computationmanager.java

java实现的隐马尔科夫模型

/* jahmm package - v0.3.1 */

/*
 *  Copyright (c) 2004, Jean-Marc Francois.
 *
 *  This file is part of Jahmm.
 *  Jahmm 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.
 *
 *  Jahmm 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 Jahmm; if not, write to the Free Software
 *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */

package be.ac.ulg.montefiore.run.jahmm.apps;

import java.util.*;

import be.ac.ulg.montefiore.run.jahmm.*;
import be.ac.ulg.montefiore.run.jahmm.gui.*;
import be.ac.ulg.montefiore.run.jahmm.learn.*;
import be.ac.ulg.montefiore.run.jahmm.toolbox.*;


/**
 * Computes new (state or observation) sequences and HMM.
 */
public class ComputationManager {
    
    static public Hmm learn(String name, ObservationSequences sequences,
			    OpdfFactory opdfFactory, int nbStates) {
	Vector<?> nsequences = GenericsLayer.convertSequences(sequences);
	
	KMeansLearner learner = new KMeansLearner(nbStates,
						  opdfFactory,
						  nsequences);
	
	return new Hmm(name, learner.learn());
    }
    
    
    static public double probability(Hmm hmm, List<Observation> sequence) {
	Vector<?> nSequence =
	    GenericsLayer.convertSequence((ArrayList<Observation>) sequence);

	return hmm.probability(nSequence);
    }
    
    
    static public double lnProbability(Hmm hmm, List<Observation> sequence) {
	Vector<?> nSequence =
	    GenericsLayer.convertSequence((ArrayList<Observation>) sequence);

	return hmm.lnProbability(nSequence);
    }

    
    static public ArrayList<Integer> mostLikelySequence(Hmm hmm,
							List<Observation>
							sequence) {
	Vector<?> nSequence =
	    GenericsLayer.convertSequence((ArrayList<Observation>) sequence);
	ArrayList<Integer> mlss = new ArrayList<Integer>();
	int[] imlss = hmm.mostLikelyStateSequence(nSequence);

	for (int i = 0; i < imlss.length; i++)
	    mlss.add(new Integer(imlss[i]));
	
	return new ArrayList<Integer>(mlss);
    }

    
    static public ObservationSequences generateSequences(String name, Hmm hmm, 
							  int nb, int length) {
	Opdf opdf = hmm.getOpdf(0);
	int type, dimension;
	
	if (opdf instanceof OpdfInteger) { // Discrete distribution
	    type = ObservationSequences.INTEGER;
	    dimension = ((OpdfInteger) opdf).nbEntries();
	} else { // Multi gaussian distribution
	    type = ObservationSequences.VECTOR;
	    dimension = ((OpdfMultiGaussian) opdf).dimension();
	}
	
	ObservationSequences sequences =
	    new ObservationSequences(name, type, dimension);
	MarkovGenerator generator = new MarkovGenerator(hmm);
	Vector<Vector<Observation>> vector = new Vector<Vector<Observation>>();
	
	for (int i = 0; i < nb; i++) 
	    vector.add(generator.observationSequence(length));
	
	sequences.addAll(GenericsLayer.convertVector(vector));
	
	return sequences;
    }
}