dmlc / dgl

Python package built to ease deep learning on graph, on top of existing DL frameworks.
http://dgl.ai
Apache License 2.0
13.55k stars 3.02k forks source link

Apply Edges for Edges with features #4007

Closed zohairhadi closed 2 years ago

zohairhadi commented 2 years ago

🚀 Feature

I want to use edges which have a feature set. Currently the apply_edges(). function does not cater to edge features.

Motivation

I am trying to use a scene graph for edge classification problem which has edge features. To create an edge classification GNN, I need a GNN layer to classify edges using its features. That is why i need this functionality.

Pitch

The apply_edges() function should have a optional parameter for taking in edge_features and use them in classification and prediction tasks.

jermainewang commented 2 years ago

You can save the edge features to g.edata and access them in edges.data in the UDF provided to apply_edges.

github-actions[bot] commented 2 years ago

This issue has been automatically marked as stale due to lack of activity. It will be closed if no further activity occurs. Thank you