search

emilydolson / phylotrackpy

Python phylogeny tracker for in silico evolution experiments
MIT License
9 stars 2 forks source link

Fix windows build #55

Closed emilydolson closed 7 months ago

emilydolson commented 7 months ago

Attempting to fix #54

codecov-commenter commented 7 months ago

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Comparison is base (ed7f9b8) 100.00% compared to head (6a66f8d) 100.00%.

Additional details and impacted files ```diff @@ Coverage Diff @@ ## taxontype #55 +/- ## =========================================== Coverage 100.00% 100.00% =========================================== Files 1 1 Lines 5 5 =========================================== Hits 5 5 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

emilydolson commented 7 months ago

It works! Had to gut optional _throw a bit (we should eventually try to add back the functionality I just removed), but I'd say it's functional.

If you're good with removing NumPy as a dependency, feel free to merge this. Also happy to be convinced to keep numpy

mmore500 commented 7 months ago

I suspect that numpy-data-as-taxon-info is a potential use case. I'm thinking of how DEAP has a whole section on Numpy genetic representations https://deap.readthedocs.io/en/master/tutorials/advanced/numpy.html

If reverting back NumPy breaks tests, though, let's axe it.

emilydolson commented 7 months ago

How does this look to you? It tests the numpy functionality without actually adding it as a dependency or testing it in the wheels (where build systems get messier)

emilydolson commented 7 months ago

I believe the numpy test case is already be marked with the nowheel and running. Pinning numpy is good with me (although I think it should be less fragile in the test environment than in the wheel environment)

mmore500 commented 7 months ago

Ah, I see that one of the tests is set up, put the other one together in f55acab. Merging!