Lack of Dataset about caster HOT 3 CLOSED

Besides, it lacks the used drugbank dataset, could you upload this dataset or give a download link?

from caster.

here is the dataset: https://drive.google.com/file/d/1drgMxl3BakcSpUQuyN-KmUdxiUuWFeJ7/view?usp=sharing

from caster.

Thanks! Besides, I want to compare the baseline fairly, so I would ask for some details.

1. Whether this Drugbank dataset splits like the BIOSNAP (train 7: valid 1: test 2).

2. And whether the shuffled seed is 1 in this Drugbank dataset
   in your code "dde_torch.py" for BIOSNAP:
   torch.manual_seed(1) # reproducible np.random.seed(1)

So maybe I can write the following code to split the dataset?

from random import Random
Random(1).shuffle(drugbank)
train_size = int(0.7 * len(drugbank))
valid_size = int(0.1 * len(drugbank))
train = drugbank[: train_size]
valid = drugbank[train_size: train_size + valid_size]
test = drugbank[train_size + valid_size:]

from caster.