com.github.brandtg.stl

Class StlDecomposition

java.lang.Object

com.github.brandtg.stl.StlDecomposition

public class StlDecomposition
extends Object

This package contains an implementation of STL: A Seasonal-Trend Decomposition Procedure based on Loess.

Robert B. Cleveland et al., "STL: A Seasonal-Trend Decomposition Procedure based on Loess," in Journal of Official Statistics Vol. 6 No. 1, 1990, pp. 3-73

Constructor Summary

Constructors

Constructor and Description
StlDecomposition(int numberOfObservations) Constructs an STL function that can de-trend data.

Method Summary

Modifier and Type	Method and Description
StlResult	decompose(double[] times, double[] series) Computes the STL decomposition of a times series.
StlResult	decompose(List<Number> times, List<Number> series) A convenience method to use objects.
StlConfig	getConfig()
static void	main(String[] args) Runs STL on a CSV of time,measure.

Methods inherited from class java.lang.Object

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

Constructor Detail

StlDecomposition

public StlDecomposition(int numberOfObservations)

Constructs an STL function that can de-trend data.

n.b. The Java Loess implementation only does linear local polynomial regression, but R supports linear (degree=1), quadratic (degree=2), and a strange degree=0 option.

Also, the Java Loess implementation accepts "bandwidth", the fraction of source points closest to the current point, as opposed to integral values.

Parameters:

numberOfObservations - The number of observations in a season.

Method Detail

getConfig

public StlConfig getConfig()

Returns:

The configuration used by this function for fine tuning.

decompose

public StlResult decompose(List<Number> times,
                           List<Number> series)

A convenience method to use objects.

Parameters:

times - A sequence of time values.

series - A dependent variable on times.

Returns:

The STL decomposition of the time series.

decompose

public StlResult decompose(double[] times,
                           double[] series)

Computes the STL decomposition of a times series.

Parameters:

times - A sequence of time values.

series - A dependent variable on times.

Returns:

The STL decomposition of the time series.

main

public static void main(String[] args)
                 throws Exception

Runs STL on a CSV of time,measure.

Outputs a CSV of time,measure,trend,seasonal,remainder.

Parameters:

args - args[0] = numberOfObservations

Throws:

Exception - If could not process data