MultiAggregation and aggregation_resolver

[lightaime]

• Add MulAggregation, MultiAggregation and aggregation_resolver

[rusty1s]

Thank you. This mostly looks good to me, and I left a few comments.

Notably, this PR groups a bunch of separate functionality (MulAggregation, MultiAggregation, MessagePassing integration, ...) into a single PR, which makes it overly complex. Next time, I think we should work on separating these functionalities into separate PRs to ease reviewing.

[lightaime]

Thanks for the helpful comments @rusty1s! The PR is ready for another review.

[Padarn]

Looks good to me so far. Happy to review again once no longer WIP.

[rusty1s]

Thanks @lightaime. In favor of smaller PRs, I removed the MessagePassing integration such that this PR solely adds support for MulAggregation, MultiAggregation and aggregation_resolver. Can you please re-send your modifications in message_passing.py and test_message_passing.py in a separate PR? Here is the last commit: https://github.com/pyg-team/pytorch_geometric/tree/2b1ac91569c4ea4de2b603e883e6e831cc2c0c0e

[codecov]

Codecov Report

Merging #4749 (325143b) into master (893aca5) will decrease coverage by 1.86%.
The diff coverage is 97.87%.

@@            Coverage Diff             @@
##           master    #4749      +/-   ##
==========================================
- Coverage   84.39%   82.53%   -1.87%     
==========================================
  Files         324      325       +1     
  Lines       17332    17369      +37     
==========================================
- Hits        14628    14336     -292     
- Misses       2704     3033     +329     

Impacted Files	Coverage Δ
torch_geometric/nn/resolver.py	90.00% <93.33%> (-5.46%)	⬇️
torch_geometric/nn/aggr/__init__.py	100.00% <100.00%> (ø)
torch_geometric/nn/aggr/basic.py	100.00% <100.00%> (ø)
torch_geometric/nn/aggr/multi.py	100.00% <100.00%> (ø)
torch_geometric/nn/models/dimenet_utils.py	0.00% <0.00%> (-75.52%)	⬇️
torch_geometric/nn/models/dimenet.py	14.51% <0.00%> (-53.00%)	⬇️
torch_geometric/nn/conv/utils/typing.py	81.25% <0.00%> (-17.50%)	⬇️
torch_geometric/nn/inits.py	67.85% <0.00%> (-7.15%)	⬇️
torch_geometric/io/tu.py	93.90% <0.00%> (-2.44%)	⬇️
... and 7 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 893aca5...325143b. Read the comment docs.