuLjL",
"outputId": "59bdd3da-e5da-442e-b415-ad56e48e96db"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 640x480 with 4 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"shap.plots.scatter(shap_values[:,\"sepal length (cm)\"])"
],
"metadata": {
"id": "In-eiIZjPPqJ",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 470
},
"outputId": "d830f7a7-a422-4b51-e749-8c6052d28a08"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 600x500 with 2 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"source": [
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"shap.plots.scatter(shap_values[:,\"sepal length (cm)\"], show=False)\n",
"# First get the index of the alcohol feature\n",
"idx = model.explain_global().data()['names'].index('sepal length (cm)')\n",
"# extract the relevant data from the tree-based GAM\n",
"explain_data = model.explain_global().data(idx)\n",
"# the alcohol feature values\n",
"x_data = explain_data[\"names\"]\n",
"# the part of the prediction function for alcohol\n",
"y_data = explain_data[\"scores\"]\n",
"y_data = np.r_[y_data, y_data[np.newaxis, -1]]\n",
"plt.plot(x_data, y_data, color='red')\n",
"plt.show()"
],
"metadata": {
"id": "9b3OIgfjPR-h",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 470
},
"outputId": "13052aae-f8c0-4465-bd3c-f6e684689f5a"
},
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 600x500 with 2 Axes>"
],
"image/png": "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\n"
},
"metadata": {}
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "2xkjKorMNRCT"
},
"source": [
"from sklearn.ensemble import RandomForestRegressor\n",
"\n",
"rf = Pipeline([\n",
" ('preprocess', preprocessing),\n",
" ('regressor', RandomForestRegressor(n_estimators=100, n_jobs=-1, random_state=42))\n",
"])"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"scrolled": false,
"id": "pAckw-G4NRCf",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 192
},
"outputId": "89a37b81-709d-4c1c-9f78-0a05de5131c9"
},
"source": [
"rf.fit(X_train, y_train)"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"Pipeline(steps=[('preprocess',\n",
" ColumnTransformer(transformers=[('cat',\n",
" Pipeline(steps=[('onehot',\n",
" OneHotEncoder())]),\n",
" ['class']),\n",
" ('num',\n",
" Pipeline(steps=[('scaler',\n",
" StandardScaler())]),\n",
" ['sepal length (cm)',\n",
" 'petal length (cm)',\n",
" 'petal width (cm)'])])),\n",
" ('regressor',\n",
" RandomForestRegressor(n_jobs=-1, random_state=42))])"
],
"text/html": [
"<style>#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 {color: black;background-color: white;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 pre{padding: 0;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-toggleable {background-color: white;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-estimator:hover {background-color: #d4ebff;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-item {z-index: 1;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-parallel-item:only-child::after {width: 0;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-6e6a82bc-5b2a-459b-87c1-a4ea86d8a213\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;,\n",
" &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])),\n",
" (&#x27;regressor&#x27;,\n",
" RandomForestRegressor(n_jobs=-1, random_state=42))])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"c633e83b-a0b1-4563-8ca9-0b9db2cf05aa\" type=\"checkbox\" ><label for=\"c633e83b-a0b1-4563-8ca9-0b9db2cf05aa\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;,\n",
" &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])),\n",
" (&#x27;regressor&#x27;,\n",
