Skip to content

parameters for lbfgsb (function for optimization)

3 messages · Kyeongmi Cheon, Bill Venables, Katie C

Bill Venables
# 
Whoa!  If you want to optimise in R, you need to write an R function to
define your objective.  If that function is already written in C, as
appears to be the case, you need to write an interfact to make it
available to R.  This is not exactly something novices would take on
lightly.

If you wish to optimise directly in C, then you should look for software
support to do it that was intended for the purpose.  The R source code
is not like that.  You can use it if you want, but you must expect it to
be difficult and very puzzling, since it is not intended to be
free-standing, but part of a large, complex system.

Best of luck. 


Bill Venables
CSIRO Laboratories
PO Box 120, Cleveland, 4163
AUSTRALIA
Office Phone (email preferred): +61 7 3826 7251
Fax (if absolutely necessary):  +61 7 3826 7304
Mobile:                         +61 4 8819 4402
Home Phone:                     +61 7 3286 7700
mailto:Bill.Venables at csiro.au
http://www.cmis.csiro.au/bill.venables/ 

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
On Behalf Of Kyeongmi Cheon
Sent: Friday, 7 March 2008 3:06 PM
To: r-help at r-project.org
Subject: [R] parameters for lbfgsb (function for optimization)

Can anyone help me with lbfgsb (function for optimization)?

It takes the following parameters:

void lbfgsb (int n, int lmm, double *x, double *lower,
double *upper, int *nbd, double *Fmin, optimfn fn,
optimgr gr, int *fail, void *ex, double factr,
double pgtol, int *fncount, int *grcount,
int maxit, char *msg, int trace, int nREPORT);

What do I put for parameter ex (11th parameter)? I looked at optim.c
codes
at R sites and it's a structure that has bunch of objects such as SEXP
R_fcall,  SEXP R_gcall, SEXP R_env, double* ndeps, etc. I cannot figure
out
what it is about.
How about fncount, rcount? R-ext.pdf or R help does not explain it in
detail. Any comments would help. Thank you.


______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Katie C
# 
I think I did not make myself clear when I asked about lbfgsb (function for
optimization) yesterday. I'm writing my functions in C but actual
calculation will be done in R by using .Call.
To be able to use lbfgsb (using the interfaces defined in header
'R_ext/Applic.h') , I need to supply my function and gradient function. And
the instruction in R-ext.pdf (p97) says that my function should have three
parameters(int n, double *par, void *ex) and gradient function should have
four parameters(int n, double *par, double *gr, void *ex).
My question is :  what is *ex (the last parameter) for? and what values
should I put for it? How about fncount and rcount (14 and 15th parameter of
lbfgsb)? Thank you in advance.
Kyeongmi
Bill.Venables wrote: