zhouhaoyi / etdataset Goto Github PK

3q very much for your sharing!
may I ask why don't load all the large data sets(ETT-large / ETT-full)?
Is there any problem with the company's permission to use them and not to disclose them?

Could you please elaborate on what are those features in the dataset, and how they were measured?
Are they High/Mean/Low values of active/reactive power delivered to the transformer load over 15 minute period?

Hello, congratulations on your nice work! It is possible to have access to the full and large datasets?

Hello,
thank you very much for your great work on this dataset.
Are there any plans to release the ETT-large and ETT-full datasets (mentioned in the readme) too, and if so, is there already an estimated date for that?

EDIT: sorry for the repost (just saw the already closed issues) - is there nevertheless maybe any update on this? You also mentioned that there is additional work to make the datasets available - I might be able to help with preparation if useful/possible.

BR Andreas

Why every 31th have same value?

Hi! May I ask the results presented in the Informer@AAAI2021 paper are based on which dataset (small, large or full)? Thanks!

大佬，请问一下如何获取large和full这个数据集？

Hello Authors,

Really nice piece of work!! I wonder if some more information about the dataset can be made available, e.g.

• the dataset is for which counties in China?
• more details about the sensors transmitting the time series value at each timestamp, manufacturer (even if it is in anonyms manner it will help, like encoded to integers etc.), sensor model number etc.

看了数据好像是每15分钟记录一次。

仅使用了small数据集进行训练，未找到full

when I run the code the TypeError: Invalid type for initial value: <class 'tensorflow.python.framework.ops.Tensor'> (expected Python scalar, list or tuple of values, or numpy.ndarray). show up, is this about the version of tensorflow and keras?