Using exogenous variables with crossvalidation in mlforecast

[mkrech]

Hi,
thank you for your fantastic library.

My question:
Is there example or a way to combine exogenous variables eg fourier (future_df, transformed_df) to combine with crossvalidation.

I found backtestsplits in utilsforecast. I could be done "by hand" :). Is there another way?

regards
Michael

[jmoralez]

Hey @mkrech, thanks for using mlforecast. You can merge the transformed_df onto your dataframe and set static_features to a list with your static features (or[] if you don't have any) and mlforecast will automatically assign the future exogenous features during cross validation.