search

yuyangw / MolCLR

Implementation of MolCLR: "Molecular Contrastive Learning of Representations via Graph Neural Networks" in PyG.
MIT License
233 stars 57 forks source link

Random split for QM9 #14

Closed yeondukim closed 1 year ago

yeondukim commented 2 years ago

Hi @yuyangw :)

I've found the valid and test indexes are shared in the case of the random split (especially for QM9). https://github.com/yuyangw/MolCLR/blob/dd87b5f4ab8719616ba3a3d2e64759299ae46e18/dataset/dataset_test.py#L201

I guess it should be this manner. valid_idx, test_idx, train_idx = indices[:split], indices[split:split+split2], indices[split+split2:]

yuyangw commented 1 year ago

Hi @yeondukim,

Thanks for pointing this out. I've fixed this in the latest update.

Best, Yuyang