Displaying 1 result from an estimated 1 matches for "fitm1".
Did you mean:
fit1
2006 Sep 19
0
How to interpret these results from a simple gamma-frailty model
...uced frailty v = ",var(v)))
id=rep(seq(1:(n/2)),2); # Frailty id's,
c <- rep(1,n); # censoring flags
r <- runif(n,0,1); # For random variate generation
t <- (-1*log(r)) / (lambda*v*exp(age*beta[1]+tc*beta[2]+hdlc*beta[3]+sbp*beta[4]+diab*beta[5]+smok*beta[6]))
fitM1=coxph(Surv(t,c)~age+tc+hdlc+sbp+diab+smok); # Model 1, no frailty
cf[i,1,] <- round(fitM1$coef,6) # Model 1 coefficients
fitM2=coxph(Surv(t,c)~age+tc+hdlc+sbp+diab+smok+
frailty(id,dist="gamma",sparse=T,method="em")) # Model 2, frailty
cf[i,2,] <- round(fitM2$coe...