Explicitly specify domain for AffineTransform in ScaledLogitTransform

[dylanhmorris]

This will prevent bugs if it is used to create TransformedDistributions See pyro-ppl/numpyro#1756

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.18%. Comparing base (0fc0c8d) to head (289f83b).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #378   +/-   ##
=======================================
  Coverage   93.17%   93.18%           
=======================================
  Files          39       39           
  Lines         909      910    +1     
=======================================
+ Hits          847      848    +1     
  Misses         62       62           

Flag	Coverage Δ
unittests	93.18% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[gvegayon]

Not sure how the constraints work exactly. I've been trying to set it to be within 0-1, but I still get weird values. I suggest creating an integration test where you call the random variable via model sampling.

[dylanhmorris]

Was this comment supposed to be in the simplex issue? I think for the inference to work you'd need to constrain the domain of the transformation to a simplex, not just constrain the components of the input vector to the unit interval