Scipy virgin here: Is it reasonable to ignore a "mild looking" FAILED scipy.test and get on with learning to use scipy? The story: After fairly careful linux install, scipy.test(level=1,verbosity=2) gives

...

...

...

...

BEGIN OUTPUT

... [snip] FAIL: check_cdf (scipy.stats.distributions.test_distributions.test_fatiguelife) ---------------------------------------------------------------------- Traceback (most recent call last): File "<string>", line 10, in check_cdf AssertionError: D = 0.362327598842; pval = 0.000245540127335; alpha = 0.01 args = (1.0213303917789236,) ---------------------------------------------------------------------- Ran 743 tests in 5.665s FAILED (failures=1, errors=15) <unittest.TextTestRunner object at 0x4053998c> <<<<<OUTPUT END All the errrors except one have to do with special functions (including orthogonal polynomials). ---- A related question: several packages had no test found, for example,

...

...

...

...

BEGIN OUTPUT

!! No test file 'test_quadpack.py' found for <module 'scipy.integrate.quadpack' from '...s/scipy/integrate/quadpack.pyc'> <<<<<OUTPUT END (I picked this missing test file example B/C I am planning to use quadpack a lot.) Thank you for your patience, -- Nicolas Robidoux Departement de mathematiques et d'informatique Universite Laurentienne