prtEvalPfAtPd   Returns the probability of false alarm at a desired
  probability of detection on the receiver operating curve.
 
    pf = prtEvalPfAtPd(prtClassifier, prtDataSet,PD) returns the probabilty
    of false alarm pf.  prtDataSet must be a labeled, binary
    prtDataSetStandard object. prtClassifier must be a prtClass object. PD
    is the desired probability of detection and must be between 0 and 1.
 
    pf = prtEvalPfAtPd(prtClassifier, prtDataSet,PD, nFolds) returns the
    probabilty of false alarm pf on the receiver operating curve with
    K-fold cross-validation.  prtDataSet must be a labeled, binary
    prtDataSetStandard object. prtClassifier must be a prtClass object. PD
    is the desired probability of detection and must be between 0 and 1.
    nFolds is the number of folds in the K-fold cross-validation.
 
    pf = prtEvalPfAtPd(prtClassifier, prtDataSet, PD, xValInds) same as
    above, but use crossValidation with specified indices instead of random
    folds.
 
    Example:
    dataSet = prtDataGenSpiral;
    classifier = prtClassDlrt;
    pf =  prtEvalPfAtPd(classifier, dataSet,.9)