Some problems with LightGCN example code

[TheAllUnknown]

Hey, I found that in the lightGCN example: https://github.com/pyg-team/pytorch_geometric/blob/master/examples/lightgcn.py, in the line 42, I think there might be some problems with the negative sampling, as the code itself is doing random sampling, rather than negative random sampling, am I right? I think a better way might be using the structure_negative_sampling