examples/mllib/JavaBinaryClassificationMetricsExample.java

0001 /*
0002  * Licensed to the Apache Software Foundation (ASF) under one or more
0003  * contributor license agreements.  See the NOTICE file distributed with
0004  * this work for additional information regarding copyright ownership.
0005  * The ASF licenses this file to You under the Apache License, Version 2.0
0006  * (the "License"); you may not use this file except in compliance with
0007  * the License.  You may obtain a copy of the License at
0008  *
0009  *    http://www.apache.org/licenses/LICENSE-2.0
0010  *
0011  * Unless required by applicable law or agreed to in writing, software
0012  * distributed under the License is distributed on an "AS IS" BASIS,
0013  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
0014  * See the License for the specific language governing permissions and
0015  * limitations under the License.
0016  */
0017
0018 package org.apache.spark.examples.mllib;
0019
0020 // $example on$
0021 import scala.Tuple2;
0022
0023 import org.apache.spark.api.java.*;
0024 import org.apache.spark.mllib.classification.LogisticRegressionModel;
0025 import org.apache.spark.mllib.classification.LogisticRegressionWithLBFGS;
0026 import org.apache.spark.mllib.evaluation.BinaryClassificationMetrics;
0027 import org.apache.spark.mllib.regression.LabeledPoint;
0028 import org.apache.spark.mllib.util.MLUtils;
0029 // $example off$
0030 import org.apache.spark.SparkConf;
0031 import org.apache.spark.SparkContext;
0032
0033 public class JavaBinaryClassificationMetricsExample {
0034   public static void main(String[] args) {
0035     SparkConf conf = new SparkConf().setAppName("Java Binary Classification Metrics Example");
0036     SparkContext sc = new SparkContext(conf);
0037     // $example on$
0038     String path = "data/mllib/sample_binary_classification_data.txt";
0039     JavaRDD<LabeledPoint> data = MLUtils.loadLibSVMFile(sc, path).toJavaRDD();
0040
0041     // Split initial RDD into two... [60% training data, 40% testing data].
0042     JavaRDD<LabeledPoint>[] splits =
0043       data.randomSplit(new double[]{0.6, 0.4}, 11L);
0044     JavaRDD<LabeledPoint> training = splits[0].cache();
0045     JavaRDD<LabeledPoint> test = splits[1];
0046
0047     // Run training algorithm to build the model.
0048     LogisticRegressionModel model = new LogisticRegressionWithLBFGS()
0049       .setNumClasses(2)
0050       .run(training.rdd());
0051
0052     // Clear the prediction threshold so the model will return probabilities
0053     model.clearThreshold();
0054
0055     // Compute raw scores on the test set.
0056     JavaPairRDD<Object, Object> predictionAndLabels = test.mapToPair(p ->
0057       new Tuple2<>(model.predict(p.features()), p.label()));
0058
0059     // Get evaluation metrics.
0060     BinaryClassificationMetrics metrics =
0061       new BinaryClassificationMetrics(predictionAndLabels.rdd());
0062
0063     // Precision by threshold
0064     JavaRDD<Tuple2<Object, Object>> precision = metrics.precisionByThreshold().toJavaRDD();
0065     System.out.println("Precision by threshold: " + precision.collect());
0066
0067     // Recall by threshold
0068     JavaRDD<?> recall = metrics.recallByThreshold().toJavaRDD();
0069     System.out.println("Recall by threshold: " + recall.collect());
0070
0071     // F Score by threshold
0072     JavaRDD<?> f1Score = metrics.fMeasureByThreshold().toJavaRDD();
0073     System.out.println("F1 Score by threshold: " + f1Score.collect());
0074
0075     JavaRDD<?> f2Score = metrics.fMeasureByThreshold(2.0).toJavaRDD();
0076     System.out.println("F2 Score by threshold: " + f2Score.collect());
0077
0078     // Precision-recall curve
0079     JavaRDD<?> prc = metrics.pr().toJavaRDD();
0080     System.out.println("Precision-recall curve: " + prc.collect());
0081
0082     // Thresholds
0083     JavaRDD<Double> thresholds = precision.map(t -> Double.parseDouble(t._1().toString()));
0084
0085     // ROC Curve
0086     JavaRDD<?> roc = metrics.roc().toJavaRDD();
0087     System.out.println("ROC curve: " + roc.collect());
0088
0089     // AUPRC
0090     System.out.println("Area under precision-recall curve = " + metrics.areaUnderPR());
0091
0092     // AUROC
0093     System.out.println("Area under ROC = " + metrics.areaUnderROC());
0094
0095     // Save and load model
0096     model.save(sc, "target/tmp/LogisticRegressionModel");
0097     LogisticRegressionModel.load(sc, "target/tmp/LogisticRegressionModel");
0098     // $example off$
0099
0100     sc.stop();
0101   }
0102 }