To: Jim Lindsey <jlindsey@alpha.luc.ac.be>
Subject: Re: [R] a bug and a question
From: Peter Dalgaard BSA <p.dalgaard@biostat.ku.dk>
Date: 13 Jan 1999 11:15:00 +0100
In-Reply-To: Jim Lindsey's message of Wed, 13 Jan 1999 08:58:10 +0100 (MET)
Jim Lindsey <jlindsey@alpha.luc.ac.be> writes:
> The question: I am getting various reports that R under MS-Windows and
> X-Windows/Linux on the same machine behaves very differently for nlm
> convergence in my libraries. Under Linux, convergence is easy with
> many starting values, but under MS they have to be very close to the
> MLE. Does anyone have experience with this and or an explanation?
> Is the math used different in the two systems?
> Jim
Not entirely unlikely. The FPU should be the same of course, but the C
and math libraries link to a Microsoft DLL, which I have heard people
speak badly about before.
Particularly the IEEE exception handling could be causing trouble if
an iterative method wanders near the boundary of a functions domain.
That is pure speculation, though, I suppose Guido and Brian will know
more.
-- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk) FAX: (+45) 35327907 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request@stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._