when i use sageconv to process a graph,for example
src_ids = torch.tensor([1, 1, 3, 3])
dst_ids = torch.tensor([2, 2, 2, 4])
g = dgl.graph((src_ids, dst_ids))
What I want to know is whether Node 1 will aggregate the features of Node 2 twice
That’s important to me!
Yes it will. In fact, you can verify it by giving an all-one node values.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.