How to subset variable in gtsummary

I would like to subset a specific variable (not the entire dataset) in gtsummary.

In the following example, how could I subset gear to remove '5' - only show proportion of cars with gear of '3' and '4' ? I would want to include all patients in mpg however.

library(gt)
library(dplyr) 

mtcars %>%
          select(cyl, mpg, gear) %>%
          tbl_summary(
                    by = cyl ### how do i say for gear, filter gear != 5  ???
          )

Solution

You'll need to build two separate tables with tbl_summary() then stack them. Example below!

library(gtsummary)
packageVersion("gtsummary")
#> [1] '1.5.0'

tbl_full_data <-
  mtcars %>%
  select(cyl, mpg) %>%
  tbl_summary(by = cyl) %>%
  # removing Ns from header, since they won't be correct for gear
  modify_header(all_stat_cols() ~ "**{level}**")

tbl_gear_subset <-
  mtcars %>%
  select(cyl, gear) %>%
  dplyr::filter(gear != 5) %>%
  tbl_summary(by = cyl) 

# stack tables together
list(tbl_full_data, tbl_gear_subset) %>%
  tbl_stack() %>%
  as_kable() # convert to kable to it'll print on SO
#> i Column headers among stacked tables differ. Headers from the first table are
#> used. Use `quiet = TRUE` to supress this message.

Characteristic	4	6	8
mpg	26.0 (22.8, 30.4)	19.7 (18.6, 21.0)	15.2 (14.4, 16.2)
gear
3	1 (11%)	2 (33%)	12 (100%)
4	8 (89%)	4 (67%)	0 (0%)

^{Created on 2021-10-25 by the reprex package (v2.0.1)}