Package Details: r-mkl 4.4.2-1

Git Clone URL: https://aur.archlinux.org/r-mkl.git (read-only, click to copy)
Package Base: r-mkl
Description: Language and environment for statistical computing and graphics, linked to the Intel(R) MKL.
Upstream URL: http://www.r-project.org/
Keywords: hpc mathematics modelling r statistics
Licenses: GPL
Conflicts: microsoft-r-open, r
Provides: r
Submitter: giniu
Maintainer: alexanderp
Last Packager: alexanderp
Votes: 25
Popularity: 0.000077
First Submitted: 2010-05-06 00:10 (UTC)
Last Updated: 2024-11-12 19:56 (UTC)

Latest Comments

« First ‹ Previous 1 .. 14 15 16 17 18 19 20 21 Next › Last »

tklee commented on 2014-07-15 02:22 (UTC)

Sorry, I forgot to mention I was using icc.

tklee commented on 2014-07-15 02:21 (UTC)

FYI, I had to add -parallel in LDFLAGS in PKGBULID to compile. LDFLAGS="${LDFLAGS} -L${_icclibpath} -parallel"

gabx commented on 2014-04-21 11:15 (UTC)

During the last part of building, the test part, I have one error when testing mgcv.Rcheck: Intel MKL FATAL ERROR: Cannot load libmkl_avx.so or libmkl_def.so. R-mkl still built and is not broken. This error is quite weird as : 1 - $ cat /etc/ld.so.confd/intel-mkl /opt/intel/composer_xe_2013_sp1.2.144/mkl/lib/intel64 /opt/intel/composer_xe_2013_sp1.2.144/compiler/lib/intel64 2- In PKGBUILD ${LD_LIBRARY_PATH} is correctly set and path to lib is OK.

gabx commented on 2014-04-21 11:05 (UTC)

FYI, your package build fine with these optimization flags: export CFLAGS="-Wall -Ofast -funsafe-loop-optimizations -ipo -openmp -xHost -Wno-fstrict-aliasing" export CXXFLAGS="-Wall -Ofast -ipo -funsafe-loop-optimizations -openmp -xHost -Wno-fstrict-aliasing -march=native -flto" Then, I use the ICC and ICPC compiler. I need each time to hack your PKGBUILD (not a great deal thought). Would you mind adding something like this: # Build the package using the icc compiler _CC="icc" if [ $_CC = "icc" ]; then export CC="icc" export CXX="icpc" export AR="xiar" export LD="xild" fi I use this as LDFLAGS: export LDFLAGS="-L${_icclibpath} -${CXXFLAGS} -shared-intel" Last, when building R, R-mkl and R-studio, I remarked my home Rprofile.r is screawing the build. So I add these lines too and have a cleaner build: # unset user Rprofile.r variable for building if [ -n $R_PROFILE_USER ]; then unset R_PROFILE_USER fi TY for your package.

jdarch commented on 2014-01-03 14:32 (UTC)

libimf is linked in now, this should usually speed up some mathematical operations.

jdarch commented on 2013-12-24 10:05 (UTC)

zeltak, I am not aware of any issues. You should be able to build 'lme4', it's prerequisites 'Matrix' and 'lattice' should be available. ((lme4 is working fine here) Note Matrix has a capital 'M'. The procedure to update packages is described in the R manual: http://cran.r-project.org/doc/manuals/R-admin.html#Updating-packages .

zeltak commented on 2013-12-22 12:48 (UTC)

thx you for this great package. i have a neewb Q. i need to use lme4 in my scripts. i see after installing r-mkl there isnt matrix 1.1 package which latest lme4 needs. i installed in through R. Does that mean i wont get mkl using lme4 since the matrix package wasnt compiled with mkl? is there a way to update matrix to latest release? thanks alot again Z

aberkoke commented on 2013-10-24 13:11 (UTC)

jdarch: Ok, it is done!

« First ‹ Previous 1 .. 14 15 16 17 18 19 20 21 Next › Last »