{ "metadata": { "name": "", "signature": "sha256:99cc1a6b9611fdd27664cd9d8e31fed84971dcfbb70169cd5366b19e1d37faf2" }, "nbformat": 3, "nbformat_minor": 0, "worksheets": [ { "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Breakout: Exploring Machine Learning\n", "\n", "In this breakout, we'll explore the use of scikit-learn on some astronomical data.\n", "\n", "This will use [astroML](http://astroML.org), a package for machine learning in astronomy which includes several convenient loaders for various astronomical datasets.\n", "\n", "Recall that if you don't have ``astroML`` installed, you can install it easily from the command-line with ``pip``:\n", "\n", "```\n", "$ pip install astroML\n", "```" ] }, { "cell_type": "code", "collapsed": false, "input": [ "%matplotlib inline\n", "import numpy as np\n", "import matplotlib.pyplot as plt\n", "from scipy import stats\n", "\n", "# use seaborn plotting defaults\n", "# If this causes an error, you can comment it out.\n", "import seaborn as sns\n", "sns.set()" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 1 }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 1. Classification: Labeling Photometric Sources\n", "\n", "Here we'll do some automated classification of photometric sources.\n", "First the data, which can be fetched via ``astroML``. If you don't have\n", "astroML installed, use ``pip install astroML``" ] }, { "cell_type": "code", "collapsed": false, "input": [ "from astroML.datasets import fetch_rrlyrae_combined\n", "from sklearn.cross_validation import train_test_split\n", "\n", "X, y = fetch_rrlyrae_combined()\n", "\n", "# For now, we'll only fit the first two colors\n", "X_train, X_test, y_train, y_test = train_test_split(X, y)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 2 }, { "cell_type": "code", "collapsed": false, "input": [ "N_plot = 5000\n", "plt.scatter(X[-N_plot:, 0], X[-N_plot:, 1], c=y[-N_plot:],\n", " edgecolors='none', cmap='RdBu')\n", "plt.xlabel('u-g color')\n", "plt.ylabel('g-r color')\n", "plt.xlim(0.7, 1.4)\n", "plt.ylim(-0.2, 0.4);" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAFkCAYAAAAnu5JEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", "text": [ "" ] } ], "prompt_number": 3 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Blue points are RR-Lyrae, Red points are main sequence stars.\n", "\n", "Now we'll do a simple and fast $K$-neighbors classification.\n", "\n", "Note that we train on part of the data and test on another part. Otherwise,\n", "an estimator which simply remembers the entire dataset would obtain a perfect\n", "classification!" ] }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn.neighbors import KNeighborsClassifier\n", "clf = KNeighborsClassifier(n_neighbors=5)\n", "clf.fit(X_train, y_train)\n", "y_pred = clf.predict(X_test)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 4 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now let's check how we're doing. You might be tempted to simply check the accuracy: i.e. in what fraction of points the predictions match the true value. Let's see this:" ] }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn.metrics import accuracy_score\n", "accuracy_score(y_test, y_pred)" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 5, "text": [ "0.99643562655672935" ] } ], "prompt_number": 5 }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn.metrics import classification_report\n", "print(classification_report(y_test, y_pred,\n", " target_names=['MS star', 'RR Lyrae']))" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ " precision recall f1-score support\n", "\n", " MS star 1.00 1.00 1.00 23154\n", " RR Lyrae 0.69 0.67 0.68 132\n", "\n", "avg / total 1.00 1.00 1.00 23286\n", "\n" ] } ], "prompt_number": 6 }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Note on the meaning of these terms:\n", "\n", "$$\n", "{\\rm precision} \\equiv \\frac{\\rm true~positives}{\\rm true~positives + false~positives}\n", "$$\n", "\n", "$$\n", "{\\rm recall} \\equiv \\frac{\\rm true~positives}{\\rm true~positives + false~negatives}\n", "$$\n", "\n", "$$ F_1 \\equiv \\frac{\\rm precision \\cdot recall}{\\rm precision + recall}\n", "$$\n", "\n", "The range for precision, recall, and F1 score is 0 to 1, with 1 being perfect.\n", "Often in astronomy, we call the recall the *completeness*, and (1 - precision) the *contamination*.