Do not propagate dims to observed component of imputed variable

[ricardoV94]

Closes #6177

Also did some refactoring and more informative errors for when imputation is not supported

Major / Breaking Changes

• ...

Bugfixes / New features

• Fix bug where dims of partially observed variables would be propagated to the smaller observed sub-component.

Docs / Maintenance

• ...

[codecov]

Codecov Report

Merging #6263 (ebcdf0c) into main (48664c0) will decrease coverage by 5.61%.
The diff coverage is 75.00%.

❗ Current head ebcdf0c differs from pull request most recent head f196c0e. Consider uploading reports for the commit f196c0e to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6263      +/-   ##
==========================================
- Coverage   94.19%   88.57%   -5.62%     
==========================================
  Files         102      102              
  Lines       21477    21480       +3     
==========================================
- Hits        20230    19026    -1204     
- Misses       1247     2454    +1207     

Impacted Files	Coverage Δ
pymc/model.py	79.72% <75.00%> (-9.75%)	⬇️
pymc/tests/step_methods/test_compound.py	0.00% <0.00%> (-100.00%)	⬇️
pymc/tests/distributions/test_timeseries.py	0.00% <0.00%> (-99.51%)	⬇️
pymc/tests/step_methods/hmc/test_quadpotential.py	0.00% <0.00%> (-95.82%)	⬇️
pymc/distributions/timeseries.py	28.46% <0.00%> (-66.00%)	⬇️
pymc/model_graph.py	15.29% <0.00%> (-63.53%)	⬇️
pymc/data.py	47.96% <0.00%> (-32.12%)	⬇️
pymc/step_methods/hmc/quadpotential.py	73.76% <0.00%> (-6.94%)	⬇️
pymc/distributions/logprob.py	91.09% <0.00%> (-6.17%)	⬇️
... and 3 more

[michaelosthege]

Just a bit of nitpicks about our testing code..

[ricardoV94]

Uh, it seems tests don't rerun when the PR has already been approved? Had to re-request a review to trigger them