The combination of survreg + gamma frailty = invalid model, i.e., the example
that you quote.
I did not realize that this had been added to the survreg help file until
very
recently. I will try to fix the oversight. Other, more detailed documentation
states that Gaussian frailty + AIC is the only valid random effects choice for
survreg.
Details: frailty(x) with no optional arguments defaults to "mle" as the
method
of choosing the variance of the random effect, via a calculation that is
specific to the Cox model.
Terry Therneau
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
