test_orthogonal.test_j_roots fails on i386

[Graham Inggs]

The following test fails during the build, but test failures are ignored.
It is then picked up as an autopkgtest failure.

======================================================================
FAIL: test_orthogonal.test_j_roots
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/nose/case.py", line 197, in runTest
    self.test(*self.arg)
  File "/usr/lib/python2.7/dist-packages/scipy/special/tests/test_orthogonal.py", line 320, in test_j_roots
    vgq(rf(0.9, 2), ef(0.9, 2), wf(0.9, 2), -1., 1., 100, atol=2e-13)
  File "/usr/lib/python2.7/dist-packages/scipy/special/tests/test_orthogonal.py", line 286, in verify_gauss_quad
    assert_allclose(vv, np.eye(N), rtol, atol)
  File "/usr/lib/python2.7/dist-packages/numpy/testing/utils.py", line 1411, in assert_allclose
    verbose=verbose, header=header, equal_nan=equal_nan)
  File "/usr/lib/python2.7/dist-packages/numpy/testing/utils.py", line 796, in assert_array_compare
    raise AssertionError(msg)
AssertionError:
Not equal to tolerance rtol=1e-15, atol=2e-13

(mismatch 0.38%)
 x: array([[ 1.000000e+00, 3.169501e-16, -1.335986e-15, ...,
          1.063756e-16, 7.590167e-16, -4.361854e-17],
       [ 3.143985e-16, 1.000000e+00, 5.329291e-16, ...,...
 y: array([[ 1., 0., 0., ..., 0., 0., 0.],
       [ 0., 1., 0., ..., 0., 0., 0.],
       [ 0., 0., 1., ..., 0., 0., 0.],...

----------------------------------------------------------------------
Ran 947 tests in 12.197s

FAILED (KNOWNFAIL=8, SKIP=159, failures=1)
Running unit tests for scipy.special
NumPy version 1.12.1
NumPy relaxed strides checking option: True
NumPy is installed in /usr/lib/python2.7/dist-packages/numpy
SciPy version 0.18.1
SciPy is installed in /usr/lib/python2.7/dist-packages/scipy
Python version 2.7.14rc1 (default, Sep 5 2017, 18:19:52) [GCC 7.2.0]
nose version 1.3.7
failed: test_orthogonal.test_j_roots
#errors: 1

[Graham Inggs (ginggs)]

Similar failure occurred previously:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=758022

See also upstream issue:
https://github.com/scipy/scipy/issues/5363
In particular, pv's comment:

> Much of scipy.special regards 1e-12 relative accuracy
> to be sufficient, so I don't see particular problems
> in bumping the tolerances here."

[Launchpad Janitor (janitor)]

This bug was fixed in the package python-scipy - 0.18.1-2ubuntu5

---------------
python-scipy (0.18.1-2ubuntu5) artful; urgency=medium

  * Bump test_orthogonal.test_j_roots tolerance (LP: #1717767)

 -- Graham Inggs <email address hidden> Sun, 17 Sep 2017 12:15:46 +0000