Package com.bayesserver.data.sampling

Class DataSampler

java.lang.Object

com.bayesserver.data.sampling.DataSampler

public final class DataSampler
extends Object

Generates samples from a Bayesian network or Dynamic Bayesian network. See takeSample(com.bayesserver.inference.Evidence, com.bayesserver.RandomNumberGenerator, com.bayesserver.data.sampling.DataSamplingOptions) for sample code.

Constructor Summary

Constructors

Constructor	Description
DataSampler(Network network)	Initializes a new instance of the DataSampler class.
DataSampler(Network network, Evidence fixedData)	Initializes a new instance of the DataSampler class.

Method Summary

Modifier and Type	Method	Description
Network	getNetwork()	Gets the Bayesian network or Dynamic Bayesian network that was used in the constructor.
void	setFixedData(Evidence value)	Sets any evidence that should be fixed for each sample.
void	takeSample(Evidence sampleData, RandomNumberGenerator random, DataSamplingOptions options)	Generates sample data from the Bayesian network or Dynamic Bayesian network.

Methods inherited from class java.lang.Object

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

Constructor Detail

DataSampler

public DataSampler(Network network)

Initializes a new instance of the DataSampler class.

Parameters:

network - The Bayesian network or Dynamic Bayesian network the samples will be taken from.

DataSampler

public DataSampler(Network network,
                   Evidence fixedData)

Initializes a new instance of the DataSampler class.

Parameters:

network - The Bayesian network or Dynamic Bayesian network the samples will be taken from.

fixedData - Any data that is fixed. Can be null.

Method Detail

getNetwork

public Network getNetwork()

Gets the Bayesian network or Dynamic Bayesian network that was used in the constructor.

Returns:

The Bayesian network or Dynamic Bayesian network.

takeSample

public void takeSample(Evidence sampleData,
                       RandomNumberGenerator random,
                       DataSamplingOptions options)
                throws InconsistentEvidenceException

Generates sample data from the Bayesian network or Dynamic Bayesian network. Each call to takeSample(com.bayesserver.inference.Evidence, com.bayesserver.RandomNumberGenerator, com.bayesserver.data.sampling.DataSamplingOptions) generates a single sample. Call repeatedly to generate multiple samples.

For Dynamic Bayesian networks, see DataSamplingOptions.getSequenceLength() to change the length of generated sequences.

Parameters:

sampleData - Evidence instance to hold the sample data.

random - Random number generator.

options - The sampling options.

Throws:

InconsistentEvidenceException - Raised if the generation of a sample leads to inconsistent evidence.

setFixedData

public void setFixedData(Evidence value)

Sets any evidence that should be fixed for each sample.