I am currently trying to build a predictive model based on logistic regression. I made ROC curve and calibration plot for the original model with val.prob.ci() function from rms package, then I validated the model with validate(fit, method="bootstrap", B=1000) from rms package.
I have two questions.
1) Is it correct to get the validated coefficients of independent variables by multiplying the "original" ones by calibration slope?
2) By the function validate I got the validated parameters of the model (c-index, calibration slope, etc). But how to draw ROC curve and calibration plot of this validated model? I would like to use the function val.prob.ci() again, because it makes a more complete calibration plot (i.e. with deciles).