org.graphstream.algorithm.measure
Class NormalizedMutualInformation

java.lang.Object
  org.graphstream.stream.SinkAdapter
      org.graphstream.algorithm.measure.CommunityMeasure
          org.graphstream.algorithm.measure.CommunityRelativeMeasure
              org.graphstream.algorithm.measure.NormalizedMutualInformation

All Implemented Interfaces:

Algorithm, DynamicAlgorithm, org.graphstream.stream.AttributeSink, org.graphstream.stream.ElementSink, org.graphstream.stream.Sink

Direct Known Subclasses:

VariationOfInformation

public class NormalizedMutualInformation
extends CommunityRelativeMeasure

Computes and updated the current Normalized Mutual Information (NMI) measure between a dynamically-performed community assignment on a graph as it evolves and a fixed assignment, known as reference.

Author:

Guillaume-Jean Herbiet

Scientific Reference :

L. Danon, A. Diaz-Guilera, J. Duch, and A. Arenas, “Comparing community structure identification,” Journal of Statistical Mechanics: Theory and Experiment, vol. 2005, no. 09, pp. P09008+, September 2005.

Constructor Summary

NormalizedMutualInformation(String marker)
New NMI measure, using the given marker for the dynamically performed assignment, and the default marker for the reference assignment.

NormalizedMutualInformation(String marker, String referenceMarker)
New NMI measure, using the given marker for the dynamically performed assignment, and the given refrenceMarker for the reference assignment.

Method Summary

void	compute() Compute the new NMI measure value.

Methods inherited from class org.graphstream.algorithm.measure.CommunityRelativeMeasure

nodeAdded, nodeAttributeChanged, nodeRemoved

Methods inherited from class org.graphstream.algorithm.measure.CommunityMeasure

edgeAdded, edgeRemoved, getLastComputedValue, getMeasure, graphCleared, init, nodeAttributeAdded, terminate

Methods inherited from class org.graphstream.stream.SinkAdapter

edgeAttributeAdded, edgeAttributeChanged, edgeAttributeRemoved, graphAttributeAdded, graphAttributeChanged, graphAttributeRemoved, nodeAttributeRemoved, stepBegins

Methods inherited from class java.lang.Object

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

Constructor Detail

NormalizedMutualInformation

public NormalizedMutualInformation(String marker)

New NMI measure, using the given marker for the dynamically performed assignment, and the default marker for the reference assignment.

Parameters:

marker - name of the attribute marking the computed communities.

NormalizedMutualInformation

public NormalizedMutualInformation(String marker,
                                   String referenceMarker)

New NMI measure, using the given marker for the dynamically performed assignment, and the given refrenceMarker for the reference assignment.

Parameters:

marker - name of the attribute marking the computed communities.

referenceMarker - name of the attribute marking the reference communities.

Method Detail

compute

public void compute()

Compute the new NMI measure value.

Specified by:

compute in interface Algorithm

Specified by:

compute in class CommunityRelativeMeasure

See Also:

Algorithm.init(Graph)

Computational Complexity :

O(2*C^2 + 6*C), where C is the expected number of communities in the graph.