Reputation: 1
Can someone explain to me why this error message keeps coming up? My previous time-to-event analyses have all worked perfectly with the same commands. Tank you.
My skript contains the following:
dfit <- survfit2(Surv(time, status) ~ intervention, data = Drugs_discon)
summary(dfit)
ggsurvfit(dfit)
Error in ggsurvfit()
:
! ggsurvfit()
cannot be used to plot objects of class <survfit2/survfitms/survfit>.
ℹ Use ggcuminc()
for competing risks cumulative incidence plotting.
Run rlang::last_trace()
to see where the error occurred.
Upvotes: 0
Views: 332
Reputation: 81
I can't confirm without a reproducible example but I'm fairly sure the problem is that status
is considered a factor in your dataframe.
You can put
dfit <- survfit2(Surv(time, as.numeric(status)) ~ intervention, data = Drugs_discon)
summary(dfit)
ggsurvfit(dfit)
This should produce the required plot
Upvotes: 1