ZHANG Zhi

The code below looks confusing to me, as it would likely result in a batch containing samples of the same class: ``` train_loader = DataLoader(train_dataset, batch_size=64, shuffle=False) ``` Could you...

Happy New Year~ You can try the following code: ``` from torcheeg.datasets import SEEDDataset from torcheeg.model_selection import KFoldPerSubjectGroupbyTrial from torcheeg import transforms dataset = SEEDDataset( io_path=f'./SEED_preprocc/DE_feat', root_path=path_to_db, offline_transform=transforms.BandDifferentialEntropy( band_dict={ "delta":...

I've updated an internal version to meet your requirement. You can install the latest version of TorchEEG by running: ``` pip install git+https://github.com/torcheeg/torcheeg.git ``` Use `train_test_split_per_subject_cross_trial` to divide the dataset...

I checked the code and found that it runs well. I couldn't reproduce your problem. I removed the irrelevant functionality. Can you check what's different between your code and the...

[Note] This is merely a proposal and relevant matters will be discussed and updated within this issue.

Yes, we are working on this feature. We agree that multimodality and especially physiological signals contribute to EEG modeling. Our current solution is based on [transforms](https://torcheeg.readthedocs.io/en/latest/torcheeg.transforms.html#label-transforms) and reads GSR from...

This functionality is already provided: https://github.com/torcheeg/torcheeg/blob/main/torcheeg/transforms/numpy/to.py#L110-L129 I am not sure if it should exist as a transform. If you need, you can call the reverse method.

This issue arises from the third-party package spectrum, and unfortunately, we are unable to resolve it on the torcheeg side. We need to wait for the resolution of https://github.com/cokelaer/spectrum/issues/83. However,...

Thanks for starting the project, it really works as magic. However, I think there seem to be more automated prompts when generating tutorial documents. I suggest the following way: Objective:...

After a simple try, I found that a lot of labor is still necessary. BTW, I suppose a more effective way may be to write comments and tutorial documents first,...