[R] Date: Wed, 27 Nov 2002 09:23:31 -0600
ripley@stats.ox.ac.uk
ripley at stats.ox.ac.uk
Wed Dec 4 18:57:02 CET 2002
You could try reading the rw-FAQ: --max-mem-size=2G won't work, but
something like 1700M might. But on a 256Mb machine this is not much
point.
I successfully did a regression of that size in 42Mb. Something is wrong.
Maybe the `columns' are factors with vast numbers of levels, for example.
You really haven't told us anything like enough to go on.
On Wed, 4 Dec 2002, Chris Yoo wrote:
> As for the R memory problem I posted, I discussed it with a consultant but
> we could not find solution.
>
> Can anyone possibly answer this question?
>
> As I told you briefly, I have a large data set (1017 Mb and about3,600 kb in
> csv format). The data contains 14,013 rows (observation) and 44 columns.
> I am using Dell Notebook I8000 (got it last year). It has 251,600 KB RAM and
> C-drive where R folder is attached has 13.5GB free space.
>
>
> I am able to read the data but whenever I try the following, I got error
> message.
>
> > test <- lm(d.tdc2~d.mktval)
> >Error: cannot allocate vector of size 1017272 Kb
> >In addition: Warning message:
> >Reached total allocation of 1000Mb: see help(memory.size)
> >>
>
> I even tried to increase memory assigned by doing
>
> > memory.limit(2000)
> NULL
> > test <- lm(d.tdc2~d.mktval)
> Error: cannot allocate vector of size 1017272 Kb
> > memory.limit(3000)
> NULL
> > test <- lm(d.tdc2~d.mktval)
> Error: cannot allocate vector of size 1017272 Kb
> >
>
> I also tried
> TARGET C:\R\rw1061\bin\Rgui.exe --max-mem-size=2G
>
> Thanks a lot,
>
> Chris
>
> -----Original Message-----
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> http://www.stat.math.ethz.ch/mailman/listinfo/r-help
>
--
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
More information about the R-help
mailing list