Re: [R] ROCR for combination of markers

Hi Eik or other who might help: I got this error: "Error in roc.formula(form = y1 ~ x + z, plot = "ROC") : Invalid formula: exactly 1 predictor is required in a formula of type response~predictor. " when I ran "out=ROC( form = y1 ~ x + z, plot="ROC") " from your code. How to fix it? Thanks. -

Re: [R] ROCR for combination of markers

Hi R users, I am sort of new on those. In the following logistic regression function "tigol( 0.3 + 3*x + 5*z + 7*w ) ", why the intercept was set 0.3, and the coefficients were set 3 ,5 and 7? Thanks. -- View this message in context: http://r.789695.n4.nabble.com/ROCR-for-combination-of-marke