I use the variance estimates obtained in a coxme
output for a further procedure:
model = coxme(Surv(time, status) ~ (0 + A | B ) + (1| B ) + strata(A) , data)
NUM = sum(unlist(model$vcoef))
However, I need the sum of the two variances estimates (NUM
in the example above) to be above 0.01.
Is there any way to add bounds/condition for the variance estimates in the coxme
function?
I thought I could do something in the control
argument by setting optpar = list(method = "L-BFGS-B", lower=0.005)
but I just got errors or strange results.
Thanks in advance for your help!