Dear R users,
pROC is a package to compare, visualize, and smooth receiver operating
characteristic (ROC) curves.
The package provides the following features:
* Partial or full area under the curve (AUC) computation
* Comparison of two ROC curves (curves and AUC)
* Calculating and plotting confidence intervals
* Smoothing of the ROC curve
* Coordinates extraction ('coords' function).
The main feature of pROC is the comparison between two ROC curves. Three
methods are currently implemented for both paired and unpaired ROC curves:
* Bootstrap for full and partial AUC and smoothed ROC curves
* DeLong [1] method for full AUC
* Venkatraman [2, 3].
Confidence intervals can be computed with bootstrap for both empirical
or smoothed ROC curves:
* partial or full AUC (also with DeLong [1] method for full AUC)
* ROC coordinates (thresholds, sensitivity or specificity).
You can find more information in our paper [4] and on pROC website:
http://www.expasy.org/tools/pROC/
Hope you'll find it useful!
Xavier Robin
--
References:
[1] DeLong ER, DeLong DM, Clarke-Pearson DL (1988) Comparing the areas
under two or more correlated receiver operating characteristic curves: a
nonparametric approach. Biometrics 44, 837?845.
[2] Venkatraman ES,Begg CB (1996) A distribution-free procedure for
comparing receiver operating characteristic curves from a paired
experiment. Biometrika 83, 835?848.
[3] Venkatraman ES (2000) A Permutation Test to Compare Receiver
Operating Characteristic Curves. Biometrics 56, 1134?1138.
[4] Robin X, Turck N, Hainard A, et al. (2011). pROC: an open-source
package for R and S+ to analyze and compare ROC curves. BMC
Bioinformatics, 12, 77. http://dx.doi.org/10.1186/1471-2105-12-77
_______________________________________________
R-packages mailing list
R-packages at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-packages