" RandomForestRegressor(n_jobs=-1, random_state=42))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4ae69dd9-1ed8-41ca-b2e8-418b94941707\" type=\"checkbox\" ><label for=\"4ae69dd9-1ed8-41ca-b2e8-418b94941707\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">preprocess: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;, OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a8318421-2b62-411d-a869-9883341f5277\" type=\"checkbox\" ><label for=\"a8318421-2b62-411d-a869-9883341f5277\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">cat</label><div class=\"sk-toggleable__content\"><pre>[&#x27;class&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"8438c20c-e881-4905-a358-e34e71257049\" type=\"checkbox\" ><label for=\"8438c20c-e881-4905-a358-e34e71257049\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">OneHotEncoder</label><div class=\"sk-toggleable__content\"><pre>OneHotEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"dd0a9c83-cce7-4a85-93e4-e6e1b0403009\" type=\"checkbox\" ><label for=\"dd0a9c83-cce7-4a85-93e4-e6e1b0403009\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num</label><div class=\"sk-toggleable__content\"><pre>[&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;, &#x27;petal width (cm)&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"09a1dbce-f69b-4922-868f-261ab19aa8c8\" type=\"checkbox\" ><label for=\"09a1dbce-f69b-4922-868f-261ab19aa8c8\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"2b0db141-667a-4c1a-8ea9-18f72d14e5b3\" type=\"checkbox\" ><label for=\"2b0db141-667a-4c1a-8ea9-18f72d14e5b3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(n_jobs=-1, random_state=42)</pre></div></div></div></div></div></div></div>"
]
},
"metadata": {},
"execution_count": 52
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "6onlZOMeNRCj",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "ee3d5a1c-7797-428f-a76c-74e9b909b5e6"
},
"source": [
"from sklearn.metrics import median_absolute_error\n",
"\n",
"print(\"train error: %0.3f, test error: %0.3f\" %\n",
" (median_absolute_error(y_train, rf.predict(X_train)),\n",
" median_absolute_error(y_test, rf.predict(X_test))))"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"train error: 0.060, test error: 0.275\n"
]
}
]
},
{
"cell_type": "code",
"metadata": {
"scrolled": false,
"id": "xLGOVM3TNRCs",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 391
},
"outputId": "d25268a9-2662-46aa-f44c-19d6ba541e82"
},
"source": [
"scatter_predictions(rf.predict(X_test), y_test)"
],
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 432x432 with 1 Axes>"
],
"image/png": "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\n"
},
"metadata": {
"needs_background": "light"
}
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "EqZqJRZmNRC1",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "63b466ef-65e5-46ee-b1c9-e6cf4d41d1a9"
},
"source": [
"print(\"The hyper-parameters for a random forest model are:\")\n",
"for param_name in rf.get_params().keys():\n",
" print(param_name)"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"The hyper-parameters for a random forest model are:\n",
"memory\n",
"steps\n",
"verbose\n",
"preprocess\n",
"regressor\n",
"preprocess__n_jobs\n",
"preprocess__remainder\n",
"preprocess__sparse_threshold\n",
"preprocess__transformer_weights\n",
"preprocess__transformers\n",
"preprocess__verbose\n",
"preprocess__verbose_feature_names_out\n",
"preprocess__cat\n",
"preprocess__num\n",
"preprocess__cat__memory\n",
"preprocess__cat__steps\n",
"preprocess__cat__verbose\n",
"preprocess__cat__onehot\n",
"preprocess__cat__onehot__categories\n",
"preprocess__cat__onehot__drop\n",
"preprocess__cat__onehot__dtype\n",
"preprocess__cat__onehot__handle_unknown\n",
"preprocess__cat__onehot__sparse\n",
"preprocess__num__memory\n",
"preprocess__num__steps\n",
"preprocess__num__verbose\n",
"preprocess__num__scaler\n",
"preprocess__num__scaler__copy\n",
"preprocess__num__scaler__with_mean\n",
"preprocess__num__scaler__with_std\n",
"regressor__bootstrap\n",
"regressor__ccp_alpha\n",
"regressor__criterion\n",
"regressor__max_depth\n",
"regressor__max_features\n",
"regressor__max_leaf_nodes\n",
"regressor__max_samples\n",
"regressor__min_impurity_decrease\n",
"regressor__min_samples_leaf\n",
"regressor__min_samples_split\n",
"regressor__min_weight_fraction_leaf\n",
"regressor__n_estimators\n",
"regressor__n_jobs\n",
