Search code examples
rdplyr

how to use replace_na() with defuse function arguments in R

when I create a column with walrus operator ":=" I need to reference the same column later on to replace_na() values.

When I use defuse operators, replace_na() argument doesn't work.

Any suggestions on how to get it to work?


library(tidyverse)

test_fn <- function(column_name){
  
dat1 <- tibble(
  id=1:5
  ,val=11:15
)

dat2 <- tibble(
  id=1:4
  ,"{{column_name}}":=11:14
)

out <- left_join(
  dat1
  ,dat2
  ,by=join_by(id)
) |> 
  replace_na(list("{{column_name}}"=0))

return(out)

}

test_fn(hi)

The intended output is that the "hi" column has no NA values (they are replaced by 0)

Solution

  • The syntax for replace_na is a little different, so we could use:

    ... |> mutate({{column_name}} := coalesce({{column_name}}, 0))

    or

    ... |> mutate(across({{column_name}}, ~replace_na(., 0)))