rocsurf.trin.Rd
Function for computation of the trinormal ROC surface.
rocsurf.trin(x, y, z, p = 300, plot = TRUE, saveVUS = FALSE)
Vectors containing the data of the three classes "healthy", "intermediate" and "diseased".
An integer for the precision of the surface. p
gives the
number of gridpoints per axis.
logical. If TRUE (default), the VUS is plotted using surface3d
from the package rgl
.
A logical whether to save a PNG of the VUS in your current
working directory (default is FALSE
).
A list with the following components:
The vectors \(t_-=F_-^{(c_-)}\) and \(t_+=F_+^{(c_+)}\)
The matix containing the surface values.
The original data.
This function takes three-class ROC data and computes the three
dimentional surface using the R-package rgl
. The ROC surface is
defined as
$$z = ROCs(t_-,t_+) = F_0(c_+) - F_0(c_-)=F_0(G_+^{-1}(t_+) ) -F_0(F_-^{-1}(t_-) ),$$
where \(c_-, c_+\) are the two cut-off points and \(F_-, F_0, F_+\) the cdf of the three classes with \(G = 1-F\).
Xiong, C., G. Van Belle, et al. (2006). Measuring and estimating diagnostic accuracy when there are three ordinal diagnostic groups. Statistics in Medicine 25(7), 1251–1273.