com.imsl.stat

Class TimeSeriesOperations

java.lang.Object

com.imsl.stat.TimeSeriesOperations

All Implemented Interfaces:

Serializable, Cloneable

public class TimeSeriesOperations
extends Object
implements Serializable, Cloneable

A class of operations and methods for objects of class TimeSeries.

See Also:

Example 1, Example 2, Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	TimeSeriesOperations.CombineMethod Public enum of methods for combining synchronous time series values.
static interface	TimeSeriesOperations.Function Public interface for the user-supplied function that defines how to combine two synchronous time series values.
static class	TimeSeriesOperations.MergeRule Public enum of merge rules that defines how two time series should be merged.

Constructor Summary

Constructors

Constructor and Description
TimeSeriesOperations() Constructor for TimeSeriesOperations.

Method Summary

Modifier and Type	Method and Description
TimeSeries	backshift(TimeSeries ts, int lag) Returns the backshifted version of the time series.
double[][]	getMeanCenteredSeries(TimeSeries timeSeries) Returns the mean-centered values of all the series in a TimeSeries object.
double[]	getMeanCenteredSeries(TimeSeries timeSeries, int k) Returns the mean-centered values of the k-th series in a TimeSeries object.
TimeSeries	merge(TimeSeries ts1, TimeSeries ts2) Merges two time series objects.
void	setCombineFunction(TimeSeriesOperations.Function function) Sets the combine function to a user supplied function.
void	setCombineMethod(TimeSeriesOperations.CombineMethod combineMethod) Sets the method for combining synchronous time series values.
void	setMergeRule(TimeSeriesOperations.MergeRule mergeRule) Sets the rule that defines how two time series are merged.
double[]	stack(TimeSeries ts) Stacks or vectorizes the values of a multivariate TimeSeries.

Methods inherited from class java.lang.Object

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

Constructor Detail

TimeSeriesOperations

public TimeSeriesOperations()

Constructor for TimeSeriesOperations.

Method Detail

getMeanCenteredSeries

public double[][] getMeanCenteredSeries(TimeSeries timeSeries)

Returns the mean-centered values of all the series in a TimeSeries object.

Parameters:

timeSeries - a TimeSeries object

Returns:

a double matrix containing the mean-centered values

getMeanCenteredSeries

public double[] getMeanCenteredSeries(TimeSeries timeSeries,
                                      int k)

Returns the mean-centered values of the k-th series in a TimeSeries object.

Parameters:

timeSeries - a TimeSeries object

k - an int, the column index of the series

Returns:

a double array containing the mean-centered values

setMergeRule

public void setMergeRule(TimeSeriesOperations.MergeRule mergeRule)

Sets the rule that defines how two time series are merged.

Parameters:

mergeRule - a MergeRule value.

setCombineMethod

public void setCombineMethod(TimeSeriesOperations.CombineMethod combineMethod)

Sets the method for combining synchronous time series values.

Parameters:

combineMethod - a CombineMethod value.

setCombineFunction

public void setCombineFunction(TimeSeriesOperations.Function function)

Sets the combine function to a user supplied function. Must be defined when combineMethod = CombineMethod.CUSTOM.

Parameters:

function - a Function that defines how two values should be combined.

backshift

public TimeSeries backshift(TimeSeries ts,
                            int lag)

Returns the backshifted version of the time series. For a given lag \(l \ge 0 \) the backshift operation is defined as: $$ B^ly_t = y_{t-l} $$.

Parameters:

ts - a TimeSeries object.

lag - an int specifying the lag.

Returns:

TimeSeries with values at index i the same as the original series values at position i-lag;

stack

public double[] stack(TimeSeries ts)

Stacks or vectorizes the values of a multivariate TimeSeries.

Parameters:

ts - TimeSeries

Returns:

A double[] array of length T*numberOfSeries with the index 1,...,numberOfSeries varying fastest.

merge

public TimeSeries merge(TimeSeries ts1,
                        TimeSeries ts2)

Merges two time series objects.

Parameters:

ts1 - a TimeSeries object.

ts2 - a TimeSeries object.

Returns:

a TimeSeries object resulting from merging ts1 and ts2.