Hello!
Im having a bit of a problem creating martingale-residuals for my aalen additive model. Is it even possible in R to do that? My code for the additive model is the following:
ruff <-addreg(Surv(E$stop,E$delta) ~E$Zh +E$Zr +E$PSA +E$Treatment)
summary(ruff)plot(ruff)
 
If it had been a cox model i would have just written 
 
stage.ph <- coxph( Surv(E$stop,E$delta) ~E$Zh +E$Zr +E$Treatment,  method="breslow", na.action=na.exclude)
resid <- resid(stage.ph, type="martingale")str(resid)summary(resid)
plot(E$PSA, E$resid, xlab="stage",ylab="Martingale Residuals", ylim=c(-4, 1.0))lines(lowess(E$PSA[!is.na(E$PSA) & !is.na(E$resid)],       E$resid[!is.na(E$PSA) & !is.na(E$resid)]))
So, how do I do this for the additive model?
 
/Eric
_________________________________________________________________


	[[alternative HTML version deleted]]