R help - Apr 2009 - arma model with garch errors

Dear R experts,
I am trying to estimate an ARMA 2,2 model with garch errors.
I used the following code on R 2.9.

#library
library(fGarch)
#data
data1<-ts(read.table("C:/Users/falcon/Desktop/Time
Series/exports/goods1.csv"), start=c(1992,1), frequency=12)
head(data1)

#garch
garchFit(formula.mean= ~arma(2,2),formula.var=~garch(1,1), data=data1)
but get this error:
> garchFit(formula.mean= ~arma(2,2),formula.var=~garch(1,1), data=data1)
Error in garchFit(formula.mean = ~arma(2, 2), formula.var = ~garch(1,  :
  element 1 is empty;
   the part of the args list of 'length' being evaluated was:
   (formula)

what am I doing wrong?   any help would be greatly appreciated
joe

	[[alternative HTML version deleted]]

>>>> "JM" == Joseph Magagnoli <jcm331 at gmail.com>
>>>> on Wed, 29 Apr 2009 14:54:26 -0500
   JM> > garchFit(formula.mean= ~arma(2,2),formula.var=~garch(1,1),
   JM> data=data1)
   JM> Error in garchFit(formula.mean = ~arma(2, 2), formula.var    JM>
~garch(1, :
   JM> element 1 is empty;
   JM> the part of the args list of 'length' being evaluated was:
   JM> (formula)
   JM>
   JM> what am I doing wrong?  any help would be greatly appreciated
   JM> joe


Hi joe,

If you are using a recent version of 'fGarch', you should use 

garchFit(~arma(2,2)+garch(1,1), data = data1)

More info in manual pages.

regards,
Yohan

-- 
PhD student
Swiss Federal Institute of Technology
Zurich

www.ethz.ch