RolloutResult.java

/**
 * 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.
 *  
 *  You should have received a copy of the GNU General Public License
 *  along with this program; if not, write to the Free Software
 *  Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
 */
package bots.mctsbot.ai.bots.bot.gametree.rollout;

import java.util.Collections;
import java.util.SortedMap;
import java.util.Map.Entry;

import org.apache.log4j.Logger;

import bots.mctsbot.common.util.MutableDouble;

public class RolloutResult {

	private final static Logger logger = Logger.getLogger(RolloutResult.class);

	private final SortedMap<Integer, MutableDouble> values;
	private final double totalProb;
	private final double rakeFactor;

	public RolloutResult(SortedMap<Integer, MutableDouble> values, double totalProb, double rakeFactor) {
		if (Double.isNaN(totalProb) || Double.isInfinite(totalProb) || totalProb == 0) {
			logger.error("Bad total probability: " + totalProb);
			throw new IllegalStateException();
		}
		this.values = Collections.unmodifiableSortedMap(values);
		this.totalProb = totalProb;
		this.rakeFactor = rakeFactor;
	}

	public double getMean() {
		double mean = 0;
		for (Entry<Integer, MutableDouble> entry : values.entrySet()) {
			mean += entry.getKey() * entry.getValue().getValue();
		}
		mean /= totalProb;
		return mean * rakeFactor;
	}

	public double getVariance(double mean, int nbSamples) {
		double var = 0;
		for (Entry<Integer, MutableDouble> entry : values.entrySet()) {
			double diff = mean - entry.getKey() * rakeFactor;
			var += diff * diff * entry.getValue().getValue();
		}
		var /= totalProb;
		// use sample variance because variance of samples is smaller than real
		// variance
		var *= nbSamples / (double) (nbSamples - 1);
		return var;
	}

	public double getTotalProb() {
		return totalProb;
	}

	public SortedMap<Integer, MutableDouble> getValues() {
		return values;
	}

	public double getRakeFactor() {
		return rakeFactor;
	}

}