de.lmu.ifi.dbs.elki.distance.distancefunction.correlation
Class ERiCDistanceFunction<V extends NumberVector<V,?>,P extends LocalPCAPreprocessor<V>>

java.lang.Object
  de.lmu.ifi.dbs.elki.logging.AbstractLoggable
      de.lmu.ifi.dbs.elki.distance.AbstractMeasurementFunction<O,D>
          de.lmu.ifi.dbs.elki.distance.distancefunction.AbstractDistanceFunction<O,D>
              de.lmu.ifi.dbs.elki.distance.distancefunction.AbstractPreprocessorBasedDistanceFunction<V,P,BitDistance>
                  de.lmu.ifi.dbs.elki.distance.distancefunction.correlation.ERiCDistanceFunction<V,P>

Type Parameters:

V - the type of NumberVector to compute the distances in between

P - the type of Preprocessor used

All Implemented Interfaces:

DistanceFunction<V,BitDistance>, LocalPCAPreprocessorBasedDistanceFunction<V,P,BitDistance>, PreprocessorBasedDistanceFunction<V,P,BitDistance>, MeasurementFunction<V,BitDistance>, PreprocessorBasedMeasurementFunction<V,P,BitDistance>, PreprocessorClient<P,V>, Parameterizable

public class ERiCDistanceFunction<V extends NumberVector<V,?>,P extends LocalPCAPreprocessor<V>>
extends AbstractPreprocessorBasedDistanceFunction<V,P,BitDistance>
implements LocalPCAPreprocessorBasedDistanceFunction<V,P,BitDistance>

Provides a distance function for building the hierarchy in the ERiC algorithm.

Author:

Elke Achtert

Field Summary

private double	delta Holds the value of DELTA_PARAM.
static OptionID	DELTA_ID OptionID for DELTA_PARAM
private DoubleParameter	DELTA_PARAM Parameter to specify the threshold for approximate linear dependency: the strong eigenvectors of q are approximately linear dependent from the strong eigenvectors p if the following condition holds for all stroneg eigenvectors q_i of q (lambda_q < lambda_p): q_i' * M^check_p * q_i <= delta^2, must be a double equal to or greater than 0.
private double	tau Holds the value of TAU_PARAM.
static OptionID	TAU_ID OptionID for TAU_PARAM
private DoubleParameter	TAU_PARAM Parameter to specify the threshold for the maximum distance between two approximately linear dependent subspaces of two objects p and q (lambda_q < lambda_p) before considering them as parallel, must be a double equal to or greater than 0.

Fields inherited from class de.lmu.ifi.dbs.elki.distance.AbstractMeasurementFunction

distanceFactory

Fields inherited from class de.lmu.ifi.dbs.elki.logging.AbstractLoggable

debug, logger

Constructor Summary

ERiCDistanceFunction(Parameterization config)
Constructor, adhering to Parameterizable

Method Summary

private boolean	approximatelyLinearDependent(PCAFilteredResult pca1, PCAFilteredResult pca2) Returns true, if the strong eigenvectors of the two specified pcas span up the same space.
BitDistance	distance(V v1, V v2) Note, that the pca of o1 must have equal ore more strong eigenvectors than the pca of o2.
BitDistance	distance(V v1, V v2, PCAFilteredResult pca1, PCAFilteredResult pca2) Computes the distance between two given DatabaseObjects according to this distance function.
AssociationID<?>	getAssociationID() Returns the association ID for the association to be set by the preprocessor.
Class<?>	getDefaultPreprocessorClass() Returns the name of the default preprocessor.
String	getPreprocessorDescription() Returns the description for the preprocessor parameter.
Class<P>	getPreprocessorSuperClass() Returns the super class for the preprocessor parameter.

Methods inherited from class de.lmu.ifi.dbs.elki.distance.distancefunction.AbstractPreprocessorBasedDistanceFunction

getPreprocessor, setDatabase

Methods inherited from class de.lmu.ifi.dbs.elki.distance.distancefunction.AbstractDistanceFunction

distance, distance

Methods inherited from class de.lmu.ifi.dbs.elki.distance.AbstractMeasurementFunction

getDatabase, getDistanceFactory, infiniteDistance, nullDistance, undefinedDistance, valueOf

Methods inherited from class de.lmu.ifi.dbs.elki.logging.AbstractLoggable

debugFine, debugFiner, debugFinest, exception, progress, verbose, warning

Methods inherited from class java.lang.Object

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

Methods inherited from interface de.lmu.ifi.dbs.elki.distance.distancefunction.DistanceFunction

distance, distance

Methods inherited from interface de.lmu.ifi.dbs.elki.distance.MeasurementFunction

getDistanceFactory, infiniteDistance, nullDistance, setDatabase, undefinedDistance, valueOf

Field Detail

DELTA_ID

public static final OptionID DELTA_ID

OptionID for DELTA_PARAM

DELTA_PARAM

private final DoubleParameter DELTA_PARAM

Parameter to specify the threshold for approximate linear dependency: the strong eigenvectors of q are approximately linear dependent from the strong eigenvectors p if the following condition holds for all stroneg eigenvectors q_i of q (lambda_q < lambda_p): q_i' * M^check_p * q_i <= delta^2, must be a double equal to or greater than 0.

Default value: 0.1

Key: -ericdf.delta

TAU_ID

public static final OptionID TAU_ID

OptionID for TAU_PARAM

TAU_PARAM

private final DoubleParameter TAU_PARAM

Parameter to specify the threshold for the maximum distance between two approximately linear dependent subspaces of two objects p and q (lambda_q < lambda_p) before considering them as parallel, must be a double equal to or greater than 0.

Default value: 0.1

Key: -ericdf.tau

delta

private double delta

Holds the value of DELTA_PARAM.

tau

private double tau

Holds the value of TAU_PARAM.

Constructor Detail

ERiCDistanceFunction

public ERiCDistanceFunction(Parameterization config)

Constructor, adhering to Parameterizable

Parameters:

config - Parameterization

Method Detail

getDefaultPreprocessorClass

public Class<?> getDefaultPreprocessorClass()

Returns the name of the default preprocessor.

Specified by:

getDefaultPreprocessorClass in interface PreprocessorClient<P extends LocalPCAPreprocessor<V>,V extends NumberVector<V,?>>

Returns:

the name of the default preprocessor, which is KnnQueryBasedLocalPCAPreprocessor

getPreprocessorDescription

public String getPreprocessorDescription()

Description copied from interface: PreprocessorClient

Returns the description for the preprocessor parameter.

Specified by:

getPreprocessorDescription in interface PreprocessorClient<P extends LocalPCAPreprocessor<V>,V extends NumberVector<V,?>>

Returns:

the description for the preprocessor parameter

getPreprocessorSuperClass

public Class<P> getPreprocessorSuperClass()

Description copied from interface: PreprocessorClient

Returns the super class for the preprocessor parameter.

Specified by:

getPreprocessorSuperClass in interface PreprocessorClient<P extends LocalPCAPreprocessor<V>,V extends NumberVector<V,?>>

Returns:

the super class for the preprocessor parameter, which is Preprocessor

getAssociationID

public AssociationID<?> getAssociationID()

Description copied from interface: PreprocessorClient

Returns the association ID for the association to be set by the preprocessor.

Specified by:

getAssociationID in interface PreprocessorClient<P extends LocalPCAPreprocessor<V>,V extends NumberVector<V,?>>

Returns:

the association ID for the association to be set by the preprocessor, which is AssociationID.LOCAL_PCA

distance

public BitDistance distance(V v1,
                            V v2)

Note, that the pca of o1 must have equal ore more strong eigenvectors than the pca of o2.

Specified by:

distance in interface DistanceFunction<V extends NumberVector<V,?>,BitDistance>

Parameters:

v1 - first DatabaseObject

v2 - second DatabaseObject

Returns:

the distance between two given DatabaseObjects according to this distance function

distance

public BitDistance distance(V v1,
                            V 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:

the distance between two given DatabaseObjects according to this distance function

approximatelyLinearDependent

private 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 equal ore more 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