br.otimizes.oplatool.core.learning

Class Clustering

java.lang.Object

br.otimizes.oplatool.core.learning.Clustering

All Implemented Interfaces:

java.io.Serializable

public class Clustering
extends java.lang.Object
implements java.io.Serializable

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
static org.apache.log4j.Logger	LOGGER

Constructor Summary

Constructors

Constructor and Description
Clustering()
Clustering(SolutionSet resultFront, ClusteringAlgorithm algorithm)

Method Summary

Modifier and Type	Method and Description
SolutionSet	dbscan() DBSCAN Execution Method - Observations: The only measure that changes the solution is as follows: getDBSCAN().setOptions(new String[]{"-A", "weka.core.ManhattanDistance"});
double	euclidianDistance(double[] doubles1, double[] doubles2)
double	euclidianDistance(Solution solution)
ClusteringAlgorithm	getAlgorithm()
java.util.List<Solution>	getAllSolutions()
ArffExecution	getArffExecution()
java.util.List<Solution>	getBestPerformingCluster()
weka.clusterers.AbstractClusterer	getClusterer()
weka.clusterers.ClusterEvaluation	getClusterEvaluation() Cluster Evaluation Object for analysis of br.otimizes.oplatool.core.jmetal4.results
weka.clusterers.DBSCAN	getDBSCAN() Cast the Clusterer to DBSCAN Object
weka.core.DistanceFunction	getDistanceFunction()
java.lang.Double	getEpsilon()
java.util.List<Solution>	getFilteredSolutions()
int	getGeneratedClusters()
java.util.List<java.lang.Integer>	getIdsFilteredSolutions()
double	getIndexToFilter() Get Index to Filter Set by default: 1 -> K-Means, 2 -> DBSCAN and OPTICS
weka.clusterers.SimpleKMeans	getKMeans() Cast the Clusterer to SimpleKMeans Object
double[]	getMax()
java.lang.Integer	getMaxIterations()
double[]	getMin()
double	getMinClusterByObjective(int objectiveIndex) If SimpleKMeans, returns the min centroid value by objective index, else, in case of DBSCAN that dont have Centroid values, is verified the min value in the solutions On DBSCAN, the clusterId -1 indicates noise
java.lang.Integer	getMinPoints()
int	getNumClusters() https://stats.stackexchange.com/questions/55215/way-to-determine-best-number-of-clusters-weka https://en.wikipedia.org/wiki/Determining_the_number_of_clusters_in_a_data_set
int	getNumObjectives()
weka.clusterers.OPTICS	getOPTICS() Cast the Clusterer to OPTICS Object
SolutionSet	getResultFront()
SilhouetteIndex	getSilhouetteIndex()
java.util.List<Solution>	getSolutionsByClusterId(double clusterId)
java.util.List<Solution>	getSolutionsByClusterWithMinObjective(int objectiveIndex)
SolutionSet	kMeans() K-Means Execution Method
java.lang.Double	np(java.lang.Integer num)
SolutionSet	optics() Method not completed, because in the current version of Weka, the OPTICS does not present br.otimizes.oplatool.core.jmetal4.results
SolutionSet	run() Execution Method
void	setAlgorithm(ClusteringAlgorithm algorithm)
void	setAllSolutions(java.util.List<Solution> allSolutions)
void	setArffExecution(ArffExecution arffExecution)
void	setClusterer(weka.clusterers.AbstractClusterer clusterer)
void	setDistanceFunction(weka.core.DistanceFunction distanceFunction)
void	setEpsilon(java.lang.Double epsilon)
void	setFilteredSolutions(java.util.List<Solution> filteredSolutions)
void	setIdsFilteredSolutions(java.util.List<java.lang.Integer> idsFilteredSolutions)
void	setIndexToFilter(double indexToFilter)
void	setMax(double[] max)
void	setMaxIterations(java.lang.Integer maxIterations)
void	setMin(double[] min)
void	setMinPoints(java.lang.Integer minPoints)
void	setNumClusters(java.lang.Integer numClusters)
void	setNumObjectives(int numObjectives)
void	setResultFront(SolutionSet resultFront)

Methods inherited from class java.lang.Object

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

Field Detail

LOGGER

public static final org.apache.log4j.Logger LOGGER

Constructor Detail

Clustering

public Clustering()

Clustering

public Clustering(SolutionSet resultFront,
                  ClusteringAlgorithm algorithm)

Method Detail

np

public java.lang.Double np(java.lang.Integer num)

run

public SolutionSet run()
                throws java.lang.Exception

Execution Method

Returns:

Solution Set - Best performing cluster with another solutions (filteredSolutions)

Throws:

java.lang.Exception - Default Exception

kMeans

public SolutionSet kMeans()
                   throws java.lang.Exception

K-Means Execution Method

Returns:

Solution Set

Throws:

java.lang.Exception - Default Exception

dbscan

public SolutionSet dbscan()
                   throws java.lang.Exception

DBSCAN Execution Method - Observations: The only measure that changes the solution is as follows: getDBSCAN().setOptions(new String[]{"-A", "weka.core.ManhattanDistance"});

