unbbayes.simulation.likelihoodweighting.sampling
Class LikelihoodWeightingSampling

java.lang.Object
  unbbayes.simulation.montecarlo.sampling.AMonteCarloSampling
      unbbayes.simulation.montecarlo.sampling.MatrixMonteCarloSampling
          unbbayes.simulation.likelihoodweighting.sampling.LikelihoodWeightingSampling

All Implemented Interfaces:

ILikelihoodWeightingSampling, IMonteCarloSampling, ILongTaskProgressObservable

public class LikelihoodWeightingSampling
extends MatrixMonteCarloSampling
implements ILikelihoodWeightingSampling

Likelihood Weighting sampling based on MC sampling. However, it does not sample for the evidence nodes, it just sets as the given state, and it calculates P(E|Par(E)) for each trial.

Author:

Rommel Carvalho (rommel.carvalho@gmail.com)

Field Summary

protected List<Node>	evidenceNodeList
protected float[]	probabilityEvidenceGivenParentList

Fields inherited from class unbbayes.simulation.montecarlo.sampling.AMonteCarloSampling

currentProgress, currentProgressStatus, factors, maxProgress, nTrials, pn, sampledStatesMap, sampledStatesMatrix, samplingNodeOrderQueue, timesSampled

Constructor Summary

LikelihoodWeightingSampling()

Method Summary

float[]	getCompactStatesSetWeight() Not implemented.
float[]	getFullStatesSetWeight() Return P(E|Par(E)) = ProductOf[P(Ei|Par(Ei))] for all evidences (findings).
Map<Integer,Float>	getMapStatesSetWeight() Not implemented.
protected void	simulate(int nTrial) Responsible for simulating MC for sampling.
void	start(ProbabilisticNetwork pn, int nTrials) Generates the MC sample with the given size for the given probabilistic network.

Methods inherited from class unbbayes.simulation.montecarlo.sampling.MatrixMonteCarloSampling

getSampledStatesCompactMatrix, getSampledStatesMap, getSampledStatesMatrix, getStatesSetTimesSampled

Methods inherited from class unbbayes.simulation.montecarlo.sampling.AMonteCarloSampling

addToSamplingOrderQueue, computeFactors, createSamplingOrderQueue, getCumulativeDistributionFunction, getCurrentProgress, getCurrentProgressStatus, getIndexInQueue, getLinearCoord, getMaxProgress, getMultidimensionalCoord, getParentsIndexesInQueue, getPercentageDone, getProbabilityMassFunction, getSamplingNodeOrderQueue, getState, initSamplingOrderQueue, notityObservers, registerObserver, removeObserver, updateProgress, updateProgress

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface unbbayes.simulation.montecarlo.sampling.IMonteCarloSampling

getLinearCoord, getMultidimensionalCoord, getSampledStatesCompactMatrix, getSampledStatesMap, getSampledStatesMatrix, getSamplingNodeOrderQueue, getStatesSetTimesSampled

Methods inherited from interface unbbayes.util.longtask.ILongTaskProgressObservable

getCurrentProgress, getCurrentProgressStatus, getMaxProgress, getPercentageDone, notityObservers, registerObserver, removeObserver

Field Detail

evidenceNodeList

protected List<Node> evidenceNodeList

probabilityEvidenceGivenParentList

protected float[] probabilityEvidenceGivenParentList

Constructor Detail

LikelihoodWeightingSampling

public LikelihoodWeightingSampling()

Method Detail

getFullStatesSetWeight

public float[] getFullStatesSetWeight()

Return P(E|Par(E)) = ProductOf[P(Ei|Par(Ei))] for all evidences (findings). There is one probability associated with each trial.

Specified by:

getFullStatesSetWeight in interface ILikelihoodWeightingSampling

Returns:

P(E|Par(E)) for each trial.

start

public void start(ProbabilisticNetwork pn,
                  int nTrials)

Description copied from class: MatrixMonteCarloSampling

Generates the MC sample with the given size for the given probabilistic network.

Specified by:

start in interface IMonteCarloSampling

Overrides:

start in class MatrixMonteCarloSampling

Parameters:

pn - Probabilistic network that will be used for sampling.

nTrials - Number of trials to generate.

simulate

protected void simulate(int nTrial)

Description copied from class: MatrixMonteCarloSampling

Responsible for simulating MC for sampling.

Overrides:

simulate in class MatrixMonteCarloSampling

Parameters:

nTrial - The trial number to simulate.

getCompactStatesSetWeight

public float[] getCompactStatesSetWeight()

Not implemented.

Specified by:

getCompactStatesSetWeight in interface ILikelihoodWeightingSampling

Returns:

null.

getMapStatesSetWeight

public Map<Integer,Float> getMapStatesSetWeight()

Not implemented.

Specified by:

getMapStatesSetWeight in interface ILikelihoodWeightingSampling

Returns:

null.