Skip to content

numpy 1.17.1 6 tests failures for PowerPC  #14438

@michelmno

Description

@michelmno

numpy 6 tests failures for PowerPC
as reported during package build in openSUSE env (1) and attached log file
Those traces are reported with numpy 1.17.1; while my local trial was with 1.16.2
python_numpy_standard__ppc64le_201909060820.log

$grep FAILED python_numpy_standard__ppc64le_201909060820.log
[  382s] ../core/tests/test_einsum.py::TestEinsum::test_einsum_sums_cfloat64 FAILED [ 12%]
[ 1107s] ../linalg/tests/test_linalg.py::TestSolve::test_generalized_sq_cases FAILED [ 60%]
[ 1125s] ../linalg/tests/test_linalg.py::TestInv::test_generalized_sq_cases FAILED [ 61%]
[ 1147s] ../linalg/tests/test_linalg.py::TestCond::test_sq_cases FAILED           [ 61%]
[ 1151s] ../linalg/tests/test_linalg.py::TestCond::test_generalized_sq_cases FAILED [ 61%]
[ 1189s] ../linalg/tests/test_linalg.py::TestCholesky::test_basic_property FAILED [ 63%]

standalone exemple of the first test:

$cat ~/work/test_einsum_sums_cfloat64
import numpy as np
dtype = 'c8'
n = 9
do_opt= 'False'
a = np.arange(2 * 3 * n, dtype=dtype).reshape(2, 3, n)
b = np.arange(n, dtype=dtype)
print('a: {!r}'.format(a))
print('b: {!r}'.format(b))
c = np.einsum("...i, ...i", a, b, optimize=do_opt)
d = np.inner(a, b)
print('check visually if c similar to d (not the case for PowerPC')
print('c: {!r}'.format(c))
print('d: {!r}'.format(d))

$python3 ~/work/test_einsum_sums_cfloat64
a: array([[[ 0.+0.j,  1.+0.j,  2.+0.j,  3.+0.j,  4.+0.j,  5.+0.j,  6.+0.j,
          7.+0.j,  8.+0.j],
        [ 9.+0.j, 10.+0.j, 11.+0.j, 12.+0.j, 13.+0.j, 14.+0.j, 15.+0.j,
         16.+0.j, 17.+0.j],
        [18.+0.j, 19.+0.j, 20.+0.j, 21.+0.j, 22.+0.j, 23.+0.j, 24.+0.j,
         25.+0.j, 26.+0.j]],

       [[27.+0.j, 28.+0.j, 29.+0.j, 30.+0.j, 31.+0.j, 32.+0.j, 33.+0.j,
         34.+0.j, 35.+0.j],
        [36.+0.j, 37.+0.j, 38.+0.j, 39.+0.j, 40.+0.j, 41.+0.j, 42.+0.j,
         43.+0.j, 44.+0.j],
        [45.+0.j, 46.+0.j, 47.+0.j, 48.+0.j, 49.+0.j, 50.+0.j, 51.+0.j,
         52.+0.j, 53.+0.j]]], dtype=complex64)
b: array([0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j, 4.+0.j, 5.+0.j, 6.+0.j, 7.+0.j,
       8.+0.j], dtype=complex64)
check visually if c similar to d (not the case for PowerPC)
c: array([[ 204.+0.j,  528.+0.j,  852.+0.j],
       [1176.+0.j, 1500.+0.j, 1824.+0.j]], dtype=complex64)
d: array([[ 204.+0.j,  500.+0.j,  852.+0.j],
       [1148.+0.j, 1500.+0.j, 1796.+0.j]], dtype=complex64)

>>> import sys, numpy; print(numpy.__version__, sys.version)
1.16.2 3.7.3 (default, Apr 09 2019, 05:18:21) [GCC]

The other failing tests details present in attached log

(1) https://build.opensuse.org/package/live_build_log/openSUSE:Factory:PowerPC/python-numpy/standard/ppc64le

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions