An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20120206/5c2d10fc/attachment.pl>
R enterprise for linux
10 messages · Alaios, Paul Hiemstra, R. Michael Weylandt +3 more
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20120206/f44b5554/attachment.pl>
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20120206/f88951a0/attachment.pl>
Are you sure? I just ran N <- 128 system.time(matrix(rnorm(N^2),N) %*% matrix(rnorm(N^2),N)) and it took less than 0.044s on my (old-ish) laptop while doing other things (and that includes the expensive rng calls). There might be some other issues in play here. Even N <- 1280 takes < 5 seconds with the rng call. Michael
On Mon, Feb 6, 2012 at 9:29 AM, Alaios <alaios at yahoo.com> wrote:
Actually, I would like to speed up matrix multiplication which is really too slow. MatrixA %*% MatrixB (each one is a [128,128] table) is running four hours now... and just imagine that I want to calculate many of those. Which other alternative do I have? Regards Alex
________________________________ ?From: Paul Hiemstra <paul.hiemstra at knmi.nl> Cc: "R-help at r-project.org" <R-help at r-project.org> Sent: Monday, February 6, 2012 2:30 PM Subject: Re: [R] R enterprise for linux Dear Alex, By Enterprise R, do you mean Revolution R? I can also only find red ? ?hat packages. You could send them an e-mail to ask for sources. They ? ?probably want to help you, especially when you pay. But why do you ? ?need RevoR and not standard R. Paul On 02/06/2012 12:55 PM, Alaios wrote: Dear all, I would like to ask you if it would be possible? to install R enterprise in opensuse. Actually it seems that there are only Red Hat packages and I was wondering if their source code is available so one to try and compile it. I would like to thank you in advance for your reply Regards Alex [[alternative HTML version deleted]] ______________________________________________ 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. -- Paul Hiemstra, Ph.D. Global Climate Division Royal Netherlands Meteorological Institute (KNMI) Wilhelminalaan 10 | 3732 GK | De Bilt | Kamer B 3.39 P.O. Box 201 | 3730 AE | De Bilt tel: +31 30 2206 494 http://intamap.geo.uu.nl/~paul http://nl.linkedin.com/pub/paul-hiemstra/20/30b/770 ? ? ? ?[[alternative HTML version deleted]] ______________________________________________ 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.
On Mon, 6 Feb 2012, Alaios wrote:
Actually, I would like to speed up matrix multiplication which is really too slow.
I believe that R supports openmp during configuration. This is the multi-processor library that allows threads to run on different cores or processors. In addition to Red Hat packages you can build Slackware packages using the scripts at www.slackbuilds.org. Rich
6-02-2012, 06:29 (-0800); Alaios escriu:
Actually, I would like to speed up matrix multiplication which is really too slow. MatrixA %*% MatrixB (each one is a [128,128] table) is running four hours now... and just imagine that I want to calculate many of those.
It doesn't seem normal to me... in my computer such a multiplication
takes a fraction of a second:
system.time(array(rnorm(128*128), c(128,128)) %*%
array(rnorm(128*128), c(128,128)))
user system elapsed
0.008 0.000 0.006
Am I missing something?
Cheers,
Ernest
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20120206/6966f329/attachment.pl>
On Feb 6, 2012, at 10:17 AM, Alaios wrote:
My bad.. these are the real matrices
G<-DiagonalGaussian( NumberOfImagePixels,sd=Gsd) str(G)
num [1:16384, 1:16384] 1400 0 0 0 0 ...
S<-PointSamplingMatrix(EffectiveSensors,NumberOfImagePixels) str(S)
num [1:10650, 1:16384] 0 0 0 0 0 0 0 0 0 0 ...
A<-S %*% G
You seems confused about the size of your objects. You earlier said they were [128,128] which we took to mean 128 x 128 matrices, they are far larger than that. I'm guessing that you are consuming all of your RAM and paging out to disk. But that is just a guess since you provided none of the system information that the Posting Guide suggests.
