InformationMatrixARz: Fisher Information Matrix Subset Case, ARz
Description
Computes the large-sample Fisher information matrix per observation for the
AR coefficients in a subset AR when parameterized by the partial
autocorrelations.
Usage
InformationMatrixARz(zeta, lags)
Arguments
zeta
vector of coefficients, ie. partial autocorrelations
at lags specified in the argument lags
lags
lags in subset model, same length as zeta argument
Value
a p-by-p Toeplitz matrix, p=length(zeta)
Details
The details of the computation are given in
McLeod and Zhang (2006, eqn 13).
FitAR uses InformationMatrixARz to obtain estimates
of the standard errors of the estimated parameters in the subset
AR model when partial autocorrelation parameterization is used.
References
McLeod, A.I. and Zhang, Y. (2006).
Partial autocorrelation parameterization for subset autoregression.
Journal of Time Series Analysis, 27, 599-612.