TypeError: full() received an invalid combination of arguments - got (tuple, int, device=NoneType),
Ninja Mountain: ## Cause analysis: The problem occurs on the data. The `x`, `edge_index`, `y` and other parameters of the `torch_geometric.` object should be PyTorch tensor objects, rather than Python lists or NumPy [array](/course/detail/40020?utm_source=glcblog&spm=1001.2101.3001.7020). I used before: ```undefined Data(x=[0, 1], edge_index=[[0, 1], [0, 1], y=[0, 1]) ``` All objects are created as lists in Python. ___ ## Solution: Just convert them all to Tensor: ```undefined data = Data(x=(features[i], dtype=), edge_index=(edge_indices[i], dtype=), y=(targets[i], dtype=float)) ```