I used the R package hdnom to plot a nomogram obtained from survival data, after applying the lasso method for model selection. I am trying to compare using the nomogram and the coefficients of the fitted Cox model, but I cannot figure out why the zeroes of the linear predictor and total point scale do not line up. Can anyone help explain why this is the case?

More specifically, I have the following scale:

Points per unit of linear predictor: 25.79354 
Linear predictor units per point   : 0.03876939

However, the zero of the linear predictor scale lines up with about 40 points on the total point scale. Is this due to rescaling of variables, or did I miss something else?

  • $\begingroup$ Can you tell us more about the data and the problem? Are you really looking for an error in the code or the interpretation of the code? If it is only about about the R package it is off topic for this site and you may want to migrate StackOverflow. $\endgroup$ Jan 14, 2017 at 0:19
  • $\begingroup$ The data is about survival of patients with a certain type of cancer. I think it's partly about the package, but also about the interpretation/knowledge of building a nomogram. It looks like some translation is done between the total points and linear predictor, but I cannot find out which one and why. $\endgroup$
    – Sanderr
    Jan 14, 2017 at 10:23


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.