calc.error: Calculate bias, RMSE, corellation etc of two vectors.

In ptompalski/UsefulRFunctions: A personal set of R tools.

Description

Calculates: corelation coeficient, bias (absolute and relative), RMSE (absolute and relative), and performes statistical test to check if the differences between reference and estimate are significant. Correlation coefficient as well as statistical test depend on the dist.normal value. Optionally statistics can be calculated for groups of data, specified with the by parameter.

Usage

calc.error(reference, estimate, by = NULL, noinfo = TRUE,
  dist.normal = TRUE, short = FALSE)

Arguments

reference	a vector of reference values
estimate	a vector of estimated values
by	Optional grouping variable
noinfo	Logical. Should the additional information on the calculations be displayed?
dist.normal	Logical. Is the distribution normal? If TRUE t-test and Pearson correlation is calculated. If FALSE, Wilcoxon paired test and Spearman corelation is calculated.
short	Logical. Calculate only a subset of summary statistics.

Value

summary statistics of the differences between reference and estimate.

ptompalski/UsefulRFunctions documentation built on May 26, 2019, 11:32 a.m.