FitAR (version 1.94)

GetFitARz: Exact MLE for AR(p) and Subset ARz -- Short Version

Description

Obtain the exact MLE for AR(p) or subset ARz model. This function is used by FitAR and FitARz. One might prefer to use GetFitARz for applications such as bootstrapping since it is faster.

Usage

GetFitARz(z, pvec, MeanValue=0, ...)

Arguments

z

time series

pvec

lags included in AR model. If pvec = 0, white noise model assumed.

MeanValue

by default it is assumed the mean of z is 0

...

optional arguments passed through to optim

Value

loglikelihood

value of maximized loglikelihood

zetaHat

estimated zeta parameters

phiHat

estimated phi parameters

convergence

result from optim

pvec

lags of estimated AR coefficient

algorithm

"BFGS" or "Nelder-Mead"

Details

The built-in function optim is used to obtain the MLE estimates for an AR or subset AR. First "BFGS" is tried. This usually works fine. In the rare cases where convergence is not obtained, "Nelder-Mead" is used. A warning message is given if this happens.

References

McLeod, A.I. and Zhang, Y. (2006). Partial autocorrelation parameterization for subset autoregression. Journal of Time Series Analysis, 27, 599-612.

McLeod, A.I. and Zhang, Y. (2008a). Faster ARMA Maximum Likelihood Estimation, Computational Statistics and Data Analysis 52-4, 2166-2176. DOI link: http://dx.doi.org/10.1016/j.csda.2007.07.020.

McLeod, A.I. and Zhang, Y. (2008b, Submitted). Improved Subset Autoregression: With R Package. Journal of Statistical Software.

See Also

FitAR, FitARz, FitARp, GetFitARpMLE, RacfPlot

Examples

#compare results from GetFitARz and FitAR
z<-log(lynx)
z<-z - mean(z)
GetFitARz(z, c(1,2,8))
out<-FitAR(log(lynx), c(1,2,8), ARModel="ARz")
out
coef(out)