I was experimenting a bit with the memory saving trick mentioned on Snew
recently and I've discovered that adding NULL to the end of the loop in the
function below gives me "Bus Error - core dumped" in Solaris with R
0.61.1. The
problem does not occur every time the function is called, but it is reproducible
in the sense that I twice tried a set of test I have and it occured in the the
same spot, and when I remove the NULL it does not occur. I also added a NULL in
some other loops with no problems, but no obvious time savings in R. For my
purposes I will just remove the NULL, since this is going to be very difficult
for me to isolate (it takes a hour on an ultra 10 to get to the point where the
error occurs). I thought I would mention the problem in the hope that someone
can immediately see a bug.
Paul Gilbert
________
hessian.default <- function(func, x, func.args=NULL, d=0.01, eps=1e-4,r=6)
{ D <- genD.default(func,x,func.args=func.args, d=d, eps=eps, r=r)$D
H <- diag(0,length(x))
u <- 0
for(i in 1:length(x))
{for(j in 1:i)
{u <- u + 1
H[i,j] <- D[,u]
NULL
} }
H <- H +t(H)
diag(H) <- diag(H)/2
H
}
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
