specFactmvar: Compute the spectral factorization of a given multivariate...

View source: R/specFactmvar.r

specFactmvarR Documentation

Compute the spectral factorization of a given multivariate p.d. sequence

Description

Compute the spectral factorization of a given multivariate p.d. sequence

Usage

specFactmvar(xAcf)

Arguments

xAcf

An array containing the given p.d. sequence. It has dimension N x (q+1) x N, where xAcf[,1,] corresponds to the zeroth coefficient matrix, xAcf[,2,] is the first coefficient matrix, etc.

Value

theta: the spectral factorization as a list object theta[[1]] is an array of dimension N x N x (q+1), consisting of the VMA coefficients from index q through index zero (Note that theta[[1]][,,q+1] should be an identity matrix.) theta[[2]] is the white noise covariance matrix theta[[3]] tracks the number of iterations used in this algorithm, which is based on Bauer's method, as described in McElroy (2018, JTSA)


jlivsey/sigex documentation built on March 20, 2024, 3:17 a.m.