I'm trying to analyze data where I found that no one failed during the period of the study. This is the Error I got, as I gave the command.

> fitsurv <- coxph(Surv(NumSessions, failed) ~ o + a + n + c + e, data=records)
Error in fitter(X, Y, strats, offset, init, control, weights = weights,  : 
 NA/NaN/Inf in foreign function call (arg 6)
In addition: Warning message:
In fitter(X, Y, strats, offset, init, control, weights = weights,  :
 Ran out of iterations and did not converge

is as above.

Can anyone tell me what this means?

  • 1
    $\begingroup$ Isn't it telling you that the survival function isn't defined in a case where all observations are censored? $\endgroup$ – Mike Hunter Nov 4 '15 at 17:58
  • $\begingroup$ If that is the case, could you tell me how to analyze such data? $\endgroup$ – rk567 Nov 4 '15 at 18:01
  • $\begingroup$ What other questions do you want answered about the data? You can summarize time elapsed so far, but it is contingent on an arbitrary cut-off date. In some contexts that makes sense, e.g. as length of service of people currently in post. $\endgroup$ – Nick Cox Nov 4 '15 at 18:11
  • $\begingroup$ It would no longer be a survival analysis but a multiple regression of some type, correct? $\endgroup$ – Mike Hunter Nov 4 '15 at 18:14
  • 1
    $\begingroup$ Note that your survival functions will just be 1. What are you trying to find out? $\endgroup$ – Glen_b Nov 4 '15 at 21:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.