_torch_cat_na

pytorch_forecasting.models.base._base_model._torch_cat_na(x: list[Tensor]) → Tensor

Concatenate tensor along dim=0 and add nans along dim=1 if necessary.

Allows concatenation of tensors where dim=1 are not equal. Missing values are filled up with nan.

Parameters:

x (List[torch.Tensor]) – list of tensors to concatenate along dimension 0

Returns:

concatenated tensor

Return type:

torch.Tensor