gr.demokritos.iit.jinsect.algorithms.clustering
Class AverageLinkClusterer

java.lang.Object
  gr.demokritos.iit.jinsect.algorithms.clustering.SingleLinkClusterer
      gr.demokritos.iit.jinsect.algorithms.clustering.AverageLinkClusterer

All Implemented Interfaces:

IClusterer

public class AverageLinkClusterer
extends SingleLinkClusterer

A generic use class that performs AverageLinkClustering. More in

Field Summary

 
Fields inherited from class gr.demokritos.iit.jinsect.algorithms.clustering.SingleLinkClusterer
CLUSTER_NAME_SEPARATOR
 



Constructor Summary
AverageLinkClusterer()



          Creates a new instance of AverageLinkClusterer.
 



Method Summary




protected  double
getSimilarityBetweenClusters(java.util.Set sClusterOne,
                             java.util.Set sClusterTwo,
                             SimilarityComparatorListener clDistanceCalculator)



          Calculates the similarity between two clusters.


 
Methods inherited from class gr.demokritos.iit.jinsect.algorithms.clustering.SingleLinkClusterer
calculateClusters, getHierarchy, main
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 





Constructor Detail



AverageLinkClusterer
public AverageLinkClusterer()

Creates a new instance of AverageLinkClusterer.






Method Detail



getSimilarityBetweenClusters
protected double getSimilarityBetweenClusters(java.util.Set sClusterOne,
                                              java.util.Set sClusterTwo,
                                              SimilarityComparatorListener clDistanceCalculator)

Calculates the similarity between two clusters. In this algorithm the 
 average similarity between all pairs of the two clusters is used.


Overrides:
getSimilarityBetweenClusters in class SingleLinkClusterer


Parameters:
sClusterOne - The first cluster.
sClusterTwo - The second cluster.
clDistanceCalculator - The calculator of distance between set elements.
Returns:
The similarity between the clusters.