Issue when using datasets slicing

[scottshufe]

Hi pyg team,

I meet a problem when I use the dataset slicing, my codes (follow the pyg intro doc) :

from torch_geometric.datasets import TUDataset

dataset = TUDataset(root='/tmp/ENZYMES', name='ENZYMES')

data = dataset[5:10]

And the error info:

Traceback (most recent call last):
File "D:/PycharmProjects/GCL_Privacy_Risk/Property inference attack/new.py", line 5, in
data = dataset[5:10]
File "D:\ProgramData\Anaconda3\envs\GCL_risk\lib\site-packages\torch_geometric\data\dataset.py", line 203, in getitem
return self.index_select(idx)
File "D:\ProgramData\Anaconda3\envs\GCL_risk\lib\site-packages\torch_geometric\data\dataset.py", line 238, in index_select
dataset = copy.copy(self)
File "D:\ProgramData\Anaconda3\envs\GCL_risk\lib\copy.py", line 88, in copy
return copier(x)
TypeError: copy() takes 1 positional argument but 2 were given

I'm using pytorch-geometric 2.0.2. I don't know what's wrong with my code. Could you please give me some suggestions to fix this issue?

[rusty1s]

I can't reproduce this issue on master or on 2.0.2. It's a bit weird that your code complains about copy.copy receiving 2 arguments. Is your conda environment broken? What happens if you set up a new one?

[scottshufe]

Thanks for your quick reply. I'll check my env first.

[scottshufe]

I have changed my environment and the issue disappeared...It's confusing. Anyway, thanks for your help :)

[rusty1s]

Indeed weird. Let me know if you have any clue why this has happened :)

[scottshufe]

Okk.