Test when the argument of function is a column name

Question

I have the following code and function:

library(tidyverse)

set.seed(12)
data_table <- tibble(
  a = sample(100),
  b = rep(c("group_1", "group_2", "group_3", "group_4"), each = 25),
  c = rep(c("group_A", "group_B", "group_C", "group_D"), times = 25),
)

mean_fct <- function(data, grouped_var){
  data %>% 
    group_by({{grouped_var}}) %>% 
    mutate(mean_a = mean(a)) %>% 
    ungroup() %>% 
    distinct(mean_a, {{grouped_var}})
}


grouped_by_number <- mean_fct(data_table, b)
grouped_by_letter <- mean_fct(data_table, c)

Now, I want to test a specific part of the function, by giving it some arguments, i.e. I want to do:

data <- data_table
grouped_var <-  b

data %>% 
  group_by({{grouped_var}}) %>% 
  mutate(mean_a = mean(a))

This doesn't work and the error is "Error: object 'b' not found"

I understand why it doesn't work, as the argument is the column name of the tibble, but then, how can I test such a specific part of the function, i.e. how can I temporarily store the argument?

Thanks a lot for any help. I really like dplyr answers. The code above is just an example. The question is about testing the function, not about the usefulness of that code.

Answer 1

library(tidyverse)

set.seed(12)
data_table <- tibble(
  a = sample(100),
  b = rep(c("group_1", "group_2", "group_3", "group_4"), each = 25),
  c = rep(c("group_A", "group_B", "group_C", "group_D"), times = 25),
)

mean_fct <- function(data, grouped_var){
  data %>% 
    group_by({{grouped_var}}) %>% 
    mutate(mean_a = mean(a)) %>% 
    ungroup() %>% 
    distinct(mean_a, {{grouped_var}})
}

# Define test data and arguments
data <- data_table
grouped_var <- quo(b)  # Using quo() to capture the argument as an expression

# Evaluate the expression within the test
data %>% 
  group_by(!!grouped_var) %>%  # Unquote the argument using !! (bang-bang)
  mutate(mean_a = mean(a))

Answer 2

You have to take col b as a column of the tibble

grouped_var <-  data_table$b

I get this output from the group_by mutate:

       a b       c       `<chr>` mean_a
   <int> <chr>   <chr>   <chr>    <dbl>
 1    66 group_1 group_A group_1   61.1
 2    90 group_1 group_B group_1   61.1
 3    80 group_1 group_C group_1   61.1
 4    91 group_1 group_D group_1   61.1
 5    46 group_1 group_A group_1   61.1
 6    69 group_1 group_B group_1   61.1
 7    95 group_1 group_C group_1   61.1
 8    92 group_1 group_D group_1   61.1
 9    34 group_1 group_A group_1   61.1
10     8 group_1 group_B group_1   61.1
# ℹ 90 more rows