[BioC] justGCRMA memory problem

[James W. MacDonald]

Jianping Jin wrote:
> Hi Jim,
> FYI I ran justRMA on R 2.1.0 and it was fine:
> 
> 
>>files <- dir(pattern="CEL")
>>data<- justRMA(filenames=files)
> 
> Background correcting
> Normalizing
> Calculating Expression
> 
>>memory.size(max=TRUE)
> 
> [1] 1536868352
> 
>>memory.limit()
> 
> [1] 1572864000
> 
> I ran into the same problem when I ran justGCRMA on R 2.1.0 as on R 2.2.0.

Yeah, I am a little thick today it seems. Your error message was 'cannot 
allocate vector of size 424868 Kb' and your memory.size(max = TRUE) == 
1194139648

You add those up (noting that the vector is listed in *Kb*), and you get
1618007648 bytes, which is probably less RAM than you have available for 
R (Windows XP takes quite a bit of RAM for itself).

You could always get more RAM, but you will have to make R 'large 
address aware' to be able to access more than 2 Gb (IIRC). I believe 
this means you have to build R from source as well. There is something 
in the R-FAQ for Windows about this issue.

Best,

Jim



> 
> regards,
> 
> JP-
> 

-- 
James W. MacDonald
Affymetrix and cDNA Microarray Core
University of Michigan Cancer Center
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109
734-647-5623