de.lmu.ifi.dbs.elki.algorithm.clustering.kmeans

Class KMeansCompare.Parameterizer<V extends NumberVector>

java.lang.Object

de.lmu.ifi.dbs.elki.utilities.optionhandling.AbstractParameterizer

de.lmu.ifi.dbs.elki.algorithm.AbstractNumberVectorDistanceBasedAlgorithm.Parameterizer<V>

de.lmu.ifi.dbs.elki.algorithm.clustering.kmeans.AbstractKMeans.Parameterizer<V>

de.lmu.ifi.dbs.elki.algorithm.clustering.kmeans.KMeansCompare.Parameterizer<V>

All Implemented Interfaces:

Parameterizer

Enclosing class:

KMeansCompare<V extends NumberVector>

public static class KMeansCompare.Parameterizer<V extends NumberVector>
extends AbstractKMeans.Parameterizer<V>

Parameterization class.

Author:

Erich Schubert

Field Summary

Fields inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.kmeans.AbstractKMeans.Parameterizer

initializer, k, maxiter, varstat

Fields inherited from class de.lmu.ifi.dbs.elki.algorithm.AbstractNumberVectorDistanceBasedAlgorithm.Parameterizer

distanceFunction

Constructor Summary

Constructors

Constructor and Description
Parameterizer()

Method Summary

Modifier and Type	Method and Description
protected KMeansCompare<V>	makeInstance() Make an instance after successful configuration.
protected boolean	needsMetric() Users could use other non-metric distances at their own risk; but some k-means variants make explicit use of the triangle inequality, we emit extra warnings then.

Methods inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.kmeans.AbstractKMeans.Parameterizer

getParameterDistanceFunction, getParameterInitialization, getParameterK, getParameterMaxIter, getParameterVarstat, makeOptions

Methods inherited from class de.lmu.ifi.dbs.elki.utilities.optionhandling.AbstractParameterizer

configure, make

Methods inherited from class java.lang.Object

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

Constructor Detail

Parameterizer

public Parameterizer()

Method Detail

needsMetric

protected boolean needsMetric()

Description copied from class: AbstractKMeans.Parameterizer

Users could use other non-metric distances at their own risk; but some k-means variants make explicit use of the triangle inequality, we emit extra warnings then.

Overrides:

needsMetric in class AbstractKMeans.Parameterizer<V extends NumberVector>

Returns:

true if the algorithm uses triangle inequality

makeInstance

protected KMeansCompare<V> makeInstance()

Description copied from class: AbstractParameterizer

Make an instance after successful configuration.

Specified by:

makeInstance in class AbstractKMeans.Parameterizer<V extends NumberVector>

Returns:

instance