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

Class FarthestPointsInitialMeans<O>

java.lang.Object

de.lmu.ifi.dbs.elki.algorithm.clustering.kmeans.initialization.AbstractKMeansInitialization<NumberVector>

de.lmu.ifi.dbs.elki.algorithm.clustering.kmeans.initialization.FarthestPointsInitialMeans<O>

Type Parameters:

O - Object type for kMedoids and kMedians

All Implemented Interfaces:

KMeansInitialization<NumberVector>, KMedoidsInitialization<O>

Direct Known Subclasses:

FarthestSumPointsInitialMeans

public class FarthestPointsInitialMeans<O>
extends AbstractKMeansInitialization<NumberVector>
implements KMedoidsInitialization<O>

K-Means initialization by repeatedly choosing the farthest point (by the minimum distance to earlier points). Note: this is less random than other initializations, so running multiple times will be more likely to return the same local minima.

Author:

Erich Schubert

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	FarthestPointsInitialMeans.Parameterizer<O> Parameterization class.

Field Summary

Fields

Modifier and Type	Field and Description
(package private) boolean	dropfirst Discard the first vector.

Fields inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.kmeans.initialization.AbstractKMeansInitialization

rnd

Constructor Summary

Constructors

Constructor and Description
FarthestPointsInitialMeans(RandomFactory rnd, boolean dropfirst) Constructor.

Method Summary

Methods

Modifier and Type	Method and Description
<T extends NumberVector,V extends NumberVector> List<V>	chooseInitialMeans(Database database, Relation<T> relation, int k, NumberVectorDistanceFunction<? super T> distanceFunction, NumberVector.Factory<V> factory) Choose initial means
DBIDs	chooseInitialMedoids(int k, DBIDs ids, DistanceQuery<? super O> distQ) Choose initial means

Methods inherited from class java.lang.Object

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

Field Detail

dropfirst

boolean dropfirst

Discard the first vector.

Constructor Detail

FarthestPointsInitialMeans

public FarthestPointsInitialMeans(RandomFactory rnd,
                          boolean dropfirst)

Constructor.

Parameters:

rnd - Random generator.

dropfirst - Flag to discard the first vector.

Method Detail

chooseInitialMeans

public <T extends NumberVector,V extends NumberVector> List<V> chooseInitialMeans(Database database,
                                                                         Relation<T> relation,
                                                                         int k,
                                                                         NumberVectorDistanceFunction<? super T> distanceFunction,
                                                                         NumberVector.Factory<V> factory)

Description copied from interface: KMeansInitialization

Choose initial means

Specified by:

chooseInitialMeans in interface KMeansInitialization<NumberVector>

Type Parameters:

T - Input vector type

V - Output vector type

Parameters:

database - Database context

relation - Relation

k - Parameter k

distanceFunction - Distance function

factory - Factory for output vectors.

Returns:

List of chosen means for k-means

chooseInitialMedoids

public DBIDs chooseInitialMedoids(int k,
                         DBIDs ids,
                         DistanceQuery<? super O> distQ)

Description copied from interface: KMedoidsInitialization

Choose initial means

Specified by:

chooseInitialMedoids in interface KMedoidsInitialization<O>

Parameters:

k - Parameter k

ids - Candidate IDs.

distQ - Distance function

Returns:

List of chosen means for k-means