pp_check with censored data

[andymilne]

Should it be possible to run pp_check with censored data? The following example results in the following error

fit3 <- brm(time | cens(censored) ~ age * sex + disease + (1|patient),
            data = kidney, family = lognormal())
pp_check(fit3)

Using 10 posterior samples for ppc type 'dens_overlay' by default.
Error in !cens : invalid argument type
In addition: Warning messages:
1: In drawGTree(x) : reached elapsed time limit
2: Censored responses are not shown in 'pp_check'. 

[paul-buerkner]

thanks! Should be fixed now

[lymanmark]

Seems to work with a numeric censoring variable, but a character variable ("right", "none", etc.) still has the same issue.

kidney <- kidney %>%
mutate(censored_alt = if_else(censored == 1, "right", "none"))

fit3 <- brm(time | cens(censored_alt) ~ age * sex + disease + (1|patient),
data = kidney, family = lognormal())
pp_check(fit3)

[paul-buerkner]

You are right. Should be fixed now.