ProbReco: Score Optimal Probabilistic Forecast Reconciliation

Training of reconciliation weights for probabilistic forecasts to optimise total energy (or variogram) score using Stochastic Gradient Descent with automatically differentiated gradients. See Panagiotelis, Gamakumara, Athanasopoulos and Hyndman, (2020) <> for a description of the methods.

Depends: R (≥ 3.5.0)
Imports: Rcpp (≥ 1.0.2), purrr (≥ 0.3.2), mvtnorm, Rdpack
LinkingTo: Rcpp, RcppEigen, StanHeaders (≥ 2.19.1), BH
Suggests: knitr, rmarkdown, fable, dplyr, tidyr, magrittr, stringi
Published: 2020-09-24
Author: Anastasios Panagiotelis ORCID iD [aut, cre]
Maintainer: Anastasios Panagiotelis <anastasios.panagiotelis at>
License: GPL-3
NeedsCompilation: yes
Materials: README
In views: TimeSeries
CRAN checks: ProbReco results


Reference manual: ProbReco.pdf
Vignettes: ProbReco-with-fable
Package source: ProbReco_0.1.0.1.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
macOS binaries: r-release: ProbReco_0.1.0.1.tgz, r-oldrel: ProbReco_0.1.0.1.tgz
Old sources: ProbReco archive


Please use the canonical form to link to this page.