ARToMA: Coefficients In Infinite Moving Average Expansion

In FitAR: Subset AR Model Fitting

Description

A stationary-causal AR(p) can be written as a general linear process (GLP). This function obtains the moving-average expansion out to the L-th lag, z[t] = a[t]+psi[1]*a[t-1]+...+psi[L]*a[t-L].

Usage

ARToMA(phi, lag.max)

Arguments

phi	AR Coefficient vector
lag.max	maximum lag

Details

The coefficients are computed recursively as indicated in Box and Jenkins (1970).

Value

Vector of length L+1 containing, (1,psi[1],...,psi[L])

Author(s)

A.I. McLeod and Y. Zhang

References

Box and Jenkins (1970), Time Series Analysis, Forecasting & Control

See Also

InvertibleQ

Examples

ARToMA(0.5,20)
ARToMA(c(0.2,0.5), 15)

FitAR documentation built on May 2, 2019, 3:22 a.m.