de.lmu.ifi.dbs.elki.algorithm.clustering.gdbscan

Class ERiCNeighborPredicate.Instance

java.lang.Object

de.lmu.ifi.dbs.elki.algorithm.clustering.gdbscan.AbstractRangeQueryNeighborPredicate.Instance<DBIDs,PCAFilteredResult>

de.lmu.ifi.dbs.elki.algorithm.clustering.gdbscan.ERiCNeighborPredicate.Instance

All Implemented Interfaces:

NeighborPredicate.Instance<DBIDs>

Enclosing class:

ERiCNeighborPredicate<V extends NumberVector>

public class ERiCNeighborPredicate.Instance
extends AbstractRangeQueryNeighborPredicate.Instance<DBIDs,PCAFilteredResult>

Instance for a particular data set.

Author:

Erich Schubert

Field Summary

Fields

Modifier and Type	Field and Description
private Relation<? extends NumberVector>	relation Vector data relation.

Fields inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.gdbscan.AbstractRangeQueryNeighborPredicate.Instance

ids, storage

Constructor Summary

Constructors

Constructor and Description
Instance(DBIDs ids, DataStore<PCAFilteredResult> storage, Relation<? extends NumberVector> relation) Constructor.

Method Summary

Modifier and Type	Method and Description
protected boolean	approximatelyLinearDependent(PCAFilteredResult pca1, PCAFilteredResult pca2) Returns true, if the strong eigenvectors of the two specified PCAs span up the same space.
int	dimensionality(DBIDRef id) Get the correlation dimensionality of a single object.
DBIDs	getNeighbors(DBIDRef reference) Get the neighbors of a reference object for DBSCAN.
DBIDIter	iterDBIDs(DBIDs neighbors) Add the neighbors to a DBID set
boolean	strongNeighbors(NumberVector v1, NumberVector v2, PCAFilteredResult pca1, PCAFilteredResult pca2) Computes the distance between two given DatabaseObjects according to this distance function.
boolean	weakNeighbors(double[] v1, double[] v2, PCAFilteredResult pca1, PCAFilteredResult pca2) Computes the distance between two given DatabaseObjects according to this distance function.

Methods inherited from class de.lmu.ifi.dbs.elki.algorithm.clustering.gdbscan.AbstractRangeQueryNeighborPredicate.Instance

getIDs

Methods inherited from class java.lang.Object

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

Field Detail

relation

private Relation<? extends NumberVector> relation

Vector data relation.

Constructor Detail

Instance

public Instance(DBIDs ids,
                DataStore<PCAFilteredResult> storage,
                Relation<? extends NumberVector> relation)

Constructor.

Parameters:

ids - IDs this is defined for.

storage - Stored models

Method Detail

getNeighbors

public DBIDs getNeighbors(DBIDRef reference)

Description copied from interface: NeighborPredicate.Instance

Get the neighbors of a reference object for DBSCAN.

Parameters:

reference - Reference object

Returns:

Neighborhood

iterDBIDs

public DBIDIter iterDBIDs(DBIDs neighbors)

Description copied from interface: NeighborPredicate.Instance

Add the neighbors to a DBID set

Parameters:

neighbors - Neighbors to iterate over

Returns:

iterator

strongNeighbors

public boolean strongNeighbors(NumberVector v1,
                               NumberVector v2,
                               PCAFilteredResult pca1,
                               PCAFilteredResult pca2)

Computes the distance between two given DatabaseObjects according to this distance function. Note, that the first PCA must have equal or more strong eigenvectors than the second PCA.

Parameters:

v1 - first DatabaseObject

v2 - second DatabaseObject

pca1 - first PCA

pca2 - second PCA

Returns:

true when the two vectors are close enough.

weakNeighbors

public boolean weakNeighbors(double[] v1,
                             double[] v2,
                             PCAFilteredResult pca1,
                             PCAFilteredResult pca2)

Computes the distance between two given DatabaseObjects according to this distance function. Note, that the first PCA must have equal or more strong eigenvectors than the second PCA.

Parameters:

v1 - first DatabaseObject

v2 - second DatabaseObject

pca1 - first PCA

pca2 - second PCA

Returns:

true when the two vectors are close enough.

approximatelyLinearDependent

protected boolean approximatelyLinearDependent(PCAFilteredResult pca1,
                                               PCAFilteredResult pca2)

Returns true, if the strong eigenvectors of the two specified PCAs span up the same space. Note, that the first PCA must have at least as many strong eigenvectors than the second PCA.

Parameters:

pca1 - first PCA

pca2 - second PCA

Returns:

true, if the strong eigenvectors of the two specified PCAs span up the same space

dimensionality

public int dimensionality(DBIDRef id)

Get the correlation dimensionality of a single object.

Parameters:

id - Object ID

Returns:

correlation dimensionality