Houston R Users Group Message Board › Parallel processing
|A former member||
What is the easiest (and/or best) way to do parallel processing in R? I have a Mac i5 processor, if that makes a difference. When I run R, I think that only two of the four cores are being used. It seems that perhaps I can more fully take advantage of the processor.
I haven't had a need to get into parallel processing with R since I mainly deal with smaller datasets, but you may want to look at the rmpi package (http://www.stats.uwo..... There is a tutorial at http://math.acadiau.c...
Rmpi looks like it's designed for clustered computer farms (a la Beowulf clusters) rather than multicore chips. Don't see why it wouldn't work with a multicore chip the same way, but I'm not an expert.
Another package to check out would be the multicore package. JD Long of the Chicago R Users Group has a blogpost about it that lays out the basics (http://www.cerebralma.... He also hangs out on StackOverflow and twitter (@CMastication) and is pretty quick to respond to most inquiries.
Hope that helps,