[R] a question about swap space, memory and read.table()

[Marcus Davy]

On a 32-bit windows (standard install or R) machine you cannot allocate more than 2Gigs
of memory,
"R --max-mem-size=2G"
or
"R --max-mem-size=2000M"

If you try to allocate more than 2G (eg 4000 Meg) I suspect it then defaults back to 1Gig. Check you memory limit with memory.limit(). There is lots of information about configuration in the windows FAQ, 
http://cran.r-project.org/bin/windows/base/rw-FAQ.html#There-seems-to-be-a-limit-on-the-memory-it-uses_0021

If you want 3Gigs per process you are going to have to modify the R executable to make it 
/LARGEADDRESSAWARE


marcus

>>> Hu Chen <chencheva at gmail.com> 10/12/2004 3:39:18 AM >>>
Hi all 
Two computers:
one is my desktop PC, windows2000, R 1.9.1. Physical RAM 256MB, Swap
(Virtual memory)  384Mb. When I allocate a large matrix, it firstly
uses up RAM, then use swap space. In windows' task manager, the usage
of memory could exceed my physic RAM's size.
The other machine is a remote server. Windows XP, R 1.9.1 Physical RAM 2GB.
Swap space 4GB. I use "R --max-mem-size=4000M" to start R. However
when I allocate a large matrix or data frame, it uses up all RAM then
exits with a error message" cannot allocate vector of size 7812 Kb ".
The Swap space is not used at all !
What's more, I found that the read.table() function is really a waste of memory.
> ft <- read.table("filepath")
> object.size(ft)
object.size(ft)
[1] 192000692
only 192Mb.
however, in the windows task manager it shows that this process takes
nearly 800Mb memory.
I used gc() to collect garbarge. Howerver it doesn't help.
Any guys have methods to release the wasted memory?
thank you all.
Regards

______________________________________________
R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help 
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html

______________________________________________________

The contents of this e-mail are privileged and/or confidenti...{{dropped}}