regMetrics: Calculate regression metrics

In mrfoliveira/Evaluation-procedures-for-forecasting-with-spatio-temporal-data: Evaluation procedures for forecasting with spatio-temporal data

Description

Calculate MAE, RMSE and utility-based regression evaluation metrics

Usage

regMetrics(trues, preds, y_train = NULL, norm = FALSE,
  aeStatFUN = stats::median, seStatFUN = mean, util = FALSE,
  util.parms = NULL)

Arguments

trues	a vector of true values
preds	a vector of predicted values
y_train	a vector of training values
norm	a Boolean indicating whether to calculate normalized regression metrics
aeStatFUN	a function to calculate a summary of y_train for absolute error normalization. Default is median
seStatFUN	a function to calculate a summary of y_train for squared error normalization. Default is mean
util	a Boolean indicating whether to calculate utility-based regression metrics
util.parms	a named list of parameters to use for calculating utility-based regression metrics. Should contain slots phi.parms - the result of function phi.control phi.control - if phi.parms is undefined, then phi.control can be provided with a list of named arguments to feed function phi.control using y_train. Default is list(method = "extremes", extr.type="high") loss.parms - the results of function uba::loss.control p - Default is 0.5 thr - Relevance threshold. Default is 1 beta - Beta for F-measure. Default is 1

Value

a named vector of calculated metrics

mrfoliveira/Evaluation-procedures-for-forecasting-with-spatio-temporal-data documentation built on April 11, 2021, 10:50 a.m.