info.debatty.spark.kmedoids

Class Clusterer<T>

java.lang.Object

info.debatty.spark.kmedoids.Clusterer<T>

Type Parameters:

T - the type of data to cluster

Direct Known Subclasses:

Clarans

public class Clusterer<T>
extends Object

A generic k-medoids clusterer. The concrete implementation of the neighbor generator will make the difference between CLARANS, SA and Heuristical clustering.

Author:

Thibault Debatty

Constructor Summary

Constructors

Constructor and Description
Clusterer()

Method Summary

Modifier and Type	Method and Description
Solution<T>	cluster(org.apache.spark.api.java.JavaRDD<T> input_data) Cluster the data according to defined parameters.
T	getRandomPoint()
void	setBudget(Budget budget)
void	setImbalance(double imbalance) Set imbalance : 1.0 equals perfectly balanced.
void	setK(int k)
void	setNeighborGenerator(NeighborGenerotor<T> neighbor_generator)
void	setSimilarity(Similarity<T> similarity)

Methods inherited from class java.lang.Object

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

Constructor Detail

Clusterer

public Clusterer()

Method Detail

setK

public final void setK(int k)

Parameters:

k -

setImbalance

public final void setImbalance(double imbalance)

Set imbalance : 1.0 equals perfectly balanced.

Parameters:

imbalance -

setNeighborGenerator

public final void setNeighborGenerator(NeighborGenerotor<T> neighbor_generator)

Parameters:

neighbor_generator -

setSimilarity

public final void setSimilarity(Similarity<T> similarity)

Parameters:

similarity -

setBudget

public final void setBudget(Budget budget)

Parameters:

budget -

cluster

public final Solution<T> cluster(org.apache.spark.api.java.JavaRDD<T> input_data)

Cluster the data according to defined parameters.

Parameters:

input_data -

Returns:

getRandomPoint

public final T getRandomPoint()

Returns: