search

microsoft / Graphormer

Graphormer is a general-purpose deep learning backbone for molecular modeling.
MIT License
2.08k stars 334 forks source link

I don't know the effect of adding offset to the x feature in datapreprocess #36

Closed huahuaai closed 2 years ago

zhengsx commented 2 years ago

Could you provide more detailed description about your question, including the reference code?

huahuaai commented 2 years ago

Could you provide more detailed description about your question, including the reference code?

Could you provide more detailed description about your question, including the reference code?

def convert_to_single_emb(x, offset=512): feature_num = x.size(1) if len(x.size()) > 1 else 1 feature_offset = 1 + \ torch.arange(0, feature_num * offset, offset, dtype=torch.long) x = x + feature_offset return x I found this code in wrapper.py in your project for preprocess data but I can't understand the effect of adding an offset to the atom feature

zhengsx commented 2 years ago

Please have a look at previous issues #18 #28 .

zhengsx commented 2 years ago

Close this issue due to inactive. Feel free to raise a new one or reopen this one for any further question.