jjj.asap.sas.models2.job.BuildUnicornEnsemble.java Source code

Introduction

Here is the source code for jjj.asap.sas.models2.job.BuildUnicornEnsemble.java

Source

/*
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or GITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */

/*
 * Copyright (C) 2012 James Jesensky
 */

package jjj.asap.sas.models2.job;

import java.util.ArrayList;
import java.util.Arrays;
import java.util.HashSet;
import java.util.LinkedList;
import java.util.List;
import java.util.Queue;
import java.util.Set;
import java.util.concurrent.Future;

import jjj.asap.sas.ensemble.EnsembleBuilder;
import jjj.asap.sas.ensemble.LoaderL1;
import jjj.asap.sas.ensemble.impl.CrossValidatedEnsemble;
import jjj.asap.sas.ensemble.impl.StackedClassifier;
import jjj.asap.sas.ensemble.scheme.GACommittee;
import jjj.asap.sas.util.Bucket;
import jjj.asap.sas.util.Contest;
import jjj.asap.sas.util.Job;
import jjj.asap.sas.util.Progress;
import weka.classifiers.bayes.NaiveBayes;
import weka.classifiers.functions.LinearRegression;
import weka.core.SelectedTag;

/**
 * Builds a variety of classifiers
 */

public class BuildUnicornEnsemble extends Job {

    private static final SelectedTag M5 = new SelectedTag(LinearRegression.SELECTION_M5,
            LinearRegression.TAGS_SELECTION);
    private static final SelectedTag NONE = new SelectedTag(LinearRegression.SELECTION_NONE,
            LinearRegression.TAGS_SELECTION);

    /**
     * @param args[0] is the output bucket
     * 
     * @param args[1] the output name. Actually outputs will be:
     * work/models2/t/k-name.csv and work/models2/u/k-name.csv. These will be
     * in submission format, though possibly not in the correct order.
     * 
     * @param args[2],args[3],... one or more input buckets. If more than one
     * bucket is given and duplicate items are found, only one will be processed.
     * 
     * @throws Exception 
     */
    public static void main(String[] args) throws Exception {

        // check
        if (args.length < 3) {
            throw new RuntimeException(Arrays.toString(args));
        }

        // input array
        String[] inputBuckets = new String[args.length - 2];
        System.arraycopy(args, 2, inputBuckets, 0, inputBuckets.length);

        Job job = new BuildUnicornEnsemble(args[0], args[1], inputBuckets);
        Job.log("ARGS", Arrays.toString(args));
        job.start();

    }

    private String outputBucket;
    private String outputName;
    private String[] inputBuckets;

    public BuildUnicornEnsemble(String outputBucket, String outputName, String[] inputBuckets) {
        super();
        this.outputBucket = outputBucket;
        this.outputName = outputName;
        this.inputBuckets = inputBuckets;
    }

    @Override
    protected void run() throws Exception {

        // get all the items from the buckets
        Set<String> items = new HashSet<String>();
        for (String inputBucket : inputBuckets) {
            items.addAll(Bucket.getBucketItems("models", inputBucket));
        }

        // split the buckets up by essay set
        List<String>[] buckets = new List[10];
        for (int i = 0; i < 10; i++) {
            buckets[i] = new ArrayList<String>();
        }

        for (String item : items) {
            int index = Contest.getEssaySet(item) - 1;
            buckets[index].add(item);
        }

        // init multi-threading
        Job.startService();
        final Queue<Future<Object>> queue = new LinkedList<Future<Object>>();

        // Stacking over labels

        for (int k = 0; k < 10; k++) {

            int essaySet = k + 1;

            queue.add(Job.submit(new EnsembleBuilder(essaySet, this.outputBucket, this.outputName + "-naive-bayes",
                    new GACommittee(10, 30,
                            new CrossValidatedEnsemble(new StackedClassifier(true, new NaiveBayes()), 5), 60),
                    buckets[k], new LoaderL1())));
        }

        // wait on complete
        Progress progress = new Progress(queue.size(), this.getClass().getSimpleName());
        while (!queue.isEmpty()) {
            try {
                queue.remove().get();
            } catch (Exception e) {
                Job.log("ERROR", e.toString());
                e.printStackTrace(System.err);
            }
            progress.tick();
        }
        progress.done();
        Job.stopService();

    }

}