de.lmu.ifi.dbs.elki.varianceanalysis
Class LocalPCA<V extends RealVector<V,?>>

java.lang.Object
  de.lmu.ifi.dbs.elki.logging.AbstractLoggable
      de.lmu.ifi.dbs.elki.utilities.optionhandling.AbstractParameterizable
          de.lmu.ifi.dbs.elki.varianceanalysis.AbstractPCA
              de.lmu.ifi.dbs.elki.varianceanalysis.LocalPCA<V>

All Implemented Interfaces:

Loggable, Parameterizable, PCA

Direct Known Subclasses:

LinearLocalPCA, LocalKernelPCA

public abstract class LocalPCA<V extends RealVector<V,?>>
extends AbstractPCA

LocalPCA is a super calss for PCA-algorithms considering only a local neighborhood. LocalPCA provides some methods valid for any extending class.

Author:

Elke Achtert

Field Summary

private Matrix	adapatedStrongEigenvectors The diagonal matrix of adapted strong eigenvalues: eigenvectors * e_czech.
private double	big Holds the big value.
static OptionID	BIG_ID OptionID for BIG_PARAM
private DoubleParameter	BIG_PARAM Parameter to specify a constant big value to reset high eigenvalues, must be a double greater than 0.
private int	correlationDimension The correlation dimension (i.e. the number of strong eigenvectors) of the object to which this PCA belongs to.
private Matrix	e_czech The selection matrix of the strong eigenvectors.
private Matrix	e_hat The selection matrix of the weak eigenvectors.
private Matrix	m_czech The dissimilarity matrix.
private Matrix	m_hat The similarity matrix.
private double	small Holds the small value.
static OptionID	SMALL_ID OptionID for SMALL_PARAM
private DoubleParameter	SMALL_PARAM Parameter to specify a constant small value to reset low eigenvalues, must be a double greater than 0.

Fields inherited from class de.lmu.ifi.dbs.elki.utilities.optionhandling.AbstractParameterizable

optionHandler

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

debug

Constructor Summary

LocalPCA()
Adds parameter for big and small value to parameter map.

Method Summary

Matrix	adapatedStrongEigenvectors() Returns a copy of the adapted strong eigenvectors.
Matrix	dissimilarityMatrix() Returns a copy of the dissimilarity matrix (M_czech) of this LocalPCA.
int	getCorrelationDimension() Returns the correlation dimension (i.e. the number of strong eigenvectors) of the object to which this PCA belongs to.
protected abstract Matrix	pcaMatrix(Database<V> database, Collection<Integer> ids) Determines and returns the matrix that is used for performaing the pca.
void	run(Collection<Integer> ids, Database<V> database) Performs a LocalPCA for the object with the specified ids stored in the given database.
Matrix	selectionMatrixOfStrongEigenvectors() Returns a copy of the selection matrix of the strong eigenvectors (E_czech) of this LocalPCA.
Matrix	selectionMatrixOfWeakEigenvectors() Returns a copy of the selection matrix of the weak eigenvectors (E_hat) of the object to which this PCA belongs to.
String[]	setParameters(String[] args) Sets the attributes of the class accordingly to the given parameters.
Matrix	similarityMatrix() Returns a copy of the similarity matrix (M_hat) of this LocalPCA.

Methods inherited from class de.lmu.ifi.dbs.elki.varianceanalysis.AbstractPCA

determineEigenPairs, getAttributeSettings, getEigenvalues, getEigenvectors, getStrongEigenvalues, getStrongEigenvectors, getWeakEigenvalues, getWeakEigenvectors

Methods inherited from class de.lmu.ifi.dbs.elki.utilities.optionhandling.AbstractParameterizable

addOption, checkGlobalParameterConstraints, deleteOption, description, description, description, getParameters, getParameterValue, getPossibleOptions, inlineDescription, isSet, setParameters

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

debugFine, debugFiner, debugFinest, exception, message, progress, progress, progress, verbose, 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.utilities.optionhandling.Parameterizable

checkGlobalParameterConstraints, description, getParameters, getPossibleOptions, inlineDescription

Field Detail

BIG_ID

public static final OptionID BIG_ID

OptionID for BIG_PARAM

SMALL_ID

public static final OptionID SMALL_ID

OptionID for SMALL_PARAM

BIG_PARAM

private final DoubleParameter BIG_PARAM

Parameter to specify a constant big value to reset high eigenvalues, must be a double greater than 0.

