Error dimension specified as 0 but tensor has no dimensions in Step 2 - SoftTCNLearning_Supervised.py

[Pacomito]

Hello,

Now that I have the input in the right place, I manage to run Step 0 and 1 successfully.

Now, in the Step 2 - SoftTCNLearning_Supervised.py when running this line : model = Net(dataset.num_features, dataset.num_classes).to(device)

I get :

---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
Cell In[57], line 2
      1 device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
----> 2 model = Net(dataset.num_features, dataset.num_classes).to(device)  #Initialize model for each fold.
      3 optimizer = torch.optim.Adam(model.parameters(), lr=LearningRate)
      5 FoldFolderName = TimeFolderName + "/Fold" + str(num_fold)

File ~/anaconda3/envs/CytoCommunity/lib/python3.10/site-packages/torch_geometric/data/dataset.py:114, in Dataset.num_features(self)
    110 @property
    111 def num_features(self) -> int:
    112     r"""Returns the number of features per node in the dataset.
    113     Alias for :py:attr:`~num_node_features`."""
--> 114     return self.num_node_features

File ~/anaconda3/envs/CytoCommunity/lib/python3.10/site-packages/torch_geometric/data/dataset.py:103, in Dataset.num_node_features(self)
    100 @property
    101 def num_node_features(self) -> int:
    102     r"""Returns the number of features per node in the dataset."""
--> 103     data = self[0]
    104     data = data[0] if isinstance(data, tuple) else data
    105     if hasattr(data, 'num_node_features'):

File ~/anaconda3/envs/CytoCommunity/lib/python3.10/site-packages/torch_geometric/data/dataset.py:198, in Dataset.__getitem__(self, idx)
    193 if (isinstance(idx, (int, np.integer))
    194         or (isinstance(idx, Tensor) and idx.dim() == 0)
    195         or (isinstance(idx, np.ndarray) and np.isscalar(idx))):
    197     data = self.get(self.indices()[idx])
--> 198     data = data if self.transform is None else self.transform(data)
    199     return data
    201 else:

File ~/anaconda3/envs/CytoCommunity/lib/python3.10/site-packages/torch_geometric/transforms/to_dense.py:51, in ToDense.__call__(self, data)
     48     size = [num_nodes - data.pos.size(0)] + list(data.pos.size())[1:]
     49     data.pos = torch.cat([data.pos, data.pos.new_zeros(size)], dim=0)
---> 51 if data.y is not None and (data.y.size(0) == orig_num_nodes):
     52     size = [num_nodes - data.y.size(0)] + list(data.y.size())[1:]
     53     data.y = torch.cat([data.y, data.y.new_zeros(size)], dim=0)

IndexError: dimension specified as 0 but tensor has no dimensions

It seems that accessing any of the attributes: dataset.num_classes, dataset.num_features, dataset.num_node_features or dataset.num_edge_features produces the same error and it seems to be related to accessing data.y.size(0)

Best, Pacôme