search

pyg-team / pytorch_geometric

Graph Neural Network Library for PyTorch
https://pyg.org
MIT License
21.14k stars 3.63k forks source link

label propagation question #2619

Open CnBDM-Su opened 3 years ago

CnBDM-Su commented 3 years ago

From now on, we recommend using our discussion forum (https://github.com/rusty1s/pytorch_geometric/discussions) for general questions.

❓ Questions & Help

I used the package nn.correct and smooth. But I found the edge_weight parameter seems not working. I have tried use a random tensor as weight, but I always get the same result as without edge_weight. Could you check if there an error in the class? Thanks.

rusty1s commented 3 years ago

Are you using the SparseTensor class for propagation? Then you can directly add edge weights to this representation, i.e.:

adj_t = adj_t.set_value(edge_weight, layout='coo')