antoinecarme / pyaf

PyAF is an Open Source Python library for Automatic Time Series Forecasting built on top of popular pydata modules.
BSD 3-Clause "New" or "Revised" License
459 stars 72 forks source link

Support for External Probabilistic forecasts Reconciliation #207

Closed Manjubn777 closed 1 year ago

Manjubn777 commented 2 years ago

Hi @antoinecarme , I would like to know if the package supports reconciliation of forecasts after generating at different hierarchy and also if it supports external probabilistic forecast reconciliation. I appreciate your quick response to my inquiry for information.

antoinecarme commented 2 years ago

Hi @Manjubn777,

Thanks for using PyAF.

PyAF supports some forecasting reconciliation methods : Bottom-up, Top-Down, Middle-out and Optimal combinations methods. These methods are described in depth in this excellent book :

https://otexts.com/fpp2/hierarchical.html

PyAF generates additional columns in the forecast output dataframe, one for each reconciliation method and at each node of the hierarchy.

The last paragraph of the following jupyter notebook gives an example of using these methods :

https://github.com/antoinecarme/pyaf/blob/master/docs/PyAF_Hierarchical_FrenchWineExportation.ipynb

Hope this helps.

antoinecarme commented 1 year ago

Closing. 3 months with no activity. Reopen if needed.