Hello, I am curious about this "unique_id" column from the AirPassengersDF. All entries in the unique_id column are 1.0. Isn't it supposed to be not the same?

The unique_id column is the id of the specific time series. If you have multiple time series, this column is used to distinguish them. Let's say you have two time series in a dataframe, all datapoints from the first time series will have unique_id 1 and all entries from the second time series will have unique_id 2. See it as a way to be able to say 'this datapoint belongs to time series 1 and this datapoint belongs to time series 2'. The AirPassengersDF only has 1 time serie in it, meaning all entries have unique_id 1!

Thank you!

@Didier Merk So is the unique_id column a way to make predictions on multiple y series ? Instead of building 1 model per y series, this unique_id helps make the pipelining process easier?

@Arun Bharadwaj Exactly!