[SOLVED] Using vector with strings in mutate with paste

Using vector with strings in mutate with paste

I am trying to build an ID in a tibble with "paste" and a vector with the column names as strings:

library(tidyverse)

df <- tribble(
  ~PERIOD, ~CODE, ~VALUE,
  1995, "A", 10,
  1996, "B", 20
)
df

mygroup <- c("PERIOD", "CODE")

I would like to have this result

# A tibble: 2 x 4
  PERIOD CODE  VALUE ID    
   <dbl> <chr> <dbl> <chr> 
1   1995 A        10 1995_A
2   1996 B        20 1996_B

Of course the following code doesn't do what I want

df %>% 
  mutate(ID = paste(mygroup, collapse = "_"))

as it produces

  PERIOD CODE  VALUE ID         
   <dbl> <chr> <dbl> <chr>      
1   1995 A        10 PERIOD_CODE
2   1996 B        20 PERIOD_CODE

I have tried with !! and curly braces, but nothing worked.

Solution

Using base R. The code is concise and tractable but will be inefficient if mygroup is very long. Use do.call() instead.

df$ID <- Reduce(\(x, y) paste(x, y, sep = "_"), df[mygroup])

#   PERIOD CODE  VALUE ID    
#    <dbl> <chr> <dbl> <chr> 
# 1   1995 A        10 1995_A
# 2   1996 B        20 1996_B

Another concise but inefficient solution:

df$ID <- apply(df[mygroup], 1, paste, collapse = "_")