Package Details: python-numpy-mkl 2.1.3-1

Git Clone URL: https://aur.archlinux.org/python-numpy-mkl.git (read-only, click to copy)
Package Base: python-numpy-mkl
Description: Scientific tools for Python, compiled with Intel MKL
Upstream URL: https://numpy.org
Keywords: numpy
Licenses: BSD-3-Clause
Conflicts: python-numpy
Provides: python-numpy
Submitter: bred
Maintainer: petronny (AutoUpdateBot, carlosal1015)
Last Packager: carlosal1015
Votes: 38
Popularity: 0.000000
First Submitted: 2012-08-25 16:55 (UTC)
Last Updated: 2024-11-18 16:05 (UTC)

Required by (1913)

Sources (1)

Pinned Comments

petronny commented on 2022-09-28 08:28 (UTC)

Prebuilt binaries of this package can be found in arch4edu.

Latest Comments

« First ‹ Previous 1 2 3 4 5 6 7 8 9 Next › Last »

Morganamilo commented on 2018-05-21 15:03 (UTC)

Replaces should not be used here, this is an alternative to the original package. That's like having wine-staging replaces wine.

Even if replaces was justified I don't think AUR packages should ever use replaces unless it is copied from an existing repo package.

petronny commented on 2018-05-11 08:36 (UTC) (edited on 2018-05-11 08:38 (UTC) by petronny)

intel-compiler-base should be a dependency

ImportError: libmkl_rt.so: cannot open shared object file: No such file or directory

moetayuko commented on 2018-05-02 11:24 (UTC) (edited on 2018-05-02 11:25 (UTC) by moetayuko)

To workaround build error with icc 18.0.1 + glibc2.27, uncomment __INTEL_PRE_CFLAGS part and change it to

export __INTEL_PRE_CFLAGS="$__INTEL_PRE_CFLAGS -D__PURE_INTEL_C99_HEADERS__ -D_Float32=float -D_Float64=double -D_Float128=\"long double\" -D_Float32x=_Float64 -D_Float64x=_Float128"

ref: https://software.intel.com/en-us/forums/intel-c-compiler/topic/742701#comment-1921418

bred commented on 2017-12-26 11:20 (UTC)

In the icc 18.0.1 the bug has been fixed. So the env variable __INTEL_PRE_CFLAGS is no longer needed.

If you are still using the icc 18.0.0 uncomment the lines 71/72/73 in the PKGBUILD.

arvidsaur commented on 2017-11-29 16:02 (UTC)

python-numpy gives me a pass each time I run it. python -c "from numpy import test; test('full')" Running unit tests for numpy NumPy version 1.13.3 NumPy relaxed strides checking option: True NumPy is installed in /usr/lib/python3.6/site-packages/numpy Python version 3.6.3 (default, Oct 24 2017, 14:48:20) [GCC 7.2.0] nose version 1.3.7 Ran 6823 tests in 99.402s OK (KNOWNFAIL=7, SKIP=13)

bred commented on 2017-11-14 10:48 (UTC) (edited on 2017-11-14 12:19 (UTC) by bred)

Yes with python-numpy-mkl I've the same error as you. Try to run the test for two time with python-numpy. On a new try I've not seen the error on the first run, but it has appeared on the second run (with python-numpy). Strange!

arvidsaur commented on 2017-11-13 11:52 (UTC)

That is weird first that we do not have the same failed test, and secondly that python-numpy has no errors for me. Did you get the chance to test python-numpy-mkl as well?

bred commented on 2017-11-08 15:41 (UTC)

In python-numpy (normal archlinux package) I've this error: ====================================================================== FAIL: test_refcount_vdot (test_regression.TestRegression) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/lib/python3.6/site-packages/numpy/core/tests/test_regression.py", line 659, in test_refcount_vdot _assert_valid_refcount(np.vdot) File "/usr/lib/python3.6/site-packages/numpy/testing/utils.py", line 1336, in _assert_valid_refcount assert_(sys.getrefcount(i) >= rc) File "/usr/lib/python3.6/site-packages/numpy/testing/utils.py", line 92, in assert_ raise AssertionError(smsg) AssertionError ---------------------------------------------------------------------- Ran 6610 tests in 55.926s FAILED (KNOWNFAIL=7, SKIP=13, failures=1) It seems to be a bug in numpy. I'll test also the numpy with MKL.

arvidsaur commented on 2017-11-08 14:32 (UTC)

Running numpy.test() with this pkgbuild I get the following failure: FAIL: test_noncommutative_reduce_accumulate (test_regression.TestRegression) ---------------------------------------------------------------------- Traceback (most recent call last): File "/usr/lib/python3.6/site-packages/numpy/core/tests/test_regression.py", line 674, in test_noncommutative_reduce_accumulate np.array([2., 4., 16.])) File "/usr/lib/python3.6/site-packages/numpy/testing/utils.py", line 854, in assert_array_equal verbose=verbose, header='Arrays are not equal') File "/usr/lib/python3.6/site-packages/numpy/testing/utils.py", line 778, in assert_array_compare raise AssertionError(msg) AssertionError: Arrays are not equal (mismatch 33.33333333333333%) x: array([ 2., 4., 0.]) y: array([ 2., 4., 16.]) ---------------------------------------------------------------------- Ran 6610 tests in 21.417s FAILED (KNOWNFAIL=7, SKIP=14, failures=1) with the 'normal' python-numpy this test passes, anyone else have this issue?