Fix tests - #430
Conversation
|
@brainiak/committers, please review. The changes are unrelated to each other, but all necessary to have the tests pass again. |
|
A couple of references for the The SciPy pinning is the same as in |
|
@danielsuo, @gdoubleyew, any chance you could review? |
| with pytest.raises(ValueError): | ||
| es_invalid.model_prior(K-1) | ||
| with pytest.raises(ValueError, message="#Events < K should cause error"): | ||
| pytest.fail("T < K should cause error") |
There was a problem hiding this comment.
It wasn't obvious to me why adding pytest.fail() here or how that would work until reading the link you added in the commit message.
Maybe add a comment above pytest.fail(), something like:
"If we reach here the expected exception wan't thrown, call pytest.fail() to display custom message"
| @@ -1,13 +1,5 @@ | |||
| #!/bin/bash | |||
|
|
|||
| if [[ "$OSTYPE" == "darwin"* ]]; then | |||
There was a problem hiding this comment.
Why did this block go away?
There was a problem hiding this comment.
MACOSX_DEPLOYMENT_TARGET is set by conda-build based on the SDK we pass. The compiler variables are not used; instead, conda-build installs and uses compilers based on its configuration.
Overlooked when pinning NumPy in PR brainiak#430.
Overlooked when pinning NumPy in PR #430.
See PyTest docs:
https://docs.pytest.org/en/4.6-maintenance/deprecations.html#message-parameter-of-pytest-raises
And NumPy bug:
numpy/numpy#14189