Error on tutorial for graph generation

[Takaogahara]

Description

Hello, I'm trying to run the Tutorial for Graph Generation example available on the documentation. However, I'm having an error when running the code described.

I don't know if it could be something related to the version of the libraries used or something else.

Here is a code to reproduce the error:

import json
from rdkit import RDLogger
from rdkit.Chem import Draw
from dig.ggraph.method import GraphDF
from dig.ggraph.dataset import ZINC250k
from dig.ggraph.evaluation import RandGenEvaluator
from torch_geometric.loader import DenseDataLoader

RDLogger.DisableLog("rdApp.*")


conf = json.load(open("rand_gen_zinc250k_config_dict.json"))
dataset = ZINC250k(one_shot=False, use_aug=True)
loader = DenseDataLoader(dataset, batch_size=conf["batch_size"], shuffle=True)

runner = GraphDF()
lr = 0.001
wd = 0
max_epochs = 10
save_interval = 1
save_dir = "rand_gen_zinc250k"

runner.train_rand_gen(loader=loader, lr=lr, wd=wd, max_epochs=max_epochs,
                      model_conf_dict=conf["model"],
                      save_interval=save_interval, save_dir=save_dir)

This is the Traceback I received.

/home/takaogahara/virtualenvs/dig/lib/python3.10/site-packages/torch_geometric/data/in_memory_dataset.py:284: UserWarning: It is not recommended to directly access the internal storage format `data` of an 'InMemoryDataset'. If you are absolutely certain what you are doing, access the internal storage via `InMemoryDataset._data` instead to suppress this warning. Alternatively, you can access stacked individual attributes of every graph via `dataset.{attr_name}`.
  warnings.warn(msg)
Traceback (most recent call last):
  File "/media/takaogahara/storage1/MolGen/teste.py", line 23, in <module>
    runner.train_rand_gen(loader=loader, lr=lr, wd=wd, max_epochs=max_epochs,
  File "/home/takaogahara/virtualenvs/dig/lib/python3.10/site-packages/dig/ggraph/method/GraphDF/graphdf.py", line 67, in train_rand_gen
    for batch, data_batch in enumerate(loader):
  File "/home/takaogahara/virtualenvs/dig/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 630, in __next__
    data = self._next_data()
  File "/home/takaogahara/virtualenvs/dig/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 674, in _next_data
    data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
  File "/home/takaogahara/virtualenvs/dig/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 51, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/home/takaogahara/virtualenvs/dig/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 51, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/home/takaogahara/virtualenvs/dig/lib/python3.10/site-packages/torch_geometric/data/dataset.py", line 263, in __getitem__
    data = self.get(self.indices()[idx])
  File "/home/takaogahara/virtualenvs/dig/lib/python3.10/site-packages/dig/ggraph/dataset/PygDataset.py", line 171, in get
    for key in self.data.keys:
TypeError: 'method' object is not iterable

Environment

• PyG version: 2.4.0
• PyTorch version: 2.1.0
• DIG version: 1.1.0
• OS: Ubuntu 22.04
• Python version: 3.10
• CUDA/cuDNN version: cu121

[irumeria]

for key in self.data.keys: -> for key in self.data.keys()

works for me

[lyzustc]

@Takaogahara Thank you for pointing this bug! Actually ''for key in self.data.keys'' follows the implementaion of InMemoryDataset class in pyg 1.x, but pyg 2.x updates the implementation and I guess ''keys'' is no longer an attribute in ''self.data'' in pyg 2.x now. @irumeria Thank you very much for providing fix suggestion! I have changed ''self.data.keys'' to ''self.data.keys()''.