myroc: Compute ROC and other performance measures for binomial model
In relgam: Reluctant Generalized Additive Models

Given a vector of true outcomes and a vector of predictions, returns a list containing performance measures.

1	myroc(ytest, rit, N = 100)

`ytest`	True test outcome: vector of 0s and 1s.
`rit`	Predictions for the true outcome. Should be vector of continuous variables between 0 and 1.
`N`	Number of breakpoints where we evaluate the performance measures. Default is 100.

We currently evaluate the performance measures at 100 quantiles of the predicted values; this can be adjusted via the N option.

A list of performance measures and intermediate computations.

`sens`	Vector of sensitivity values.
`spec`	Vector of specificity values.
`ppv`	Vector of PPV values.
`npv`	Vector of NPV values
`area`	Area under ROC curve (AUC).
`se`	Standard error for AUC.
`cutp`	Cut points at which the performance measures were computed.
`cutp.max`	Cut point which maximizes (sens + spec) / 2.