com.rapidminer.tools.math.similarity.numerical
Class DTWDistance

java.lang.Object
  com.rapidminer.tools.math.similarity.DistanceMeasure
      com.rapidminer.tools.math.similarity.numerical.DTWDistance

All Implemented Interfaces:

java.io.Serializable

public class DTWDistance
extends DistanceMeasure

A distance measure based on "Dynamic Time Warping". The DTW distance is mapped to a similarity measure using f(x)= 1 - (x / (1 + x)). Feature weights are also supported.

Author:

Piotr Kasprzak, Sebastian Land

See Also:

Serialized Form

Constructor Summary

DTWDistance()

Method Summary

double	calculateDistance(double[] value1, double[] value2) This method does the calculation of the distance between two double arrays.
double	calculateSimilarity(double[] value1, double[] value2) This method does the similarity of the distance between two double arrays.
void	init(ExampleSet exampleSet) Before using a similarity measure, it is needed to initialize.
protected double	pointDistance(int i, int j, double[] ts1, double[] ts2)
java.lang.String	toString()

Methods inherited from class com.rapidminer.tools.math.similarity.DistanceMeasure

calculateDistance, calculateDistance, calculateSimilarity, calculateSimilarity, init, installAdditionalPorts, isDistance, isSimilarity, uninstallAdditionalPorts

Methods inherited from class java.lang.Object

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

Constructor Detail

DTWDistance

public DTWDistance()

Method Detail

pointDistance

protected double pointDistance(int i,
                               int j,
                               double[] ts1,
                               double[] ts2)

calculateDistance

public double calculateDistance(double[] value1,
                                double[] value2)

Description copied from class: DistanceMeasure

This method does the calculation of the distance between two double arrays. The meanings of the double values might be remembered from the init method.

Specified by:

calculateDistance in class DistanceMeasure

Returns:

the distance

calculateSimilarity

public double calculateSimilarity(double[] value1,
                                  double[] value2)

Description copied from class: DistanceMeasure

This method does the similarity of the distance between two double arrays. The meanings of the double values might be remembered from the init method.

Specified by:

calculateSimilarity in class DistanceMeasure

Returns:

the distance

init

public void init(ExampleSet exampleSet)
          throws OperatorException

Description copied from class: DistanceMeasure

Before using a similarity measure, it is needed to initialize. Subclasses might use initializing for remembering the exampleset properties like attribute type or test if applicable to exampleSet at all. Please note that it might be necessary to also override the other init methods if this measure should make use of parameters or other IOObjects.

Specified by:

init in class DistanceMeasure

Parameters:

exampleSet - the exampleset

Throws:

OperatorException

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object