Learn R Programming

wevid (version 0.6.2)

Quantifying Performance of a Binary Classifier Through Weight of Evidence

Description

The distributions of the weight of evidence (log Bayes factor) favouring case over noncase status in a test dataset (or test folds generated by cross-validation) can be used to quantify the performance of a diagnostic test (McKeigue (2019), ). The package can be used with any test dataset on which you have observed case-control status and have computed prior and posterior probabilities of case status using a model learned on a training dataset. To quantify how the predictor will behave as a risk stratifier, the quantiles of the distributions of weight of evidence in cases and controls can be calculated and plotted.

Copy Link

Version

Install

install.packages('wevid')

Monthly Downloads

148

Version

0.6.2

License

GPL-3

Maintainer

Marco Colombo

Last Published

September 12th, 2019

Functions in wevid (0.6.2)

wevid.datasets

Example datasets
prop.belowthreshold

Proportions of cases and controls below a threshold of weight of evidence
Wdensities

Compute densities of weights of evidence in cases and controls
Wdensities.crude

Calculate the crude smoothed densities of W in cases and in controls
wevid-package

Quantifying performance of a diagnostic test using the sampling distribution of the weight of evidence favouring case over noncase status
weightsofevidence

Calculate weights of evidence in natural log units
summary-densities

Summary evaluation of predictive performance
recalibrate.p

Recalibrate posterior probabilities
plotWdists

Plot the distribution of the weight of evidence in cases and in controls
plotroc

Plot crude and model-based ROC curves
plotcumfreqs

Plot the cumulative frequency distributions in cases and in controls