"regressor__oob_score\n",
"regressor__random_state\n",
"regressor__verbose\n",
"regressor__warm_start\n"
]
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "Xeklg8zNNRDA",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 193
},
"outputId": "8f27e800-8030-4525-8553-5e8aae37c8c8"
},
"source": [
"from sklearn.model_selection import GridSearchCV\n",
"\n",
"param_grid = {\n",
" 'regressor__max_features': (2, 3, 4),\n",
" 'regressor__max_depth': (2, 3, 5),\n",
" 'regressor__min_samples_leaf': (1, 3, 5),\n",
"}\n",
"\n",
"model_grid_search = GridSearchCV(rf, param_grid=param_grid,\n",
" n_jobs=-1, cv=3)\n",
"model_grid_search.fit(X_train, y_train)"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"GridSearchCV(cv=3,\n",
" estimator=Pipeline(steps=[('preprocess',\n",
" ColumnTransformer(transformers=[('cat',\n",
" Pipeline(steps=[('onehot',\n",
" OneHotEncoder())]),\n",
" ['class']),\n",
" ('num',\n",
" Pipeline(steps=[('scaler',\n",
" StandardScaler())]),\n",
" ['sepal '\n",
" 'length '\n",
" '(cm)',\n",
" 'petal '\n",
" 'length '\n",
" '(cm)',\n",
" 'petal '\n",
" 'width '\n",
" '(cm)'])])),\n",
" ('regressor',\n",
" RandomForestRegressor(n_jobs=-1,\n",
" random_state=42))]),\n",
" n_jobs=-1,\n",
" param_grid={'regressor__max_depth': (2, 3, 5),\n",
" 'regressor__max_features': (2, 3, 4),\n",
" 'regressor__min_samples_leaf': (1, 3, 5)})"
],
"text/html": [
"<style>#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d {color: black;background-color: white;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d pre{padding: 0;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-toggleable {background-color: white;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-estimator:hover {background-color: #d4ebff;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-item {z-index: 1;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-parallel-item:only-child::after {width: 0;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-dccfe0c3-2ebb-4508-a6b6-a1741249642d\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=3,\n",
" estimator=Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal &#x27;\n",
" &#x27;length &#x27;\n",
" &#x27;(cm)&#x27;,\n",
" &#x27;petal &#x27;\n",
" &#x27;length &#x27;\n",
" &#x27;(cm)&#x27;,\n",
" &#x27;petal &#x27;\n",
" &#x27;width &#x27;\n",
" &#x27;(cm)&#x27;])])),\n",
" (&#x27;regressor&#x27;,\n",
" RandomForestRegressor(n_jobs=-1,\n",
" random_state=42))]),\n",
" n_jobs=-1,\n",
" param_grid={&#x27;regressor__max_depth&#x27;: (2, 3, 5),\n",
" &#x27;regressor__max_features&#x27;: (2, 3, 4),\n",
" &#x27;regressor__min_samples_leaf&#x27;: (1, 3, 5)})</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"6c075c67-d6dc-40c4-8bf7-de8a9d2c220f\" type=\"checkbox\" ><label for=\"6c075c67-d6dc-40c4-8bf7-de8a9d2c220f\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=3,\n",
" estimator=Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal &#x27;\n",
" &#x27;length &#x27;\n",
" &#x27;(cm)&#x27;,\n",
" &#x27;petal &#x27;\n",
" &#x27;length &#x27;\n",
" &#x27;(cm)&#x27;,\n",
" &#x27;petal &#x27;\n",
" &#x27;width &#x27;\n",
" &#x27;(cm)&#x27;])])),\n",
" (&#x27;regressor&#x27;,\n",
" RandomForestRegressor(n_jobs=-1,\n",
" random_state=42))]),\n",
" n_jobs=-1,\n",
" param_grid={&#x27;regressor__max_depth&#x27;: (2, 3, 5),\n",
" &#x27;regressor__max_features&#x27;: (2, 3, 4),\n",
" &#x27;regressor__min_samples_leaf&#x27;: (1, 3, 5)})</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"816267fb-1988-467c-afb4-f45ae0b73c1b\" type=\"checkbox\" ><label for=\"816267fb-1988-467c-afb4-f45ae0b73c1b\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">preprocess: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;, OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"54d01f43-9062-456c-a7a9-02b45b6fad6d\" type=\"checkbox\" ><label for=\"54d01f43-9062-456c-a7a9-02b45b6fad6d\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">cat</label><div