Hi Hegang, Windows can handle 2Gb of RAM by default. It's possible to get it to allocate 3Gb, but I've never tried it. 32bit linux handles up to 4Gb and 64bit linux is basically only limited by physical memory. I don't use affy chips personally, maybe other people could help you further. I would find it suspicious that justGCRMA fail for large number of arrays when enough memory is available (ie error other than "unable to allocate vector of size ..."). You might want to post the result of sessionInfo() and the exact code and error message that you're getting. Francois On Thu, 2006-11-02 at 11:49 -0800, Henry Chen wrote: > > Hi Francois, > > I indeed used justGCRMA. It worked for justRMA. > It appears that justGCRMA does not work for a large number > of HG-U133plus2 chips. I could try GCRMA by running Linux computer. > Do you think that would work? > > Hegang > > > Francois Pepin <fpepin at="" cs.mcgill.ca=""> wrote: > Hi Hegang, > > This is happening because you're running out of memory. > > Using justGCRMA instead of GCRMA should help solve your > problem. R is > kind of bad at memory management and the justGCRMA function is > coded in > C to get around this. > > Otherwise you could also try to work on a computer with more > RAM, but > keep in mind that windows has limitation as to how much memory > can be > used. > > There also is a lot of discussion in the list archives if you > want more > details. > > hope this helps, > > Francois > > On Thu, 2006-11-02 at 10:46 -0800, Henry Chen wrote: > > Hi all, > > > > I have faced a problem on running GCRMA. > > I have 67 chips of hgu133plus2 and try to calculate > expression measures > > and normalize them by GCRMA. However, there is an error as > > follows > > > > "Adjusting for optical effect..Error: cannot allocate vector > of size 297408 Kb" > > > > Does anyone know how to solve this problem. Thanks! > > > > > > > > > > Hegang H. Chen, Ph.D. > > Associate Professor > > Division of Biostatistics and Bioinformatics > > Dept of Epidemiology and Preventive Medicine > > University of Maryland School of Medicine > > 660 W. Redwood > > Baltimore, MD 21201 > > U.S.A. > > > > --------------------------------- > > > > [[alternative HTML version deleted]] > > > > _______________________________________________ > > Bioconductor mailing list > > Bioconductor at stat.math.ethz.ch > > https://stat.ethz.ch/mailman/listinfo/bioconductor > > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > > > > > > > > ______________________________________________________________________ > Cheap Talk? Check out Yahoo! Messenger's low PC-to-Phone call rates.