ai.libs.jaicore.ml.core.dataset.sampling.inmemory.factories

Class KmeansSamplingFactory<I extends INumericLabeledAttributeArrayInstance<? extends java.lang.Number>,D extends IDataset<I>>

java.lang.Object

ai.libs.jaicore.ml.core.dataset.sampling.inmemory.factories.KmeansSamplingFactory<I,D>

All Implemented Interfaces:

IRerunnableSamplingAlgorithmFactory<D,KmeansSampling<I,D>>, ISamplingAlgorithmFactory<D,KmeansSampling<I,D>>

public class KmeansSamplingFactory<I extends INumericLabeledAttributeArrayInstance<? extends java.lang.Number>,D extends IDataset<I>>
extends java.lang.Object
implements IRerunnableSamplingAlgorithmFactory<D,KmeansSampling<I,D>>

Constructor Summary

Constructors

Constructor and Description
KmeansSamplingFactory()

Method Summary

Modifier and Type	Method and Description
KmeansSampling<I,D>	getAlgorithm(int sampleSize, D inputDataset, java.util.Random random) After the necessary config is done, this method returns a fully configured instance of a sampling algorithm.
void	setClusterSeed(long clusterSeed) Set the seed the clustering will use for initialization.
void	setDistanceMeassure(org.apache.commons.math3.ml.distance.DistanceMeasure distanceMeassure) Set the distance measure for the clustering.
void	setK(int k) Set how many clusters shall be created.
void	setPreviousRun(KmeansSampling<I,D> previousRun) Set the previous run of the sampling algorithm, if one occurred, can be set here to get data from it.

Methods inherited from class java.lang.Object

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

Constructor Detail

KmeansSamplingFactory

public KmeansSamplingFactory()

Method Detail

setPreviousRun

public void setPreviousRun(KmeansSampling<I,D> previousRun)

Description copied from interface: IRerunnableSamplingAlgorithmFactory

Set the previous run of the sampling algorithm, if one occurred, can be set here to get data from it.

Specified by:

setPreviousRun in interface IRerunnableSamplingAlgorithmFactory<D extends IDataset<I>,KmeansSampling<I extends INumericLabeledAttributeArrayInstance<? extends java.lang.Number>,D extends IDataset<I>>>

Parameters:

previousRun - Algorithm object of the previous of the sampling algorithm.

setK

public void setK(int k)

Set how many clusters shall be created. Default is the sample size;

Parameters:

k - Parameter k of k-means.

setClusterSeed

public void setClusterSeed(long clusterSeed)

Set the seed the clustering will use for initialization. Default is without a fix seed and the system time instead.

Parameters:

clusterSeed -

setDistanceMeassure

public void setDistanceMeassure(org.apache.commons.math3.ml.distance.DistanceMeasure distanceMeassure)

Set the distance measure for the clustering. Default is the Manhattan distance.

Parameters:

distanceMeassure -

getAlgorithm

public KmeansSampling<I,D> getAlgorithm(int sampleSize,
                                        D inputDataset,
                                        java.util.Random random)

Description copied from interface: ISamplingAlgorithmFactory

After the necessary config is done, this method returns a fully configured instance of a sampling algorithm.

Specified by:

getAlgorithm in interface ISamplingAlgorithmFactory<D extends IDataset<I>,KmeansSampling<I extends INumericLabeledAttributeArrayInstance<? extends java.lang.Number>,D extends IDataset<I>>>

Parameters:

sampleSize - Desired size of the sample that will be created.

inputDataset - Dataset where the sample will be drawn from.

random - Random object to make samples reproducible.

Returns:

Configured sampling algorithm object.