[Python-modules-team] Bug#906198: [python-scipy] causes FTBFS on of imexam: scipy.optimize.leastsq fails on S390x

[Ole Streicher]

Package: python-scipy
Version: 1.1.0-1+b1
Severity: serious
Affects: python3-astropy, python-astropy, python-imexam, python3-imexam

Dear maintainer,

The following simple test fails on s390x:

---------------------------8<------------------------------
import numpy as np
from scipy.optimize import leastsq

y = np.array([0., 1., 1., 2., 1., 1., 0.])
x = np.arange(len(y))

def func(par):
    return par[2] * np.exp(-(x - par[0])**2/par[1]) - y

print(leastsq(func, [0,1,0]))
---------------------------8<------------------------------

The expected result is

(array([3.        , 4.42280548, 1.67210345]), 1)

while on S390x, one gets

(array([0.        , 1.        , 0.34027645]), 3)

This caused FTBFS of astropy [1], where I had to disable
the failing tests in astropy. Note that the original version
(1.1.0-1) did not have this problem [2] (the other problems are
irrelevant here).

Currently that issue causes an FTBFS on imexam [3], which is
the reason for the severity.

[1] https://buildd.debian.org/status/fetch.php?pkg=astropy&arch=s390x&ver=3.0.3-4&stamp=1533023784&raw=0
[2] https://buildd.debian.org/status/fetch.php?pkg=astropy&arch=s390x&ver=3.0.3-3%2Bb1&stamp=1532416286&raw=0
[3] https://buildd.debian.org/status/fetch.php?pkg=imexam&arch=s390x&ver=0.8.0-2%2Bb1&stamp=1534266846&raw=0

Best regards

Ole