Re: [Rd] matrix multiplication speed R

2011-05-10 Thread Prof Brian Ripley
On Mon, 9 May 2011, Sharpie wrote: aftar wrote: Hi Can we use BLAS in R X64 for windows? Regards Aftar You are already using BLAS in R as R includes its own BLAS library. On Windows the 64-bit DLL is located at R_HOME\bin\x64\Rblas.dll. If you are asking about swapping that out for an

Re: [Rd] matrix multiplication speed R

2011-05-10 Thread aftar
Hi Can we use BLAS in R X64 for windows? Regards Aftar -- View this message in context: http://r.789695.n4.nabble.com/matrix-multiplication-speed-R-tp3217257p3509596.html Sent from the R devel mailing list archive at Nabble.com. __

[Rd] Which higher-order function

2011-05-10 Thread Hadley Wickham
Would it be appropriate to add a Which higher-order function? Which - function(f, x) { ind - vapply(x, f, logical(1)) !is.na(ind) ind } Then Filter would be: Filter - function(f, x) x[Which(f, x)] I think the use of vapply is slightly better than the current as.logical(sapply(x, f))

[Rd] Debugging warnings

2011-05-10 Thread Hadley Wickham
Hi all, One thing that's currently a little tricky in R is debugging warnings. One solution is to turn them into errors and then use the usual error tracking mechanisms (e.g. options(error = recover)). But it should be possible to use withCallingHandlers to provide more flexibility. The

[Rd] Time-machine typo

2011-05-10 Thread Dirk Eddelbuettel
Just noticed this in R 2.13.0, it is also still present in R-devel in SVN: %%% Rd.sty ... Style for printing the R manual %%% Part of the R package, http://www.R-project.org %%% Copyright (C) 2003-20010 The R Foundation ^ Very forward-looking ... Dirk --

Re: [Rd] Time-machine typo

2011-05-10 Thread Ravi Varadhan
Indeed, Dirk, but realistic, I might add! BTW, I enjoyed some of the Gaussian Facts. My favorite: It only takes Gauss 4 minutes to sing Aleph-Null Bottles of Beer on the Wall. Best, Ravi. --- Ravi Varadhan, Ph.D. Assistant Professor,

Re: [Rd] cerr and cout not working calling c++ from R

2011-05-10 Thread Sean Robert McGuffee
Just a short follow up to the list. It turns out that the strange behavior I was experiencing had nothing to do with cout and cerr, but ostreams in general. For some reason, at a minimum, the operators: ostream operator (long val ); ostream operator (double val ); were not functioning properly

Re: [Rd] anova.lm fails with test=Cp

2011-05-10 Thread John Maindonald
For unknown sigma^2, a version that is a modification of AIC may be preferred, i.e. n log(RSS/n) + 2p - n I notice that this is what is given in Maindonald and Braun (2010) Data Analysis Graphics Using R, 3rd edition. Cf: Venables and Ripley, MASS, 4th edn, p.174. VR do however stop short of

[Rd] Reference Classes copy(shallow=FALSE) unexpected behavior.

2011-05-10 Thread Manuel Castejón Limas
Dear all, I've just discovered the 'Reference Classes'. In a previous attempt ---a year ago--- to re-implement in a Object Oriented fashion the AMORE package using S4 classes I strongly felt the need of such capability. It's great to have the Reference Classes now available. Along with the

Re: [Rd] Fortran Symbol Name not in Load Table

2011-05-10 Thread Duncan Murdoch
On 11-05-09 9:28 PM, vioravis wrote: I used the DLL export viewer to what is the table name being exported. It is showing as VALUEAHROPTIMIZE_. This is the name of the function we have used plus the underscore. I believe R will be looking for the lower case version of that name, which is

Re: [Rd] Fortran Symbol Name not in Load Table

2011-05-10 Thread vioravis
Thank you Sharpie and Duncan. I am using Compaq Visual Fortran compiler on a Windows machine. I tried with all small letters for the function names and created the DLLs. But somehow the DLL created only had all CAPS. But the DLL worked when I changed the function name to myxmean. Not really sure