How to define the RHS of case_when() with existing and newly defined columns passed as arguments in user defined function?

Question

Sample code:

library(tidyverse)

iris <- iris

test_tidyeval <- function(data, col_name, col_name_2, column1) {
  
  mutate(
    data,
    {{col_name}} := case_when(Species == "setosa" ~ column1 + Sepal.Width + Petal.Length,
                TRUE ~ column1),
    {{col_name_2}} := case_when(Species == "setosa" ~ {{col_name}} + 100,
                TRUE ~ {{col_name}} + 500))
  
}

iris %>% test_tidyeval("new_column_test", "new_column_test_2", Sepal.Length)

I'm sure this is a tidyeval/nse issue which I can never get my head around.

What I basically want is for new_column_test to be created where if the row Species == "setosa" then for this to be the sum of Sepal.Length, which we're passing to column1 in the user-defined function, Sepal.Width and Petal.length, else just return the value from Sepal.Length, then for new_column_test_2 to add 100 to new_column_test with the same logical condition used previously and 500 to non setosa species.

I can seem to manipulate the LHS of case_when okay but I'm stuck on the RHS statements.

score 2 · Accepted Answer · answered Nov 07 '22 at 14:14

You need to be careful when mixing strings and symbols. They behave differently. You use {{ }} when working with symbols and .data[[]] when working with strings. This should work

test_tidyeval <- function(data, col_name, col_name_2, column1) {
  
  mutate(
    data,
    "{col_name}" := case_when(Species == "setosa" ~ {{column1}} + Sepal.Width + Petal.Length,
                            TRUE ~ {{column1}}),
    "{col_name_2}":= case_when(Species == "setosa" ~ .data[[col_name]] + 100,
                                TRUE ~ .data[[col_name]] + 500))
  
}
iris %>% test_tidyeval("new_column_test", "new_column_test_2", Sepal.Length) %>% str()
#'data.frame':  150 obs. of  7 variables:
# $ Sepal.Length     : num  5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
# $ Sepal.Width      : num  3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
# $ Petal.Length     : num  1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
# $ Petal.Width      : num  0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
# $ Species          : Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...
# $ new_column_test  : num  10 9.3 9.2 9.2 10 11 9.4 9.9 8.7 9.5 ...
# $ new_column_test_2: num  110 109 109 109 110 ...

If you passed everything as symbols, it would look like this

test_tidyeval <- function(data, col_name, col_name_2, column1) {
  
  mutate(
    data,
    "{{col_name}}" := case_when(Species == "setosa" ~ {{column1}} + Sepal.Width + Petal.Length,
                            TRUE ~ {{column1}}),
    "{{col_name_2}}":= case_when(Species == "setosa" ~ {{col_name}} + 100,
                                TRUE ~ {{col_name}} + 500))
  
}

iris %>% test_tidyeval(new_column_test, new_column_test_2, Sepal.Length)

Andy Baxter · Answer 2 · 2022-11-07T14:10:41.510

A few tweaks and this should get you what you are looking for:

library(tidyverse)
library(rlang)


test_tidyeval <- function(data, col_name, col_name_2, column1) {
  
  mutate(
    data,
    {{col_name}} := case_when(Species == "setosa" ~ !!enquo(column1) + Sepal.Width + Petal.Length,
                              TRUE ~ !!enquo(column1)),
    {{col_name_2}} := case_when(Species == "setosa" ~ !!parse_expr(col_name) + 100,
                                TRUE ~ !!parse_expr(col_name) + 500)
    )
  
}

iris %>% 
  test_tidyeval("new_column_test", "new_column_test_2", Sepal.Length) %>%
  head()
#>   Sepal.Length Sepal.Width Petal.Length Petal.Width Species new_column_test
#> 1          5.1         3.5          1.4         0.2  setosa            10.0
#> 2          4.9         3.0          1.4         0.2  setosa             9.3
#> 3          4.7         3.2          1.3         0.2  setosa             9.2
#> 4          4.6         3.1          1.5         0.2  setosa             9.2
#> 5          5.0         3.6          1.4         0.2  setosa            10.0
#> 6          5.4         3.9          1.7         0.4  setosa            11.0
#>   new_column_test_2
#> 1             110.0
#> 2             109.3
#> 3             109.2
#> 4             109.2
#> 5             110.0
#> 6             111.0

As brief explanations (of tools used from here):

!!enquo(column1) first captures the non-character column1 argument (without evaluating it) and then !! evaluates it in RHS of case_when
!!parse_expr(col_name) takes the col_name string and parses it, again then evaluating it using !!

That tidyeval cheatsheet is a bit out of data and not really current best practices when using dplyr. A more up-to-date reference is [Programming with dplyr](https://dplyr.tidyverse.org/articles/programming.html). — MrFlick, Nov 07 '22 at 14:17

G. Grothendieck · Answer 3 · 2022-11-07T15:35:18.240

The problems are:

column1 is passed unevaluated so we need to use {{column1}} in the function
col_name is passed as a character string so use !!sym(col_name) in the function or c_across(col_name) or as already mentioned in another answer .data[[col_name]]
note that since each of the case_when's have only 2 arms we could have written this more compactly using if_else statements
note that only a dplyr library call is needed

Keeping this as close as we can to the code in the question this gives

library(dplyr)

test_tidyeval <- function(data, col_name, col_name_2, column1) {
  
  mutate(
    data,
    {{col_name}} := case_when(Species == "setosa" ~ {{column1}} +
                    Sepal.Width + Petal.Length,
                TRUE ~ {{column1}}),
    {{col_name_2}} := case_when(Species == "setosa" ~ !!sym(col_name) + 100,
                TRUE ~ !!sym(col_name) + 500))
  
}

iris %>% test_tidyeval("new_column_test", "new_column_test_2", Sepal.Length)

How to define the RHS of case_when() with existing and newly defined columns passed as arguments in user defined function?

3 Answers3