roc_sqe: Squared error for fit of empirical by parametric ROC curve
In PeterVogel1991/betaROC:

Squared error for fit of empirical by parametric ROC curve

1	roc_sqe(pars, empROC, MDE_info)

`pars`	Parameter vector corresponding to `MDE_info$method`
`empROC`	`data.frame` with true positive (TPR) and false positive (FPR) values of the empirical ROC curve
`MDE_info`	`list` containing information about the required minimum distance estimation (MDE) fit

Computes the sum of the squared errors between an empirical ROC curve and a parametric one specified by pars and MDE_info, evaluated at the FPR values of the empirical ROC curve. In case a concave binormal ROC curve is specified, the parameters have to be adjusted in that sigma = 1 is inserted into the pars vector.

PeterVogel1991/betaROC documentation built on May 14, 2019, 4:01 a.m.