Take default values from model in `plot_surv_area()`

[DanChaltiel]

Hi,

Thanks for this very nice package.

Would it be possible to add default values in plot_surv_area()?

You will often use the same values as the ones you used in the model.

Therefore, you could write a much simpler code to get the same output.

If you think about putting model as the first argument, your function even becomes pipeable! (At the price of backward compatibility, though)

library(survival)
library(contsurvplot)
#> Warning: le package 'contsurvplot' a été compilé avec la version R 4.4.3
#> Please cite as: 
#> 
#> Denz R, Timmesfeld N (2023). "Visualizing the (Causal) Effect of a Continuous Variable on a Time-To-Event Outcome." Epidemiology, 34 (5). doi: 10.1097/EDE.0000000000001630.


plot_surv_area2 = function(model, variable, data, time, status, ...) {
  if(missing(data)) data <- eval(model$call$data, envir=parent.frame())
  surv_var = all.vars(formula(model))
  if(missing(time)) time=surv_var[1]
  if(missing(status)) status=surv_var[2]
  if(missing(variable)) variable=surv_var[3]
  
  plot_surv_area(time=time,
                 status=status,
                 variable=variable,
                 data=data,
                 model=model, ...)
}

coxph(Surv(time, status) ~ age + sex + nodes, data=colon, x=TRUE) %>% 
  plot_surv_area2(variable="nodes")

^{Created on 2025-03-05 with reprex v2.1.1}

[RobinDenz1]

Thanks for the suggestion. I haven't done this yet because it only really works for coxph models, but I would write something like a waiver() function or something similar that does this. I will keep this open, but the contsurvplot package is unfortunately pretty low on my priority list right now.