Package Details: r-mkl 4.0.3-1

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=4.0.3
Submitter: giniu
Maintainer: alexanderp
Last Packager: alexanderp
Votes: 21
Popularity: 1.11
First Submitted: 2010-05-06 00:10
Last Updated: 2020-11-21 09:07

Dependencies (25)

Required by (375)

Sources (5)

Latest Comments

1 2 3 4 5 6 ... Next › Last »

alexanderp commented on 2021-01-20 14:22

@LuisDamiano, in my opinion, distributing a binary of this package defeats the purpose of the additional CPU performance optimizations obtained by compiling on a target system.

Svenstaro commented on 2021-01-19 08:09

@LuisDamiano you could talk to the maintainers of the r package whether they'd be interested in pulling this in.

LuisDamiano commented on 2021-01-18 16:44

Thanks for the great work! It takes quite a bit to compile locally, would it be possible to get r-mkl added to the Extra official repo? Having the binaries would be quite useful.

k07k454n commented on 2020-11-24 18:30

Had issues for 4.0.2 and 4.0.3 like this during build:

Error: package or namespace load failed for ‘nnet’ in dyn.load(file, DLLpath = DLLpath, ...):                                                                                                                                                                                     
 unable to load shared object '/home/user/aur/r-mkl/src/R-4.0.3/library/nnet/libs/nnet.so':                                                                                                                                                                                       
  /home/user/aur/r-mkl/src/R-4.0.3/library/nnet/libs/nnet.so: undefined symbol: _ZGVbN2v_exp                                                                                                                                                                                      
Error: loading failed 
** testing if installed package can be loaded                                                                                            
Error: package or namespace load failed for ‘MASS’ in dyn.load(file, DLLpath = DLLpath, ...):                                                      
 unable to load shared object '/home/user/aur/r-mkl/src/R-4.0.3/library/MASS/libs/MASS.so':                                                        
  /home/user/aur/r-mkl/src/R-4.0.3/library/MASS/libs/MASS.so: undefined symbol: _ZGVbN2v_log                                                       
Error: loading failed

..so to build and install successfully had to first do 'export SHLIB_LDFLAGS=-lmvec' before makepkg

alexanderp commented on 2020-08-24 19:12

@jalapeno, if intel-mkl was updated, try recompiling the package.

jalapeno commented on 2020-08-20 19:55

When attempting to install a package in r-mkl, I get this error:

install.packages("mnormt")
[truncated]
/usr/bin/ld: cannot find -lmkl_gf_lp64
/usr/bin/ld: cannot find -lmkl_gnu_thread
/usr/bin/ld: cannot find -lmkl_core
collect2: error: ld returned 1 exit status
make: *** [/usr/share/R//make/shlib.mk:6: tmvnsim.so] Error 1
ERROR: compilation failed for package ‘tmvnsim’

I was unable to find a solution here, but managed to just copy lmkl_gf_lp64.so lmkl_gf_thread and lmkl_core.so' from/opt/intel/mkl/lib/intel64to/usr/lib/R/lib`.

Is there some other way that this ought to be done? Or, is this the only work-around?

tealeaf commented on 2020-07-26 19:51

As a follow-up to my previous comment: downgrading to intel-mkl-2020.1.217-6 resolved this for the moment.

tealeaf commented on 2020-07-26 15:28

Not sure if this belongs here, in intel-mkl, or elsewhere, but I'm having a problem running R-INLA under this on a new machine. I can install with:

install.packages("INLA", repos=c(getOption("repos"), INLA="https://inla.r-inla-download.org/R/testing"), dep=TRUE)

as normal, but running:

library(INLA)
inla.pardiso.check()

I get:

/home/joss/.R/packages/INLA/bin/linux/64bit/inla: symbol lookup error: /opt/intel/mkl/lib/intel64/libmkl_gf_lp64.so: undefined symbol: mkl_lapack_scombssq
Error in system(paste(shQuote(inla.call.no.remote()), "-m pardiso"), intern = TRUE) : error in running command

The strange thing is that this is working on my old machine with an almost identical setup.

sxyzy1016 commented on 2020-06-26 10:32

using icc to build it by uncomment the line # _CC="icc" and get: 

...
icc -I../../src/extra -I/usr/include/tirpc -I. -I../../src/include -I../../src/include  -D_FORTIFY_SOURCE=2 -I../../src/nmath -DHAVE_CONFIG_H   -fopenmp -fpic  -O3 -fPIC -m64 -march=native -fp-model precise -fp-model source -I/opt/intel/mkl/include  -c arithmetic.c -o arithmetic.o
arithmetic.c(61): warning #274: declaration is not visible outside of function
  int matherr(struct exception *exc)
                     ^

arithmetic.c(63): error: pointer to incomplete class type is not allowed
      switch (exc->type) {
              ^

arithmetic.c(64): error: identifier "DOMAIN" is undefined
      case DOMAIN:
           ^

arithmetic.c(65): error: identifier "SING" is undefined
      case SING:
           ^

arithmetic.c(68): error: identifier "OVERFLOW" is undefined
      case OVERFLOW:
           ^

arithmetic.c(71): error: identifier "UNDERFLOW" is undefined
      case UNDERFLOW:
           ^

arithmetic.c(72): error: pointer to incomplete class type is not allowed
    exc->retval = 0.0;
    ^

compilation aborted for arithmetic.c (code 2)
...

in build(). EDIT: according to this, use

sed -i '/^#define HAVE_MATHERR 1/d' src/include/config.h

between configure and make will make it work.

Svenstaro commented on 2020-06-15 17:22

Alright but can you check the new intel-mkl that blue includes a shim script for mklvars.sh? It should work.