Torch dispatch for scipy-like functions and Softplus

[Ch0ronomato]

Description

I was goofing around with pymc using the pytorch backend. I implemented a beta binomial model and had an issue when supporting some scipy methods, and was missing softplus. I wrote some small shims. Let me know your thoughts on the importlib stuff.

Related Issue

• Closes #
• Related to BUG: Pymc receiving torch tensors causes sampling failures #1065

Checklist

• Checked that the pre-commit linting/style checks pass
• Included tests that prove the fix is effective or that the new feature works
• Added necessary documentation (docstrings and/or example notebooks)
• If you are a pro: each commit corresponds to a relevant logical change

Type of change

• New feature / enhancement
• Bug fix
• Documentation
• Maintenance
• Other (please specify):

---

📚 Documentation preview 📚: https://pytensor--1066.org.readthedocs.build/en/1066/

[Ch0ronomato]

I'm okay with closing this, just want to get your thoughts on it in relation to #1065

[codecov]

Codecov Report

Attention: Patch coverage is 86.36364% with 3 lines in your changes missing coverage. Please review.

Project coverage is 82.12%. Comparing base (ae66e82) to head (7fec045).
Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
pytensor/link/pytorch/dispatch/elemwise.py	72.72%	2 Missing and 1 partial ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1066   +/-   ##
=======================================
  Coverage   82.12%   82.12%           
=======================================
  Files         183      183           
  Lines       48016    48035   +19     
  Branches     8648     8650    +2     
=======================================
+ Hits        39433    39450   +17     
- Misses       6417     6418    +1     
- Partials     2166     2167    +1     

Files with missing lines	Coverage Δ
pytensor/link/pytorch/dispatch/scalar.py	71.87% <100.00%> (+11.00%)	⬆️
pytensor/link/pytorch/dispatch/elemwise.py	66.66% <72.72%> (+1.28%)	⬆️

[ricardoV94]

I'm okay with closing this, just want to get your thoughts on it in relation to #1065

This looks fine why did you ask about closing it?

Can you add some small tests for one of the scipy ops and the softplus?

[ricardoV94]

Looks great, just a nit in the tests

[ricardoV94]

@Ch0ronomato can you fix the conflicts, should be simple