using quasiquotation in a function with summarize in dplyr

Question

I'm trying to write a function that can take column names as strings assigned to variables and produce a summarized output for them like so

my_function <- function(my_df, columnA,columnB){
    

    summ_df <- my_df %>% 
        group_by(cyl) %>% 
        summarise(base_mean = mean(columnA),
                  contrast_mean = mean(columnB)) 
    
    return(summ_df)
    }


    base = "drat"
    cont = "wt"
    my_function(mtcars,base,cont)

What I would want is that the above function would return the same thing as

mtcars %>% 
    group_by(cyl) %>% 
    summarise(base_mean = mean(drat),
              contrast_mean = mean(wt)) 

I'm sure it's some combination of enexpr, or ensym, and !! but i keep getting NA values

Answer 1

Use ensym with !! so that it can take both unquoted and quoted actual arguments

my_function <- function(my_df, columnA,columnB){
    

    my_df %>% 
        group_by(cyl) %>% 
        summarise(base_mean = mean(!! ensym(columnA)),
                  contrast_mean = mean(!! ensym(columnB)), .groups = 'drop' )
    
   
    }

-testing

> my_function(mtcars, !!base, !!cont)
# A tibble: 3 × 3
    cyl base_mean contrast_mean
  <dbl>     <dbl>         <dbl>
1     4      4.07          2.29
2     6      3.59          3.12
3     8      3.23          4.00