[R] Error in using coxph()

Prof Brian Ripley ripley at stats.ox.ac.uk
Sun May 16 11:52:19 CEST 2004


Note you said you used 1-F but the output said you used F. You also say
`1-F is a vector of '0's' so I think you may have declared that all
observations are right-censored.  (NB if you give Surv an event vector of
all 1's it is ambiguous, so don't do this.)

If you did declare all observations as censored, then it seems the coxph
algorithm does not cope (although I did not try to run it for 1e10
iterations).  That's really an undetected user error, as there is no 
information in the partial likelihood about the coefficient (and no 
recorded events).

On Sun, 16 May 2004, Siew Leng TENG wrote:

> I am getting errors of the following kind. I can't
> seem to point the source of the error. I would greatly
> appreciate any advice.

Please read the posting guide and try to avoid us having to guess what you 
actually did.

> Error message :
> ----------------
> "Ran out of iterations and did not converge in:
> fitter(X, Y, strats, offset, init, control, weights =
> weights,..."
> 
> Details :
> ---------
> E is a vector of survival times (or censored times),
> 1-F is a vector of '0's, and cov is a matrix of 1
> column. There are altogether 75 observations.
> 
> Tried :
> ------
> P <- try(surv.cox <- coxph(Surv(E, 1-F) ~ cov,
> coxph.control(iter.max=1e10, eps = 1e-03)))
> 
> Output (from above command) :
> -----------------------------
> Call:
> coxph(formula = Surv(E, F) ~ cov, data =
> coxph.control(iter.max = 1e+10, 
>     eps = 0.001))
> 
> 
>       coef exp(coef) se(coef)     z     p
> cov -0.233     0.793     0.11 -2.11 0.035
> 
> Likelihood ratio test=4.51  on 1 df, p=0.0336  n= 75 
> 
> 
> Machine Specs :
> ----------------
> Unix/ Windows XP
> R-1.8.1
> Survival package

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595




More information about the R-help mailing list