On Jan 16, 2011, at 10:41 PM, Simon Wotherspoon wrote:
>
> I believe there is a memory leak in isoreg in the current version of R, as
I believe the following shows
>
> > gc()
> used (Mb) gc trigger (Mb) max used (Mb)
> Ncells 120405 3.3 350000 9.4 350000 9.4
> Vcells 78639 0.6 786432 6.0 392463 3.0
> > for(k in 1:100) {
> +
> + y <- runif(10000)
> + isoreg(x,y)
> + }
> > rm(x)
> > rm(y)
> > gc()
> used (Mb) gc trigger (Mb) max used (Mb)
> Ncells 121637 3.3 350000 9.4 350000 9.4
> Vcells 578615 4.5 1300721 10.0 1300642 10.0
> ^^^
>
>
> Looking at the C code, I believe the problem arises as a consequence of
using SETLENGTH to resize the result near the very end of isoreg.c,
> and the solution is to make a copy of iKnots.
>
AFAICS this is more of a reporting issue than a memory leak -
ReleaseLargeFreeVectors() uses LENGTH() to determine the number by which to
reduce R_LargeVallocSize - and it will be smaller than the actually released
memory if SETLENGTH was used. I'm not sure how to fix it, because AFAICS the
real size is not recorded anywhere (unless we use truelength for that but
I'm not sure I know the implications). The good news is that it's not
really a memory leak - the bad news is that the memory usage stats are rather an
upper bound of the reality ;).
Cheers,
Simon
>
> SEXP R_isoreg(SEXP y)
> {
> int n = LENGTH(y), i, ip, known, n_ip;
> double tmp, slope;
> SEXP yc, yf, iKnots, ans;
> const char *anms[] = {"y", "yc", "yf",
"iKnots", ""};
>
> /* unneeded: y = coerceVector(y, REALSXP); */
>
> PROTECT(ans = mkNamed(VECSXP, anms));
>
> SET_VECTOR_ELT(ans, 0, y = y);
> SET_VECTOR_ELT(ans, 1, yc = allocVector(REALSXP, n+1));
> SET_VECTOR_ELT(ans, 2, yf = allocVector(REALSXP, n));
> SET_VECTOR_ELT(ans, 3, iKnots= allocVector(INTSXP, n));
>
> ... calculation ...
>
> SETLENGTH(iKnots, n_ip);
> UNPROTECT(1);
> return(ans);
> }
>
>
> But if this is the problem, I am at a bit of a loss as to what SETLENGTH is
actually for in general.
>
> Clearly my understanding of how allocation/gc works is a bit off here, but
I can't see how else the leak may occur. Hope this is more use than
nuisance.
>
> Simon.
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
>