Make rvs_to_values work with non-RandomVariables

[ricardoV94]

What is this PR about?
Cleanup some logic that assumed only RandomVariables can be valued

Related to #6072
Related to aesara-devs/aeppl#174

Checklist

• Explain important implementation details 👆
• Make sure that the pre-commit linting/style checks pass.
• Link relevant issues (preferably in nice commit messages)
• Are the changes covered by tests and docstrings?
• Fill out the short summary sections 👇

Major / Breaking Changes

• ...

Bugfixes / New features

• rvs_to_values now works with non-RandomVariables and unvalued RandomVariables

Docs / Maintenance

• ...

[codecov]

Codecov Report

Merging #6101 (29500af) into main (761f77d) will increase coverage by 0.02%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #6101      +/-   ##
==========================================
+ Coverage   92.08%   92.10%   +0.02%     
==========================================
  Files          86       86              
  Lines       17861    17848      -13     
==========================================
- Hits        16447    16439       -8     
+ Misses       1414     1409       -5     

Impacted Files	Coverage Δ
pymc/aesaraf.py	93.45% <100.00%> (+0.54%)	⬆️
pymc/gp/util.py	96.55% <100.00%> (ø)
pymc/model.py	88.23% <100.00%> (ø)
pymc/step_methods/metropolis.py	83.40% <100.00%> (ø)
pymc/tests/distributions/test_logprob.py	100.00% <100.00%> (ø)
pymc/variational/opvi.py	87.04% <100.00%> (ø)
pymc/step_methods/hmc/base_hmc.py	89.76% <0.00%> (-0.79%)	⬇️
pymc/distributions/censored.py	100.00% <0.00%> (+7.50%)	⬆️

[lucianopaz]

Looks ok @ricardoV94 . I’m entirely sure about how the model graphs are walked and the replacements are being set. But I think it’s just that I’m tired and I’ll have to review it a second time. Anyway, I wanted to ask you to change a type hint that I think is wrong and maybe also explain why you don’t raise an exception instead of returning Nones