Add a sum method to _sparse_py_operators

[aerubanov]

This PR add sum method for sparse tensors according to #522. So one can call x.sum() instead of aesara.sparse.sp_sum(x).

I think this change should also be reflected in the documentation, but I could not find a section with the methods of sparse tensors. We need to decide something about what to do with it.

[brandonwillard]

Oddly, there's a test that asserts the lack of a sum method. That's what's causing the errors in CI.

Don't forget that you can run just the sparse tests locally (e.g. pytest tests/sparse) if you want to debug; those shouldn't take too long to run by themselves.

[codecov]

Codecov Report

Merging #745 (8e30d88) into main (3edbbc4) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main     #745   +/-   ##
=======================================
  Coverage   78.16%   78.17%           
=======================================
  Files         152      152           
  Lines       47652    47663   +11     
  Branches    10880    10881    +1     
=======================================
+ Hits        37249    37260   +11     
  Misses       7846     7846           
  Partials     2557     2557           

Impacted Files	Coverage Δ
aesara/sparse/basic.py	82.11% <100.00%> (+0.02%)	⬆️
aesara/graph/opt.py	65.74% <0.00%> (ø)
aesara/tensor/type_other.py	82.75% <0.00%> (+1.98%)	⬆️

[aerubanov]

I just deleted this strange assert.

[brandonwillard]

Looks great, thanks!

[brandonwillard]

By the way, if you're interested in revolutionizing the sparse matrix support in Aesara, take a look at #303. It's actually pretty close to done, but I haven't had the time to get back to it. You're more than welcome to take that over, though.

[aerubanov]

@brandonwillard, Thanks for the information! I'll be sure to take a look at that pool-request.