examples/ml/JavaFMRegressorExample.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.ml;
0019
0020 // $example on$
0021 import org.apache.spark.ml.Pipeline;
0022 import org.apache.spark.ml.PipelineModel;
0023 import org.apache.spark.ml.PipelineStage;
0024 import org.apache.spark.ml.evaluation.RegressionEvaluator;
0025 import org.apache.spark.ml.feature.MinMaxScaler;
0026 import org.apache.spark.ml.feature.MinMaxScalerModel;
0027 import org.apache.spark.ml.regression.FMRegressionModel;
0028 import org.apache.spark.ml.regression.FMRegressor;
0029 import org.apache.spark.sql.Dataset;
0030 import org.apache.spark.sql.Row;
0031 import org.apache.spark.sql.SparkSession;
0032 // $example off$
0033
0034 public class JavaFMRegressorExample {
0035   public static void main(String[] args) {
0036     SparkSession spark = SparkSession
0037         .builder()
0038         .appName("JavaFMRegressorExample")
0039         .getOrCreate();
0040
0041     // $example on$
0042     // Load and parse the data file, converting it to a DataFrame.
0043     Dataset<Row> data = spark.read().format("libsvm").load("data/mllib/sample_libsvm_data.txt");
0044
0045     // Scale features.
0046     MinMaxScalerModel featureScaler = new MinMaxScaler()
0047         .setInputCol("features")
0048         .setOutputCol("scaledFeatures")
0049         .fit(data);
0050
0051     // Split the data into training and test sets (30% held out for testing).
0052     Dataset<Row>[] splits = data.randomSplit(new double[] {0.7, 0.3});
0053     Dataset<Row> trainingData = splits[0];
0054     Dataset<Row> testData = splits[1];
0055
0056     // Train a FM model.
0057     FMRegressor fm = new FMRegressor()
0058         .setLabelCol("label")
0059         .setFeaturesCol("scaledFeatures")
0060         .setStepSize(0.001);
0061
0062     // Create a Pipeline.
0063     Pipeline pipeline = new Pipeline().setStages(new PipelineStage[] {featureScaler, fm});
0064
0065     // Train model.
0066     PipelineModel model = pipeline.fit(trainingData);
0067
0068     // Make predictions.
0069     Dataset<Row> predictions = model.transform(testData);
0070
0071     // Select example rows to display.
0072     predictions.select("prediction", "label", "features").show(5);
0073
0074     // Select (prediction, true label) and compute test error.
0075     RegressionEvaluator evaluator = new RegressionEvaluator()
0076         .setLabelCol("label")
0077         .setPredictionCol("prediction")
0078         .setMetricName("rmse");
0079     double rmse = evaluator.evaluate(predictions);
0080     System.out.println("Root Mean Squared Error (RMSE) on test data = " + rmse);
0081
0082     FMRegressionModel fmModel = (FMRegressionModel)(model.stages()[1]);
0083     System.out.println("Factors: " + fmModel.factors());
0084     System.out.println("Linear: " + fmModel.linear());
0085     System.out.println("Intercept: " + fmModel.intercept());
0086     // $example off$
0087
0088     spark.stop();
0089   }
0090 }