[R] Memory use

[Prof Brian Ripley]

On Wed, 14 Jan 2009, Albrecht, Dr. Stefan (AIM SE) wrote:

> Dear R-help list,
>
> I am just trying to increase my memory for R, but for some reason I 
> cannot really increase the memory size as much as I would like. I 
> have 3 GB of RAM on a Windows XP (R 2.8.1)
>
>> memory.size()
>
> [1] 879.4667
>
>> memory.size(T)
>
> [1] 1162.75
>
>> memory.limit()
>
> [1] 3072
>
>> gc()
>
> used (Mb) gc trigger (Mb) max used (Mb)
>
> Ncells 421697 11.3 1476915 39.5 1476915 39.5
>
> Vcells 45503399 347.2 147241270 1123.4 147023097 1121.7

Yes, so you didn't use all the memory available: but to do so you 
probably need very many small objects.

> What am I missing?

The rw-FAQ?  Setting the limit to more than your address space (which 
you probably) have does not increase the latter.  Of your 3GB, Windows 
normally only gives 2GB to any one process.  See the rw-FAQ for the 
exceptions.

Also ?'Memory-limits'.

> With many thanks and best regards,
> Stefan
>
> ____________________________________
> Dr. Stefan Albrecht, CFA
> Allianz Investment Management SE
> AIM-GI-ALM
> Königinstr. 28 | 80802 Munich | Germany

-- 
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 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595