Home Is there a way to loop R functions across multiple columns when the function is determined by the column name?

Questions

Is there a way to loop R functions across multiple columns when the function is determined by the column name?

January 24, 2022

I have a dataset that I am converting from long to wide format. My original dataset contains countries of occurrence for each species ("code") as one of the columns. I have created new columns for all country codes (e.g., "FI", "NO"), and I now need to translate the data in "code" to these columns (e.g., for df$FI, if df$Name == "FI" then YES, otherwise NA). Is there a way of automating this across all of these new columns (so that a YES value will be returned if the value in df$Name is equivalent to the column title, otherwise an NA value will be returned), or do I have to do this manually for each new column created? Thanks.

   Name        code
1   ABC         FI
2   ABC         NO
3   ABC         RU
4   ABC         SE
5   BCD         IT
6   BCD         RO

>Solution :

Here is a somewhat strange solution

library(tidyverse)


example_data <- read_table('Name        code
ABC         NO
ABC         RU
ABC         SE
BCD         IT
BCD         RO')

example_data |> 
  pivot_wider(names_from = code,values_from = code) |> 
  mutate(across(-Name,.fns = ~ if_else(.x |> is.na(),
                                       NA_character_,
                                       'YES')))
#> # A tibble: 2 x 6
#>   Name  NO    RU    SE    IT    RO   
#>   <chr> <chr> <chr> <chr> <chr> <chr>
#> 1 ABC   YES   YES   YES   <NA>  <NA> 
#> 2 BCD   <NA>  <NA>  <NA>  YES   YES

^{Created on 2022-01-24 by the reprex package (v2.0.1)}