[Rd] f2c problem
Wed, 20 Feb 2002 11:27:46 +0100 (CET)
On Wed, 20 Feb 2002, Stefano Iacus wrote:
> In building princurve package for MacOS I need to use f2c on fortran
> Here is the output of f2c for getlam.f
> Error on line 116 of getlam.f: Declaration error for v: adjustable
> dimension on non-argument
> Error on line 116 of getlam.f: wr_ardecls: nonconstant array size
> what I found strange is that both newlam and get lam have definitions
> subroutine newlam(n,p,sx,lambda,tag)
> integer n,p,tag(n)
> double precision sx(n,p),lambda(n),lami
> but for some reason f2c complains only with the last lamix subroutine
> subroutine lamix(ns,p,x,s,lambda,dismin,temps)
> integer ns,p
> double precision lambda,x(p),s(ns,p),dismin,temps(p)
> double precision v(2,p),d1sqr,d2sqr,d12,dsqr, d1,w
> probably because v is not passed as an argument.
Definitely because v is local; this construct is a gcc extension to the
standard, I think. Solutions: 1) pass v as an argument and reserv space
in the calling function. 2) Instead of p, use 1000 (or some other
large number, which p never will exceed). 3) Use gcc.
> Any idea on how to manage this in f2c ?
> 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: email@example.com
Göran Broström tel: +46 90 786 5223
professor fax: +46 90 786 6614
Department of Statistics http://www.stat.umu.se/egna/gb/
SE-90187 Umeå, Sweden e-mail: firstname.lastname@example.org
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: email@example.com