bayesnet.jayes.factor

Class FactorFactory

java.lang.Object

bayesnet.jayes.factor.FactorFactory

public class FactorFactory
extends java.lang.Object

Constructor Summary

Constructors

Constructor and Description
FactorFactory()

Method Summary

Modifier and Type	Method and Description
AbstractFactor	create(java.util.List<java.lang.Integer> vars, java.util.List<AbstractFactor> multiplicationPartners) creates a factor, the class of which is dependent on different criteria defined in the concrete subclasses.
static FactorFactory	defaultFactory()
BayesNet	getReferenceNetwork()
void	setFloatingPointType(java.lang.Class<?> contentType) sets the floating point precision to use.
void	setReferenceNetwork(BayesNet net)
void	setUseLogScale(boolean useLogScale) using logarithmic values for computation is a lot slower, but also a lot more numerically stable

Methods inherited from class java.lang.Object

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

Constructor Detail

FactorFactory

public FactorFactory()

Method Detail

setFloatingPointType

public void setFloatingPointType(java.lang.Class<?> contentType)

sets the floating point precision to use.

Parameters:

contentType - possible values: double.class, Double.class, float.class, Float.class

setReferenceNetwork

public void setReferenceNetwork(BayesNet net)

getReferenceNetwork

public BayesNet getReferenceNetwork()

setUseLogScale

public void setUseLogScale(boolean useLogScale)

using logarithmic values for computation is a lot slower, but also a lot more numerically stable

create

public AbstractFactor create(java.util.List<java.lang.Integer> vars,
                             java.util.List<AbstractFactor> multiplicationPartners)

creates a factor, the class of which is dependent on different criteria defined in the concrete subclasses. The default behavior is to return a DenseFactor.

Parameters:

vars -

multiplicationPartners -

Returns:

defaultFactory

public static FactorFactory defaultFactory()