tsdecomp (version 0.2)

acov2ma: Convert Autocovariances to Coefficients of a Moving Average

Description

Convert autocovariances to coefficients of a moving average.

Usage

acov2ma.init(x, tol = 0.00001, maxiter = 100)
acov2ma(x, tol = 1e-16, maxiter = 100, init = NULL)

Arguments

x

a numeric vector containing the autocovariances.

tol

numeric, convergence tolerance.

maxiter

numeric, maximum number of iterations.

init

numeric, vector of initial coefficients.

Value

A list containing the vector of coefficients and the variance of the innovations in the moving average model; convergence code and number of iterations.

Details

acov2ma.init is based on procedure (17.35) described in Pollock (1999). acov2ma is the Newton-Raphson procedure (17.39) described in the same reference.

References

Pollock, D. S. G. (1999) A Handbook of Time-Series Analysis Signal Processing and Dynamics. Academic Press. Chapter 17. \Sexpr[results=rd,stage=build]{tools:::Rd_expr_doi("#1")}10.1016/B978-012560990-6/50002-6http://doi.org/10.1016/B978-012560990-6/50002-6doi:\ifelse{latex}{\out{~}}{ }latex~ 10.1016/B978-012560990-6/50002-6

Examples

set.seed(123)
x <- arima.sim(n=200, model=list(ma=c(0.7,-0.3)))
#sample autocovariances
a <- c(var(x), cov(x[-1], x[-200]), cov(x[-c(1,2)], x[-c(199,200)]))
#inferred coefficients and variance
acov2ma(a, init=acov2ma.init(a, maxit=10)$macoefs)
#compare with maximum-likelihood
arima(x, order=c(2,0,0), include.mean=FALSE)