search

scheike / timereg

timereg R package, survival analysis methods
30 stars 5 forks source link

Warning message: In Event(futime, cause) : exit times must be >0 #8

Open acmilannesta opened 5 years ago

acmilannesta commented 5 years ago

It seems there is a convergence problem in comp.risk 

cr.out = comp.risk(Event(futime, cause)~const(proc_type), data =cand2, cause = 1)
summary(cr.out)

After run the above, I got following

Warning message:
In Event(futime, cause) : exit times must be >0
Competing risks Model 

No test for non-parametric terms
Error in if (sum(abs(object$score) > 1e-06)) cat("Did not converge, allow more iterations\n\n") : 
  argument is not interpretable as logical

Even if I increase 'Nit' and decrease 'conv', the problem still exists.

scheike commented 5 years ago

Here there is not very much to go on. This must be a convergence issue for the simple NR algorithm, and luckily there are other ways of fitting the FG model.

On Thu, 29 Aug 2019, acmilannesta wrote:

It seems there is a convergence problem in comp.risk

cr.out = comp.risk(Event(futime, cause)~const(proc_type), data =cand2, cause = 1)
summary(cr.out)

After run the above, I got following

Warning message:
In Event(futime, cause) : exit times must be >0
Competing risks Model

No test for non-parametric terms
Error in if (sum(abs(object$score) > 1e-06)) cat("Did not converge, allow more iterations\n\n") :
 argument is not interpretable as logical

Even if I increase 'Nit' and decrease 'conv', the problem still exists.

-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/scheike/timereg/issues/8