Package dev.brachtendorf.clustering

Class KMeans

java.lang.Object

dev.brachtendorf.clustering.KMeans

All Implemented Interfaces:

ClusterAlgorithm

Direct Known Subclasses:

KMeansPlusPlus

public class KMeans
extends Object
implements ClusterAlgorithm

Author:

Kilian

Field Summary

Fields

Modifier and Type	Field	Description
protected DistanceFunction	distanceFunction	Function to calculate distance between individual data points
protected int	k	The number of cluster the data will be partitioned into
protected int	lastIterationCount

Constructor Summary

Constructors

Constructor	Description
KMeans(int clusters)	Create a KMeans clusterer with k clusters and EuclideanDistance.
KMeans(int clusters, DistanceFunction distanceFunction)	Create a KMeans clusterer

Method Summary

Modifier and Type	Method	Description
ClusterResult	cluster(double[][] data)
protected void	computeKMeans(DoubleSummaryStatistics[][] clusterMeans, double[][] data, int[] cluster, int dataDimension)
protected DoubleSummaryStatistics[][]	computeStartingClusters(double[][] data, int k, int dataDimension)
int	iterations()

Methods inherited from class java.lang.Object

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

Field Detail

k

protected int k

The number of cluster the data will be partitioned into

distanceFunction

protected DistanceFunction distanceFunction

Function to calculate distance between individual data points

lastIterationCount

protected int lastIterationCount

Constructor Detail

KMeans

public KMeans(int clusters)

Create a KMeans clusterer with k clusters and EuclideanDistance.

Parameters:

clusters - the number of cluster to partition the data into

KMeans

public KMeans(int clusters,
              DistanceFunction distanceFunction)

Create a KMeans clusterer

Parameters:

clusters - the number of cluster to partition the data into

distanceFunction - the distanceFunction used to compute the distance between data points

Method Detail

cluster

public ClusterResult cluster(double[][] data)

Specified by:

cluster in interface ClusterAlgorithm

computeStartingClusters

protected DoubleSummaryStatistics[][] computeStartingClusters(double[][] data,
                                                              int k,
                                                              int dataDimension)

computeKMeans

protected void computeKMeans(DoubleSummaryStatistics[][] clusterMeans,
                             double[][] data,
                             int[] cluster,
                             int dataDimension)

iterations

public int iterations()

Returns:

the number of iterations used to cluster the data