de.lmu.ifi.dbs.elki.normalization
Class AttributeWiseMinMaxNormalization<V extends NumberVector<V,?>>

java.lang.Object
  de.lmu.ifi.dbs.elki.logging.AbstractLoggable
      de.lmu.ifi.dbs.elki.normalization.AbstractNormalization<V>
          de.lmu.ifi.dbs.elki.normalization.AttributeWiseMinMaxNormalization<V>

Type Parameters:

V - vector type

All Implemented Interfaces:

Normalization<V>, Parameterizable

public class AttributeWiseMinMaxNormalization<V extends NumberVector<V,?>>
extends AbstractNormalization<V>

Class to perform and undo a normalization on real vectors with respect to given minimum and maximum in each dimension.

Author:

Elke Achtert

Field Summary

private double[]	maxima Stores the maximum in each dimension.
static OptionID	MAXIMA_ID OptionID for MAXIMA_PARAM
private DoubleListParameter	MAXIMA_PARAM Parameter for maximum.
private double[]	minima Stores the minimum in each dimension.
static OptionID	MINIMA_ID OptionID for MINIMA_PARAM
private DoubleListParameter	MINIMA_PARAM Parameter for minimum.

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

debug, logger

Constructor Summary

AttributeWiseMinMaxNormalization(Parameterization config)
Constructor, adhering to Parameterizable

Method Summary

private void	determineMinMax(List<Pair<V,Associations>> objectAndAssociationsList) Determines the minima and maxima values in each dimension of the given featureVectors.
private void	determineMinMax(V[] featureVectors) Determines the minima and maxima values in each dimension of the given featureVectors.
private double	factor(int dimension) Returns a factor for normalization in a certain dimension.
private void	initMinMax(int dimensionality) Initializes the min and max array.
List<V>	normalize(List<V> featureVectors) Performs a normalization on a set of feature vectors.
List<Pair<V,Associations>>	normalizeObjects(List<Pair<V,Associations>> objectAndAssociationsList) Performs a normalization on a list of database objects and their associations.
List<V>	restore(List<V> featureVectors) Transforms a set of feature vectores to the original attribute ranges.
V	restore(V featureVector) Transforms a feature vector to the original attribute ranges.
String	toString(String pre) Returns a string representation of this normalization.
LinearEquationSystem	transform(LinearEquationSystem linearEquationSystem) Transforms a linear equation system describing linear dependencies derived on the normalized space into a linear equation system describing linear dependencies quantitatively adapted to the original space.
private void	updateMinMax(V featureVector) Updates the min and max array according to the specified feature vector.

Methods inherited from class de.lmu.ifi.dbs.elki.normalization.AbstractNormalization

toString

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, wait, wait, wait

Field Detail

MINIMA_ID

public static final OptionID MINIMA_ID

OptionID for MINIMA_PARAM

MAXIMA_ID

public static final OptionID MAXIMA_ID

OptionID for MAXIMA_PARAM

MINIMA_PARAM

private final DoubleListParameter MINIMA_PARAM

Parameter for minimum.

MAXIMA_PARAM

private final DoubleListParameter MAXIMA_PARAM

Parameter for maximum.

maxima

private double[] maxima

Stores the maximum in each dimension.

minima

private double[] minima

Stores the minimum in each dimension.

Constructor Detail

AttributeWiseMinMaxNormalization

public AttributeWiseMinMaxNormalization(Parameterization config)

Constructor, adhering to Parameterizable

Parameters:

config - Parameterization

Method Detail

normalizeObjects

public List<Pair<V,Associations>> normalizeObjects(List<Pair<V,Associations>> objectAndAssociationsList)
                                                                      throws NonNumericFeaturesException

Description copied from interface: Normalization

Performs a normalization on a list of database objects and their associations.

Parameters:

objectAndAssociationsList - the list of database objects and their associations

Returns:

a list of normalized database objects and their associations corresponding to the given list

Throws:

NonNumericFeaturesException - if feature vectors differ in length or values are not suitable to normalization

normalize

public List<V> normalize(List<V> featureVectors)
                                            throws NonNumericFeaturesException

Description copied from interface: Normalization

Performs a normalization on a set of feature vectors.

Parameters:

featureVectors - a set of feature vectors to be normalized

Returns:

a set of normalized feature vectors corresponding to the given feature vectors but being different objects

Throws:

NonNumericFeaturesException - if feature vectors differ in length or values are not suitable to normalization

restore

public V restore(V featureVector)
                                    throws NonNumericFeaturesException

Description copied from interface: Normalization

Transforms a feature vector to the original attribute ranges.

Parameters:

featureVector - a feature vector to be transformed into original space

Returns:

a feature vector transformed into original space corresponding to the given feature vector

Throws:

NonNumericFeaturesException - feature vector is not compatible with values initialized during normalization

restore

public List<V> restore(List<V> featureVectors)
                                          throws NonNumericFeaturesException

Description copied from interface: Normalization

Transforms a set of feature vectores to the original attribute ranges.

Parameters:

featureVectors - a set of feature vectors to be transformed into original space

Returns:

a set of feature vectors transformed into original space corresponding to the given feature vectors

Throws:

NonNumericFeaturesException - if feature vectors differ in length or are not compatible with values initialized during normalization

transform

public LinearEquationSystem transform(LinearEquationSystem linearEquationSystem)

Description copied from interface: Normalization

Transforms a linear equation system describing linear dependencies derived on the normalized space into a linear equation system describing linear dependencies quantitatively adapted to the original space.

Parameters:

linearEquationSystem - the linear equation system to be transformed

Returns:

a linear equation system describing linear dependencies derived on the normalized space transformed into a linear equation system describing linear dependencies quantitatively adapted to the original space

toString

public String toString(String pre)

Returns a string representation of this normalization. The specified prefix pre will be the prefix of each new line. This method is used to write the parameters of a normalization to a result of an algorithm using this normalization.

Parameters:

pre - the prefix of each new line

Returns:

a string representation of this normalization

factor

private double factor(int dimension)

Returns a factor for normalization in a certain dimension.

The provided factor is the maximum-minimum in the specified dimension, if these two values differ, otherwise it is the maximum if this value differs from 0, otherwise it is 1.

Parameters:

dimension - the dimension to get a factor for normalization

Returns:

a factor for normalization in a certain dimension

determineMinMax

private void determineMinMax(V[] featureVectors)

Determines the minima and maxima values in each dimension of the given featureVectors.

Parameters:

featureVectors - the list of feature vectors

determineMinMax

private void determineMinMax(List<Pair<V,Associations>> objectAndAssociationsList)

Determines the minima and maxima values in each dimension of the given featureVectors.

Parameters:

objectAndAssociationsList - the list of feature vectors and their associations

initMinMax

private void initMinMax(int dimensionality)

Initializes the min and max array.

Parameters:

dimensionality - the dimensionality of the feature vectors to be normalized

updateMinMax

private void updateMinMax(V featureVector)

Updates the min and max array according to the specified feature vector.

Parameters:

featureVector - the feature vector