{ "cells": [ { "attachments": {}, "cell_type": "markdown", "metadata": {}, "source": [ "# Machine learning classification\n", "\n", "Below is an example for using the `ml.classificator` class to classify spectral data.\n", "It wraps scikit-learn algorithms in a similar way compared to the `rp.mlregression` class.\n", " \n", "## Fake data generation\n", "\n", "First we generate 5 fake signals that can be measured modulo some noise. More complex examples can be generated if wanted." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Number of samples = 500\n", "Number of labels = 500\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "%matplotlib inline\n", "import numpy as np\n", "np.random.seed(42) # fixing the seed\n", "import matplotlib.pyplot as plt\n", "import rampy as rp\n", "from sklearn.model_selection import train_test_split\n", "\n", "# The X axis\n", "x = np.arange(0, 1000, 1.0)\n", "\n", "# The perfect 5 signals\n", "spectra_1 = rp.gaussian(x, 10.0, 300., 25.) + rp.lorentzian(x, 15., 650., 50.)\n", "spectra_2 = rp.gaussian(x, 20.0, 350., 25.) + rp.gaussian(x, 25.0, 380., 20.) + rp.lorentzian(x, 15., 630., 50.)\n", "spectra_3 = rp.gaussian(x, 10.0, 500., 50.) + rp.lorentzian(x, 15.0, 520., 10.) + rp.gaussian(x, 25., 530., 3.)\n", "spectra_4 = rp.gaussian(x, 10.0, 100., 5.) + rp.lorentzian(x, 30.0, 110., 3.) + rp.gaussian(x, 5., 900., 10.)\n", "spectra_5 = rp.gaussian(x, 10.0, 600., 200.)\n", "\n", "# the number of observations of each signal\n", "number_of_spectra = 100\n", "\n", "# generating a dataset (will be shuffled later during the train-test split)\n", "dataset = np.hstack((np.ones((len(x),number_of_spectra))*spectra_1.reshape(-1,1),\n", " np.ones((len(x),number_of_spectra))*spectra_2.reshape(-1,1),\n", " np.ones((len(x),number_of_spectra))*spectra_3.reshape(-1,1),\n", " np.ones((len(x),number_of_spectra))*spectra_4.reshape(-1,1),\n", " np.ones((len(x),number_of_spectra))*spectra_5.reshape(-1,1)\n", " )).T\n", "\n", "# add noise\n", "noise_scale = 2.0\n", "dataset = dataset + np.random.normal(scale=noise_scale,size=(len(dataset),len(x)))\n", "\n", "# create numeric labels\n", "labels = np.vstack((np.tile(np.array([1]).reshape(-1,1),number_of_spectra),\n", " np.tile(np.array([2]).reshape(-1,1),number_of_spectra),\n", " np.tile(np.array([3]).reshape(-1,1),number_of_spectra),\n", " np.tile(np.array([4]).reshape(-1,1),number_of_spectra),\n", " np.tile(np.array([5]).reshape(-1,1),number_of_spectra),\n", " )).reshape(-1,1)\n", "\n", "print('Number of samples = {}'.format(dataset.shape[0]))\n", "print('Number of labels = {}'.format(labels.shape[0]))\n", "\n", "# Do figure\n", "plt.figure(figsize=(8,4))\n", "\n", "plt.subplot(1,2,1)\n", "plt.title('(a) The 5 \"perfect\" signals')\n", "plt.plot(x, spectra_1, label='signal 1')\n", "plt.plot(x, spectra_2, label='signal 2')\n", "plt.plot(x, spectra_3, label='signal 3')\n", "plt.plot(x, spectra_4, label='signal 4')\n", "plt.plot(x, spectra_5, label='signal 5')\n", "plt.xlabel('X axis')\n", "plt.ylabel('Ideal signals')\n", "plt.legend()\n", "\n", "plt.subplot(1,2,2)\n", "plt.title('(b) Noisy observations')\n", "plt.plot(x, dataset[0*number_of_spectra:1*number_of_spectra,:].T, color=\"C0\",alpha=0.1)\n", "plt.plot(x, dataset[1*number_of_spectra:2*number_of_spectra,:].T, color=\"C1\",alpha=0.1)\n", "plt.plot(x, dataset[2*number_of_spectra:3*number_of_spectra,:].T, color=\"C2\",alpha=0.1)\n", "plt.plot(x, dataset[3*number_of_spectra:4*number_of_spectra,:].T, color=\"C3\",alpha=0.1)\n", "plt.plot(x, dataset[4*number_of_spectra:5*number_of_spectra,:].T, color=\"C4\",alpha=0.1)\n", "\n", "plt.xlabel('X axis')\n", "plt.ylabel('Noisy observed signals (1000 times each)')\n", "\n", "plt.tight_layout()" ] }, { "attachments": {}, "cell_type": "markdown", "metadata": {}, "source": [ "## Machine Learning example of treatment\n", "\n", "Below a quick example of how you will do a classification of the signals using scikit-learn functions." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "- Classifier:Nearest Neighbors is scoring 1.0.\n", "\n", "- Classifier:SVC is scoring 1.0.\n", "\n", "- Classifier:Gaussian Process is scoring 0.19393939393939394.\n", "\n", "- Classifier:Decision Tree is scoring 0.9757575757575757.\n", "\n", "- Classifier:Random Forest is scoring 1.0.\n", "\n", "- Classifier:Neural Net is scoring 1.0.\n", "\n", "- Classifier:AdaBoost is scoring 0.9818181818181818.\n", "\n", "- Classifier:Naive Bayes is scoring 1.0.\n", "\n", "- Classifier:QDA is scoring 1.0.