model para issue..

[zyd98123]

Hello, I get this error AttributeError: 'IAGCN' object has no attribute 'nodevec1' when I run your program, have you encountered it?

[wtl52656]

hi,
I don't meet this error, but I guess if you set the parameter "gcn_bool" and "addaptadj" as True, the error could be addressed.
I have updated the config command in the train.py and command.py, you can download it and run the code.

[zyd98123]

Thank you for your answer, I have solved this problem now, but when I was training, I found that there is a conv1d as a forgetting gate in the model, but my data dimensions are not aligned, I followed your operation step by step, I suspect it is my data problem, please what is the first x.shape of your data.npy file

Below is the interface where I follow the command of your program

issue ：gate = self.gate_convsi

[wtl52656]

@zyd98123
The first X.shape of data.npy file is the time dimension, your input format is right.
The reason for this problem is due to the PyTorch version. You can update Pytorch to 1.9.1. Or change the code in line 200 of the model.py from nn.Conv1d to nn.Conv2d.

[zyd98123]

Thank you very much for your help, it has worked successfully