CODEWITHSUNDEEP
pythontestingtravis-cisympy
Hamish
Hamish

Reputation: 1015

Travis-ci build fails but tests pass locally

I've submitted a (pretty small really) pull request to sympy on github. All the tests pass locally, but the build has failed in a module that doesn't seem to have anything to do with what I have changed.

My pull request: https://github.com/sympy/sympy/pull/8281

The failing travis-ci build: https://travis-ci.org/sympy/sympy/builds/38577918

This is the output:

________________________________________________________________________________
________________________ sympy.stats.crv_types.Uniform _________________________
File "/home/travis/virtualenv/python3.4.1/lib/python3.4/site-packages/sympy-0.7.5_git-py3.4.egg/sympy/stats/crv_types.py", line 2217, in sympy.stats.crv_types.Uniform
Failed example:
    cdf(X)(z)
Expected:
    -a/(-a + b) + z/(-a + b)
Got:
    z/(-a + b) - Min(a, z)/(-a + b)

The code is only failing when it tries to run the a test in python 3.4 (but passes for other versions of python).

My change is to a unit test (sympy.physics.quantum.tests.test_qubit.py) and I can't find any connection to the failing code (sympy.stats.crv_types.Uniform).

The weird thing is, I submitted this change a few hours earlier and this test passed (but failed a lint test) - after removing some offending white space, this now fails.

This is the first pull request I've done with a Travis-ci build - am I missing something obvious here? Can I rerun the build myself to see if it happens again? (I don't know why it wouldn't .. but then I can't think why it's failing!)

Cheers!

Upvotes: 0

Views: 3066

Answers (1)

smichr
smichr

Reputation: 18939

You should raise such questions on the PR itself. What you are seeing is just a random failure that has not been tracked down yet. See https://github.com/sympy/sympy/issues/7702.

Upvotes: 3

Related Questions