Observe the amazing obscurity of 100% "functional" R code!

functional_madness.R

## Some sort of list comprehension using only functional programming 
## techniques that returns a list holding the coefficients of variation
## (i.e. the mean divided by the standard deviation) for each column in the 
## mtcars data set. 

## This isn't the clearest code (i.e. its really hard to tell its dividing
## the mean by the standard deviation) but it uses absolutely no imperative control 
## flow structures and no vectorized R functions, so if you can read and understand
## this you can do you some functional programming for great good =). I can't believe I
## actually figured this out, much less that I figured it out in like 5 minutes, 
## so I decided to preserve it for posteriry in a gist. yay posterity!

a <- lapply(mtcars, 
            function(col) {
              Reduce(`/`, lapply(list(mean,sd),
                                 function(f) {
                                   f(col,na.rm=TRUE)
                                 }))
            })

## You can even write a function to do this for any data frame/list!
coef_of_var <- function(list_struct = mtcars, na.rm=TRUE) {
  
  valid_cols <- vapply(list_struct, function(v) is.numeric(v) || is.logical(v),
                       logical(1))
  a <- lapply(list_struct[valid_cols],
              function(col) {
                Reduce(`/`, lapply(list(mean,sd),
                                   function(f) {
                                     f(col,na.rm=na.rm)
                                   }))
              })
  return(a)
}