dev/notebooks/auto_examples/classification/plot_digits_classification.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "\n# Recognizing hand-written digits\n\nThis example shows how scikit-learn can be used to recognize images of\nhand-written digits, from 0-9.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "# Author: Gael Varoquaux <gael dot varoquaux at normalesup dot org>\n# License: BSD 3 clause\n\n# Standard scientific Python imports\nimport matplotlib.pyplot as plt\n\n# Import datasets, classifiers and performance metrics\nfrom sklearn import datasets, metrics, svm\nfrom sklearn.model_selection import train_test_split"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Digits dataset\n\nThe digits dataset consists of 8x8\npixel images of digits. The ``images`` attribute of the dataset stores\n8x8 arrays of grayscale values for each image. We will use these arrays to\nvisualize the first 4 images. The ``target`` attribute of the dataset stores\nthe digit each image represents and this is included in the title of the 4\nplots below.\n\nNote: if we were working from image files (e.g., 'png' files), we would load\nthem using :func:`matplotlib.pyplot.imread`.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "digits = datasets.load_digits()\n\n_, axes = plt.subplots(nrows=1, ncols=4, figsize=(10, 3))\nfor ax, image, label in zip(axes, digits.images, digits.target):\n    ax.set_axis_off()\n    ax.imshow(image, cmap=plt.cm.gray_r, interpolation=\"nearest\")\n    ax.set_title(\"Training: %i\" % label)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Classification\n\nTo apply a classifier on this data, we need to flatten the images, turning\neach 2-D array of grayscale values from shape ``(8, 8)`` into shape\n``(64,)``. Subsequently, the entire dataset will be of shape\n``(n_samples, n_features)``, where ``n_samples`` is the number of images and\n``n_features`` is the total number of pixels in each image.\n\nWe can then split the data into train and test subsets and fit a support\nvector classifier on the train samples. The fitted classifier can\nsubsequently be used to predict the value of the digit for the samples\nin the test subset.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "# flatten the images\nn_samples = len(digits.images)\ndata = digits.images.reshape((n_samples, -1))\n\n# Create a classifier: a support vector classifier\nclf = svm.SVC(gamma=0.001)\n\n# Split data into 50% train and 50% test subsets\nX_train, X_test, y_train, y_test = train_test_split(\n    data, digits.target, test_size=0.5, shuffle=False\n)\n\n# Learn the digits on the train subset\nclf.fit(X_train, y_train)\n\n# Predict the value of the digit on the test subset\npredicted = clf.predict(X_test)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Below we visualize the first 4 test samples and show their predicted\ndigit value in the title.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "_, axes = plt.subplots(nrows=1, ncols=4, figsize=(10, 3))\nfor ax, image, prediction in zip(axes, X_test, predicted):\n    ax.set_axis_off()\n    image = image.reshape(8, 8)\n    ax.imshow(image, cmap=plt.cm.gray_r, interpolation=\"nearest\")\n    ax.set_title(f\"Prediction: {prediction}\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        ":func:`~sklearn.metrics.classification_report` builds a text report showing\nthe main classification metrics.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "print(\n    f\"Classification report for classifier {clf}:\\n\"\n    f\"{metrics.classification_report(y_test, predicted)}\\n\"\n)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "We can also plot a `confusion matrix <confusion_matrix>` of the\ntrue digit values and the predicted digit values.\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "disp = metrics.ConfusionMatrixDisplay.from_predictions(y_test, predicted)\ndisp.figure_.suptitle(\"Confusion Matrix\")\nprint(f\"Confusion matrix:\\n{disp.confusion_matrix}\")\n\nplt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "If the results from evaluating a classifier are stored in the form of a\n`confusion matrix <confusion_matrix>` and not in terms of `y_true` and\n`y_pred`, one can still build a :func:`~sklearn.metrics.classification_report`\nas follows:\n\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "# The ground truth and predicted lists\ny_true = []\ny_pred = []\ncm = disp.confusion_matrix\n\n# For each cell in the confusion matrix, add the corresponding ground truths\n# and predictions to the lists\nfor gt in range(len(cm)):\n    for pred in range(len(cm)):\n        y_true += [gt] * cm[gt][pred]\n        y_pred += [pred] * cm[gt][pred]\n\nprint(\n    \"Classification report rebuilt from confusion matrix:\\n\"\n    f\"{metrics.classification_report(y_true, y_pred)}\\n\"\n)"
      ]
    }
  ],
  "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": "3.9.16"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}