Package Details: python-tinyarray 1.2.0-1

Git Clone URL: https://aur.archlinux.org/python-tinyarray.git (read-only)
Package Base: python-tinyarray
Description: Similar to NumPy arrays, but optimized for small sizes
Upstream URL: http://git.kwant-project.org/tinyarray/about/
Licenses: BSD
Submitter: lobachevsky
Maintainer: lobachevsky
Last Packager: lobachevsky
Votes: 0
Popularity: 0.000000
First Submitted: 2015-12-19 16:09
Last Updated: 2017-05-12 14:09

Latest Comments

lobachevsky commented on 2018-09-09 12:27

Hello aoowweenn,

the 1.2.0 release of tinyarray uses nose, whereas later upstream moved to pytest. Therefore the PKGBUILD for python-tinyarray checkdepends on python-nose, but python-tinyarray-git depends on python-pytest.

Hope this helps!

aoowweenn commented on 2018-09-07 18:50

Hi lobachevsky,

I think that tinyarray is designed to be only nose tests, but the python-pytest package should be added to checkdepends.

lobachevsky commented on 2017-03-27 10:15

Hi salviati,

could you please test the new git-packages?

best,
lobachevsky

salviati commented on 2017-03-26 20:33

Hi lobachevsky,

Thanks a lot for the info. Actually, kwant itself fails too. So I'll wait for the -git packages.

lobachevsky commented on 2017-03-22 22:33

Hi salviati,

the bug you encountered is probably [1], which was due to a change Numpy introduced upstream [2] in version and which was fixed in the current Numpy 1.12.1.

I just tested that I can reproduce your build failure of tinyarray 1.1.0 against the currrent Numpy 1.12.1. The error puzzles me a bit, because it seems to be the same as [1], but that one should not occur anymore.

For now you have three options:
1. Ignore the error and build the package with --nocheck. This is not insensible, because the error in [1] was simply a bug that let the check fail and this should be, in principle, the same error. It will not interfere with the functionality of Kwant; as long as Kwant's tests pass, you are fine.
2. Downgrade Numpy to 1.11.3. tinyarray builds successfully against it. I just tested this.
3. Wait for me to upload python-tinyarray-git and python-kwant-git this week. The current git head of tinyarray builds without error against Numpy 1.12.1

best,
lobachevsky

[1] https://gitlab.kwant-project.org/kwant/tinyarray/issues/11
[2] https://github.com/numpy/numpy/pull/8602

salviati commented on 2017-03-22 20:38

Build fails with

g++ -pthread -shared -Wl,-O1,--sort-common,--as-needed,-z,relro -Wl,-O1,--sort-common,--as-needed,-z,relro -Wl,-O1,--sort-common,--as-needed,-z,relro -march=native -mtune=native -O2 -pipe -fstack-protector-strong -D_FORTIFY_SOURCE=2 build/temp.linux-x86_64-3.6/src/arithmetic.o build/temp.linux-x86_64-3.6/src/array.o build/temp.linux-x86_64-3.6/src/functions.o -L/usr/lib -lpython3.6m -o build/lib.linux-x86_64-3.6/tinyarray.cpython-36m-x86_64-linux-gnu.so
copying build/lib.linux-x86_64-3.6/tinyarray.cpython-36m-x86_64-linux-gnu.so ->
test_tinyarray.test_array ... ok
test_tinyarray.test_matrix ... ok
test_tinyarray.test_conversion ... ok
test_tinyarray.test_special_constructors ... ok
test_tinyarray.test_dot ... /tmp/yaourt-tmp-user/aur-python-tinyarray/./PKGBUILD: line 31: 19250 Segmentation fault (core dumped) python setup.py test