Thinking of Version 2.0 Ideas

[rickeycarter]

Some working ideas for a "2.0" version of the package (or a CRAN 1.0 package):

--Write to be compatible with CRAN?

--Add tests and plotting capabilities for >1 ROC curve. Include an inset that gives a matrix of results including the AUCs with 95% CIs, Differences in AUCs between discriminators, Tests for differences in AUCs

--Adjust the dx_roc to change the way the summary statistics are shown on the figure. We introduce LRT+ and LRT- sometime back and this moved the OR and F1 statistic off the page

--Add some prevalence sensitivity analysis capabilities: plots of NPV & PPV over a range of prevalence values along with tabulation