Switch to InverseFunctions.inverse#33
Conversation
Codecov Report
@@ Coverage Diff @@
## main #33 +/- ##
=======================================
Coverage 91.70% 91.70%
=======================================
Files 14 14
Lines 1181 1181
=======================================
Hits 1083 1083
Misses 98 98 Continue to review full report at Codecov.
|
|
Thanks, Ian! The changes look good, I just want to check something- @oschulz should we change the definition here to use Transits.jl/src/distributions.jl Line 58 in 16fc5c9 |
No, I think it can stay as it is. If you implement both the InverseFunctions.jl and ChangesOfVariables.jl API, you may be able to not depend on Bijectors.jl at all, though. |
Sounds like that's worth trying in a future PR! Thanks for the comments Oliver, and thanks for the contribution Ian! |
|
Sounds good! Sort of a bookkeeping question, but is there a preference on how we want keep tabs on things like this? I just added it as a discussion, but maybe it would be better to track as a |
|
For bookkeeping I have a "rent-free" unit in the back of my head 😂 I think discussion is a good spot for planning contributions, so that's appropriate, then it can move to a PR when someone finds the time. |
Just following up from here TuringLang/Bijectors.jl#212