{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"%matplotlib inline"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"\n# Plotting Learning Curves\n\nIn the first column, first row the learning curve of a naive Bayes classifier\nis shown for the digits dataset. Note that the training score and the\ncross-validation score are both not very good at the end. However, the shape\nof the curve can be found in more complex datasets very often: the training\nscore is very high at the beginning and decreases and the cross-validation\nscore is very low at the beginning and increases. In the second column, first\nrow we see the learning curve of an SVM with RBF kernel. We can see clearly\nthat the training score is still around the maximum and the validation score\ncould be increased with more training samples. The plots in the second row\nshow the times required by the models to train with various sizes of training\ndataset. The plots in the third row show how much time was required to train\nthe models for each training sizes.\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"print(__doc__)\n\nimport numpy as np\nimport matplotlib.pyplot as plt\nfrom sklearn.naive_bayes import GaussianNB\nfrom sklearn.svm import SVC\nfrom sklearn.datasets import load_digits\nfrom sklearn.model_selection import learning_curve\nfrom sklearn.model_selection import ShuffleSplit\n\n\ndef plot_learning_curve(estimator, title, X, y, axes=None, ylim=None, cv=None,\n n_jobs=None, train_sizes=np.linspace(.1, 1.0, 5)):\n \"\"\"\n Generate 3 plots: the test and training learning curve, the training\n samples vs fit times curve, the fit times vs score curve.\n\n Parameters\n ----------\n estimator : object type that implements the \"fit\" and \"predict\" methods\n An object of that type which is cloned for each validation.\n\n title : string\n Title for the chart.\n\n X : array-like, shape (n_samples, n_features)\n Training vector, where n_samples is the number of samples and\n n_features is the number of features.\n\n y : array-like, shape (n_samples) or (n_samples, n_features), optional\n Target relative to X for classification or regression;\n None for unsupervised learning.\n\n axes : array of 3 axes, optional (default=None)\n Axes to use for plotting the curves.\n\n ylim : tuple, shape (ymin, ymax), optional\n Defines minimum and maximum yvalues plotted.\n\n cv : int, cross-validation generator or an iterable, optional\n Determines the cross-validation splitting strategy.\n Possible inputs for cv are:\n\n - None, to use the default 5-fold cross-validation,\n - integer, to specify the number of folds.\n - :term:`CV splitter`,\n - An iterable yielding (train, test) splits as arrays of indices.\n\n For integer/None inputs, if ``y`` is binary or multiclass,\n :class:`StratifiedKFold` used. If the estimator is not a classifier\n or if ``y`` is neither binary nor multiclass, :class:`KFold` is used.\n\n Refer :ref:`User Guide ` for the various\n cross-validators that can be used here.\n\n n_jobs : int or None, optional (default=None)\n Number of jobs to run in parallel.\n ``None`` means 1 unless in a :obj:`joblib.parallel_backend` context.\n ``-1`` means using all processors. See :term:`Glossary `\n for more details.\n\n train_sizes : array-like, shape (n_ticks,), dtype float or int\n Relative or absolute numbers of training examples that will be used to\n generate the learning curve. If the dtype is float, it is regarded as a\n fraction of the maximum size of the training set (that is determined\n by the selected validation method), i.e. it has to be within (0, 1].\n Otherwise it is interpreted as absolute sizes of the training sets.\n Note that for classification the number of samples usually have to\n be big enough to contain at least one sample from each class.\n (default: np.linspace(0.1, 1.0, 5))\n \"\"\"\n if axes is None:\n _, axes = plt.subplots(1, 3, figsize=(20, 5))\n\n axes[0].set_title(title)\n if ylim is not None:\n axes[0].set_ylim(*ylim)\n axes[0].set_xlabel(\"Training examples\")\n axes[0].set_ylabel(\"Score\")\n\n train_sizes, train_scores, test_scores, fit_times, _ = \\\n learning_curve(estimator, X, y, cv=cv, n_jobs=n_jobs,\n train_sizes=train_sizes,\n return_times=True)\n train_scores_mean = np.mean(train_scores, axis=1)\n train_scores_std = np.std(train_scores, axis=1)\n test_scores_mean = np.mean(test_scores, axis=1)\n test_scores_std = np.std(test_scores, axis=1)\n fit_times_mean = np.mean(fit_times, axis=1)\n fit_times_std = np.std(fit_times, axis=1)\n\n # Plot learning curve\n axes[0].grid()\n axes[0].fill_between(train_sizes, train_scores_mean - train_scores_std,\n train_scores_mean + train_scores_std, alpha=0.1,\n color=\"r\")\n axes[0].fill_between(train_sizes, test_scores_mean - test_scores_std,\n test_scores_mean + test_scores_std, alpha=0.1,\n color=\"g\")\n axes[0].plot(train_sizes, train_scores_mean, 'o-', color=\"r\",\n label=\"Training score\")\n axes[0].plot(train_sizes, test_scores_mean, 'o-', color=\"g\",\n label=\"Cross-validation score\")\n axes[0].legend(loc=\"best\")\n\n # Plot n_samples vs fit_times\n axes[1].grid()\n axes[1].plot(train_sizes, fit_times_mean, 'o-')\n axes[1].fill_between(train_sizes, fit_times_mean - fit_times_std,\n fit_times_mean + fit_times_std, alpha=0.1)\n axes[1].set_xlabel(\"Training examples\")\n axes[1].set_ylabel(\"fit_times\")\n axes[1].set_title(\"Scalability of the model\")\n\n # Plot fit_time vs score\n axes[2].grid()\n axes[2].plot(fit_times_mean, test_scores_mean, 'o-')\n axes[2].fill_between(fit_times_mean, test_scores_mean - test_scores_std,\n test_scores_mean + test_scores_std, alpha=0.1)\n axes[2].set_xlabel(\"fit_times\")\n axes[2].set_ylabel(\"Score\")\n axes[2].set_title(\"Performance of the model\")\n\n return plt\n\n\nfig, axes = plt.subplots(3, 2, figsize=(10, 15))\n\nX, y = load_digits(return_X_y=True)\n\ntitle = \"Learning Curves (Naive Bayes)\"\n# Cross validation with 100 iterations to get smoother mean test and train\n# score curves, each time with 20% data randomly selected as a validation set.\ncv = ShuffleSplit(n_splits=100, test_size=0.2, random_state=0)\n\nestimator = GaussianNB()\nplot_learning_curve(estimator, title, X, y, axes=axes[:, 0], ylim=(0.7, 1.01),\n cv=cv, n_jobs=4)\n\ntitle = r\"Learning Curves (SVM, RBF kernel, $\\gamma=0.001$)\"\n# SVC is more expensive so we do a lower number of CV iterations:\ncv = ShuffleSplit(n_splits=10, test_size=0.2, random_state=0)\nestimator = SVC(gamma=0.001)\nplot_learning_curve(estimator, title, X, y, axes=axes[:, 1], ylim=(0.7, 1.01),\n cv=cv, n_jobs=4)\n\nplt.show()"
]
}
],
"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": "3.8.5"
}
},
"nbformat": 4,
"nbformat_minor": 0
}