How to solve TypeError: forward() got an unexpected keyword argument 'edge_index'

[yung1231]

I try to explain whole graph and then execute the following command

for data in data_loader:
  explainer = GNNExplainer(model, epochs=3)
  print('explainer: ', explainer)
  edge_index = data.edge_index
  node_feat_mask, edge_mask = explainer.explain_graph(data.x, data.edge_index)  # return node_feat_mask, edge_mask

and get the error

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
/tmp/ipykernel_1834/3877013572.py in <module>
      5   # data = data.to(device)
      6   # edge_mask, _ = explainer.explain_graph(data)  # return node_feat_mask, edge_mask
----> 7   node_feat_mask, edge_mask = explainer.explain_graph(data.x, data.edge_index)  # return node_feat_mask, edge_mask
      8   # ax, G = explainer.visualize_subgraph(None, data.edge_index, edge_mask, y=data.y)
      9   # plt.show()

~/CFG/lib/python3.7/site-packages/torch_geometric/nn/models/gnn_explainer.py in explain_graph(self, x, edge_index, **kwargs)
    164             optimizer.zero_grad()
    165             h = x * self.node_feat_mask.sigmoid()
--> 166             out = self.model(x=h, edge_index=edge_index, batch=batch, **kwargs)
    167             loss = self.get_loss(out, prediction, None)
    168             loss.backward()

~/CFG/lib/python3.7/site-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
   1128         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks
   1129                 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1130             return forward_call(*input, **kwargs)
   1131         # Do not call functions when jit is used
   1132         full_backward_hooks, non_full_backward_hooks = [], []

TypeError: forward() got an unexpected keyword argument 'edge_index'

I couldn't find the answer on the internet.

[rusty1s]

How does your model's forward function look like? Does it expect both x and edge_index as input arguments?

[yung1231]

my forward is written like this forward(self, x, edges, batch), and then I changed it to forward(self, x, edge_index, batch) and it works fine.

Thanks.

[rusty1s]

Ok, cool :) It is a current limitation to require same naming at the moment. Will look into it.