Parallelize the LineGraph implementation to speed up in multi-core machines
Open
feature request
- Dominant language
- Python
- Stars
- 14.3k
- Forks
- 3.1k
- PR merge metrics
- No merged PRs in 30d
Description
https://github.com/dmlc/dgl/blob/230b886ec55b9927bc806848e30c3a9b33bbcd6e/python/dgl/transforms/module.py#L764
Current LineGraph implementation takes long time in large graphs (15M edges-8M nodes). Could we speed up the implementation of this transform?
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.