\n" ] } ], "source": [ "from sklearn.neural_network import MLPClassifier\n", "from sklearn.neighbors import KNeighborsClassifier\n", "from sklearn.svm import SVC\n", "from sklearn.gaussian_process import GaussianProcessClassifier\n", "from sklearn.gaussian_process.kernels import RBF\n", "from sklearn.tree import DecisionTreeClassifier\n", "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier\n", "from sklearn.naive_bayes import GaussianNB\n", "from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis\n", "\n", "X = dataset\n", "y = labels\n", "\n", "# shufling\n", "from sklearn.utils import shuffle\n", "X, y = shuffle(X, y, random_state=42)\n", "\n", "#\n", "# TRain/Test split\n", "#\n", "\n", "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)\n", "\n", "#\n", "# Initiate classifiers\n", "#\n", "\n", "# The classifiers with default parameters\n", "classifiers = [\n", " KNeighborsClassifier(),\n", " SVC(),\n", " GaussianProcessClassifier(),\n", " DecisionTreeClassifier(),\n", " RandomForestClassifier(),\n", " MLPClassifier(),\n", " AdaBoostClassifier(),\n", " GaussianNB(), QuadraticDiscriminantAnalysis(solver='eigen',shrinkage='auto')]\n", "\n", "# Their names\n", "names = [\"Nearest Neighbors\", \"SVC\", \"Gaussian Process\",\n", " \"Decision Tree\", \"Random Forest\", \"Neural Net\", \"AdaBoost\",\n", " \"Naive Bayes\", \"QDA\"]\n", "\n", "# iterate over classifiers\n", "for name, clf in zip(names, classifiers):\n", " clf.fit(X_train, y_train.ravel())\n", " score = clf.score(X_test, y_test.ravel())\n", " print('\\n- Classifier:'+name+' is scoring '+str(score)+'.')" ] }, { "attachments": {}, "cell_type": "markdown", "metadata": {}, "source": [ "## ML classification with rampy\n", "\n", "Now we use the rampy.mlclassificator to do the exact same thing as above, but in a more concise manner!" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "- Classifier:Nearest Neighbors is scoring (test:) 1.0.\n", "\n", "- Classifier:SVC is scoring (test:) 1.0.\n", "\n", "- Classifier:Gaussian Process is scoring (test:) 0.19393939393939394.\n", "\n", "- Classifier:Decision Tree is scoring (test:) 0.9757575757575757.\n", "\n", "- Classifier:Random Forest is scoring (test:) 1.0.\n", "\n", "- Classifier:Neural Net is scoring (test:) 1.0.\n", "\n", "- Classifier:AdaBoost is scoring (test:) 0.9818181818181818.\n", "\n", "- Classifier:Naive Bayes is scoring (test:) 1.0.\n", "\n", "- Classifier:QDA is scoring (test:) 1.0.\n" ] } ], "source": [ "# initiate model\n", "MLC = rp.mlclassificator(X_train, y_train, X_test = X_test, y_test=y_test, scaling=False)\n", "\n", "# iterate over classifiers\n", "for name in names:\n", " MLC.algorithm = name\n", " if MLC.algorithm == \"QDA\":\n", " MLC.params_ = {'solver':'eigen','shrinkage':'auto'}\n", " MLC.fit()\n", " score = MLC.model.score(MLC.X_test, MLC.y_test)\n", " print('\\n- Classifier:'+name+' is scoring (test:) '+str(score)+'.')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We obtain the same results, great! Now using mlclassificator, you can do automatic train/test split and scaling. Let's see if this improves the results from the Decision Tree classifier:" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "- Classifier SVC is scoring (test:) 1.0.\n" ] } ], "source": [ "# initiate model\n", "MLC = rp.mlclassificator(X, y, \n", " scaling=True, \n", " test_size=0.33, \n", " random_state=42, \n", " algorithm = 'SVC')\n", "\n", "MLC.fit()\n", "\n", "# if you ask for scaling, do not forget to pass the scaled data to the model when scoring!!\n", "score = MLC.model.score(MLC.X_test_sc, MLC.y_test)\n", "print('\\n- Classifier SVC is scoring (test:) '+str(score)+'.')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can also try to tune the hyperparameters of other algorithms, such as Decision Tree:" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "- Classifier Decision Tree is scoring (test:) 0.9818181818181818.\n" ] } ], "source": [ "# we define the _params as we would do with the scikit-learn class:\n", "params = {'max_depth': 10}\n", "\n", "# initiate model\n", "MLC = rp.mlclassificator(X, y, \n", " scaling=True, \n", " test_size=0.33, \n", " random_state=42, \n", " algorithm = 'Decision Tree',\n", " params_ = params)\n", "MLC.fit()\n", "score = MLC.model.score(MLC.X_test_sc, MLC.y_test)\n", "print('\\n- Classifier Decision Tree is scoring (test:) '+str(score)+'.')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can also try to improve the Gaussian Process classifier using for instance the [Matern kernel](https://scikit-learn.org/stable/modules/gaussian_process.html#gp-kernels):" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "- Classifier GP is scoring (test:) 1.0.\n" ] } ], "source": [ "# we define the _params as we would do with the scikit-learn GP class:\n", "# we actually will try a linear kernel\n", "from sklearn.gaussian_process.kernels import Matern\n", "params = {'kernel': 1.0*Matern()}\n", "\n", "# initiate model\n", "MLC = rp.mlclassificator(X, y, \n", " scaling=True, \n", " test_size=0.33, \n", " random_state=42, \n", " algorithm = 'Gaussian Process',\n", " params_ = params)\n", "MLC.fit()\n", "score = MLC.model.score(MLC.X_test_sc, MLC.y_test)\n", "print('\\n- Classifier GP is scoring (test:) '+str(score)+'.')" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "test", "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": "3.12.13" } }, "nbformat": 4, "nbformat_minor": 4 }