Feature Request: TimeSeriesDataset retrieve column/series names

[LuisPerezVazquez]

• PyTorch-Forecasting version: 0.10.1 PyTorch version: 1.11.0+cu113 Python version: 3.7.13 Operating System:

Expected behavior

Be able to easily retrieve time series name/ group_id after prediction

Actual behavior

Right now it is not clear or easy to retrieve column names from torch tensor after transforming data into pytorch forecasting TimeSeriesDataset. This is crucial to be able to use this library and interpret the results in a more straightforward way.

Code to reproduce the problem

Any model.predict(val_dataloader) will give a bunch of columns (one per time series) with their prediction, but it is not trivial to retrieve which column is which times series (group_id)

[DominikPKaiser]

Hi Luis. I just use the built-in filter function to make specific predictions.

raw_prediction, x = model.predict(validation.filter(lambda x: (x.COMPANY_ID == "APPLE")), mode="raw", return_x=True, )