Problem with function "garchFit" (fSeries)
Mathias Slansky wrote:
I have a problem with the function "garchFit" of the package "fSeries". When I try to fit the garch model, the system goes to an endless loop and I don?t get a result. here?s what I wrote: garchFit(formula= ~garch(1,1), series = corb, include.mean=T) hope you con help me... thanks!
garchFit can take a very long time with a large series: long enough that I sometimes use it on a large fast server rather than on my laptop. Have you tried with include.mean=FALSE ? This will speed things up, and may let you know if you are on the right track, and it is worth letting the computer think for potentially a very long time. Have you tried with a different algorithm? The algorithms have different performance characteristics. Have you read the garch paper posted on the RMetrics website? It has some additional suggestions on speeding the process. Regards, - Brian