class=\"sk-toggleable__content\"><pre>[&#x27;class&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"d21a9f13-e085-4de0-ba2c-b6eb1b2f6012\" type=\"checkbox\" ><label for=\"d21a9f13-e085-4de0-ba2c-b6eb1b2f6012\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">OneHotEncoder</label><div class=\"sk-toggleable__content\"><pre>OneHotEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"92f07b94-ae84-4f3d-8b98-b02fd3e76dcc\" type=\"checkbox\" ><label for=\"92f07b94-ae84-4f3d-8b98-b02fd3e76dcc\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num</label><div class=\"sk-toggleable__content\"><pre>[&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;, &#x27;petal width (cm)&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"f3be7a87-9255-43b8-99e8-ba325b6b5f98\" type=\"checkbox\" ><label for=\"f3be7a87-9255-43b8-99e8-ba325b6b5f98\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"59b1b5d8-d034-4453-b2c0-d01db98b9f88\" type=\"checkbox\" ><label for=\"59b1b5d8-d034-4453-b2c0-d01db98b9f88\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(n_jobs=-1, random_state=42)</pre></div></div></div></div></div></div></div></div></div></div></div></div>"
]
},
"metadata": {},
"execution_count": 56
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "T_pgWdWVNRDG",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "3c8eff42-1db3-49e5-f438-3a0d51e4369f"
},
"source": [
"print(\"train error: %0.3f, test error: %0.3f\" %\n",
" (median_absolute_error(y_train, model_grid_search.predict(X_train)),\n",
" median_absolute_error(y_test, model_grid_search.predict(X_test))))"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"train error: 0.120, test error: 0.201\n"
]
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "Fmxbr-noNRDL",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "085498bf-40b1-49a9-8249-27c759c90e7b"
},
"source": [
"print(f\"The best set of hyperparameters is: \"\n",
" f\"{model_grid_search.best_params_}\")"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"The best set of hyperparameters is: {'regressor__max_depth': 5, 'regressor__max_features': 3, 'regressor__min_samples_leaf': 3}\n"
]
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "28wkVwv2NRDU"
},
"source": [
"rf_best = Pipeline([\n",
" ('preprocess', preprocessing),\n",
" ('regressor', RandomForestRegressor(\n",
" n_estimators=100, max_depth=5, max_features=3, min_samples_leaf=3, n_jobs=-1, random_state=42))\n",
"])"
],
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {
"id": "uvSX8FmHNRDZ",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 192
},
"outputId": "13ded2fb-db7f-45cc-fbdc-d305a2a587a6"
},
"source": [
"rf_best.fit(X_train, y_train)"
],
"execution_count": null,
"outputs": [
{
"output_type": "execute_result",
"data": {
"text/plain": [
"Pipeline(steps=[('preprocess',\n",
" ColumnTransformer(transformers=[('cat',\n",
" Pipeline(steps=[('onehot',\n",
" OneHotEncoder())]),\n",
" ['class']),\n",
" ('num',\n",
" Pipeline(steps=[('scaler',\n",
" StandardScaler())]),\n",
" ['sepal length (cm)',\n",
" 'petal length (cm)',\n",
" 'petal width (cm)'])])),\n",
" ('regressor',\n",
" RandomForestRegressor(max_depth=5, max_features=3,\n",
" min_samples_leaf=3, n_jobs=-1,\n",
" random_state=42))])"
],
"text/html": [
"<style>#sk-28126540-c632-420e-a8f5-7a3690bd7822 {color: black;background-color: white;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 pre{padding: 0;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-toggleable {background-color: white;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-estimator:hover {background-color: #d4ebff;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-item {z-index: 1;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-parallel-item:only-child::after {width: 0;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-28126540-c632-420e-a8f5-7a3690bd7822 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-28126540-c632-420e-a8f5-7a3690bd7822\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;,\n",
" &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])),\n",