Returns:

Solution Set

Throws:

java.lang.Exception

optics

public SolutionSet optics()
                   throws java.lang.Exception

Method not completed, because in the current version of Weka, the OPTICS does not present br.otimizes.oplatool.core.jmetal4.results

Returns:

Nothing

Throws:

java.lang.Exception - Default Exception

euclidianDistance

public double euclidianDistance(double[] doubles1,
                                double[] doubles2)

euclidianDistance

public double euclidianDistance(Solution solution)

getClusterEvaluation

public weka.clusterers.ClusterEvaluation getClusterEvaluation()
                                                       throws java.lang.Exception

Cluster Evaluation Object for analysis of br.otimizes.oplatool.core.jmetal4.results

Returns:

Clustes Evaluation Objetc

Throws:

java.lang.Exception - Default Exception

getSilhouetteIndex

public SilhouetteIndex getSilhouetteIndex()
                                   throws java.lang.Exception

Throws:

java.lang.Exception

getKMeans

public weka.clusterers.SimpleKMeans getKMeans()

Cast the Clusterer to SimpleKMeans Object

Returns:

SimpleKMeans Object

getDBSCAN

public weka.clusterers.DBSCAN getDBSCAN()

Cast the Clusterer to DBSCAN Object

Returns:

DBSCAN Object

getOPTICS

public weka.clusterers.OPTICS getOPTICS()

Cast the Clusterer to OPTICS Object

Returns:

OPTICS Object

getResultFront

public SolutionSet getResultFront()

setResultFront

public void setResultFront(SolutionSet resultFront)

getAlgorithm

public ClusteringAlgorithm getAlgorithm()

setAlgorithm

public void setAlgorithm(ClusteringAlgorithm algorithm)

getClusterer

public weka.clusterers.AbstractClusterer getClusterer()

setClusterer

public void setClusterer(weka.clusterers.AbstractClusterer clusterer)

getArffExecution

public ArffExecution getArffExecution()

setArffExecution

public void setArffExecution(ArffExecution arffExecution)

getNumClusters

public int getNumClusters()

https://stats.stackexchange.com/questions/55215/way-to-determine-best-number-of-clusters-weka https://en.wikipedia.org/wiki/Determining_the_number_of_clusters_in_a_data_set

Returns:

number of clusters

getGeneratedClusters

public int getGeneratedClusters()
                         throws java.lang.Exception

Throws:

java.lang.Exception

setNumClusters

public void setNumClusters(java.lang.Integer numClusters)

getMinPoints

public java.lang.Integer getMinPoints()

setMinPoints

public void setMinPoints(java.lang.Integer minPoints)

getEpsilon

public java.lang.Double getEpsilon()

setEpsilon

public void setEpsilon(java.lang.Double epsilon)

getFilteredSolutions

public java.util.List<Solution> getFilteredSolutions()

getBestPerformingCluster

public java.util.List<Solution> getBestPerformingCluster()

setFilteredSolutions

public void setFilteredSolutions(java.util.List<Solution> filteredSolutions)

getIndexToFilter

public double getIndexToFilter()

Get Index to Filter Set by default: 1 -> K-Means, 2 -> DBSCAN and OPTICS

Returns:

Index to Filter Value

setIndexToFilter

public void setIndexToFilter(double indexToFilter)

getIdsFilteredSolutions

public java.util.List<java.lang.Integer> getIdsFilteredSolutions()

setIdsFilteredSolutions

public void setIdsFilteredSolutions(java.util.List<java.lang.Integer> idsFilteredSolutions)

getDistanceFunction

public weka.core.DistanceFunction getDistanceFunction()

setDistanceFunction

public void setDistanceFunction(weka.core.DistanceFunction distanceFunction)

getAllSolutions

public java.util.List<Solution> getAllSolutions()

setAllSolutions

public void setAllSolutions(java.util.List<Solution> allSolutions)

getSolutionsByClusterWithMinObjective

public java.util.List<Solution> getSolutionsByClusterWithMinObjective(int objectiveIndex)

getSolutionsByClusterId

public java.util.List<Solution> getSolutionsByClusterId(double clusterId)

getMinClusterByObjective

public double getMinClusterByObjective(int objectiveIndex)

If SimpleKMeans, returns the min centroid value by objective index, else, in case of DBSCAN that dont have Centroid values, is verified the min value in the solutions On DBSCAN, the clusterId -1 indicates noise

Parameters:

objectiveIndex - Objective array index

Returns:

Min value by objective

getNumObjectives

public int getNumObjectives()

setNumObjectives

public void setNumObjectives(int numObjectives)

getMin

public double[] getMin()

setMin

public void setMin(double[] min)

getMax

public double[] getMax()

setMax

public void setMax(double[] max)

getMaxIterations

public java.lang.Integer getMaxIterations()

setMaxIterations

public void setMaxIterations(java.lang.Integer maxIterations)