[R] memory usage

[Prof Brian D Ripley]

On Fri, 21 Sep 2001 Setzer.Woodrow at epamail.epa.gov wrote:

> Does the following indicate that I have a memory leak?
>
> > gc(TRUE)
> Garbage collection 22891 = 21012+1557+322 (level 2) ...
> 483257 cons cells free (59%)
> 11.0 Mbytes of heap free (82%)
>          used (Mb) gc trigger (Mb)
> Ncells 334906  9.0     818163 21.9
> Vcells 309342  2.4    1746173 13.4
> > memory.size()/1024/1024 ## in MB
> [1] 643.8978
> > sum(sapply(ls(all.names=TRUE),object.size))
> [1] 1776
> > memory.limit()/1024/1024
> [1] 600
>
> I was repeatedly using try(eval(substitute(nlme(..)))), with lots of error
> aborts from nlme.

I don't think so.  From CHANGES:

  The Windows memory manager proved to be (far) too slow to work with
  the new generational garbage collector, so it has been replaced.  The
  current memory manager rarely releases memory back to Windows (but
  reuses memory very efficiently) so total memory usage will often grow
  slowly over the course of a session.  Memory used for transient large
  objects will normally be returned at garbage collection.  (Once memory
  usage has gone over the 256Mb of address space that is initially
  reserved, it is unlikely to be reduced much, but the allocated memory
  will be reused.)  You can compile R to use the Windows memory manager
  if you need to.

This behaviour is not ideal, but given the way Windows manages memory is
not really compatible with the Unix model of a single large block that can
be grown or shrunk by calls to sbrk, it is the best we can do with the
human resources available.  Getting this far was hard enough.

(The speed hit with the Windows memory manager was often three or more.)

>
> --please do not edit the information below--
>
> Version:
>  platform = i386-pc-mingw32
>  arch = x86
>  os = Win32
>  system = x86, Win32
>  status =
>  major = 1
>  minor = 3.1
>  year = 2001
>  month = 08
>  day = 31
>  language = R
>
> Windows 98 SE 4.10 (build 2222)  A
>
> Search Path:
>  .GlobalEnv, package:ctest, package:RBMDS, package:nls, package:nlme,
> Autoloads, package:base
>
> R. Woodrow Setzer, Jr.                                            Phone:
> (919) 541-0128
> Experimental Toxicology Division                       Fax:  (919) 541-5394
> Pharmacokinetics Branch
> NHEERL MD-74; US EPA; RTP, NC 27711
>
> -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
> r-help 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-help-request at stat.math.ethz.ch
> _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
>

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._