Passing dataframe column to custom function for both analysis and title

Question

I have a custom function that takes as inputs both a dataframe and a specific column of that dataframe. I'd like to use the column name to auto-generate a title, based on a separate labeller argument. I've figured out a way to make it work by using colnames() to isolate the column variable and turn its name into a string, but is there a more efficient way to do this? I am trying to learn how to set up NSE functions properly, but keep running against the limits of my understanding.

Here is a reproducible example:

library(dplyr)
library(ggplot2)

#sample data
data(msleep)

#labels for title generation
fancylabs <- as_labeller(c('awake' = "Awake", 
                        'sleep_rem' = "REM"))

#function to plot two columns and generate a title based on y
quickplot <- function(df, x, y){
  
  #use colnames to convert y to a string - IS THERE A LESS CLUNKY WAY TO DO THIS?
  yStr <- df %>% select({{y}}) %>% colnames()
  
  #plot and use labeller to convert string to fancy label
  df %>% 
    ggplot(aes({{x}}, {{y}})) +
    geom_point() +
    labs(title=paste0("Total Sleep Time vs. ", fancylabs(yStr), " Time"))
}

quickplot(df=msleep, x=sleep_total, y=awake)

Answer 1

As I understand your function is to make titles for graphs, I can offer you my idea.

I have created a function that converts the variable name to spelling text, it works only for logically named variables, as in most test databases in R, but it can also be modernized

library(dplyr)
library(ggplot2)
library(rlang)

# Sample data
data(msleep)

#Function to rename the variables
rename_variable <- function(var_name) {
  var_name <- gsub("_", " ", var_name)  # Replace underscores with spaces
  var_name <- tools::toTitleCase(var_name)  # Change the first letter to a capital letter
  return(var_name)
}

quickplot <- function(df, x, y) {
  # Get the names of x and y columns
  x_name <- deparse(substitute(x))
  y_name <- deparse(substitute(y))
  #Rename the variables
  x_title  <- rename_variable(x_name)
  y_title  <- rename_variable(y_name)
  
  df %>%
    ggplot(aes({{x}}, {{y}})) +
    geom_point() +
    labs(title = paste0(x_title, " vs ", y_title))
}

# Example usage
quickplot(df = msleep, x = sleep_total, y = awake)

You can also add two more parameters x_name and y_name to your function so that the user can enter the name itself, and if they ignore them when entering the function, the rename_variable function will be used