[SciPy-dev] statistic tests failures

Arnd Baecker arnd.baecker at web.de
Tue Nov 22 01:57:01 EST 2005


Moin,

>From time to time one of the statistics tests will fail.
Would it be reasonable/possible to automatically run those tests
a couple of times to reduce the chance of failing?


======================================================================
FAIL: check_normal (scipy.stats.morestats.test_morestats.test_anderson)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"/home/abaecker/BUILDS2/Build_81/inst_scipy_newcore/lib/python2.4/site-packages/scipy/stats/tests/test_morestats.py",
line 51, in check_normal
    assert_array_less(A, crit[-2:])
  File
"/home/abaecker/BUILDS2/Build_81//inst_scipy_newcore/lib/python2.4/site-packages/scipy/test/testing.py",
line 782, in assert_array_less
    assert cond,\
AssertionError:
Arrays are not less-ordered (mismatch 50.0%):
        Array 1: 0.94024324379482493
        Array 2: [ 0.858               1.0210000000000001]

Not that this point is really worrying me a lot,
but it might confuse newcomers ...

Best, Arnd




More information about the SciPy-Dev mailing list