"p" is "predicted" ... View more

proc glimmix data=simulation() method=laplace; class strata; effect myspline=spline(t/knotmethod=list(1,3,5,7,9) NATURALCUBIC); model y(event="1")= intercept*myspline gender*t gender*myspline/dist=binary link=logit noint ; random strata; store mystore; run; *generate with values for t - this will be used drawing the spline; data template; intercept=0; do gender=1; do t=0 to 10 by 0.01; output; end; end; run; *calculate the predicated values for each value of t; proc plm restore=mystore; score data=template out=plot predicted=predicted lclm=lclm uclm=uclm; run; data plot; set plot; by t; *transform to oddsratio scale; oddsratio=exp(predicted); upper=exp(uclm); lower=exp(lclm); 'p' means 'predicted' but how can I get odds ratio with more than one covariates ... View more

ep=exp(p); elower=exp(lower); eupper=exp(upper); Does ep output "odds" or " odds ratio"? ... View more

Hello! How can I get odds ratio instead of odds by "effect"? Thank you! ... View more