Package weka.classifiers.lazy.AM.data

Class AMResults

java.lang.Object

weka.classifiers.lazy.AM.data.AMResults

public class AMResults
extends java.lang.Object

The results of running AM, containing the analogical effects of the individual training instances as well as the relevant supracontexts and overall class likelihoods.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	AMResults.Judgement
static class	AMResults.PointerCountingStrategry

Constructor Summary

Constructors

Constructor	Description
AMResults(Lattice lattice, SubcontextList subList, weka.core.Instance testItem, boolean linear, Labeler labeler)

Method Summary

Modifier and Type	Method	Description
weka.core.Instance	getClassifiedEx()
java.util.Map<java.lang.String,java.math.BigDecimal>	getClassLikelihood()
java.util.Map<java.lang.String,java.math.BigInteger>	getClassPointers()
java.math.BigDecimal	getClassProbability()
java.util.Map<weka.core.Instance,java.math.BigDecimal>	getExemplarEffectMap()
java.util.Map<weka.core.Instance,java.math.BigInteger>	getExemplarPointers()
java.lang.String	getExpectedClassName()
java.util.List<GangEffect>	getGangEffects()
AMResults.Judgement	getJudgement()
Labeler	getLabeler()
AMResults.PointerCountingStrategry	getPointerCountingStrategy()
java.util.Set<java.lang.String>	getPredictedClasses()
java.util.Set<Subcontext>	getSubcontexts()
SubcontextList	getSubList()
java.util.Set<Supracontext>	getSupraList()
java.math.BigInteger	getTotalPointers()
java.lang.String	toString()

Methods inherited from class java.lang.Object

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

Constructor Detail

AMResults

public AMResults(Lattice lattice,
                 SubcontextList subList,
                 weka.core.Instance testItem,
                 boolean linear,
                 Labeler labeler)

Parameters:

lattice - filled lattice, which contains the data for calculating the analogical set

testItem - Exemplar being classified

linear - True if counting of pointers should be done linearly; false if quadratically.

labeler - The labeler that was used to assign contextual labels; this is made available for printing purposes.

Method Detail

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object

getExemplarEffectMap

public java.util.Map<weka.core.Instance,java.math.BigDecimal> getExemplarEffectMap()

Returns:

A mapping between exemplars and their analogical effect (decimal percentage)

getExemplarPointers

public java.util.Map<weka.core.Instance,java.math.BigInteger> getExemplarPointers()

Returns:

Mapping of exemplars in the analogical set to the number of pointers to it

getTotalPointers

public java.math.BigInteger getTotalPointers()

Returns:

The total number of pointers in this analogical set

getClassPointers

public java.util.Map<java.lang.String,java.math.BigInteger> getClassPointers()

Returns:

A mapping between a class value index the number of pointers pointing to it

getClassLikelihood

public java.util.Map<java.lang.String,java.math.BigDecimal> getClassLikelihood()

Returns:

A mapping between the class name and its selection probability

getClassifiedEx

public weka.core.Instance getClassifiedEx()

Returns:

The exemplar which was classified

getClassProbability

public java.math.BigDecimal getClassProbability()

Returns:

Probability of the predicted class

getPredictedClasses

public java.util.Set<java.lang.String> getPredictedClasses()

Returns:

Index of the predicted class attribute value

getSupraList

public java.util.Set<Supracontext> getSupraList()

Returns:

The Supracontexts that comprise the analogical set.

getSubcontexts

public java.util.Set<Subcontext> getSubcontexts()

Returns:

All subcontexts contained in all of the supracntexts of the analogical set.

getGangEffects

public java.util.List<GangEffect> getGangEffects()

Returns:

The gang effects, sorted by size of the effect and then alphabetically by the subcontext display label

getLabeler

public Labeler getLabeler()

Returns:

The Labeler object that was used to assign all of the contextual labels.

getExpectedClassName

public java.lang.String getExpectedClassName()

getSubList

public SubcontextList getSubList()

getJudgement

public AMResults.Judgement getJudgement()

Returns:

a judgement of the prediction

getPointerCountingStrategy

public AMResults.PointerCountingStrategry getPointerCountingStrategy()