moa.classifiers.lazy.neighboursearch

Class EuclideanDistance

java.lang.Object

moa.classifiers.lazy.neighboursearch.NormalizableDistance

moa.classifiers.lazy.neighboursearch.EuclideanDistance

All Implemented Interfaces:

Cloneable, DistanceFunction

public class EuclideanDistance
extends NormalizableDistance
implements Cloneable

Implementing Euclidean distance (or similarity) function.

One object defines not one distance but the data model in which the distances between objects of that data model can be computed.

Attention: For efficiency reasons the use of consistency checks (like are the data models of the two instances exactly the same), is low.

For more information, see:

Wikipedia. Euclidean distance. URL http://en.wikipedia.org/wiki/Euclidean_distance.

BibTeX:

 @misc{missing_id,
    author = {Wikipedia},
    title = {Euclidean distance},
    URL = {http://en.wikipedia.org/wiki/Euclidean_distance}
 }
 

Valid options are:

 -D
  Turns off the normalization of attribute 
  values in distance calculation.

 -R <col1,col2-col4,...>
  Specifies list of columns to used in the calculation of the 
  distance. 'first' and 'last' are valid indices.
  (default: first-last)

 -V
  Invert matching sense of column indices.

Version:

$Revision: 8034 $

Author:

Gabi Schmidberger (gabi@cs.waikato.ac.nz), Ashraf M. Kibriya (amk14@cs.waikato.ac.nz), FracPete (fracpete at waikato dot ac dot nz)

Field Summary

Fields inherited from class moa.classifiers.lazy.neighboursearch.NormalizableDistance

m_ActiveIndices, m_Data, m_DontNormalize, m_Ranges, m_Validated, R_MAX, R_MIN, R_WIDTH

Constructor Summary

Constructors

Constructor and Description
EuclideanDistance() Constructs an Euclidean Distance object, Instances must be still set.
EuclideanDistance(weka.core.Instances data) Constructs an Euclidean Distance object and automatically initializes the ranges.

Method Summary

Methods

Modifier and Type	Method and Description
int	closestPoint(weka.core.Instance instance, weka.core.Instances allPoints, int[] pointList) Returns the index of the closest point to the current instance.
double	distance(weka.core.Instance first, weka.core.Instance second) Calculates the distance between two instances.
double	getMiddle(double[] ranges) Returns value in the middle of the two parameter values.
String	globalInfo() Returns a string describing this object.
void	postProcessDistances(double[] distances) Does post processing of the distances (if necessary) returned by distance(distance(Instance first, Instance second, double cutOffValue).
double	sqDifference(int index, double val1, double val2) Returns the squared difference of two values of an attribute.
protected double	updateDistance(double currDist, double diff) Updates the current distance calculated so far with the new difference between two attributes.
boolean	valueIsSmallerEqual(weka.core.Instance instance, int dim, double value) Returns true if the value of the given dimension is smaller or equal the value to be compared with.

Methods inherited from class moa.classifiers.lazy.neighboursearch.NormalizableDistance

attributeIndicesTipText, difference, distance, dontNormalizeTipText, getAttributeIndices, getDontNormalize, getInstances, getInvertSelection, getRanges, initialize, initializeAttributeIndices, initializeRanges, initializeRanges, initializeRanges, initializeRangesEmpty, inRanges, invalidate, invertSelectionTipText, isMissingValue, norm, rangesSet, setAttributeIndices, setDontNormalize, setInstances, setInvertSelection, toString, update, updateRanges, updateRanges, updateRanges, updateRangesFirst, validate

Methods inherited from class java.lang.Object

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

Constructor Detail

EuclideanDistance

public EuclideanDistance()

Constructs an Euclidean Distance object, Instances must be still set.

EuclideanDistance

public EuclideanDistance(weka.core.Instances data)

Constructs an Euclidean Distance object and automatically initializes the ranges.

Parameters:

data - the instances the distance function should work on

Method Detail

globalInfo

public String globalInfo()

Returns a string describing this object.

Specified by:

globalInfo in class NormalizableDistance

Returns:

a description of the evaluator suitable for displaying in the explorer/experimenter gui

distance

public double distance(weka.core.Instance first,
              weka.core.Instance second)

Calculates the distance between two instances.

Specified by:

distance in interface DistanceFunction

Overrides:

distance in class NormalizableDistance

Parameters:

first - the first instance

second - the second instance

Returns:

the distance between the two given instances

updateDistance

protected double updateDistance(double currDist,
                    double diff)

Updates the current distance calculated so far with the new difference between two attributes. The difference between the attributes was calculated with the difference(int,double,double) method.

Specified by:

updateDistance in class NormalizableDistance

Parameters:

currDist - the current distance calculated so far

diff - the difference between two new attributes

Returns:

the update distance

See Also:

NormalizableDistance.difference(int, double, double)

postProcessDistances

public void postProcessDistances(double[] distances)

Does post processing of the distances (if necessary) returned by distance(distance(Instance first, Instance second, double cutOffValue). It is necessary to do so to get the correct distances if distance(distance(Instance first, Instance second, double cutOffValue) is used. This is because that function actually returns the squared distance to avoid inaccuracies arising from floating point comparison.

Specified by:

postProcessDistances in interface DistanceFunction

Overrides:

postProcessDistances in class NormalizableDistance

Parameters:

distances - the distances to post-process

sqDifference

public double sqDifference(int index,
                  double val1,
                  double val2)

Returns the squared difference of two values of an attribute.

Parameters:

index - the attribute index

val1 - the first value

val2 - the second value

Returns:

the squared difference

getMiddle

public double getMiddle(double[] ranges)

Returns value in the middle of the two parameter values.

Parameters:

ranges - the ranges to this dimension

Returns:

the middle value

closestPoint

public int closestPoint(weka.core.Instance instance,
               weka.core.Instances allPoints,
               int[] pointList)
                 throws Exception

Returns the index of the closest point to the current instance. Index is index in Instances object that is the second parameter.

Parameters:

instance - the instance to assign a cluster to

allPoints - all points

pointList - the list of points

Returns:

the index of the closest point

Throws:

Exception - if something goes wrong

valueIsSmallerEqual

public boolean valueIsSmallerEqual(weka.core.Instance instance,
                          int dim,
                          double value)

Returns true if the value of the given dimension is smaller or equal the value to be compared with.

Parameters:

instance - the instance where the value should be taken of

dim - the dimension of the value

value - the value to compare with

Returns:

true if value of instance is smaller or equal value