fGarch: possible bug in .garchLLH

[chalabi at phys.ethz.ch]

michal miklovic <mmiklovic at yahoo.com> writes:

hi Michal,

thanks to point this out!

By the way, have you noticed that we moved all Rmetrics development packages
to R-Forge. You can access the new subversion server with the following
commands :

svn checkout svn://svn.r-forge.r-project.org/svnroot/rmetrics

regards,
Yohan

> Hi,
>
> I wanted to estimate a garch-gjr model without leverage (which is, in fact, a garch model) using the following code:
>
> arma11gjr11n <- garchFit(~ arma(1,1) + aparch(1,1), data, delta=2,
>             include.delta=FALSE, leverage=FALSE, algorithm="nlminb", hessian="rcd")
> summary(arma11gjr11n)
>
> and obtained the following error:
>
> --- START OF TRACE ---
>
> R coded nlminb Solver: 
>
> Error in gamma[i] : object is not subsettable
>
>
> I would suggest to copy the below line from .garchInitParameters into .garchLLH:
>
> if(p > 0 & !leverage) gamma = rep(0, times = p)
>
> I tried this, recompiled the fGarch package and it worked fine.
>
>
> Best regards,
>
> Michal Miklovic
>
>
>
>
>
>
>       ____________________________________________________________________________________
>
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> Rmetrics-core mailing list
> Rmetrics-core at r-project.org
> https://stat.ethz.ch/mailman/listinfo/rmetrics-core

-- 


The 2nd International R/Rmetrics User and Developer Workshop ...
[http://www.rmetrics.org]