\n", "\n", "Because this is an *unbalanced* dataset, we see that the RR Lyrae stars are completely overwhelmed by the larger number of normal Main Sequence stars." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Classification Exercise\n", "\n", "Use another classifier and try to improve on this precision/recall score.\n", "\n", "There are several you might try:\n", "\n", "- ``sklearn.naive_bayes.GaussianNB`` (fast but inaccurate)\n", "- ``sklearn.lda.LDA`` (fast but inaccurate)\n", "- ``sklearn.svm.SVC`` (slow but accurate)\n", "- ``sklearn.svm.SVC`` with ``kernel='rbf'`` (slow but accurate)\n", "- ``sklearn.ensemble.RandomForestClassifier`` (fast & potentially accurate with tuning)\n", "\n", "For the slower algorithms, it might be a good idea to use only part of the dataset as you experiment, i.e. when training do something like the following:\n", "\n", "```\n", "clf.fit(X_train[::5], y_train[::5])\n", "```\n", "\n", "Once you're happy, you can run the training on the full dataset.\n", "\n", "What's the best precision/recall you can obtain for the RR-Lyrae data?" ] }, { "cell_type": "code", "collapsed": false, "input": [], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 6 }, { "cell_type": "code", "collapsed": false, "input": [], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 6 }, { "cell_type": "code", "collapsed": false, "input": [], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 6 }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 2. Regression: Photometric Redshifts\n", "\n", "The photometric redshift problem is a classic *Regression* problem" ] }, { "cell_type": "code", "collapsed": false, "input": [ "from astroML.datasets import fetch_sdss_specgals" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 7 }, { "cell_type": "code", "collapsed": false, "input": [ "data = fetch_sdss_specgals()\n", "\n", "# put magnitudes in a matrix\n", "X = np.vstack([data['modelMag_%s' % f] for f in 'ugriz']).T\n", "y = data['z']\n", "\n", "# Split into training and testing data\n", "X_train, X_test, y_train, y_test = train_test_split(X, y)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 8 }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn.linear_model import LinearRegression\n", "est = LinearRegression()\n", "est.fit(X_train, y_train)\n", "y_pred = est.predict(X_test)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 9 }, { "cell_type": "code", "collapsed": false, "input": [ "plt.plot(y_test, y_pred, ',k')\n", "plt.plot([0, 1], [0, 1], ':k')\n", "plt.xlim(0, 0.6)\n", "plt.ylim(0, 0.6)" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "pyout", "prompt_number": 10, "text": [ "(0, 0.6)" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAecAAAFVCAYAAADVDycqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCs4iIiI5o+AsIiKSM/8fB7Rkmt6wPOIAAAAASUVORK5CYII=\n", "text": [ "" ] } ], "prompt_number": 10 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Evidently, a simple linear model is not a very good fit!\n", "\n", "Let's compute the RMS deviation to see how poor this actually is:" ] }, { "cell_type": "code", "collapsed": false, "input": [ "rms = np.sqrt(np.mean((y_test - y_pred) ** 2))\n", "print(rms)" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "0.0322454288984\n" ] } ], "prompt_number": 11 }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Regression Exercise\n", "\n", "Try to improve on this result using another regression algorithm. Some potential choices:\n", "\n", "- ``sklearn.neighbors.KNeighborsRegressor`` (fast-ish, but often inaccurate)\n", "- ``sklearn.svm.SVR`` (potentially accurate, but slow)\n", "- ``sklearn.ensemble.RandomForestRegressor`` (fast, and accurate when well-tuned)\n", "- ``sklearn.ensemble.GradientBoostingRegressor`` (accurate with tuning, can be very slow depending on parameters)\n", "\n", "Again, sub-sampling the data can help as you explore some of the slower techniques." ] }, { "cell_type": "code", "collapsed": false, "input": [ "from sklearn.ensemble import RandomForestClassifier\n", "\n", "# Use a question mark in IPython to find help:\n", "RandomForestClassifier?" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 12 }, { "cell_type": "code", "collapsed": false, "input": [], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 12 }, { "cell_type": "code", "collapsed": false, "input": [], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 12 } ], "metadata": {} } ] }