org.apache.mahout.clustering.fuzzykmeans
Class FuzzyKMeansClusterer

java.lang.Object
  org.apache.mahout.clustering.fuzzykmeans.FuzzyKMeansClusterer

public class FuzzyKMeansClusterer
extends Object

Constructor Summary

FuzzyKMeansClusterer()

FuzzyKMeansClusterer(org.apache.hadoop.conf.Configuration conf)

FuzzyKMeansClusterer(DistanceMeasure measure, double convergenceDelta, double m)
Init the fuzzy k-means clusterer with the distance measure to use for comparison.

Method Summary

protected void	addPointToClusters(List<SoftCluster> clusterList, Vector point)
static List<List<SoftCluster>>	clusterPoints(Iterable<Vector> points, List<SoftCluster> clusters, DistanceMeasure measure, double threshold, double m, int numIter) This is the reference k-means implementation.
boolean	computeConvergence(Cluster cluster) Return if the cluster is converged by comparing its center and centroid.
Vector	computePi(Collection<SoftCluster> clusters, List<Double> clusterDistanceList)
double	computeProbWeight(double clusterDistance, Iterable<Double> clusterDistanceList) Computes the probability of a point belonging to a cluster
void	emitPointProbToCluster(Vector point, List<SoftCluster> clusters, org.apache.hadoop.mapreduce.Mapper.Context context) Emit the point and its probability of belongingness to each cluster
void	emitPointToClusters(VectorWritable point, List<SoftCluster> clusters, org.apache.hadoop.mapreduce.Mapper.Context context)
void	emitPointToClusters(VectorWritable point, List<SoftCluster> clusters, org.apache.hadoop.io.SequenceFile.Writer writer)
double	getM()
DistanceMeasure	getMeasure()
protected static boolean	runFuzzyKMeansIteration(Iterable<Vector> points, List<SoftCluster> clusterList, FuzzyKMeansClusterer clusterer) Perform a single iteration over the points and clusters, assigning points to clusters and returning if the iterations are completed.
protected boolean	testConvergence(Iterable<SoftCluster> clusters)

Methods inherited from class java.lang.Object

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

Constructor Detail

FuzzyKMeansClusterer

public FuzzyKMeansClusterer(DistanceMeasure measure,
                            double convergenceDelta,
                            double m)

Init the fuzzy k-means clusterer with the distance measure to use for comparison.

FuzzyKMeansClusterer

public FuzzyKMeansClusterer(org.apache.hadoop.conf.Configuration conf)

FuzzyKMeansClusterer

public FuzzyKMeansClusterer()

Method Detail

clusterPoints

public static List<List<SoftCluster>> clusterPoints(Iterable<Vector> points,
                                                    List<SoftCluster> clusters,
                                                    DistanceMeasure measure,
                                                    double threshold,
                                                    double m,
                                                    int numIter)

This is the reference k-means implementation. Given its inputs it iterates over the points and clusters until their centers converge or until the maximum number of iterations is exceeded.

Parameters:

points - the input List of points

clusters - the initial List of clusters

measure - the DistanceMeasure to use

threshold - the double convergence threshold

m - the double "fuzzyness" argument (>1)

numIter - the maximum number of iterations

Returns:

a List> of clusters produced per iteration

runFuzzyKMeansIteration

protected static boolean runFuzzyKMeansIteration(Iterable<Vector> points,
                                                 List<SoftCluster> clusterList,
                                                 FuzzyKMeansClusterer clusterer)

Perform a single iteration over the points and clusters, assigning points to clusters and returning if the iterations are completed.

Parameters:

points - the List having the input points

clusterList - the List clusters

emitPointProbToCluster

public void emitPointProbToCluster(Vector point,
                                   List<SoftCluster> clusters,
                                   org.apache.hadoop.mapreduce.Mapper.Context context)
                            throws IOException,
                                   InterruptedException

Emit the point and its probability of belongingness to each cluster

Parameters:

point - a point

clusters - a List

context - the Context to emit into

Throws:

IOException

InterruptedException

computeProbWeight

public double computeProbWeight(double clusterDistance,
                                Iterable<Double> clusterDistanceList)

Computes the probability of a point belonging to a cluster

computeConvergence

public boolean computeConvergence(Cluster cluster)

Return if the cluster is converged by comparing its center and centroid.

Returns:

if the cluster is converged

getM

public double getM()

getMeasure

public DistanceMeasure getMeasure()

emitPointToClusters

public void emitPointToClusters(VectorWritable point,
                                List<SoftCluster> clusters,
                                org.apache.hadoop.mapreduce.Mapper.Context context)
                         throws IOException,
                                InterruptedException

Throws:

IOException

InterruptedException

computePi

public Vector computePi(Collection<SoftCluster> clusters,
                        List<Double> clusterDistanceList)

addPointToClusters

protected void addPointToClusters(List<SoftCluster> clusterList,
                                  Vector point)

testConvergence

protected boolean testConvergence(Iterable<SoftCluster> clusters)

emitPointToClusters

public void emitPointToClusters(VectorWritable point,
                                List<SoftCluster> clusters,
                                org.apache.hadoop.io.SequenceFile.Writer writer)
                         throws IOException

Throws:

IOException