David. > _______________________________ > From: R. Michael Weylandt <michael.weylandt at gmail.com> > > Cc: Paul Hiemstra <paul.hiemstra at knmi.nl>; "R-help at r-project.org" <R-help at r-project.org > > > Sent: Monday, February 6, 2012 3:39 PM > Subject: Re: [R] R enterprise for linux > > Are you sure? > > I just ran > > N <- 128 > system.time(matrix(rnorm(N^2),N) %*% matrix(rnorm(N^2),N)) > > and it took less than 0.044s on my (old-ish) laptop while doing other > things (and that includes the expensive rng calls). There might be > some other issues in play here. Even N <- 1280 takes < 5 seconds with > the rng call. > > Michael > > >> Actually, I would like to speed up matrix multiplication >> >> which is really too slow. >> MatrixA %*% MatrixB (each one is a [128,128] table) is running four >> hours now... and just imagine that I want to calculate many of those. >> >> Which other alternative do I have? >> >> Regards >> Alex >> >> >> ________________________________ >> From: Paul Hiemstra <paul.hiemstra at knmi.nl> >> >> Cc: "R-help at r-project.org" <R-help at r-project.org> >> Sent: Monday, February 6, 2012 2:30 PM >> Subject: Re: [R] R enterprise for linux >> >> >> Dear Alex, >> >> By Enterprise R, do you mean Revolution R? I can also only find red >> hat packages. You could send them an e-mail to ask for sources. >> They >> probably want to help you, especially when you pay. But why do you >> need RevoR and not standard R. >> >> Paul >> >> On 02/06/2012 12:55 PM, Alaios wrote: >> Dear all, >> I would like to ask you if it would be possible to install R >> enterprise in opensuse. >> Actually it seems that there are only Red Hat packages and I was >> wondering if their source code is available so one to try and >> compile it. I would like to thank you in advance for your reply >> Regards >> Alex [[alternative HTML version deleted]] >>> >> ______________________________________________ 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. >> >> >> -- >> Paul Hiemstra, Ph.D. >> Global Climate Division >> Royal Netherlands Meteorological Institute (KNMI) >> Wilhelminalaan 10 | 3732 GK | De Bilt | Kamer B 3.39 >> P.O. Box 201 | 3730 AE | De Bilt >> tel: +31 30 2206 494 http://intamap.geo.uu.nl/~paul >> http://nl.linkedin.com/pub/paul-hiemstra/20/30b/770 >> [[alternative HTML version deleted]] >> >> >> ______________________________________________ >> 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. >> > [[alternative HTML version deleted]] > > ______________________________________________ > 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. David Winsemius, MD West Hartford, CT
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20120206/4828c68b/attachment.pl>
I think there is some support for multi-processor matrix multiplication (google around for it) but in your case, it might suffice to use an optimized BLAS and R's builtin parallel facilities. As I said in your other thread -- if you are working with diagonal/sparse matrices, it's also possible to use special cases for efficient manipulations. Michael
On Mon, Feb 6, 2012 at 10:47 AM, Alaios <alaios at yahoo.com> wrote:
That was my bad. (doing two things at same time) I am not out of memory by no means.... it is just that it only runs at one core. IF somehow R was able to see the many cores in my system. Regards Alex
________________________________ From: David Winsemius <dwinsemius at comcast.net> To: Alaios <alaios at yahoo.com> Cc: R. Michael Weylandt <michael.weylandt at gmail.com>; "R-help at r-project.org" <R-help at r-project.org> Sent: Monday, February 6, 2012 4:31 PM Subject: Re: [R] R enterprise for linux On Feb 6, 2012, at 10:17 AM, Alaios wrote: My bad.. these are the real matrices G<-DiagonalGaussian( NumberOfImagePixels,sd=Gsd) str(G) ? num [1:16384, 1:16384] 1400 0 0 0 0 ... S<-PointSamplingMatrix(EffectiveSensors,NumberOfImagePixels) str(S) ? num [1:10650, 1:16384] 0 0 0 0 0 0 0 0 0 0 ... A<-S %*% G You seems confused about the size of your objects. You earlier said they were? [128,128] which we took to mean 128 x 128 matrices, they are far larger than that. I'm guessing that you are consuming all of your RAM and paging out to disk. But that is just a guess since you provided none of the system information that the Posting Guide suggests. --David. _______________________________ From: R. Michael Weylandt <michael.weylandt at gmail.com> Cc: Paul Hiemstra <paul.hiemstra at knmi.nl>; "R-help at r-project.org" <R-help at r-project.org> Sent: Monday, February 6, 2012 3:39 PM Subject: Re: [R] R enterprise for linux Are you sure? I just ran N <- 128 system.time(matrix(rnorm(N^2),N) %*% matrix(rnorm(N^2),N)) and it took less than 0.044s on my (old-ish) laptop while doing other things (and that includes the expensive rng calls). There might be some other issues in play here. Even N <- 1280 takes < 5 seconds with the rng call. Michael Actually, I would like to speed up matrix multiplication which is really too slow. MatrixA %*% MatrixB (each one is a [128,128] table) is running four hours now... and just imagine that I want to calculate many of those. Which other alternative do I have? Regards Alex ________________________________ ? From: Paul Hiemstra <paul.hiemstra at knmi.nl> Cc: "R-help at r-project.org" <R-help at r-project.org> Sent: Monday, February 6, 2012 2:30 PM Subject: Re: [R] R enterprise for linux Dear Alex, By Enterprise R, do you mean Revolution R? I can also only find red ? ? hat packages. You could send them an e-mail to ask for sources. They ? ? probably want to help you, especially when you pay. But why do you ? ? need RevoR and not standard R. Paul On 02/06/2012 12:55 PM, Alaios wrote: Dear all, I would like to ask you if it would be possible? to install R enterprise in opensuse. Actually it seems that there are only Red Hat packages and I was wondering if their source code is available so one to try and compile it. I would like to thank you in advance for your reply Regards Alex [[alternative HTML version deleted]] ______________________________________________ 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. -- Paul Hiemstra, Ph.D. Global Climate Division Royal Netherlands Meteorological Institute (KNMI) Wilhelminalaan 10 | 3732 GK | De Bilt | Kamer B 3.39 P.O. Box 201 | 3730 AE | De Bilt tel: +31 30 2206 494 http://intamap.geo.uu.nl/~paul http://nl.linkedin.com/pub/paul-hiemstra/20/30b/770 ? ? ? ? [[alternative HTML version deleted]] ______________________________________________ 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. ??? [[alternative HTML version deleted]] ______________________________________________ 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. David Winsemius, MD West Hartford, CT