jminhep.algorithms

Class KMeansAlg

java.lang.Object

jminhep.algorithms.KMeansAlg

public class KMeansAlg
extends java.lang.Object

The standard K-means algorithm.

Constructor Summary

Constructors

Constructor and Description
KMeansAlg(DataHolder data) Initialize the standard K-means algorithm

Method Summary

Modifier and Type	Method and Description
void	delete() Clear
void	generateSeed() Generate random seeds for initial centroids
DataHolder	getCenters() Returns cluster centers
int	getClusters() Get number of clusters
double	getCompactness() This method returns the Compactness
int	getError() Return error
java.lang.String	getName() Name for this algorithm.
int[]	getNumberPoints() Returns the number of points in each cluster
double[][]	getSeed() Return seeds
DataHolder	getSeedHolder() Get positions of seeds
long	getSize() This method returns the estimated size (steps) for this task.
void	run() Run classic K-Means clustering algorithm.
void	run(int Iter) Runs the k-means several iterations to find best seed values Only iteration with the best compactness is taken.
void	runKM() Classic K-Means clustering algorithm.
void	setClusters(int numClusters) Set the designed number of clusters
void	setOptions(int maxIterations, double epsilon) Set initial options for calculations
void	setSeed(double[][] seed) Set seeds by hand

Methods inherited from class java.lang.Object

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

Constructor Detail

KMeansAlg

public KMeansAlg(DataHolder data)

Initialize the standard K-means algorithm

Parameters:

data - input data

Method Detail

setClusters

public void setClusters(int numClusters)

Set the designed number of clusters

Parameters:

numClusters - expected number of clusters

getClusters

public int getClusters()

Get number of clusters

Returns:

number of clusters

generateSeed

public void generateSeed()

Generate random seeds for initial centroids

setSeed

public void setSeed(double[][] seed)

Set seeds by hand

Parameters:

seed - positions of seeds

getSeed

public double[][] getSeed()

Return seeds

Returns:

seeds

setOptions

public void setOptions(int maxIterations,
                       double epsilon)

Set initial options for calculations

Parameters:

maxIterations - the maximum number of iterations.

epsilon - a small value used to verify if clustering has converged.

delete

public void delete()

Clear

getCenters

public DataHolder getCenters()

Returns cluster centers

Returns:

DataHolder with center positions

getSeedHolder

public DataHolder getSeedHolder()

Get positions of seeds

Returns:

seed positions

getNumberPoints

public int[] getNumberPoints()

Returns the number of points in each cluster

Returns:

int[] getNumberPoints

run

public void run()

Run classic K-Means clustering algorithm. Seeds will be generated automatically

runKM

public void runKM()

Classic K-Means clustering algorithm. Seeds are not set. Instead, consider the method run()

getCompactness

public double getCompactness()

This method returns the Compactness

Returns:

compactness value

run

public void run(int Iter)

Runs the k-means several iterations to find best seed values Only iteration with the best compactness is taken.

Parameters:

Iter - Number of iterations

getError

public int getError()

Return error

Returns:

zero if no errors

getSize

public long getSize()

This method returns the estimated size (steps) for this task.

getName

public java.lang.String getName()

Name for this algorithm.

Returns: