Raise exception on shape error in CI

[rlouf]

This PR is to track the warning found in #29. We set the on_shape_error config variable to raise in the CI; the tests will fail until the underlying issue is fixed.

[codecov]

Codecov Report

Merging #42 (712cbfb) into main (8a634df) will not change coverage.
The diff coverage is n/a.

@@            Coverage Diff            @@
##              main       #42   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           10        12    +2     
  Lines          385       442   +57     
  Branches        17        23    +6     
=========================================
+ Hits           385       442   +57     

Impacted Files	Coverage Δ
aehmc/hmc.py	100.00% <0.00%> (ø)
aehmc/nuts.py	100.00% <0.00%> (ø)
aehmc/metrics.py	100.00% <0.00%> (ø)
aehmc/trajectory.py	100.00% <0.00%> (ø)
aehmc/step_size.py	100.00% <0.00%> (ø)
aehmc/window_adaptation.py	100.00% <0.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 8a634df...712cbfb. Read the comment docs.

[rlouf]

I am not sure why this is not raising an exception ? Running pytest tests/test_hmc.py locally will raise an exception as expected, but running pytest or pytest tests won't. print(aesara.config.on_shape_error) always returns raise.

[rlouf]

I added on_opt_error=raise and still get the same behavior.

[brandonwillard]

I added on_opt_error=raise and still get the same behavior.

Which test is expected to fail again? Everything seems to pass for me locally as well.

[rlouf]

test_hmc.py is supposed to fail. It does when I run pytest tests/test_hmc.py but not when I simply run pytest.

[brandonwillard]

test_hmc.py is supposed to fail. It does when I run pytest tests/test_hmc.py but not when I simply run pytest.

Oh, interesting; I'll try that locally.