Skip to content

Is this a bug in R?

2 messages · Revilla,AJ (pgt), Brian Ripley

#
Dear all,

I am trying to fit a nonlinear model with a autocorrelation term, but everytime I type in the command, I got an error message from Winwows and R closes itself.
The command line is as follows:

mod1<-nlme(V~A*exp(-B*A.O)*Vac.t.1.,data,fixed=A+B~1,random=A+B~1|ORDINAL,+
correlation=corCAR1(0.3179,~A.O|ORDINAL,TRUE),start=c(A=1.2,B=0.2))

I have already fitted this model allowing Phi to vary while optimizing, and it was fine, but as soon as I try to keep it fixed (argument "TRUE"), I simply can't

I don't get any error message from R, just a Windows error seying something like "R for windows GUI front-end has detected a problem and has to close". And that?s it, R is over!

I don't know if I am doing anything wrong, or if it has to be with my system (I have Windows XP Pro), but it looks like a bug in R.

Do you know anything else about this. Thank you very much,

Antonio
#
You are in fact using the contributed package 'nlme', not just R.

Please read both the section on BUGS in the FAQ and the posting guide, and 
send a reproducible example to the nlme maintainer.

One thing the posting guide asks for is a useful subject line.  Something 
like

 	`A crash when using nlme'.
On Thu, 28 Apr 2005, Revilla,AJ (pgt) wrote: