[R] .Rdata questions

Lorenz Gygax lgygax at access.unizh.ch
Mon Feb 15 08:58:16 CET 1999

Dear all,

in a current project I have a pretty huge .Rdata. Thus I was working with
R --vsize 100 --nsize 1000000. Today when I tried to restart R I get the
following error message:

Error: a read error occured
Fatal error: unable to restore saved data
 (remove .RData or increase memory)

I increased memory up to --vsize 180 and --nsize 2000000, but the error
reoccurrs. Is there a way to know whether this error results from a
corrupt .Rdata or whether it is really a question of memory?

And, are there good strategies for not loosing big .Rdata (except for
backups between sessions and regular save.image() within sessions) ?

That is, I guess, one of the few things I prefer in S-plus: firstly, there
is no big file containing all the objects, but the objects are in a
directory .Data (so they can be managed there and if an object is
corrupted you don't loose all your objects) and secondly, that every
object is saved as soon as it is created.

I am mostly working with ESS and it happened to me that after q () I quit
emacs too quickly and lost an .Rdata and some other time I did an C-x k on
the R buffer (which is, admittedly, pretty stupid). Are there ways to get
emacs to need a confirmation if one is trying to kill an active buffer
(well, actually, it needs one normally, doesn't it)? 

Thanks for any help and/or pointers!
--			~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Lorenz Gygax		 LGygax at amath.unizh.ch;       room: 36-L-40
			 Department of Applied Mathematics
			 University of Zuerich-Irchel
			 Winterthurerstr. 190; CH-8057 Zurich
			 voice: 41-1-635-58-52  fax: 41-1-635-57-05

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

More information about the R-help mailing list