cc.mallet.topics
Class TopicModelDiagnostics

java.lang.Object
  cc.mallet.topics.TopicModelDiagnostics

public class TopicModelDiagnostics
extends java.lang.Object

Nested Class Summary

class

TopicModelDiagnostics.TopicScores

Field Summary

static double[]	DEFAULT_DOC_PROPORTIONS
static int	FIFTY_PERCENT_INDEX
static int	TWO_PERCENT_INDEX

Constructor Summary

TopicModelDiagnostics(ParallelTopicModel model, int numTopWords)

Method Summary

void	collectDocumentStatistics()
int[][]	getCodocumentMatrix(int topic)
TopicModelDiagnostics.TopicScores	getCoherence()
TopicModelDiagnostics.TopicScores	getDistanceFromCorpus() Low-quality topics may be very similar to the global distribution.
TopicModelDiagnostics.TopicScores	getDistanceFromUniform()
TopicModelDiagnostics.TopicScores	getDocumentEntropy(int[] tokensPerTopic)
TopicModelDiagnostics.TopicScores	getDocumentPercent(int i)
TopicModelDiagnostics.TopicScores	getDocumentPercentRatio(int numeratorIndex, int denominatorIndex)
TopicModelDiagnostics.TopicScores	getEffectiveNumberOfWords()
TopicModelDiagnostics.TopicScores	getRank1Percent()
TopicModelDiagnostics.TopicScores	getTokenDocumentDiscrepancies()
TopicModelDiagnostics.TopicScores	getTokensPerTopic(int[] tokensPerTopic)
TopicModelDiagnostics.TopicScores	getWordLengthScores() Low-quality topics often have lots of unusually short words.
TopicModelDiagnostics.TopicScores	getWordLengthStandardDeviation() Low-quality topics often have lots of unusually short words.
static void	main(java.lang.String[] args)
java.lang.String	toString()
java.lang.String	toXML()

Methods inherited from class java.lang.Object

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

Field Detail

TWO_PERCENT_INDEX

public static final int TWO_PERCENT_INDEX

See Also:

Constant Field Values

FIFTY_PERCENT_INDEX

public static final int FIFTY_PERCENT_INDEX

See Also:

Constant Field Values

DEFAULT_DOC_PROPORTIONS

public static final double[] DEFAULT_DOC_PROPORTIONS

Constructor Detail

TopicModelDiagnostics

public TopicModelDiagnostics(ParallelTopicModel model,
                             int numTopWords)

Method Detail

collectDocumentStatistics

public void collectDocumentStatistics()

getCodocumentMatrix

public int[][] getCodocumentMatrix(int topic)

getTokensPerTopic

public TopicModelDiagnostics.TopicScores getTokensPerTopic(int[] tokensPerTopic)

getDocumentEntropy

public TopicModelDiagnostics.TopicScores getDocumentEntropy(int[] tokensPerTopic)

getDistanceFromUniform

public TopicModelDiagnostics.TopicScores getDistanceFromUniform()

getEffectiveNumberOfWords

public TopicModelDiagnostics.TopicScores getEffectiveNumberOfWords()

getDistanceFromCorpus

public TopicModelDiagnostics.TopicScores getDistanceFromCorpus()

Low-quality topics may be very similar to the global distribution.

getTokenDocumentDiscrepancies

public TopicModelDiagnostics.TopicScores getTokenDocumentDiscrepancies()

getWordLengthScores

public TopicModelDiagnostics.TopicScores getWordLengthScores()

Low-quality topics often have lots of unusually short words.

getWordLengthStandardDeviation

public TopicModelDiagnostics.TopicScores getWordLengthStandardDeviation()

Low-quality topics often have lots of unusually short words.

getCoherence

public TopicModelDiagnostics.TopicScores getCoherence()

getRank1Percent

public TopicModelDiagnostics.TopicScores getRank1Percent()

getDocumentPercentRatio

public TopicModelDiagnostics.TopicScores getDocumentPercentRatio(int numeratorIndex,
                                                                 int denominatorIndex)

getDocumentPercent

public TopicModelDiagnostics.TopicScores getDocumentPercent(int i)

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object

toXML

public java.lang.String toXML()

main

public static void main(java.lang.String[] args)
                 throws java.lang.Exception

Throws:

java.lang.Exception