pharmaverse / ggsurvfit

http://www.danieldsjoberg.com/ggsurvfit/
Other
67 stars 19 forks source link

ggsurvfit::ggcuminc - customized statistics with risktable_stats not working #190

Closed devanto86 closed 4 months ago

devanto86 commented 4 months ago

I'm trying to plot cumulative incidence curves for a competing risk scenario with ggsurvfit::ggcuminc. I'd like to add a risk table, customized to show n.risk and cum.event in the same line (not different line), as explained in the vignette https://www.danieldsjoberg.com/ggsurvfit/reference/add_risktable.html#ref-examples (see third example).

set.seed(1)
summary(time <- rnorm(200, 50, 10))
summary(event <- factor(sample(0:2, 200, replace = TRUE, prob = c(.4,.5,.1)), levels = 0:2, labels = c("censor", "event", "comp.event")))
table(female <- sample(0:1, 200, replace = TRUE), useNA = "ifany")
data <- tibble(time, event, female)

tidycmprsk::cuminc(Surv(time, event)~female, data=data) %>% 
  ggcuminc(outcome = "event") + add_risktable(risktable_stats = "{n.risk} ({cum.event})")

However, i always get this error

Error in match.arg(rev(risktable_stats), choices = c("n.risk", "cum.censor", : 'arg' should be one of “n.risk”, “cum.censor”, “cum.event”, “n.censor”, “n.event”

Can you help in understanding why this is not working, as shown in the vignette?

https://stackoverflow.com/questions/77981972/ggsurvfitggcuminc-customized-statistics-with-risktable-stats-not-working

ddsjoberg commented 4 months ago

Hi @devanto86 , stackoverflow.com is indeed the correct place to ask questions about the package. Thanks!