ai.libs.jaicore.ml.tsc.filter.derivate

Class ADerivateFilter

java.lang.Object

ai.libs.jaicore.ml.tsc.filter.derivate.ADerivateFilter

All Implemented Interfaces:

IFilter

Direct Known Subclasses:

BackwardDifferenceDerivate, ForwardDifferenceDerivate, GulloDerivate, KeoghDerivate

public abstract class ADerivateFilter
extends java.lang.Object
implements IFilter

Abstract superclass for all derivate filters.

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	withBoundaries Flag that states wheter the filter should add a padding to the derivate assure that is has the same length as the origin time series or not.

Constructor Summary

Constructors

Constructor and Description
ADerivateFilter()
ADerivateFilter(boolean withBoundaries)

Method Summary

Modifier and Type	Method and Description
protected abstract double[]	derivate(double[] t) Calculates the derivate of a time series.
protected abstract double[]	derivateWithBoundaries(double[] t) Calcuates the derivates of a time series.
void	fit(double[] input) The function only fits a single instance of the dataset
void	fit(double[][] input)
void	fit(TimeSeriesDataset input) the function computes the needed information for the transform function.
double[]	fitTransform(double[] input) the function fit and transforms a single instance
double[][]	fitTransform(double[][] input)
TimeSeriesDataset	fitTransform(TimeSeriesDataset input) a utility function to avoid the added effort of calling the fit and transform function separate
double[]	transform(double[] input) This function transforms only a single instance.
double[][]	transform(double[][] input)
TimeSeriesDataset	transform(TimeSeriesDataset input) represents a function working on a dataset by transforming the dataset itself.

Methods inherited from class java.lang.Object

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

Field Detail

withBoundaries

protected boolean withBoundaries

Flag that states wheter the filter should add a padding to the derivate assure that is has the same length as the origin time series or not.

Constructor Detail

ADerivateFilter

public ADerivateFilter(boolean withBoundaries)

ADerivateFilter

public ADerivateFilter()

Method Detail

derivate

protected abstract double[] derivate(double[] t)

Calculates the derivate of a time series.

Parameters:

t - The time series to calculate the derivate for.

Returns:

The derivate of the time series.

derivateWithBoundaries

protected abstract double[] derivateWithBoundaries(double[] t)

Calcuates the derivates of a time series. In contrast to the normal derivate calculation, this method is guaranteed to return a derivate that has the same length than the original time series. This is accomplished via padding.

Parameters:

t - The time series to calculate the derivate for.

Returns:

The, possibly padded, derivate of the time series.

transform

public TimeSeriesDataset transform(TimeSeriesDataset input)

Description copied from interface: IFilter

represents a function working on a dataset by transforming the dataset itself.

Specified by:

transform in interface IFilter

Parameters:

input - the data set that is to transform

Returns:

the transformt dataset

transform

public double[] transform(double[] input)

Description copied from interface: IFilter

This function transforms only a single instance.

Specified by:

transform in interface IFilter

Parameters:

input - the to transform instance

Returns:

the transformed instance

transform

public double[][] transform(double[][] input)

Specified by:

transform in interface IFilter

fit

public void fit(TimeSeriesDataset input)

Description copied from interface: IFilter

the function computes the needed information for the transform function.

Specified by:

fit in interface IFilter

Parameters:

input - the dataset that is to transform

fit

public void fit(double[] input)

Description copied from interface: IFilter

The function only fits a single instance of the dataset

Specified by:

fit in interface IFilter

Parameters:

input - The to fit instance

fit

public void fit(double[][] input)

Specified by:

fit in interface IFilter

fitTransform

public TimeSeriesDataset fitTransform(TimeSeriesDataset input)

Description copied from interface: IFilter

a utility function to avoid the added effort of calling the fit and transform function separate

Specified by:

fitTransform in interface IFilter

Parameters:

input - the dataset that is to be transfromed

Returns:

the transformed dataset

fitTransform

public double[] fitTransform(double[] input)

Description copied from interface: IFilter

the function fit and transforms a single instance

Specified by:

fitTransform in interface IFilter

Parameters:

input - the to fit and transform instance

Returns:

the transformed instance

fitTransform

public double[][] fitTransform(double[][] input)

Specified by:

fitTransform in interface IFilter