by in tbl_summary fails if passed column named "variable"

[berg-michael]

Hi,

I've noticed that using the by argument in tbl_summary won't work if the column I am summarizing by is named "variable". I'm assuming this is some kind of parsing issue. Things work fine when I change the variable name to group, see reprex below. I haven't taken a close look as to why this might be occurring. Thought an issue may be relevant as the default output of data.table's melt function (from wide to long) includes a column entitled variable.

library(gtsummary)


df <- data.frame(group = c(rep("A", 5), rep("B", 5)), value= 1:10)

tbl1 <- df |> 
  tbl_summary(
  by = "group"
)


df <- data.frame(variable = c(rep("A", 5), rep("B", 5)), value= 1:10)

tbl2 <- df |> 
  tbl_summary(
  by = "variable"
)
#> Error in `mutate()`:
#> ! Problem while computing `df_stats = pmap(...)`.
#> Caused by error in `set_names()`:
#> ! The size of `nm` (2) must be compatible with the size of `x` (1).

^{Created on 2022-05-03 by the reprex package (v2.0.1)}

[ddsjoberg]

Thank you so much for the reprex and the bug report. I used the column name variable SO much within the internals....it's surely an issue with one too many columns running around named variable.

Depending on how tricky it is to fix, I may just add a check that the by variable is not named variable for the time being.

[ddsjoberg]

Thank you for the report @berg-michael ! Luckily it was an easy fix :)

[berg-michael]

Of course, thank you taking the time to fix it!