download dataset failure

[didi226]

`import os.path as osp from mne import get_config from moabb.utils import set_download_dir from moabb.datasets import Lee2019_MI,Lee2019_ERP from moabb.paradigms import LeftRightImagery from moabb import benchmark, set_log_level original_path = get_config("MNE_DATA") print(f"The download directory is currently {original_path}")

new_path = r"D:\data\bci_peformance" set_download_dir(new_path) dataset = Lee2019_ERP() cc = dataset.get_data() dataset2 = Lee2019_MI cc = dataset2.get_data() error: C:\software\Anaconda\envs\py3.10\lib\site-packages\moabb\pipelines__init__.py:26: ModuleNotFoundError: Tensorflow is not installed. You won't be able to use these MOABB pipelines if you attempt to do so. warn( Downloading data from 'ftp://parrot.genomics.cn/gigadb/pub/10.5524/100001_101000/100542/session1/s1/sess01_subj01_EEG_ERP.mat' to file 'D:\data\bci_peformance\MNE-lee2019-erp-data\gigadb\pub\10.5524\100001_101000\100542\session1\s1\sess01_subj01_EEG_ERP.mat'. The download directory is currently D:\data\bci_peformance Traceback (most recent call last): File "D:\code_cloud\new_eeg_deep\scut_eeg_dl\scut_ssvep_aperiod\src\scut_ssvep_aperiod\predict_resting_state\load_dataset_erp.py", line 13, in cc = dataset.get_data() File "C:\software\Anaconda\envs\py3.10\lib\site-packages\moabb\datasets\base.py", line 342, in get_data data[subject] = self._get_single_subject_data_using_cache( File "C:\software\Anaconda\envs\py3.10\lib\site-packages\moabb\datasets\base.py", line 436, in _get_single_subject_data_using_cache sessions_data = self._get_single_subject_data(subject) File "C:\software\Anaconda\envs\py3.10\lib\site-packages\moabb\datasets\Lee2019.py", line 168, in _get_single_subject_data file_path_list = self.data_path(subject) File "C:\software\Anaconda\envs\py3.10\lib\site-packages\moabb\datasets\Lee2019.py", line 213, in data_path data_path = dl.data_dl(url, self.code, path, force_update, verbose) File "", line 12, in data_dl File "C:\software\Anaconda\envs\py3.10\lib\site-packages\moabb\datasets\download.py", line 155, in data_dl dlpath = retrieve( File "C:\software\Anaconda\envs\py3.10\lib\site-packages\pooch\core.py", line 239, in retrieve stream_download(url, full_path, known_hash, downloader, pooch=None) File "C:\software\Anaconda\envs\py3.10\lib\site-packages\pooch\core.py", line 807, in stream_download downloader(url, tmp, pooch) File "C:\software\Anaconda\envs\py3.10\lib\site-packages\pooch\downloaders.py", line 341, in call ftp.connect(host=parsed_url["netloc"], port=self.port) File "C:\software\Anaconda\envs\py3.10\lib\ftplib.py", line 158, in connect self.sock = socket.create_connection((self.host, self.port), self.timeout, File "C:\software\Anaconda\envs\py3.10\lib\socket.py", line 845, in create_connection raise err File "C:\software\Anaconda\envs\py3.10\lib\socket.py", line 833, in create_connection sock.connect(sa) ` my net is right, but I can't download the dataset.

[bruAristimunha]

It could be two things:

• version issue, or
• Issue with windows. Can you provide your moabb version, @didi226?

[didi226]

my version python==3.10, moabb ==1.0.0 the service for 'ftp://parrot.genomics.cn/gigadb/pub/10.5524/100001_101000/100542/session1/s1/sess01_subj01_EEG_ERP.mat' have been closed.

[CheHumbleProgger]

I have the same issue considering the Lee2019ERP dataset

My code is as follows

from braindecode.datasets import MOABBDataset

subj_ids = [n for n in range(2, 5)]
lee_erp = MOABBDataset(dataset_name='Lee2019_ERP', subject_ids=subj_ids)

And it returns the following:

Downloading data from 'ftp://parrot.genomics.cn/gigadb/pub/10.5524/100001_101000/100542/session1/s2/sess01_subj02_EEG_ERP.mat' to file 'C:\Users\user\mne_data\MNE-lee2019-erp-data\gigadb\pub\10.5524\100001_101000\100542\session1\s2\sess01_subj02_EEG_ERP.mat'.

Then the WinError 10061 is thrown. This makes me think that the problem is on the server side

Braindecode version: 0.8.1 MOABB version: 1.0.0 Python version: 3.9

Thank you in advance for your support.

[bruAristimunha]

Hi @didi226 and @CheHumbleProgger!

Thank you for reporting.

Can you please both of you upgrade the moabb version, we release a small patch to fix this issue.

[bruAristimunha]

And indeed, the problem is on the server side

[CheHumbleProgger]

I can load data from Lee2019_ERP successfuly. Thank you for fixing it so fast

[didi226]

Thank you very much, I have got data from "https://s3.ap-northeast-1.wasabisys.com/gigadb-datasets/live/pub/10.5524/100001_101000/100542/". Now, I can also load data from Lee2019_ERP successfuly by moabb.