i have already got the answer from my friends when you load big data into memory, u often meet the memory problem. but in fact, you have enough physical memeory to load the data. that is because the linux use much of your free memory to cache the files becaues u do calculation you should run such command to set them free memory problem echo 3 > /proc/sys/vm/drop_caches -- shan gao Room 231(Dr.Fei lab) Boyce Thompson Institute Cornell University Tower Road, Ithaca, NY 14853-1801 Office phone: 1-607-254-1267(day) Official email:sg839 at cornell.edu Facebook:http://www.facebook.com/profile.php?id=100001986532253

On Sat, Jan 14, 2012 at 10:17 AM, wang peter <wng.peter at="" gmail.com=""> wrote: > i have already got the answer from my friends > > when you load big data into memory, u often meet the memory problem. > but in fact, you have enough physical memeory to load the data. > that is because the linux use much of your free memory to cache the files > becaues u do calculation > you should run such command to set them free > > memory problem > > echo 3 > /proc/sys/vm/drop_caches Hi, Shao Gan. Thank you for sharing, but I do not think this is either necessary or relevant for successful use of bioconductor. There is not a need to do this on modern linux systems, as the cache is automatically downsized as RAM for running programs is allocated. I would not expect the command above to in any way affect the amount of memory available to R/Bioconductor. In fact, emptying the cache may, in some cases, be detrimental to performance if one is, for example, reading or writing files. I just wanted to be clear that what you suggest is not a recommended approach to dealing with memory issues in R and is not likely to be successful (since the cache is not the problem). Sean