histUev: Diagnostic distributions plot for a 'trainOcc' object and...

In benmack/oneClass: One-class classification in the absence of test data

Description

The histogram of the predicted unlabeled data is shown together with the hold-out predictions of the positive and unlabeled traning data (boxplots). The threshold dependent accuracy (user's/producer's accuracy and kappa) is superimposed.

Usage

histUev(model, ev, yLimits = NULL, ...)

Arguments

model	an object of class trainOcc.
ev	an object of class ModelEvaluation.
yLimits	specify the minimum and maximum points in the density scale of to which the accuracy scale should be projected. This is useful if you add the plot to another one, e.g. a histogram plot (hist.trainOcc), whose y-axis scale is not suitable for the accuracy plot (i.e. if it is not between 0 and 1).
...	arguments that can be passed to hist.trainOcc, e.g. pred

Details

histUev

Value

yLimits (invisible) and the diagnostic plot with threshold depenent accuracy.

benmack/oneClass documentation built on Dec. 15, 2020, 7:38 p.m.