de.lmu.ifi.dbs.elki.index.tree.metrical.mtreevariants.strategies.split

Class RandomSplit<O,N extends AbstractMTreeNode<O,N,E>,E extends MTreeEntry>

java.lang.Object

de.lmu.ifi.dbs.elki.index.tree.metrical.mtreevariants.strategies.split.MTreeSplit<O,N,E>

de.lmu.ifi.dbs.elki.index.tree.metrical.mtreevariants.strategies.split.RandomSplit<O,N,E>

Type Parameters:

O - the type of DatabaseObject to be stored in the M-Tree

N - the type of AbstractMTreeNode used in the M-Tree

E - the type of MetricalEntry used in the M-Tree

@Reference(authors="P. Ciaccia, M. Patella, P. Zezula",
           title="M-tree: An Efficient Access Method for Similarity Search in Metric Spaces",
           booktitle="VLDB\'97, Proceedings of 23rd International Conference on Very Large Data Bases, August 25-29, 1997, Athens, Greece",
           url="http://www.vldb.org/conf/1997/P426.PDF")
public class RandomSplit<O,N extends AbstractMTreeNode<O,N,E>,E extends MTreeEntry>
extends MTreeSplit<O,N,E>

Encapsulates the required methods for a split of a node in an M-Tree. The routing objects are chosen according to the RANDOM strategy. Note: only the routing objects are chosen at random, this is not a random assignment! Reference:

P. Ciaccia, M. Patella, P. Zezula
M-tree: An Efficient Access Method for Similarity Search in Metric Spaces
In Proceedings of 23rd International Conference on Very Large Data Bases (VLDB'97), August 25-29, 1997, Athens, Greece

Since:

0.2

Author:

Elke Achtert

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	RandomSplit.Parameterizer<O,N extends AbstractMTreeNode<O,N,E>,E extends MTreeEntry> Parameterization class.

Field Summary

Fields

Modifier and Type	Field and Description
private Random	random Random generator.

Constructor Summary

Constructors

Constructor and Description
RandomSplit(RandomFactory rnd) Creates a new split object.

Method Summary

Methods

Modifier and Type	Method and Description
Assignments<E>	split(AbstractMTree<O,N,E,?> tree, N node) Selects two objects of the specified node to be promoted and stored into the parent node.

Methods inherited from class de.lmu.ifi.dbs.elki.index.tree.metrical.mtreevariants.strategies.split.MTreeSplit

balancedPartition, balancedPartition, computeDistanceMatrix

Methods inherited from class java.lang.Object

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

Field Detail

random

private Random random

Random generator.

Constructor Detail

RandomSplit

public RandomSplit(RandomFactory rnd)

Creates a new split object.

Method Detail

split

public Assignments<E> split(AbstractMTree<O,N,E,?> tree,
                   N node)

Selects two objects of the specified node to be promoted and stored into the parent node. The m-RAD strategy considers all possible pairs of objects and, after partitioning the set of entries, promotes the pair of objects for which the sum of covering radiuses is minimum.

Specified by:

split in class MTreeSplit<O,N extends AbstractMTreeNode<O,N,E>,E extends MTreeEntry>

Parameters:

tree - Tree to use

node - the node to be split

Returns:

the assignments of this split