Tidy evaluation with user-specified column names

user-tidy-eval.R

# Passing user-specified column names to tidyverse functions inside your own functions
# Based on: https://dplyr.tidyverse.org/articles/programming.html

library(tidyverse)
str(mtcars)

# How you might think you write a function that takes a column name as argument
# and uses tidyverse functions to do something based on that column name...
my_first_group_summarise_fn <- function(df, col_name) {
  
  df_grouped_summarised <- df %>%
    group_by(col_name) %>%
    summarise(freq = n())
  
  return(df_grouped_summarised)
  
}

# But this doesn't work...
my_first_group_summarise_fn(df = mtcars, col_name = "gear") # pass as string (doesn't work)
my_first_group_summarise_fn(df = mtcars, col_name = gear) # pass using tidy evaluation (doesn't work)

# group_by is looking for a column called col_name

# so instead... if you want to pass the column name as a string...

my_second_group_summarise_fn <- function(df, col_name) {
  
  df_grouped_summarised <- df %>%
    group_by(.data[[col_name]]) %>%
    summarise(freq = n())
  
  return(df_grouped_summarised)
  
}

my_second_group_summarise_fn(df = mtcars, col_name = "gear") # pass as string (works)
my_second_group_summarise_fn(df = mtcars, col_name = gear) # pass using tidy evaluation (doesn't work)

# And if you want to be able to use tidyverse semantics for your own function...

my_third_group_summarise_fn <- function(df, col_name) {
  
  df_grouped_summarised <- df %>%
    group_by({{ col_name }}) %>%
    summarise(freq = n())
  
  return(df_grouped_summarised)
  
}

my_third_group_summarise_fn(df = mtcars, col_name = "gear") # pass as string (doesn't work)
my_third_group_summarise_fn(df = mtcars, col_name = gear) # pass using tidy evaluation (works)

# Some more operations

my_fourth_group_summarise_fn <- function(df, col_name) {
  
  print(class(df %>% pull({{ col_name }})))
  df <- df %>% mutate("{{ col_name }}" := {{ col_name }} + 1)
  
  df_grouped_summarised <- df %>%
    group_by({{ col_name }}) %>%
    summarise(freq = n())
  
  df_grouped_summarised2 <- df_grouped_summarised %>%
    rename(passed_col = {{ col_name }})
  
  colplot <- ggplot(df_grouped_summarised, aes(x = {{ col_name }}, y = freq)) +
    geom_bar(stat = "identity")
  
  return(colplot)
  
}

my_fourth_group_summarise_fn(df = mtcars, col_name = "gear") # pass as string (doesn't work)
my_fourth_group_summarise_fn(df = mtcars, col_name = gear) # pass using tidy evaluation (works)
my_fourth_group_summarise_fn(df = mtcars, col_name = cyl) # pass using tidy evaluation (works)