Using dplyr quosures with mutate()

Question

Reading the Programming with dplyr guide, I get a unexpected error.

In fact, trying the examples with summarize works quite right, but if I try similar things with mutate, the code won't run.

Here is my code :

df = data.frame(A=c(464,3465,48,3415,357,21,657), B=c(12,15,985,35,67,13,467))

df %>% 
  mutate(x = A-B) %>%
  pull(x)
#returns the good output :
#[1]  452 3450 -937 3380  290    8  190

get.diff = function(var1, var2, data=df){
  var1 = enquo(var1)
  var2 = enquo(var2)
  data %>% 
    mutate(x = (!!var1 - !!var2)) %>%
    pull(x)
}
get.diff(A, B)
returns an error

Here is the error :

Error in !var2 : invalid argument type

I tried a lot of thing, but this is the closest I got to the guide samples.

What did I miss ?

Answer 1

You just need parentheses around each "bang bang" / unquoting (!!varX) :

mutate(x = ((!!var1) - (!!var2))) %>%

Full code:

df = data.frame(A=c(464,3465,48,3415,357,21,657), B=c(12,15,985,35,67,13,467))

  df %>%
    dplyr::mutate(x = A-B) %>%
    pull(x)
  #returns the good output :
  #[1]  452 3450 -937 3380  290    8  190

  get.diff = function(var1, var2, data=df){
    var1 = enquo(var1)
    var2 = enquo(var2)
    data %>%
      dplyr::mutate(x = ((!!var1) - (!!var2))) %>%
      pull(x)
  }
  get.diff(A, B)

  [1]  452 3450 -937 3380  290    8  190