I'm not trying to start a Windows vs. Linux debate, but I've been using R on a Windows machine for a while, and was recently wondering if R's performance would be faster on a Linux machine. And similarly, if any incremental increase in processing speed would be worth the time it would take me to migrate my entire system to Linux (including a database that I access via an R package.) I don't know how much it matters what R is doing - but I've got R pulling a large amount data from a database, performing many complex computations on that data, and then writing output data to a database. Thanks so much for the input, Brigid
On Fri, Mar 18, 2011 at 11:00 PM, Brigid Mooney <bkmooney at gmail.com> wrote:> I'm not trying to start a Windows vs. Linux debate, but I've been > using R on a Windows machine for a while, and was recently wondering > if R's performance would be faster on a Linux machine. ?And similarly, > if any incremental increase in processing speed would be worth the > time it would take me to migrate my entire system to Linux (including > a database that I access via an R package.) >It would help to know what is your system. A sessionInfo() would be a start. Liviu> I don't know how much it matters what R is doing - but I've got R > pulling a large amount data from a database, performing many complex > computations on that data, and then writing output data to a database. > > Thanks so much for the input, > Brigid > > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >-- Do you know how to read? http://www.alienetworks.com/srtest.cfm http://goodies.xfce.org/projects/applications/xfce4-dict#speed-reader Do you know how to write? http://garbl.home.comcast.net/~garbl/stylemanual/e.htm#e-mail
Hi Brigid, I haven't used it yet, but I would suggest profiling your code with 'Rprof' on both a Windows and Linux machine. ?Rprof for details. Hope this helps, ~Caitlin On Fri, Mar 18, 2011 at 3:00 PM, Brigid Mooney <bkmooney@gmail.com> wrote:> I'm not trying to start a Windows vs. Linux debate, but I've been > using R on a Windows machine for a while, and was recently wondering > if R's performance would be faster on a Linux machine. And similarly, > if any incremental increase in processing speed would be worth the > time it would take me to migrate my entire system to Linux (including > a database that I access via an R package.) > > I don't know how much it matters what R is doing - but I've got R > pulling a large amount data from a database, performing many complex > computations on that data, and then writing output data to a database. > > Thanks so much for the input, > Brigid > > ______________________________________________ > R-help@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide > http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >[[alternative HTML version deleted]]
One thing that Linux makes trivially easy is to interpolate R with C++ through the Rcpp package. The GCC compiler collection is part of all mainstream Linux distro. This is, however, not the case with Windows: you may be able to do it eventually (not sure about this point), but it takes quite some tweaks... Shige On Fri, Mar 18, 2011 at 6:00 PM, Brigid Mooney <bkmooney at gmail.com> wrote:> I'm not trying to start a Windows vs. Linux debate, but I've been > using R on a Windows machine for a while, and was recently wondering > if R's performance would be faster on a Linux machine. ?And similarly, > if any incremental increase in processing speed would be worth the > time it would take me to migrate my entire system to Linux (including > a database that I access via an R package.) > > I don't know how much it matters what R is doing - but I've got R > pulling a large amount data from a database, performing many complex > computations on that data, and then writing output data to a database. > > Thanks so much for the input, > Brigid > > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >
To try to answer the actual question .... I run x86_64 Linux and both 32-bit Windows XP and x64 Windows 7 on my home desktop. So from years of experience of using the same hardware on those OSes: 1) Generally Linux will be a bit faster, mainly because I do not hobble it with --enable-R-shlib (the only way R can be built on Windows). But the difference is small (5-20%). 2) Linux handles large amounts of memory (and especially swapping) better, so the differences will be more if what you do is using nearly all the address space or physical RAM of your computer. 3) Linux's process launch and disc access is many times faster, which makes a big difference to R development work.>From your described usage I would expect you to see a worthwhileperformance gain -- it does matter what you are doing with R. It also matters which database system. On Fri, 18 Mar 2011, Brigid Mooney wrote:> I'm not trying to start a Windows vs. Linux debate, but I've been > using R on a Windows machine for a while, and was recently wondering > if R's performance would be faster on a Linux machine. And similarly, > if any incremental increase in processing speed would be worth the > time it would take me to migrate my entire system to Linux (including > a database that I access via an R package.) > > I don't know how much it matters what R is doing - but I've got R > pulling a large amount data from a database, performing many complex > computations on that data, and then writing output data to a database. > > Thanks so much for the input, > Brigid > > ______________________________________________ > R-help at r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. >-- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595