News
Version Updates
0.3.0
- CRAN release.
- Add Kundel dataset.
- Check for and report partially paired factorial designs.
- Converted FORTRAN code for conventional and likelihood-ratio
binormal ROC curves to R.
- Remove deprecated argument
method from
mrmc(), srmc(), and stmc().
0.2.1
- Order tests and readers lexically in results from
mrmc().
- Rename
bichisquare column to bichisquared
in binormalLR_curve.
0.2.0
- Implement fast algorithm for unbiased covariance estimation.
- Remove deprecated functions
proproc_roc(),
proproc_sens(), and proproc_spec().
- Rename argument
method to cov in
mrmc(), srmc(), and stmc().
0.1.16
- Add functions
OR_to_RMH() and RMH_to_OR()
for conversion from Obuchowski-Rockette to Roe, Metz & Hillis model
parameters and vice versa.
0.1.15
- Implement ROC expected utility metrics
binormal_eu(),
binormalLR_eu(), and empirical_eu().
0.1.14
- Add progress bar for unbiased covariance calculation.
- Truncate confidence intervals only for
sens,
spec, and auc metrics.
- Fix direction of ratings in
pROC calculation of
empirical ROC AUC to allow values < 0.5.
0.1.13
- Add
mrmc() support for single readers.
- Rename
proproc functions to
binormalLR.
- Add
parameters() method function for mrmc
output.
- Add
roc_curves() method function for srmc
output.
- Add
normalize argument to auc()
metrics.
- Redefine
srmc() function for single-reader (multi-test)
multi-case analysis.
- Add
stmc() function for single-reader (single-test)
multi-case analysis.
0.1.12
- Fix
proproc_sens() and proproc_spec() to
use proproc instead of empirical curves.
0.1.11
- Add support for readers nested within tests.
- Add
smmc() for single reader multiple case
analysis.
- Add
parameters() function to extract ROC curve
parameters from roc_curves() results.
- Fix partial AUC calculation when specificity ranges are given.
- Add plot option to overlay empirical ROC points on parametric
curves.
0.1.10
- Subset data in calls to
mrmc() by event/non-event cases
for binary_sens()/binary_spec() metrics.
- Add
coord_fixed argument to plot()
methods.
- Add
mean() method for binormal_curves to
average over their ROC curve parameters.
0.1.9
- Add
binary_sens() and binary_spec()
metrics.
0.1.8
- Support sensitivity values in
points() methods.
- Implement conventional binormal ROC curves and metrics.
- Add package citation.
0.1.7
- Add
roc_curves() function for construction of ROC
curves and method functions points() for calculating points
on the curves, mean() for averaging multiple curves, and
plot() for displaying them graphically.
- Deprecate
roc() and proproc().
0.1.6
- Fix unbiased covariance calculation.
0.1.5
0.1.4
0.1.3
- Add metric functions for sensitivity, specificity, and partial
AUC.
0.1.2