The bugs are fixed. This was due to hard-coded path in check(), and Python in the official repos has been updated to 3.9.
For scipy, I think you have to use python-scipy-mkl...
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: | http://www.numpy.org/ |
Keywords: | numpy |
Licenses: | custom |
Conflicts: | python-numpy |
Provides: | python-numpy |
Submitter: | bred |
Maintainer: | petronny (AutoUpdateBot) |
Last Packager: | AutoUpdateBot |
Votes: | 38 |
Popularity: | 0.000000 |
First Submitted: | 2012-08-25 16:55 (UTC) |
Last Updated: | 2024-02-06 08:01 (UTC) |
The bugs are fixed. This was due to hard-coded path in check(), and Python in the official repos has been updated to 3.9.
For scipy, I think you have to use python-scipy-mkl...
Does this work with the python-scipy package in the official arch repo? I installed this and then the python-scipy package from the official repo. But it seems that scipy is linked to lapack. When I try to import scipy.ndimage it gives error that it cannot find liblapack.so. Any solutions?
Okay I found a solution. In the PKGBUILD check() section I made the following changes.
- PATH="$PWD/usr/bin:$PATH" PYTHONPATH="$PWD/usr/lib/python3.8/site-packages:$PYTHONPATH" python -c 'import numpy; numpy.test()'
+ PATH="$PWD/usr/bin:$PATH" PYTHONPATH="$PWD/usr/lib/python3.9/site-packages:$PYTHONPATH" python -c 'import numpy; numpy.test()'
I got this error :( What to do?
no previously-included directories found matching 'doc/build'
no previously-included directories found matching 'doc/source/generated'
no previously-included directories found matching 'benchmarks/env'
no previously-included directories found matching 'benchmarks/results'
no previously-included directories found matching 'benchmarks/html'
no previously-included directories found matching 'benchmarks/numpy'
warning: no previously-included files matching '*.pyo' found anywhere in distribution
warning: no previously-included files matching '*.pyd' found anywhere in distribution
warning: no previously-included files matching '*.swp' found anywhere in distribution
warning: no previously-included files matching '*.bak' found anywhere in distribution
warning: no previously-included files matching '*~' found anywhere in distribution
warning: no previously-included files found matching 'LICENSES_bundled.txt'
writing manifest file 'numpy.egg-info/SOURCES.txt'
removing '/home/ibraheemmoosa/AUR/python-numpy-mkl/src/numpy-1.19.4/tmp_install/usr/lib/python3.9/site-packages/numpy-1.19.4-py3.9.egg-info' (and everything under it)
Copying numpy.egg-info to /home/ibraheemmoosa/AUR/python-numpy-mkl/src/numpy-1.19.4/tmp_install/usr/lib/python3.9/site-packages/numpy-1.19.4-py3.9.egg-info
running install_scripts
Installing f2py script to /home/ibraheemmoosa/AUR/python-numpy-mkl/src/numpy-1.19.4/tmp_install/usr/bin
Installing f2py3 script to /home/ibraheemmoosa/AUR/python-numpy-mkl/src/numpy-1.19.4/tmp_install/usr/bin
Installing f2py3.9 script to /home/ibraheemmoosa/AUR/python-numpy-mkl/src/numpy-1.19.4/tmp_install/usr/bin
Traceback (most recent call last):
File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'numpy'
==> ERROR: A failure occurred in check().
Aborting...
@RunAndTumble Do you want to add icc back?
Building numpy with icc is introduced in https://software.intel.com/content/www/us/en/develop/articles/numpyscipy-with-intel-mkl.html and I think it's an official document.
We should always use the official way to build every package unless we have to patch it.
@RunAndTumble Sorry, I thought you were too busy for that. icc support has been dropped because it doesn't accelerate much, the heavy computing parts are already using mkl. Also we don't need to deal with icc's license.
Eh, that was unexpected.
Sorry, folks, I'm finishing my thesis and barely had enough time for anything else. v1.17.4 based on python3.8 and the latest icc. I completely removed python2 support. One of the tests fails, I observed a similar issue before, I'll try to find the problem.
Since python upgraded to 3.8.0, this package no longer works. I wonder if it will be updated soon?
v 1.16.3 is out, all test are fine
Pinned Comments
petronny commented on 2022-09-28 08:28 (UTC)
Prebuilt binaries of this package can be found in arch4edu.