RuntimeError: expected scalar type Float but found Double From GCNConv to GraphConv

[navidaftabi]

Hello, I am getting the following error even when I change the data type before feeding the output from GCNConv to GraphConve

RuntimeError: expected scalar type Float but found Double

The DataBatch's are like DataBatch(edge_index=[2, 41034], n_id=[1644, 100], edge_attr=[41034, 1], num_nodes=1644, y=[64], batch=[1644], ptr=[65]).

and, my code is like the following:

idConv = pyg_nn.GCNConv(in_channels=100,  out_channels=64)
conv1 = pyg_nn.GraphConv(in_channels=64, out_channels=64)

n_id = data.n_id.type(torch.float32)            #data is a DataBatch
edge_index = data.edge_index
edge_weight = data.edge_attr

x = idConv(n_id, edge_index, edge_weight)
x = x.relu()
x = conv1(x, edge_index, edge_weight)

But I get the following error:

Traceback (most recent call last):

  File "C:\Users\...\AppData\Local\Temp\ipykernel_5036\1539429924.py", line 1, in <module>
    x = conv1(x, edge_index, edge_weight)

  File "C:\Users\...\Anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1110, in _call_impl
    return forward_call(*input, **kwargs)

  File "C:\Users\...\Anaconda3\lib\site-packages\torch_geometric\nn\conv\graph_conv.py", line 82, in forward
    out = self.lin_rel(out)

  File "C:\Users\...\Anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1110, in _call_impl
    return forward_call(*input, **kwargs)

  File "C:\Users\...\Anaconda3\lib\site-packages\torch_geometric\nn\dense\linear.py", line 118, in forward
    return F.linear(x, self.weight, self.bias)

RuntimeError: expected scalar type Float but found Double

I tried x = self.conv1(x.float(), edge_index, edge_weight) but I got the same error. I would appreciate if you could help me to resolve this error.

[navidaftabi]

I understand what was the problem. I used edge_weight.float() and it is resolved.