Package com.bayesserver.inference

Class LikelihoodSamplingQueryOptions

java.lang.Object

com.bayesserver.inference.LikelihoodSamplingQueryOptions

All Implemented Interfaces:

QueryOptions, QuerySamplingOptions

public final class LikelihoodSamplingQueryOptions
extends Object
implements QueryOptions, QuerySamplingOptions

Options that govern the calculations performed by Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput).

Constructor Summary

Constructors

Constructor	Description
LikelihoodSamplingQueryOptions()	Initializes a new instance of the LikelihoodSamplingQueryOptions class.

Method Summary

Modifier and Type	Method	Description
Cancellation	getCancellation()	Allows cancellation of a query.
CausalEffectKind	getCausalEffectKind()	Gets the kind of effect to calculate.
InferenceFactory	getCausalInferenceFactory()	Factory that can create inference engines, used by estimation (adjustment) algorithms to perform queries on the Bayesian network.
boolean	getConflict()	Gets a value indicating whether the conflict measure should be calculated in a call to Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput).
DecisionAlgorithm	getDecisionAlgorithm()	Gets the algorithm to use when a network contains Decision nodes.
InconsistentEvidenceMode	getInconsistentEvidenceMode()	Determines when an InconsistentEvidenceException is raised.
boolean	getLogLikelihood()	Gets a value indicating whether the log-likelihood of a case should be calculated in a call to Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput).
PropagationMethod	getPropagation()	Gets the propagation method to be used during inference.
QueryEvidenceMode	getQueryEvidenceMode()	Determines whether evidence is retracted for each query.
Integer	getSampleCount()	Gets a value indicating how many samples cases to generate in order to approximate the current query.
Integer	getSeed()	Gets an optional seed for the random number generator.
Integer	getTerminalTime()	Gets the terminal time, which is the time at which any terminal nodes in a Dynamic Bayesian Network connect to temporal nodes.
void	setCancellation(Cancellation value)	Allows cancellation of a query.
void	setCausalEffectKind(CausalEffectKind value)	Sets the kind of effect to calculate.
void	setCausalInferenceFactory(InferenceFactory value)	Factory that can create inference engines, used by estimation (adjustment) algorithms to perform queries on the Bayesian network.
void	setConflict(boolean value)	Sets a value indicating whether the conflict measure should be calculated in a call to Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput).
void	setDecisionAlgorithm(DecisionAlgorithm value)	Sets the algorithm to use when a network contains Decision nodes.
void	setInconsistentEvidenceMode(InconsistentEvidenceMode value)	Determines when an InconsistentEvidenceException is raised.
void	setLogLikelihood(boolean value)	Sets a value indicating whether the log-likelihood of a case should be calculated in a call to Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput).
void	setPropagation(PropagationMethod value)	Sets the propagation method to be used during inference.
void	setQueryEvidenceMode(QueryEvidenceMode value)	Determines whether evidence is retracted for each query.
void	setSampleCount(Integer value)	Sets a value indicating how many samples cases to generate in order to approximate the current query.
void	setSeed(Integer value)	Sets an optional seed for the random number generator.
void	setTerminalTime(Integer value)	Sets the terminal time, which is the time at which any terminal nodes in a Dynamic Bayesian Network connect to temporal nodes.

Methods inherited from class java.lang.Object

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

Constructor Detail

LikelihoodSamplingQueryOptions

public LikelihoodSamplingQueryOptions()

Initializes a new instance of the LikelihoodSamplingQueryOptions class.

Method Detail

getCausalEffectKind

public CausalEffectKind getCausalEffectKind()

Gets the kind of effect to calculate. e.g. Total/Direct. Not all algorithms support direct effect calculations.

Specified by:

getCausalEffectKind in interface QueryOptions

setCausalEffectKind

public void setCausalEffectKind(CausalEffectKind value)

Sets the kind of effect to calculate. e.g. Total/Direct. Not all algorithms support direct effect calculations.

Specified by:

setCausalEffectKind in interface QueryOptions

getCausalInferenceFactory

public InferenceFactory getCausalInferenceFactory()

Factory that can create inference engines, used by estimation (adjustment) algorithms to perform queries on the Bayesian network.

Specified by:

getCausalInferenceFactory in interface QueryOptions

setCausalInferenceFactory

public void setCausalInferenceFactory(InferenceFactory value)

Factory that can create inference engines, used by estimation (adjustment) algorithms to perform queries on the Bayesian network.

Specified by:

setCausalInferenceFactory in interface QueryOptions

getSampleCount

public Integer getSampleCount()

Gets a value indicating how many samples cases to generate in order to approximate the current query. The higher the value, the more accurate the predictions will be, at the expense of query time.

When null, the algorithm being used will decide how many samples to use.

Specified by:

getSampleCount in interface QuerySamplingOptions

setSampleCount

public void setSampleCount(Integer value)

Sets a value indicating how many samples cases to generate in order to approximate the current query. The higher the value, the more accurate the predictions will be, at the expense of query time.

When null, the algorithm being used will decide how many samples to use.

Specified by:

setSampleCount in interface QuerySamplingOptions

getSeed

public Integer getSeed()

Gets an optional seed for the random number generator.

setSeed

public void setSeed(Integer value)

Sets an optional seed for the random number generator.

getTerminalTime

public Integer getTerminalTime()

Gets the terminal time, which is the time at which any terminal nodes in a Dynamic Bayesian Network connect to temporal nodes.

Specified by:

getTerminalTime in interface QueryOptions

setTerminalTime

public void setTerminalTime(Integer value)

Sets the terminal time, which is the time at which any terminal nodes in a Dynamic Bayesian Network connect to temporal nodes.

Specified by:

setTerminalTime in interface QueryOptions

getPropagation

public PropagationMethod getPropagation()

Gets the propagation method to be used during inference. Default is PropagationMethod.SUM.

Specified by:

getPropagation in interface QueryOptions

setPropagation

public void setPropagation(PropagationMethod value)

Sets the propagation method to be used during inference. Default is PropagationMethod.SUM.

Specified by:

setPropagation in interface QueryOptions

getDecisionAlgorithm

public DecisionAlgorithm getDecisionAlgorithm()

Gets the algorithm to use when a network contains Decision nodes.

Specified by:

getDecisionAlgorithm in interface QueryOptions

setDecisionAlgorithm

public void setDecisionAlgorithm(DecisionAlgorithm value)

Sets the algorithm to use when a network contains Decision nodes.

Specified by:

setDecisionAlgorithm in interface QueryOptions

getCancellation

public Cancellation getCancellation()

Allows cancellation of a query. When cancelled, a CancellationException will be raised.

If Cancellation.getCancel() has been set to true, it should be reset to false, before a subsequent query.

Specified by:

getCancellation in interface QueryOptions

setCancellation

public void setCancellation(Cancellation value)

Allows cancellation of a query. When cancelled, a CancellationException will be raised.

If Cancellation.getCancel() has been set to true, it should be reset to false, before a subsequent query.

Specified by:

setCancellation in interface QueryOptions

getLogLikelihood

public boolean getLogLikelihood()

Gets a value indicating whether the log-likelihood of a case should be calculated in a call to Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput). If calculated the result can be obtained from QueryOutput.getLogLikelihood().

This should only be requested if required.

The natural log (base e) is calculated.

Specified by:

getLogLikelihood in interface QueryOptions

Returns:

true if the log-likelihood should be calculated; otherwise, false. Defaults to false.

setLogLikelihood

public void setLogLikelihood(boolean value)

Sets a value indicating whether the log-likelihood of a case should be calculated in a call to Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput). If calculated the result can be obtained from QueryOutput.getLogLikelihood().

This should only be requested if required.

The natural log (base e) is calculated.

Specified by:

setLogLikelihood in interface QueryOptions

Parameters:

value - true if the log-likelihood should be calculated; otherwise, false. Defaults to false.

getConflict

public boolean getConflict()

Gets a value indicating whether the conflict measure should be calculated in a call to Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput). If calculated the result can be obtained from QueryOutput.getConflict().

Conflict is a measure that detects evidence that is conflicting or rare. The greater the conflict value above zero, the more likely the evidence is in conflict, or rare.

Conflict = log((P(e1)P(e2)...P(ei)) / P(e)), where P(e1), P(e2) etc... are the likelihoods of each variable considered in isolation, and P(e) is the likelihood of the all the evidence together.

Specified by:

getConflict in interface QueryOptions

setConflict

public void setConflict(boolean value)

Sets a value indicating whether the conflict measure should be calculated in a call to Inference.query(com.bayesserver.inference.QueryOptions, com.bayesserver.inference.QueryOutput). If calculated the result can be obtained from QueryOutput.getConflict().

Conflict is a measure that detects evidence that is conflicting or rare. The greater the conflict value above zero, the more likely the evidence is in conflict, or rare.

Conflict = log((P(e1)P(e2)...P(ei)) / P(e)), where P(e1), P(e2) etc... are the likelihoods of each variable considered in isolation, and P(e) is the likelihood of the all the evidence together.

Specified by:

setConflict in interface QueryOptions

getQueryEvidenceMode

public QueryEvidenceMode getQueryEvidenceMode()

Determines whether evidence is retracted for each query. See QueryEvidenceMode for more information.

Specified by:

getQueryEvidenceMode in interface QueryOptions

setQueryEvidenceMode

public void setQueryEvidenceMode(QueryEvidenceMode value)

Determines whether evidence is retracted for each query. See QueryEvidenceMode for more information.

Specified by:

setQueryEvidenceMode in interface QueryOptions

getInconsistentEvidenceMode

public InconsistentEvidenceMode getInconsistentEvidenceMode()

Determines when an InconsistentEvidenceException is raised.

Specified by:

getInconsistentEvidenceMode in interface QueryOptions

setInconsistentEvidenceMode

public void setInconsistentEvidenceMode(InconsistentEvidenceMode value)

Determines when an InconsistentEvidenceException is raised.

Specified by:

setInconsistentEvidenceMode in interface QueryOptions