fix(analytical): fix the louvain algorithm that the result is not consistent with #183

[acezen]

What do these changes do?

This change try to fix the louvain algorithm that the result is not consistent with #183

Related issue number

Close #3457

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (f0441bb) 40.90% compared to head (f9a7e4d) 70.96%.

Additional details and impacted files

@@             Coverage Diff             @@
##             main    #3472       +/-   ##
===========================================
+ Coverage   40.90%   70.96%   +30.06%     
===========================================
  Files         109      110        +1     
  Lines       11343    11384       +41     
===========================================
+ Hits         4640     8079     +3439     
+ Misses       6703     3305     -3398     

Files	Coverage Δ
python/graphscope/tests/unittest/test_app.py	98.00% <100.00%> (+98.00%)	⬆️
python/graphscope/tests/unittest/test_lazy.py	93.66% <100.00%> (+93.66%)	⬆️

... and 63 files with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f0441bb...f9a7e4d. Read the comment docs.

[JackyYangPassion]

unit test #3474

expect 103 community but 10733

FAILED ../../../.local/lib/python3.8/site-packages/graphscope/tests/unittest/test_app.py::test_other_app_on_undirected_graph - assert 10733 == 103

[acezen]

unit test #3474

expect 103 community but 10733

FAILED ../../../.local/lib/python3.8/site-packages/graphscope/tests/unittest/test_app.py::test_other_app_on_undirected_graph - assert 10733 == 103

Yes, there are still something wrong and it's not related to min_progress parameter, trying to figure out

[acezen]

unit test #3474

expect 103 community but 10733

FAILED ../../../.local/lib/python3.8/site-packages/graphscope/tests/unittest/test_app.py::test_other_app_on_undirected_graph - assert 10733 == 103

I find the real problem for that, it seems that we need to do a wcc to the louvain result and then get the real community result... since the community id in graphscope can represent by any of the id in the community.

[JackyYangPassion]

unit test #3474
expect 103 community but 10733

FAILED ../../../.local/lib/python3.8/site-packages/graphscope/tests/unittest/test_app.py::test_other_app_on_undirected_graph - assert 10733 == 103

I find the real problem for that, it seems that we need to do a wcc to the louvain result and then get the real community result... since the community id in graphscope can represent by any of the id in the community.

Very cool. @acezen Can you provide the logic for subsequent processing? I can verify it based on test data. I have been watching GAE for the past two days based on the PIE model, trying to find the root cause in a difficult way.

[acezen]

unit test #3474
expect 103 community but 10733

FAILED ../../../.local/lib/python3.8/site-packages/graphscope/tests/unittest/test_app.py::test_other_app_on_undirected_graph - assert 10733 == 103

I find the real problem for that, it seems that we need to do a wcc to the louvain result and then get the real community result... since the community id in graphscope can represent by any of the id in the community.

Very cool. @acezen Can you provide the logic for subsequent processing? I can verify it based on test data. I have been watching GAE for the past two days based on the PIE model, trying to find the root cause in a difficult way.

later I will push a commit that output the correct result of Lovain, then it should not need user to do the union find.

[acezen]

hi, @shirly121, the test_gremlin_timeout failed in CI, can you help look at the reason for the failure?
https://github.com/alibaba/GraphScope/actions/runs/7496630317/job/20409902909?pr=3472