com.numericalmethod.suanshu.stats.timeseries.linear.univariate.stationaryprocess.arima.arma
Class LinearRepresentation

java.lang.Object
  com.numericalmethod.suanshu.stats.timeseries.linear.univariate.stationaryprocess.arima.arma.LinearRepresentation

public class LinearRepresentation
extends java.lang.Object

This class computes the linear representation of an Autoregressive Moving Average (ARMA) model. A linear representation writes the time series as an (infinite) AR model.

See Also:

"P. J. Brockwell and R. A. Davis, "Eqs. 3.3.3, 3.3.4, Theorem 3.1.1., Chapter 11.3 Multivariate ARMA Processes," in Time Series: Theory and Methods, Springer, 2006."

Field Summary

static int

DEFAULT_NUMBER_OF_LAGS the default number of lags

Constructor Summary

LinearRepresentation(ArmaModel arma)
Construct the linear representation of an ARMA model.

LinearRepresentation(ArmaModel arma, int nlags)
Construct the linear representation of an ARMA model.

Method Summary

double[]

psi() Get a copy of the linear representation coefficients.

Methods inherited from class java.lang.Object

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

Field Detail

DEFAULT_NUMBER_OF_LAGS

public static final int DEFAULT_NUMBER_OF_LAGS

the default number of lags

See Also:

Constant Field Values

Constructor Detail

LinearRepresentation

public LinearRepresentation(ArmaModel arma,
                            int nlags)

Construct the linear representation of an ARMA model.

Parameters:

arma - an ARMA model

nlags - number of lags

LinearRepresentation

public LinearRepresentation(ArmaModel arma)

Construct the linear representation of an ARMA model.

Parameters:

arma - an ARMA model

Method Detail

psi

public double[] psi()

Get a copy of the linear representation coefficients.

Returns:

{ψ_i}