Package kieker.analysis.generic.clustering.mtree

Interface ISplitFunction<T>

Type Parameters:

T - The type of the data objects.

All Known Implementing Classes:

ComposedSplitFunction

public interface ISplitFunction<T>

Defines an object to be used to split a node in an M-Tree. A node must be split when it has reached its maximum capacity and a new child node would be added to it.

The splitting consists in choosing a pair of "promoted" data objects from the children and then partition the set of children in two partitions corresponding to the two promoted data objects.

Since:

2.0.0

Author:

Eduardo R. D'Avila

Nested Class Summary

Nested Classes

Modifier and Type	Interface	Description
static class	ISplitFunction.SplitResult<R>	An object used as the result for the process(Set, IDistanceFunction) method.

Method Summary

Modifier and Type	Method	Description
ISplitFunction.SplitResult<T>	process(java.util.Set<T> dataSet, IDistanceFunction<? super T> distanceFunction)	Processes the splitting of a node.

Method Details

process

ISplitFunction.SplitResult<T> process(java.util.Set<T> dataSet, IDistanceFunction<? super T> distanceFunction)

Processes the splitting of a node.

Parameters:

dataSet - A set of data that are keys to the children of the node to be split.

distanceFunction - A distance function that can be used to help splitting the node.

Returns:

A ISplitFunction.SplitResult object with a pair of promoted data objects and a pair of corresponding partitions of the data objects.