Package elki.clustering.kmeans

Interface KMeans<V extends NumberVector,M extends Model>

Type Parameters:

V - Number vector type

M - Actual model type

All Superinterfaces:

Algorithm, ClusteringAlgorithm<Clustering<M>>

All Known Implementing Classes:

AbstractKMeans, AnnulusKMeans, BestOfMultipleKMeans, BetulaLloydKMeans, BisectingKMeans, CompareMeans, ElkanKMeans, EuclideanSphericalElkanKMeans, EuclideanSphericalHamerlyKMeans, EuclideanSphericalSimplifiedElkanKMeans, ExponionKMeans, GMeans, HamerlyKMeans, HartiganWongKMeans, KDTreeFilteringKMeans, KDTreePruningKMeans, KMeansMinusMinus, KMediansLloyd, LloydKMeans, MacQueenKMeans, ParallelLloydKMeans, SameSizeKMeans, ShallotKMeans, SimplifiedElkanKMeans, SingleAssignmentKMeans, SortMeans, SphericalElkanKMeans, SphericalHamerlyKMeans, SphericalKMeans, SphericalSimplifiedElkanKMeans, SphericalSimplifiedHamerlyKMeans, SphericalSingleAssignmentKMeans, XMeans, YinYangKMeans

public interface KMeans<V extends NumberVector,M extends Model>
extends ClusteringAlgorithm<Clustering<M>>

Some constants and options shared among kmeans family algorithms.

Since:

0.1

Author:

Erich Schubert

Nested Class Summary

Nested classes/interfaces inherited from interface elki.Algorithm

Algorithm.Utils

Field Summary

Fields

Modifier and Type	Field	Description
static OptionID	DISTANCE_FUNCTION_ID	OptionID for the distance function.
static OptionID	INIT_ID	Parameter to specify the initialization method
static OptionID	K_ID	Parameter to specify the number of clusters to find, must be an integer greater than 0.
static OptionID	MAXITER_ID	Parameter to specify the number of clusters to find, must be an integer greater or equal to 0, where 0 means no limit.
static OptionID	SEED_ID	Parameter to specify the random generator seed.
static OptionID	VARSTAT_ID	Flag to compute the final clustering variance statistic (for methods that employ bounds to avoid computing all distances).

Method Summary

Modifier and Type	Method	Description
NumberVectorDistance<? super V>	getDistance()	Returns the distance.
Clustering<M>	run(Relation<V> rel)	Run the clustering algorithm.
void	setDistance(NumberVectorDistance<? super V> distance)	Set the distance function to use.
void	setInitializer(KMeansInitialization init)	Set the initialization method.
void	setK(int k)	Set the value of k.

Methods inherited from interface elki.Algorithm

getInputTypeRestriction

Methods inherited from interface elki.clustering.ClusteringAlgorithm

autorun

Field Detail

DISTANCE_FUNCTION_ID

static final OptionID DISTANCE_FUNCTION_ID

OptionID for the distance function.

INIT_ID

static final OptionID INIT_ID

Parameter to specify the initialization method

K_ID

static final OptionID K_ID

Parameter to specify the number of clusters to find, must be an integer greater than 0.

MAXITER_ID

static final OptionID MAXITER_ID

Parameter to specify the number of clusters to find, must be an integer greater or equal to 0, where 0 means no limit.

SEED_ID

static final OptionID SEED_ID

Parameter to specify the random generator seed.

VARSTAT_ID

static final OptionID VARSTAT_ID

Flag to compute the final clustering variance statistic (for methods that employ bounds to avoid computing all distances).

Method Detail

run

Clustering<M> run(Relation<V> rel)

Run the clustering algorithm.

Parameters:

rel - Relation to process.

Returns:

Clustering result

setK

void setK(int k)

Set the value of k. Needed for some types of nested k-means.

Parameters:

k - K parameter

setDistance

void setDistance(NumberVectorDistance<? super V> distance)

Set the distance function to use.

Parameters:

distance - Distance function.

getDistance

NumberVectorDistance<? super V> getDistance()

Returns the distance.

Returns:

the distance

setInitializer

void setInitializer(KMeansInitialization init)

Set the initialization method.

Parameters:

init - Initialization method