edu.ucla.sspace.clustering.criterion

Class E1Function

java.lang.Object

edu.ucla.sspace.clustering.criterion.BaseFunction

edu.ucla.sspace.clustering.criterion.E1Function

All Implemented Interfaces:

CriterionFunction

public class E1Function
extends BaseFunction

This CriterionFunction measures the external differences between clusters. It gives a better score to clustering solutions with centroids that are more distant from the centroid for the data set as a whole.

Author:

Keith Stevens

Field Summary

Fields inherited from class edu.ucla.sspace.clustering.criterion.BaseFunction

assignments, centroids, clusterSizes, costs, matrix

Constructor Summary

Constructors

Constructor and Description
E1Function() Constructs a new E1Function.

Method Summary

Methods

Modifier and Type	Method and Description
protected double	getNewCentroidScore(int newCentroidIndex, DoubleVector dataPoint) Returns the new score for the cluster centroid indexed by newCentroidIndex when dataPoint is added to it.
protected double	getOldCentroidScore(DoubleVector vector, int oldCentroidIndex, int altClusterSize) Returns the new score for the cluster centroid represented by altCurrentCentroid with the new altClusterSize.
boolean	isMaximize() Returns true if this CriterionFunction tries to maximize it's score, and false otherwise.
protected void	subSetup(Matrix m) Setup any extra information needed before computing the cost values for each cluster.
protected void	updateScores(int newCentroidIndex, int oldCentroidIndex, DoubleVector vector)

Methods inherited from class edu.ucla.sspace.clustering.criterion.BaseFunction

assignments, centroids, clusterSizes, modifiedMagnitude, modifiedMagnitudeSqrd, score, setup, subtract, subtractedMagnitude, subtractedMagnitudeSqrd, update

Methods inherited from class java.lang.Object

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

Constructor Detail

E1Function

public E1Function()

Constructs a new E1Function.

Method Detail

subSetup

protected void subSetup(Matrix m)

Setup any extra information needed before computing the cost values for each cluster.

Overrides:

subSetup in class BaseFunction

getOldCentroidScore

protected double getOldCentroidScore(DoubleVector vector,
                         int oldCentroidIndex,
                         int altClusterSize)

Returns the new score for the cluster centroid represented by altCurrentCentroid with the new altClusterSize.

Specified by:

getOldCentroidScore in class BaseFunction

altClusterSize - The current updated cluster size

getNewCentroidScore

protected double getNewCentroidScore(int newCentroidIndex,
                         DoubleVector dataPoint)

Returns the new score for the cluster centroid indexed by newCentroidIndex when dataPoint is added to it. Implementations of this method should not actually add dataPoint to the centroid, but should instead use the helper functions provided to compute the new score.

Specified by:

getNewCentroidScore in class BaseFunction

Parameters:

newCentroidIndex - The index of the current alternate centroid

dataPoint - The current data point that is being reassigned

isMaximize

public boolean isMaximize()

Returns true if this CriterionFunction tries to maximize it's score, and false otherwise.

updateScores

protected void updateScores(int newCentroidIndex,
                int oldCentroidIndex,
                DoubleVector vector)

Overrides:

updateScores in class BaseFunction