Learn R Programming

auctestr (version 1.0.0)

Statistical Testing for AUC Data

Description

Performs statistical testing to compare predictive models based on multiple observations of the A' statistic (also known as Area Under the Receiver Operating Characteristic Curve, or AUC). Specifically, it implements a testing method based on the equivalence between the A' statistic and the Wilcoxon statistic. For more information, see Hanley and McNeil (1982) .

Copy Link

Version

Install

install.packages('auctestr')

Monthly Downloads

233

Version

1.0.0

License

MIT + file LICENSE

Maintainer

Last Published

November 13th, 2017

Functions in auctestr (1.0.0)

fbh_test

Apply z-test for difference between auc_1 and auc_2 using FBH method.
sample_experiment_data

Performance of several predictive models over three different datasets, using multiple cutoff points for time within each dataset.
se_auc

Compute standard error of AUC score, using its equivalence to the Wilcoxon statistic.
stouffer_z

Compute aggregate z-score using Stouffer's method.
auc_compare

Compare AUC values using the FBH method.
auctestr

auctestr: Statistical Testing for AUC data.