/*
* File: LinearMixtureModel.java
* Authors: Kevin R. Dixon
* Company: Sandia National Laboratories
* Project: Cognitive Foundry
*
* Copyright November 3, 2006, Sandia Corporation. Under the terms of Contract
* DE-AC04-94AL85000, there is a non-exclusive license for use of this work by
* or on behalf of the U.S. Government. Export of this program may require a
* license from the United States Government. See CopyrightHistory.txt for
* complete details.
*
*/
package gov.sandia.cognition.statistics.distribution;
import gov.sandia.cognition.annotation.PublicationReference;
import gov.sandia.cognition.annotation.PublicationType;
import gov.sandia.cognition.collection.CollectionUtil;
import gov.sandia.cognition.statistics.AbstractDistribution;
import gov.sandia.cognition.statistics.Distribution;
import gov.sandia.cognition.statistics.ProbabilityMassFunctionUtil;
import gov.sandia.cognition.util.ObjectUtil;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
import java.util.Random;
/**
* A linear mixture of RandomVariables, with a prior probability distribution.
* The posterior pdf is:
* p(x|this) = \sum_{y\in this} p(x|y,this)P(y|this),
* where p(x|y,this) is the pdf of the underlying RandomVariable, and
* P(y|this) is the prior probability of RandomVariable y in this.
*
* @param <DataType>
* Type of data in this mixture model
* @param <DistributionType>
* The type of the internal distributions inside the mixture.
* @author Kevin R. Dixon
* @since 1.0
*
*/
@PublicationReference(
author="Wikipedia",
title="Mixture Model",
type=PublicationType.WebPage,
year=2009,
url="http://en.wikipedia.org/wiki/Mixture_model"
)
public abstract class LinearMixtureModel<DataType, DistributionType extends Distribution<DataType>>
extends AbstractDistribution<DataType>
{
/**
* Underlying distributions from which we sample
*/
protected ArrayList<? extends DistributionType> distributions;
/**
* Weights proportionate by which the distributions are sampled
*/
protected double[] priorWeights;
/**
* Creates a new instance of LinearMixtureModel
* @param distributions
* Underlying distributions from which we sample
*/
public LinearMixtureModel(
Collection<? extends DistributionType> distributions )
{
this( distributions, null );
}
/**
* Creates a new instance of LinearMixtureModel
* @param distributions
* Underlying distributions from which we sample
* @param priorWeights
* Weights proportionate by which the distributions are sampled
*/
public LinearMixtureModel(
Collection<? extends DistributionType> distributions,
double[] priorWeights)
{
if( priorWeights == null )
{
priorWeights = new double[distributions.size()];
Arrays.fill(priorWeights, 1.0);
}
if( distributions.size() != priorWeights.length )
{
throw new IllegalArgumentException(
"Distribution count must equal number of priors" );
}
for( int i = 0; i < priorWeights.length; i++ )
{
if( priorWeights[i] < 0.0 )
{
throw new IllegalArgumentException( "weights must be >= 0.0!" );
}
}
this.setDistributions(CollectionUtil.asArrayList(distributions));
this.setPriorWeights(priorWeights);
}
@Override
@SuppressWarnings("unchecked")
public LinearMixtureModel<DataType,DistributionType> clone()
{
LinearMixtureModel<DataType,DistributionType> clone =
(LinearMixtureModel<DataType, DistributionType>) super.clone();
clone.setDistributions( ObjectUtil.cloneSmartElementsAsArrayList(
this.getDistributions() ) );
clone.setPriorWeights( ObjectUtil.cloneSmart( this.getPriorWeights() ) );
return clone;
}
@Override
public String toString()
{
StringBuilder retval = new StringBuilder(1000);
retval.append("LinearMixtureModel has " + this.getDistributionCount() + " distributions:\n");
int k = 0;
for( DistributionType distribution : this.getDistributions() )
{
retval.append( " " + k + ": Prior: " + this.getPriorWeights()[k] + ", Distribution: " + distribution + "\n" );
k++;
}
return retval.toString();
}
/**
* Getter for distributions
* @return
* Underlying distributions from which we sample
*/
public ArrayList<? extends DistributionType> getDistributions()
{
return this.distributions;
}
/**
* Setter for distributions
* @param distributions
* Underlying distributions from which we sample
*/
public void setDistributions(
ArrayList<? extends DistributionType> distributions)
{
this.distributions = distributions;
}
/**
* Gets the number of distributions in the model
* @return
* Number of distributions in the model
*/
public int getDistributionCount()
{
return this.distributions.size();
}
@Override
public DataType sample(
Random random)
{
final DistributionType d = ProbabilityMassFunctionUtil.sampleSingle(
this.getPriorWeights(), this.getDistributions(), random);
return d.sample(random);
}
@Override
public void sampleInto(
final Random random,
final int sampleCount,
final Collection<? super DataType> output)
{
// Build the cumulative distribution for batch sampling.
final int distributionCount = this.getDistributionCount();
final double[] priorWeights = this.getPriorWeights();
final double[] cumulativeWeights = new double[distributionCount];
double sum = 0.0;
for(int n = 0; n < distributionCount; n++)
{
sum += priorWeights[n];
cumulativeWeights[n] = sum;
}
// Sample each of the mixtures.
for (int i = 0; i < sampleCount; i++)
{
final DistributionType d = ProbabilityMassFunctionUtil.sample(
cumulativeWeights, this.getDistributions(), random);
output.add(d.sample(random));
}
}
/**
* Getter for priorWeights
* @return
* Weights proportionate by which the distributions are sampled
*/
public double[] getPriorWeights()
{
return this.priorWeights;
}
/**
* Getter for priorWeights
* @param priorWeights
* Weights proportionate by which the distributions are sampled
*/
public void setPriorWeights(
double[] priorWeights)
{
this.priorWeights = priorWeights;
}
/**
* Computes the sum of the prior weights
* @return
* Sum of the prior weights
*/
public double getPriorWeightSum()
{
double sum = 0.0;
final int K = this.getPriorWeights().length;
for( int k = 0; k < K; k++ )
{
sum += this.getPriorWeights()[k];
}
return (sum <= 0.0) ? 1.0 : sum;
}
}