" (&#x27;regressor&#x27;,\n",
" RandomForestRegressor(max_depth=5, max_features=3,\n",
" min_samples_leaf=3, n_jobs=-1,\n",
" random_state=42))])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"47a0de4f-104f-43c2-b2e3-2fc40484c8cf\" type=\"checkbox\" ><label for=\"47a0de4f-104f-43c2-b2e3-2fc40484c8cf\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;preprocess&#x27;,\n",
" ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;,\n",
" OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;,\n",
" StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;,\n",
" &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])),\n",
" (&#x27;regressor&#x27;,\n",
" RandomForestRegressor(max_depth=5, max_features=3,\n",
" min_samples_leaf=3, n_jobs=-1,\n",
" random_state=42))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"68debd82-1f1a-4c5e-b30d-b0697be4f2ed\" type=\"checkbox\" ><label for=\"68debd82-1f1a-4c5e-b30d-b0697be4f2ed\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">preprocess: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[(&#x27;cat&#x27;,\n",
" Pipeline(steps=[(&#x27;onehot&#x27;, OneHotEncoder())]),\n",
" [&#x27;class&#x27;]),\n",
" (&#x27;num&#x27;,\n",
" Pipeline(steps=[(&#x27;scaler&#x27;, StandardScaler())]),\n",
" [&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;,\n",
" &#x27;petal width (cm)&#x27;])])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"73cfdbb8-8855-4068-9878-1e164f3cfc4e\" type=\"checkbox\" ><label for=\"73cfdbb8-8855-4068-9878-1e164f3cfc4e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">cat</label><div class=\"sk-toggleable__content\"><pre>[&#x27;class&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"4ce30119-fcc1-409f-b3dc-2478709da57e\" type=\"checkbox\" ><label for=\"4ce30119-fcc1-409f-b3dc-2478709da57e\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">OneHotEncoder</label><div class=\"sk-toggleable__content\"><pre>OneHotEncoder()</pre></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"e74a3cc8-c602-4c9e-8126-ede33edfd876\" type=\"checkbox\" ><label for=\"e74a3cc8-c602-4c9e-8126-ede33edfd876\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num</label><div class=\"sk-toggleable__content\"><pre>[&#x27;sepal length (cm)&#x27;, &#x27;petal length (cm)&#x27;, &#x27;petal width (cm)&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"6a932f69-dc77-49cb-8cdc-8322919f761b\" type=\"checkbox\" ><label for=\"6a932f69-dc77-49cb-8cdc-8322919f761b\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"a8fca666-57a5-464a-ab01-ae535f2719b5\" type=\"checkbox\" ><label for=\"a8fca666-57a5-464a-ab01-ae535f2719b5\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor(max_depth=5, max_features=3, min_samples_leaf=3,\n",
" n_jobs=-1, random_state=42)</pre></div></div></div></div></div></div></div>"
]
},
"metadata": {},
"execution_count": 60
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "5Xryqy6lNRDh",
"colab": {
"base_uri": "https://localhost:8080/"
},
"outputId": "75870e34-e3dd-4dd1-f57e-ceea9c4fd52d"
},
"source": [
"print(\"train error: %0.3f, test error: %0.3f\" %\n",
" (median_absolute_error(y_train, rf_best.predict(X_train)),\n",
" median_absolute_error(y_test, rf_best.predict(X_test))))"
],
"execution_count": null,
"outputs": [
{
"output_type": "stream",
"name": "stdout",
"text": [
"train error: 0.120, test error: 0.201\n"
]
}
]
},
{
"cell_type": "code",
"metadata": {
"id": "qNNb1Y20NRDp",
"colab": {
"base_uri": "https://localhost:8080/",
"height": 441
},
"outputId": "679e4a78-d7d5-4d2b-8636-23e306b180e0"
},
"source": [
"boxplot_pi(rf_best, X_test, y_test)"
],
"execution_count": null,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": [
"<Figure size 576x432 with 1 Axes>"
],
"image/png": "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\n"
},
"metadata": {
"needs_background": "light"
}
}
]
},
{
"cell_type": "code",
"source": [],
"metadata": {
"id": "NOux8KgNaFvu"
},
"execution_count": null,
"outputs": []
}
]
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment