ProbCast

ProbCast is an R Package under continuous development by researchers at the University of Strathclyde. It is a collection of functions for probabilistic forecasting (mainly wrappers for qunatile and semi-parametric regression model fitting functions), cross-validation, evaluation and visualisation. Central to ProbCast is the data class MultiQR, for storing the results of multiple quantile regression, and methods for working with MultiQR objects.

Set-up

You can install the latest version of ProbCast using:

# install.packages("devtools")
library(devtools)
install_github("jbrowell/ProbCast")

The latest release is:

It may be necessary to set the following:

Sys.setenv(R_REMOTES_NO_ERRORS_FROM_WARNINGS="true")

Usage

The package includes a script Example.R which demonstrates much of ProbCast's functionallity.

Guide for Contributors

Contributors should follow the following guidelines:

• Open new branches when adding new functinoality, or making changes to exisig functions
• Raise issues when adding functionality, and identifying and fixing bugs
• Use rogygen to include documentation with each new function. Include helpful notes on all inputs, outputs, and wokrings of the function
• Include helpful comments throughout code
• Add yourself as @author to functions that you have "ownership" of
• Invite others to review pull requests (check for documentation, back compatability, confilcts...)

Acknowledgements

Thanks to everyone who has contributed: Jethro Browell, Ciaran Gilbert, Gordon McFadzean, Rosemary Tawn.

Development of ProbCast has been supported by the following grants and organisations:

• EPSRC Innovation Fellowship "System-winde probabilistic energy forecasting" (EP/R023484/1 and EP/R023484/2, 2018-2022)
• EPSRC Wind and Marine Energy Systems CDT (EP/L016680/1)
• Network Innovation Allowance project "Control REACT" (NIA_NGSO0032)
• University of Strathclyde
• University of Glasgow
• TNEI Services Ltd

References

ProbCast was introduced in the following paper, and has sinced been used in multiple academic studies and to facilitate training in probabilistic forecasting for researchers and practitioners.

• J. Browell and C. Gilbert, (2020), "ProbCast: Open-source production, evaluation and visualisation of probabilistic forecasts", DOI: 10.1109/PMAPS47429.2020.9183441
• Citing articles
• Archived versions on Zenodo