RothErevLearnerTest.java

/*
 * JASA Java Auction Simulator API
 * Copyright (C) 2013 Steve Phelps
 *
 * 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 2 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 FITNESS FOR A PARTICULAR PURPOSE.
 * See the GNU General Public License for more details.
 */

package net.sourceforge.jabm.learning;

import junit.framework.Test;
import junit.framework.TestCase;
import junit.framework.TestSuite;
import net.sourceforge.jabm.learning.NPTRothErevLearner;
import net.sourceforge.jabm.learning.RothErevLearner;
import net.sourceforge.jabm.test.PRNGTestSeeds;
import net.sourceforge.jabm.util.SummaryStats;
import cern.jet.random.engine.MersenneTwister64;
import cern.jet.random.engine.RandomEngine;

public class RothErevLearnerTest extends TestCase {

	static final int CORRECT_ACTION = 2;

	/**
	 * @uml.property name="learner1"
	 * @uml.associationEnd
	 */
	NPTRothErevLearner learner1;
	
	RandomEngine prng;

	public RothErevLearnerTest(String name) {
		super(name);
	}

	public void setUp() {
		prng = new MersenneTwister64(PRNGTestSeeds.UNIT_TEST_SEED);
		learner1 = new NPTRothErevLearner(10, 0.2, 0.2, 100.0, prng);
	}

	public void testBasic() {
		learner1.setExperimentation(0.99);
		System.out.println("testBasic()");
		SummaryStats stats = new SummaryStats("action");
		int correctActions = 0;
		for (int i = 0; i < 100; i++) {
			int action = learner1.act();
			stats.newData(action);
			if (action == CORRECT_ACTION) {
				learner1.reward(1.0);
				correctActions++;
			} else {
				learner1.reward(0);
			}
			checkProbabilities(learner1);
		}
		System.out.println("final state of learner1 = " + learner1);
		System.out.println("learner1 score = " + correctActions + "%");
		System.out.println(stats);
	}

	public void testDistribution() {
		System.out.println("\ntestDistribution()");
		double q[] = { 55, 5, 5, 5, 5, 5, 5, 5, 5, 5 };
		SummaryStats action1Data = new SummaryStats("action1");
		for (int r = 0; r < 10000; r++) {
			learner1 = new NPTRothErevLearner(10, 0.2, 0.2, 1, prng);
			learner1.setPropensities(q);
			SummaryStats choiceData = new SummaryStats("choice");
			int action1Chosen = 0;
			for (int i = 0; i < 100; i++) {
				int choice = learner1.act();
				choiceData.newData(choice);
				action1Chosen = 0;
				if (choice == 0) {
					action1Chosen = 1;
				}
				action1Data.newData(action1Chosen);
			}
		}
		System.out.println(action1Data);
		assertTrue(action1Data.getMean() <= 0.57 && action1Data.getMean() >= 0.53);
	}

	public static void checkProbabilities(RothErevLearner l) {
		double prob = 0;
		for (int i = 0; i < l.getK(); i++) {
			prob += l.getProbability(i);
		}
		if (prob > 1.001 || prob < 0.999) {
			throw new Error("Probabilities should sum to 1");
		}
	}

	public static void main(String[] args) {
		junit.textui.TestRunner.run(suite());
	}

	public static Test suite() {
		return new TestSuite(RothErevLearnerTest.class);
	}
}