FoReco and FoRecoML: A Unified Toolbox for Forecast Reconciliation in R

2026-04-30

AI Summary

This work proposes a unified R-based framework that addresses the lack of integrated support for reconciliation across cross-sectional, temporal, and cross-temporal dimensions in hierarchical time series forecasting. For the first time, it combines classical least squares, regression-based linear approaches, and machine learning–driven nonlinear reconciliation techniques within a single toolkit, comprehensively covering all three reconciliation paradigms. The framework maintains ease of use while offering highly customizable extension interfaces, thereby significantly enhancing both coherence and accuracy in multi-level time series forecasts. Its design ensures strong practical applicability alongside substantial research value, making it a versatile solution for both methodological development and real-world forecasting applications.

Abstract

Forecast reconciliation has become key to improving the accuracy and coherence of forecasts for linearly constrained multiple time series, such as hierarchical and grouped series. Yet, comprehensive software that jointly covers cross-sectional, temporal, and cross-temporal reconciliation has so far been lacking. The R packages FoReco and FoRecoML address this gap by offering a comprehensive and unified framework. The packages respectively implement classical and regression-based linear reconciliation approaches, and non-linear approaches based on machine learning for cross-sectional, temporal and cross-temporal frameworks. Designed for accessibility and flexibility, these packages provide sensible default options that allow new users to apply reconciliation methods with minimal effort, while still giving expert users full control to explore state-of-the-art extensions through customized settings. With this dual focus, FoReco and FoRecoML are versatile tools for practitioners and researchers working on forecast reconciliation.

Problem

• forecast reconciliation
• hierarchical time series
• grouped time series
• cross-sectional reconciliation
• cross-temporal reconciliation

Innovation

• forecast reconciliation
• machine learning
• hierarchical time series
• cross-temporal reconciliation
• R package