5 Functions in `R`

We’ve learned how to use built-in R functions like dnorm() and pnorm() to analyze distributions, but sometimes it’s going to be more helpful to be able to (A) do the math by hand or (B) code your function to do it. So, let’s learn how in the world you do that!

Getting Started

Packages

We’ll be using the tidyverse package, a super-package that auto-loads dplyr, ggplot2, and other common functions.

library(tidyverse)

5.1 Coding your own function!

Functions are machines that do a specific calculation using an input to produce a specific output.

Below, we’ll write an example function, called add(a, b).

This function takes two numeric values, a and b, as inputs, and adds them together.
Using function(), we’ll tell R that our function contains two inputs, a and b.
Then, using { ... }, we’ll put the action we want R to do in there.
The function can involve multiple operations inside it. But at the end, you need to print one final output, or put return() around your output.

# Make function
add <- function(a, b){
  # Compute and directly output
  a + b 
}
add(1, 2)

## [1] 3

# This also works
add <- function(a, b){
  # Assign output to a temporary object
  output <- a + b
  # Return the temporary object 'output'
  return(output)
}
# 
add(1, 2)

## [1] 3

5.2 Functions with Default Inputs

You can also assign default input values to your function. Below, we write that by default, b = 2. If we supply a different b, the default will get overwritten, but otherwise, we won’t need to supply b.

add = function(a, b = 2){
  a + b
}

Let’s try it!

# See? I only need to write 'a' now 
add(1)

## [1] 3

# But if I write 'b' too....
add(1, 2)

## [1] 3

# And if I change 'b'...
add(1, 3)

## [1] 4

# It will adjust accordingly

# clear data
remove(add)

Conclusion

Great! Let’s go make some functions!

5 Functions in R