org.apache.mahout.clustering.dirichlet.models
Class GaussianClusterDistribution

java.lang.Object
  extended by org.apache.mahout.clustering.dirichlet.models.AbstractVectorModelDistribution
      extended by org.apache.mahout.clustering.dirichlet.models.GaussianClusterDistribution
All Implemented Interfaces:
ModelDistribution<VectorWritable>

public class GaussianClusterDistribution
extends AbstractVectorModelDistribution

An implementation of the ModelDistribution interface suitable for testing the DirichletCluster algorithm. Uses a Normal Distribution to sample the prior model values. Model values have a vector standard deviation, allowing assymetrical regions to be covered by a model.


Field Summary
 
Fields inherited from class org.apache.mahout.clustering.dirichlet.models.AbstractVectorModelDistribution
MODEL_DISTRIBUTION_TYPE
 
Constructor Summary
GaussianClusterDistribution()
           
GaussianClusterDistribution(VectorWritable modelPrototype)
           
 
Method Summary
 Model<VectorWritable>[] sampleFromPosterior(Model<VectorWritable>[] posterior)
          Return a list of models sampled from the posterior
 Model<VectorWritable>[] sampleFromPrior(int howMany)
          Return a list of models sampled from the prior
 
Methods inherited from class org.apache.mahout.clustering.dirichlet.models.AbstractVectorModelDistribution
asJsonString, getModelPrototype, setModelPrototype
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

GaussianClusterDistribution

public GaussianClusterDistribution()

GaussianClusterDistribution

public GaussianClusterDistribution(VectorWritable modelPrototype)
Method Detail

sampleFromPrior

public Model<VectorWritable>[] sampleFromPrior(int howMany)
Description copied from interface: ModelDistribution
Return a list of models sampled from the prior

Parameters:
howMany - the int number of models to return
Returns:
a Model[] representing what is known apriori

sampleFromPosterior

public Model<VectorWritable>[] sampleFromPosterior(Model<VectorWritable>[] posterior)
Description copied from interface: ModelDistribution
Return a list of models sampled from the posterior

Parameters:
posterior - the Model[] after observations
Returns:
a Model[] representing what is known apriori


Copyright © 2008-2010 The Apache Software Foundation. All Rights Reserved.