Internal ncvreg functions
convexMin(b, X, penalty,gamma, l2, family, penalty.factor, a, Delta)
setupLambda(X, y, family, alpha, lambda.min, nlambda, penalty.factor)
loss.ncvreg(y,yhat,family)These are not intended for use by users. convexMin
calculates the lowest index for which the penalized objective function
is locally convex. setupLambda creates an appropriate vector
of regularization parameter values. loss.ncvreg calculates the
value of the loss function for the given predictions (used for
cross-validation).