net.sf.tweety.logics.pcl.analysis

Class MeanDistanceCulpabilityMeasure

java.lang.Object

net.sf.tweety.logics.pcl.analysis.MeanDistanceCulpabilityMeasure

All Implemented Interfaces:

CulpabilityMeasure<ProbabilisticConditional,PclBeliefSet>, SignedCulpabilityMeasure

public class MeanDistanceCulpabilityMeasure
extends java.lang.Object
implements SignedCulpabilityMeasure

This class implements the mean distance culpability measure, see [PhD thesis, Thimm].

Author:

Matthias Thimm TODO: this does not work correctly yet, fix it.

Field Summary

Fields

Modifier and Type	Field and Description
private boolean	normalized Whether this measure uses the normalized mindev measure.

Constructor Summary

Constructors

Constructor and Description
MeanDistanceCulpabilityMeasure(boolean normalized) Creates a new measure.

Method Summary

Modifier and Type	Method and Description
java.lang.Double	culpabilityMeasure(PclBeliefSet beliefSet, ProbabilisticConditional conditional) Returns the degree of responsibility of the given formula to cause inconsistency in the given belief set (NOTE: the formula should be in the given belief set).
private OptimizationProblem	getBaseProblem(PclBeliefSet beliefSet, ProbabilisticConditional conditional, double incVal) Creates the base problem (without target function set) for determining both minimal and maximal incon values.
private java.lang.Double	getMaximumValue(PclBeliefSet beliefSet, ProbabilisticConditional conditional, double incVal) Determines the maximal value for the given conditional.
private java.lang.Double	getMinimumValue(PclBeliefSet beliefSet, ProbabilisticConditional conditional, double incVal) Determines the minimal value for the given conditional.
java.lang.Double	sign(PclBeliefSet beliefSet, ProbabilisticConditional conditional) Determines the sign of the culpability of the given conditional in the given belief set, i.e.

Methods inherited from class java.lang.Object

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

Field Detail

normalized

private boolean normalized

Whether this measure uses the normalized mindev measure.

Constructor Detail

MeanDistanceCulpabilityMeasure

public MeanDistanceCulpabilityMeasure(boolean normalized)

Creates a new measure.

Parameters:

normalized - whether this measure uses the normalized mindev measure.

Method Detail

culpabilityMeasure

public java.lang.Double culpabilityMeasure(PclBeliefSet beliefSet,
                                           ProbabilisticConditional conditional)

Description copied from interface: CulpabilityMeasure

Returns the degree of responsibility of the given formula to cause inconsistency in the given belief set (NOTE: the formula should be in the given belief set).

Specified by:

culpabilityMeasure in interface CulpabilityMeasure<ProbabilisticConditional,PclBeliefSet>

Specified by:

culpabilityMeasure in interface SignedCulpabilityMeasure

Parameters:

beliefSet - a belief set.

Returns:

a Double indicating the degree of inconsistency (NOTE: if the given formula does not appear in the given belief set the degree is defined to be zero).

sign

public java.lang.Double sign(PclBeliefSet beliefSet,
                             ProbabilisticConditional conditional)

Description copied from interface: SignedCulpabilityMeasure

Determines the sign of the culpability of the given conditional in the given belief set, i.e. one of {-1,0,1}.

Specified by:

sign in interface SignedCulpabilityMeasure

Parameters:

beliefSet - a belief set

conditional - a conditional

Returns:

one of {-1,0,1}

getMinimumValue

private java.lang.Double getMinimumValue(PclBeliefSet beliefSet,
                                         ProbabilisticConditional conditional,
                                         double incVal)

Determines the minimal value for the given conditional.

Parameters:

beliefSet - a belief set

conditional - some conditional.

incVal - the value of the inconsistency

Returns:

a double

getMaximumValue

private java.lang.Double getMaximumValue(PclBeliefSet beliefSet,
                                         ProbabilisticConditional conditional,
                                         double incVal)

Determines the maximal value for the given conditional.

Parameters:

beliefSet - a belief set

conditional - some conditional.

incVal - the value of the inconsistency

Returns:

a double

getBaseProblem

private OptimizationProblem getBaseProblem(PclBeliefSet beliefSet,
                                           ProbabilisticConditional conditional,
                                           double incVal)

Creates the base problem (without target function set) for determining both minimal and maximal incon values.

Parameters:

beliefSet - a belief set

conditional - some conditional.

incVal - the value of the inconsistency

Returns:

the base optimization problem