search

SherylHYX / pytorch_geometric_signed_directed

PyTorch Geometric Signed Directed is a signed/directed graph neural network extension library for PyTorch Geometric. The paper is accepted by LoG 2023.
https://pytorch-geometric-signed-directed.readthedocs.io/en/latest/
MIT License
117 stars 16 forks source link

SGCN and SNEA optimizations #40

Closed huangjunjie-cs closed 1 year ago

huangjunjie-cs commented 1 year ago
  1. Rewrite SNEAConv and SGCNConv
  2. Extract Link Sign Loss and Structural Balance Loss
  3. Add validation set for link sign prediction
codecov[bot] commented 1 year ago

Codecov Report

Base: 99.59% // Head: 99.59% // Decreases project coverage by -0.00% :warning:

Coverage data is based on head (e94584f) compared to base (8e8c4bf). Patch coverage: 98.74% of modified lines in pull request are covered.

Additional details and impacted files ```diff @@ Coverage Diff @@ ## main #40 +/- ## ========================================== - Coverage 99.59% 99.59% -0.01% ========================================== Files 74 76 +2 Lines 3749 3709 -40 ========================================== - Hits 3734 3694 -40 Misses 15 15 ``` | [Impacted Files](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He) | Coverage Δ | | |---|---|---| | [torch\_geometric\_signed\_directed/nn/signed/SGCN.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC9ubi9zaWduZWQvU0dDTi5weQ==) | `97.50% <90.90%> (-2.50%)` | :arrow_down: | | [torch\_geometric\_signed\_directed/nn/signed/SNEA.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC9ubi9zaWduZWQvU05FQS5weQ==) | `97.67% <92.85%> (-0.85%)` | :arrow_down: | | [torch\_geometric\_signed\_directed/nn/signed/SDGNN.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC9ubi9zaWduZWQvU0RHTk4ucHk=) | `100.00% <100.00%> (ø)` | | | [...ch\_geometric\_signed\_directed/nn/signed/SGCNConv.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC9ubi9zaWduZWQvU0dDTkNvbnYucHk=) | `100.00% <100.00%> (ø)` | | | [...ch\_geometric\_signed\_directed/nn/signed/SNEAConv.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC9ubi9zaWduZWQvU05FQUNvbnYucHk=) | `100.00% <100.00%> (ø)` | | | [torch\_geometric\_signed\_directed/nn/signed/SiGAT.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC9ubi9zaWduZWQvU2lHQVQucHk=) | `100.00% <100.00%> (ø)` | | | [...geometric\_signed\_directed/utils/signed/\_\_init\_\_.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC91dGlscy9zaWduZWQvX19pbml0X18ucHk=) | `100.00% <100.00%> (ø)` | | | [...\_directed/utils/signed/create\_spectral\_features.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC91dGlscy9zaWduZWQvY3JlYXRlX3NwZWN0cmFsX2ZlYXR1cmVzLnB5) | `100.00% <100.00%> (ø)` | | | [...ric\_signed\_directed/utils/signed/link\_sign\_loss.py](https://codecov.io/gh/SherylHYX/pytorch_geometric_signed_directed/pull/40/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He#diff-dG9yY2hfZ2VvbWV0cmljX3NpZ25lZF9kaXJlY3RlZC91dGlscy9zaWduZWQvbGlua19zaWduX2xvc3MucHk=) | `100.00% <100.00%> (ø)` | | Help us with your feedback. Take ten seconds to tell us [how you rate us](https://about.codecov.io/nps?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He). Have a feature suggestion? [Share it here.](https://app.codecov.io/gh/feedback/?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=Yixuan+He)

:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.