coxph with frailty models fails when conf.int=TRUE

[matthieu-faron]

Hi, thank you for the great work on the broom package

I'am using CRAN broom 0.7.0

I'am triyng a coxph model with a frailty term. When using tidy without conf.int = TRUE, everything works properly

library(survival)
library(broom)

model <- coxph(Surv(time, status) ~ age + sex + frailty(inst), lung)
tidy(model, conf.int = FALSE)

But when using conf.int = TRUE, it returns an errors

model <- coxph(Surv(time, status) ~ age + sex + frailty(inst), lung)
tidy(model, conf.int = TRUE)

Note that if your coxphmodel only has one variable, the confidence interval will recycle when the function binds the conf int with the coefficients.

model <- coxph(Surv(time, status) ~ age + frailty(inst), lung)
tidy(model, conf.int = TRUE)

I guess the conf.int should be [NA-NA] for the frailty term ?
I may be able to contribute a solution but I'am not very used to github...
Kind regards

[alexpghayes]

I don't know enough about survival to answer this. We should find someone who does.

…

On Thu, Jul 16, 2020 at 9:08 AM Simon P. Couch ***@***.***> wrote: Thanks for catching this! This fix would probably need to address any of the penalty functions ( ridge(), pspline(), and frailty(), as far as I know) supplied by coxph. Let me know what you think the best approach here is, @alexpghayes <https://github.com/alexpghayes>. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#899 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADTBG2ZL62G7DP7PIDYHIP3R34CXDANCNFSM4O3XCIZQ> .

[simonpcouch]

Closing in favor of #966! Thanks for yalls patience here. :-)

[github-actions]

This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.