search

tskit-dev / pyslim

Tools for dealing with tree sequences coming to and from SLiM.
MIT License
27 stars 23 forks source link

more testing of population remapping #273

Closed petrelharp closed 2 years ago

petrelharp commented 2 years ago

This will close #256.

bhaller commented 2 years ago

Awesome, thanks! What's up with the multiple cancelled and one failed test in the checks above? Is that something I should worry about, or no? Also: you mentioned that a remap entry referencing a non-existent subpop ought to error and does not; do these tests flag that as a problem, or do they let it slide? (I guess they let it slide, since the tests passed on some platforms?) If we now have a test suite that works, I might try to fix that bug today; I was worried about causing new bugs if I did the fix wrong, but that risk is substantially lowered by having a working test suite, so this is great work, already paying off. :->

petrelharp commented 2 years ago

I'll look into the failing tests shortly. The test for referring to non-extant pops is currently commented out.

petrelharp commented 2 years ago

Well, the tests failed because I forgot to commit the recipe for some of them. =)