Authors: Ivan Svetunkov
Published at: Open Forecast
Abstract: There are many forecasting related packages in R with varied popularity, the most famous of all being forecast, which implements several important forecasting approaches, such as ARIMA, ETS, TBATS and others. However, the main issue with the existing functionality is the lack of flexibility for research purposes, when it comes to modifying the implemented models. The R package smooth introduces a new approach to univariate forecasting, implementing ETS and ARIMA models in Single Source of Error (SSOE) state space form and implementing an advanced functionality for experiments and time series analysis. It builds upon the SSOE model and extends it by including explanatory variables, multiple frequencies, and introducing advanced forecasting instruments. In this paper, we explain the philosophy behind the package and show how the main functions work.
How to cite: Svetunkov (2023). Smooth forecasting with the smooth package in R. OpenForecast.org
The story of the paper: This paper was rejected from the Journal of Statistical Software by a reviewer maintaining the package competing with the
smooth. Given that the paper was written specifically for that journal, and I have nowhere else to submit it, I’ve decided to upload it online and make it freely available.