Assuming I have a list of multiple time-series patterns, how would I go about forecasting the most likely next pattern using Stompy?
