tslearn-team / tslearn

The machine learning toolkit for time series analysis in Python
https://tslearn.readthedocs.io
BSD 2-Clause "Simplified" License
2.92k stars 342 forks source link

UCR_UEA_datasets().list_univariate_datasets() and UCR_UEA_datasets().load_dataset() and Failed #516

Closed LbzSteven closed 5 months ago

LbzSteven commented 6 months ago

Describe the bug When I call the function UCR_UEA_datasets().list_univariate_datasets(), it failed at: image Also, many of the datasets I could not load correctly.

Environment (please complete the following information):

LbzSteven commented 6 months ago

The tslearn version is 0.6.3

LbzSteven commented 6 months ago

I think the issue is www.timeseriesclassification.com website is down right now. Just wonder if we have other resorts.

nkchangshu commented 6 months ago

URLError: <urlopen error [Errno -2] Name or service not known> I encountered the same issue, and the URL www.timeseriesclassification.com is not accessible. Please help update the source or upload the dataset to GitHub as soon as possible. Thank you very much.

YannCabanes commented 5 months ago

Hello @LbzSteven and @nkchangshu,

Thanks for opening an issue. It seems that the URL https://www.timeseriesclassification.com/index.php is not down anymore, but it is written on the home page of the website that: "This website has been down for a while."

Using Google Colab, after installing tslearn running !pip install tslearn, I can run:

from tslearn.datasets import UCR_UEA_datasets

X_train, y_train, X_test, y_test = UCR_UEA_datasets().load_dataset("TwoPatterns")

print(X_train.shape)  # (1000, 128, 1)
print(y_train.shape)  # (1000,)

ucr_uea_list_univariate_datasets = UCR_UEA_datasets().list_univariate_datasets()
print(len(ucr_uea_list_univariate_datasets))  # 0

ucr_uea_list_multivariate_datasets = UCR_UEA_datasets().list_multivariate_datasets()
print(len(ucr_uea_list_multivariate_datasets))  # 30

It seems that the univariate dataset list is empty, but it is still possible to import the datasets using their names.

The following web pages can be useful to look for the dataset names and features:

YannCabanes commented 5 months ago

Hello @LbzSteven and @nkchangshu, This issue has been solved in the PR https://github.com/tslearn-team/tslearn/pull/517. Thanks for reporting this problem!