ggplot2, how do you use access dataframe elements from within the function?

Question

I'm trying to create plots in with ggplot2 to standardize the creation of many similar plots. Given you can stack numerous options, using theme(), stat_*(), or geom_*() with the + operator, how can you wrap this inside a function?

I'd like to create a function to plot sections of a data frame, for example :

library(ggplot2)
test<-function(fdata,fx){
  p<-ggplot(fdata,aes(x=fx,y=mpg))+geom_point()
  print(p)}; 
test(mtcars,wt)

returns an error of:

Error in eval(expr, envir, enclos) : object 'fx' not found

but changing the aesthetic definition to a locally scoped variable (e.g. x=fx to x=wt) works, but won't use the function's second option of 'fx':

library(ggplot2)
test<-function(fdata,fx){
  p<-ggplot(fdata,aes(x=wt,y=mpg))+geom_point()
  print(p)}; 
test(mtcars,wt)

How do I write a function to create a plot from different variables in a data frame?

E.g. how do I write function test so that I could call multiple times to get similar plots, like :

test(mtcars,'wt')  #scatter plot of wt x mpg
test(mtcars,'disp') #scatter plot of disp x mpg
test(mtcars,'hp')  #scatter plot of hp x mpg

Answer 1

The trick here is to use aes_string instead of using aes. The former allows you to pass strings as aesthetics. This allows you to pass a variable containing the string of the aesthetic, e.g. 'wt' in case of plotting mtcars. You need to modify your code like this:

library(ggplot2)
test <- function(fdata, fx) {
  p <- ggplot(fdata, aes_string(x = fx, y = "mpg")) + geom_point()
  print(p)
}
test(mtcars, "wt")