Symbolic derivation

[Douglas Bates]

Wenyaun Xi <xi@keynes.econ.utah.edu> writes to S-news:

> I ran into some problems when I tried to get symbolic derivation by 
> using a function deriv3(). The expression contains 18 
> parameters, of which I want to have an information matrix eventually. I 
> ran it on my 200MHz Pentium PC with 64meg RAM. Each time I kick out the 
> job, the S-plus 3.1 for Windows stops after about 2 hour of real time 
> and suggests that it requires more memory(size varies each time) in 
> addition to those already in use. I don't know if this is a memory related 
> problem or there is a whole different issue involved here other than the 
> memory.

May I suggest sending a message on this to R-devel@stat.math.ethz.ch?
I am sending a cc: of this message to that group.

R is a freely available language that behaves like S but has a
fundamentally different underlying implementation.  It is possible
that your problem that causes difficulty in S will be easier to run in 
R.  The issues are probably memory related and R does "garbage
collection" but S does not.  Also, deriv in R is implemented in the
underlying C code rather than in the S code.

If you want to install R yourself you can check the CRAN
(Comprehensive R Archive Network) sites.  One close to you is
	   ftp://ftp.biostat.washington.edu/mirrors/R/CRAN/
Otherwise, you could post the call to deriv3() that you are trying to
evaluate to the R-devel list and someone could try to run it for you.
-- 
Douglas Bates                            bates@stat.wisc.edu
Statistics Department                    608/262-2598
University of Wisconsin - Madison        http://www.stat.wisc.edu/~bates/
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel 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-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._