Default value: 1.0

Key: -localpca.big

SMALL_PARAM

private final DoubleParameter SMALL_PARAM

Parameter to specify a constant small value to reset low eigenvalues, must be a double greater than 0.

Default value: 0.0

Key: -localpca.small

big

private double big

Holds the big value.

small

private double small

Holds the small value.

correlationDimension

private int correlationDimension

The correlation dimension (i.e. the number of strong eigenvectors) of the object to which this PCA belongs to.

e_hat

private Matrix e_hat

The selection matrix of the weak eigenvectors.

e_czech

private Matrix e_czech

The selection matrix of the strong eigenvectors.

m_hat

private Matrix m_hat

The similarity matrix.

m_czech

private Matrix m_czech

The dissimilarity matrix.

adapatedStrongEigenvectors

private Matrix adapatedStrongEigenvectors

The diagonal matrix of adapted strong eigenvalues: eigenvectors * e_czech.

Constructor Detail

LocalPCA

public LocalPCA()

Adds parameter for big and small value to parameter map.

Method Detail

run

public final void run(Collection<Integer> ids,
                      Database<V> database)

Performs a LocalPCA for the object with the specified ids stored in the given database.

Parameters:

ids - the ids of the objects for which the PCA should be performed

database - the database containing the objects

setParameters

public String[] setParameters(String[] args)
                       throws ParameterException

Description copied from interface: Parameterizable

Sets the attributes of the class accordingly to the given parameters. Returns a new String array containing those entries of the given array that are neither expected nor used by this Parameterizable.

Specified by:

setParameters in interface Parameterizable

Overrides:

setParameters in class AbstractPCA

Parameters:

args - parameters to set the attributes accordingly to

Returns:

String[] an array containing the unused parameters

Throws:

ParameterException - in case of wrong parameter-setting

See Also:

Parameterizable.setParameters(String[])

getCorrelationDimension

public int getCorrelationDimension()

Returns the correlation dimension (i.e. the number of strong eigenvectors) of the object to which this PCA belongs to.

Returns:

the correlation dimension

selectionMatrixOfWeakEigenvectors

public Matrix selectionMatrixOfWeakEigenvectors()

Returns a copy of the selection matrix of the weak eigenvectors (E_hat) of the object to which this PCA belongs to.

Returns:

the selection matrix of the weak eigenvectors E_hat

selectionMatrixOfStrongEigenvectors

public Matrix selectionMatrixOfStrongEigenvectors()

Returns a copy of the selection matrix of the strong eigenvectors (E_czech) of this LocalPCA.

Returns:

the selection matrix of the weak eigenvectors E_czech

similarityMatrix

public Matrix similarityMatrix()

Returns a copy of the similarity matrix (M_hat) of this LocalPCA.

Returns:

the similarity matrix M_hat

dissimilarityMatrix

public Matrix dissimilarityMatrix()

Returns a copy of the dissimilarity matrix (M_czech) of this LocalPCA.

Returns:

the dissimilarity matrix M_hat

adapatedStrongEigenvectors

public Matrix adapatedStrongEigenvectors()

Returns a copy of the adapted strong eigenvectors.

Returns:

the adapted strong eigenvectors

pcaMatrix

protected abstract Matrix pcaMatrix(Database<V> database,
                                    Collection<Integer> ids)

Determines and returns the matrix that is used for performaing the pca.

Parameters:

database - the database holding the objects

ids - the list of the object ids for which the matrix should be determined

Returns:

he matrix that is used for performaing a pca