Hierarchical Forecasting in Python | Nixtla

[@Darokbr]

I am following this package. Very interesting.

[@tgnbeats]

Amazing Work! Well done to the Nixtla team

[@TheKinsey06]

I face these exact issues at work. I needed this video

[@Leibniz_28]

Great initiative

[@IvanRubnenkov]

Can you use hierarchical forecast methods (minTrace, ERM) for multiplicative dependencies between time series, not the additive ones? Thanks in advance

[@FindMultiBagger]

Hirachical forecast is really slow , but hats of statsforecast developers and contributers for ecosystem 🎉 😇♥️

[@gauravsukhadia638]

Hypothetically if I have let's say 4 million skus what do you think how long will it take ?

[@FindMultiBagger]

@@gauravsukhadia638 For me their are 10 Million sku , for 1 million sku it took 3/4 hrs , but their are lots of issues like , historical validation and future forecast. but I able to fix those issue myself. suggestion use OLS method for faster execution MinT is slow

[@Alfalphe]

What about Exogenous variables?

[@STEEZEAI]

please if anyone knows how to use exog variables with this, explain how you did it

[@517127]

How acess the residual or rmse for each id?