On Thu, 8 Sep 2016 08:51:56 +0200 Ole Streicher <oleb...@debian.org> wrote: > Control: reassign -1 src:python-scipy 0.18.0-1 > Control: affects -1 python-astropy > Control: forwarded -1 https://github.com/scipy/scipy/issues/6420 > Control: tags -1 patch > > > >> =================================== FAILURES > >> =================================== > >> _________________________ test_circmean_against_scipy > >> __________________________ > >> > >> @pytest.mark.skipif('not HAS_SCIPY') > >> def test_circmean_against_scipy(): > >> # testing against scipy.stats.circmean function > >> data = np.random.uniform(-np.pi, np.pi, size=(100, 5)) > >> answer = scipy.stats.circmean(data, np.pi, -np.pi) > >>> assert_equal(np.around(answer, 2), np.around(circmean(data), 2)) > >> E AssertionError: > >> E Items are not equal: > >> E ACTUAL: -4.3099999999999996 > >> E DESIRED: 1.97 > > This is actually a regression bug in scipy; see the github issue. A > patch is available at https://github.com/scipy/scipy/pull/6543
That regression is said to be fixed in scipy 0.18.1, which was uploaded on Oct 7 - could you verify this bug is actually fixed rebuilding astropy? thanks!
