examples/ml/JavaDecisionTreeClassificationExample.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 // scalastyle:off println
0018 package org.apache.spark.examples.ml;
0019 // $example on$
0020 import org.apache.spark.ml.Pipeline;
0021 import org.apache.spark.ml.PipelineModel;
0022 import org.apache.spark.ml.PipelineStage;
0023 import org.apache.spark.ml.classification.DecisionTreeClassifier;
0024 import org.apache.spark.ml.classification.DecisionTreeClassificationModel;
0025 import org.apache.spark.ml.evaluation.MulticlassClassificationEvaluator;
0026 import org.apache.spark.ml.feature.*;
0027 import org.apache.spark.sql.Dataset;
0028 import org.apache.spark.sql.Row;
0029 import org.apache.spark.sql.SparkSession;
0030 // $example off$
0031
0032 public class JavaDecisionTreeClassificationExample {
0033   public static void main(String[] args) {
0034     SparkSession spark = SparkSession
0035       .builder()
0036       .appName("JavaDecisionTreeClassificationExample")
0037       .getOrCreate();
0038
0039     // $example on$
0040     // Load the data stored in LIBSVM format as a DataFrame.
0041     Dataset<Row> data = spark
0042       .read()
0043       .format("libsvm")
0044       .load("data/mllib/sample_libsvm_data.txt");
0045
0046     // Index labels, adding metadata to the label column.
0047     // Fit on whole dataset to include all labels in index.
0048     StringIndexerModel labelIndexer = new StringIndexer()
0049       .setInputCol("label")
0050       .setOutputCol("indexedLabel")
0051       .fit(data);
0052
0053     // Automatically identify categorical features, and index them.
0054     VectorIndexerModel featureIndexer = new VectorIndexer()
0055       .setInputCol("features")
0056       .setOutputCol("indexedFeatures")
0057       .setMaxCategories(4) // features with > 4 distinct values are treated as continuous.
0058       .fit(data);
0059
0060     // Split the data into training and test sets (30% held out for testing).
0061     Dataset<Row>[] splits = data.randomSplit(new double[]{0.7, 0.3});
0062     Dataset<Row> trainingData = splits[0];
0063     Dataset<Row> testData = splits[1];
0064
0065     // Train a DecisionTree model.
0066     DecisionTreeClassifier dt = new DecisionTreeClassifier()
0067       .setLabelCol("indexedLabel")
0068       .setFeaturesCol("indexedFeatures");
0069
0070     // Convert indexed labels back to original labels.
0071     IndexToString labelConverter = new IndexToString()
0072       .setInputCol("prediction")
0073       .setOutputCol("predictedLabel")
0074       .setLabels(labelIndexer.labelsArray()[0]);
0075
0076     // Chain indexers and tree in a Pipeline.
0077     Pipeline pipeline = new Pipeline()
0078       .setStages(new PipelineStage[]{labelIndexer, featureIndexer, dt, labelConverter});
0079
0080     // Train model. This also runs the indexers.
0081     PipelineModel model = pipeline.fit(trainingData);
0082
0083     // Make predictions.
0084     Dataset<Row> predictions = model.transform(testData);
0085
0086     // Select example rows to display.
0087     predictions.select("predictedLabel", "label", "features").show(5);
0088
0089     // Select (prediction, true label) and compute test error.
0090     MulticlassClassificationEvaluator evaluator = new MulticlassClassificationEvaluator()
0091       .setLabelCol("indexedLabel")
0092       .setPredictionCol("prediction")
0093       .setMetricName("accuracy");
0094     double accuracy = evaluator.evaluate(predictions);
0095     System.out.println("Test Error = " + (1.0 - accuracy));
0096
0097     DecisionTreeClassificationModel treeModel =
0098       (DecisionTreeClassificationModel) (model.stages()[2]);
0099     System.out.println("Learned classification tree model:\n" + treeModel.toDebugString());
0100     // $example off$
0101
0102     spark.stop();
0103   }
0104 }