You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is it possible to add multiple features to a single node to a networkx graph and then pass the graph to the graph net function utils_np.networkx_to_data_dict()?
no, this is not currently supported, as utils_np.networkx_to_data_dict will only pull the features contained under the key "features". Moreover the remaining of the graph net library (e.g. the modules) assumes that the node features are single tensors.
You could create several graphs and then concatenate them down the line if that helps.
Is it possible to add multiple features to a single node to a networkx graph and then pass the graph to the graph net function
utils_np.networkx_to_data_dict()
?In the documentation this case is not explicitly stated:
https://github.com/deepmind/graph_nets/blob/master/docs/graph_nets.md#utils_npnetworkx_to_data_dictgraph_nx-node_shape_hintnone-edge_shape_hintnone-data_type_hintfloat32
Thank you
The text was updated successfully, but these errors were encountered: