[R] big model on modest computer.

[ripley@stats.ox.ac.uk]

On Fri, 29 Nov 2002, Corrin Lakeland wrote:

> On Fri, 29 Nov 2002 06:59, Ronaldo Reis Jr. wrote:
> > Hi,
> > I have a pentium III 600 with 128 ram running Linux.
>
> > I try this and fail. Memory problem.
>
> dd if=/dev/zero of=/tmp/tmpswap bs=1024 count=1000000
> mkswap /tmp/tmpswap
> swapon /tmp/tmpswap
>
> It will be _very_ slow, but it will work.

You know so?  My calculations made this a 1.9Gb design matrix with more
columns than rows, so

(a) my guess is that you will hit the 4Gb/process address limit in i686
Linux and

(b) it won't `work' in any statistically meaningful way. That model fits
all products of 0 to 4 continuous variables in each cell of a 22*14*4
(from memory) contingency table.

I think it is the statistical strategy that needs to be thought through.

-- 
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._