Bidirectional graph with dgl.heterograph

[driss]

please is it possible to create a heterogeneous bidirectional graph under DGL? containing nodes and edges of different types… if so? how can we do it? And thank you in advance.

[mufeili]

You can do something as follows.

import dgl
import torch

data_dict = {
    ('user', 'follows', 'topic'): (torch.tensor([1, 1]), torch.tensor([1, 2])),
}
g = dgl.heterograph(data_dict)
transform = dgl.AddReverse()
new_g = transform(g)