org.apache.mahout.clustering.lda
Class LDASampler
java.lang.Object
org.apache.mahout.clustering.lda.LDASampler
public class LDASampler
- extends Object
Takes in a Matrix
of topic distributions (such as generated by LDADriver
,
CVB0Driver
or
InMemoryCollapsedVariationalBayes0
, and constructs
a set of samplers over this distribution, which may be sampled from by providing a distribution
over topics, and a number of samples desired
Method Summary |
int[] |
sample(Vector topicDistribution,
int numSamples)
|
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
LDASampler
public LDASampler(Matrix model,
Random random)
sample
public int[] sample(Vector topicDistribution,
int numSamples)
- Parameters:
topicDistribution
- vector of p(topicId) for all topicId < model.numTopics()numSamples
- the number of times to sample (with replacement) from the model
- Returns:
- array of length numSamples, with each entry being a sample from the model. There
may be repeats
Copyright © 2008-2012 The Apache Software Foundation. All Rights Reserved.