global_normalization vs global_time_normalization

[Alan-Wolf-VN]

Can anyone explain what are global_normalization and global_time_normalization?

I try to build a global local model which forecasts many time series that share "some behaviour". However, the default option: global_normalization = False leads to the following error when I run m.fit:

ValueError: Dataset name 'df' missing from training data params. Set unknown_data_normalization to use global (average) normalization parameters

How to avoid this error? Why setting global_normalization = True does not lead to this error?

[LeonieFreisinger]

@Alan-Wolf-VN
Welcome to NP!
Can you please share all the model parameters you selected? Did you set bot parameters global_normalization and global_time_normalization to False?