[R] Interpreting the example given by Prof Frank Harrell in {Design} validate.cph

[Frank Harrell]

Don't worry about the sign.  When predicting relative log hazard, high hazard
means short survival time so Dxy is negative.  When predicting survival
probability (u specified), high prob. means long survival time so Dxy is
positive.  You can just reverse the sign when u is not specified.

I did not see Dxy of -.54 or .6 anywhere on the output you sent.

Specify time.inc of 30 to cph in the 2nd case.  As you did not provide code
to simulate test data, please run the second model yourself without strat
but with surv=TRUE time.inc=30 (and u=30 to validate) to see if you get the
same Dxy as for the first model, before bringing strat into the picture.

rms and Design prefer for you to store the predictors as factor()s in the
dataset instead of specifying this in the model formula.

Frank

-----
Frank Harrell
Department of Biostatistics, Vanderbilt University
-- 
View this message in context: http://r.789695.n4.nabble.com/Interpreting-the-example-given-by-Prof-Frank-Harrell-in-Design-validate-cph-tp3316820p3318189.html
Sent from the R help mailing list archive at Nabble.com.