Learn R Programming

scoringfunctions (version 1.2)

Loss Functions for Assessing Point Forecasts

Description

Implements consistent loss functions and the corresponding identification functions for point forecasts, including those of the mean, the median, quantiles, expectiles, moments and Huber functionals (Gneiting T (2011) ; Fissler T, Ziegel JF (2016) ). Pointwise losses and realised average scores are both available. Detailed documentation of the functions' properties is included for facilitating the interpretation of results.

Copy Link

Version

Install

install.packages('scoringfunctions')

Monthly Downloads

157

Version

1.2

License

GPL-3

Maintainer

Hristos Tyralis

Last Published

August 30th, 2026

Functions in scoringfunctions (1.2)

Bregman scoring function (type 4, Patton scoring function)

Expectile identification function

Realised expectile score

Asymmetric piecewise quadratic scoring function (expectile scoring function, expectile loss function)

capping_function

Capping function

Error - spread scoring function

Realised generalized piecewise linear power score (type 1)

Generalized piecewise linear power scoring function (type 1)

Generalized Huber scoring function

Realised generalized Huber score

Interval scoring function (Winkler scoring function)

Realised Huber score

LINEX scoring function

Huber mean identification function

Realised \(L_q\)-mean score

Huber scoring function

Realised LINEX score

huberquantile_if

Huber quantile identification function

Realised generalized piecewise linear power score (type 2)

Generalized piecewise linear power scoring function (type 2)

\(L_q\)-quantile scoring function

Mean absolute percentage error (MAPE)

Realised MAE-LOG score

Realised MAE-SD score

MAE-LOG scoring function

Mean identification function

MAE-SD scoring function

\(L_q\)-mean scoring function

Mean absolute error (MAE)

Realised \(L_q\)-quantile score

Mean squared relative error (MSRE)

Mean squared percentage error (MSPE)

Exp-transformed identification function

Power-transformed identification function

Mean squared error (MSE)

\(n\)-th moment identification function

Mean relative error (MRE)

Mean - variance scoring function

Mean - variance identification function

Log-transformed identification function

Nash-Sutcliffe efficiency (NSE)

powerweighted_sf

Power-weighted squared error scoring function

Realised observation-weighted score

Realised \(n\)-th moment score

\(n\)-th moment scoring function

Quantile identification function

Observation-weighted scoring function

powerweighted_if

Power-weighted identification function

Sample quantile level function

Squared error scoring function

Realised quantile score

Squared error exp scoring function

Asymmetric piecewise linear scoring function (quantile scoring function, quantile loss function)

Realised squared error exp score

Squared error log scoring function

Relative error scoring function (MAE-PROP scoring function)

Realised squared error of power transformations score

scoringfunctions-package

Overview of the functions in the scoringfunctions package

Realised squared error log score

Squared percentage error scoring function

Realised squared error of squares score

Squared relative error scoring function

Squared error of power transformations scoring function

Squared error of squares scoring function

Bregman scoring function (type 3, QLIKE scoring function)

Absolute error scoring function

Realised \(\beta\)-median score

Bregman scoring function (type 2, Patton scoring function)

Realised Bregman score (type 1)

Realised Bregman score (type 4, Patton score)

Bregman scoring function (type 1)

Realised Bregman score (type 2, Patton score)

\(\beta\)-median scoring function

Absolute percentage error scoring function