Request cold start data file about r2-ddi

[linjc16]

1. You can use drugbank_id_smiles.csv in transductive folder.
   2,3,4 The original files comes from GMPNN [1]. For our repo, you can just run [run_binarize_unseen.sh] and [run_process_ind.py] since I have uploaded train.csv and valid.csv.

[1] Nyamabo AK, Hui Y, Liu Z, et al. Drug–drug interaction predic�tion with learnable size-adaptive molecular substructures. Brief
Bioinform 2022;23(1):bbab441.

[Ngqgq]

In GMPNN-CS, through the data_preprocessing.py can only generate warm start data files, I can't find a way to generate cold start temporarily, if you can, could you provide a way to generate cold start data or tell me how to get that?