Dear Professor,
Sorry if I am being dim, I re-did the external validation part as far as I understand:
###calibration for external validation
fitlognorm<-psm(Surv(nonsurgical.med$TimeToGe, nonsurgical.med$Generalization)~ MaleGender + AChRValue + AntiMuSKAb + OAID + AgeAtOnset+ SteroidsUse, data = nonsurgical.med, dist = “lognormal”)
g<-update(fitlognorm, x=TRUE, y=TRUE)
estimates=survest(g,newdata=zzunonsurgical,times=36)$surv
w1<-val.surv(g,newdata=zzunonsurgical, S=Surv(zzunonsurgical$TimeToGe, zzunonsurgical$Generalization), est.surv = estimates, u=36)
print(w1)
plot(w1)
Output of print(w1):
Output of plot(w1)
Then I tried to get the c-index:
###Determine concordance
surv.obj=with(zzunonsurgical,Surv(TimeToGe,Generalization))
rcorr.cens(x=estimates,S=surv.obj)
The output reads as below:
If my validation is failed, I will start over again. Thanks a lot for your time, patience and kind help. 