idl.tmt.representation
Class TmtClustering

java.lang.Object
  |
  +--idl.tmt.representation.TmtClustering

public class TmtClustering
extends java.lang.Object

Builds, saves, reads from disk and provides evaluation interface for a document clustering

Author:
miles

Field Summary
(package private)  int[] clusterAssignments
           
(package private)  java.lang.String collectionName
           
(package private)  Instances data
           
(package private)  int maxClusters
           
(package private)  int maxIterations
           
(package private)  int minClusters
           
(package private)  EM myClusterer
           
(package private)  IndexedCollection myCollection
           
 
Constructor Summary
TmtClustering(java.lang.String colName, IndexedCollection col)
           
 
Method Summary
 void buildClusters()
           
 int getHardClusterForDoc(int docID)
           
 int getNumClusters()
           
 double getPrior(int clusterID)
           
 int[] hardCluster(boolean writeClusters)
           
 int[] readHardClusters()
           
 EM readSerializedModel()
           
 void serializeModel()
           
 void setMaxIterations(int m)
           
 void setMaxNumClusters(int maxClusters)
           
 void setMinNumClusters(int minClusters)
           
 void setMinStdDev(double s)
           
 void setNumClusters(int k)
           
 double[][] softCluster()
           
 Instances SoftClustersInWekaForm()
           
 int TmtClusterInstance(Instance instance)
           
 
Methods inherited from class java.lang.Object
, clone, equals, finalize, getClass, hashCode, notify, notifyAll, registerNatives, toString, wait, wait, wait
 

Field Detail

myCollection

IndexedCollection myCollection

data

Instances data

myClusterer

EM myClusterer

clusterAssignments

int[] clusterAssignments

collectionName

java.lang.String collectionName

maxIterations

int maxIterations

minClusters

int minClusters

maxClusters

int maxClusters
Constructor Detail

TmtClustering

public TmtClustering(java.lang.String colName,
                     IndexedCollection col)
Method Detail

setNumClusters

public void setNumClusters(int k)

getNumClusters

public int getNumClusters()

setMinNumClusters

public void setMinNumClusters(int minClusters)

setMaxNumClusters

public void setMaxNumClusters(int maxClusters)

setMinStdDev

public void setMinStdDev(double s)

setMaxIterations

public void setMaxIterations(int m)

buildClusters

public void buildClusters()

hardCluster

public int[] hardCluster(boolean writeClusters)

TmtClusterInstance

public int TmtClusterInstance(Instance instance)

getHardClusterForDoc

public int getHardClusterForDoc(int docID)

softCluster

public double[][] softCluster()

SoftClustersInWekaForm

public Instances SoftClustersInWekaForm()

serializeModel

public void serializeModel()

readSerializedModel

public EM readSerializedModel()

readHardClusters

public int[] readHardClusters()

getPrior

public double getPrior(int clusterID)