A function is just a command that typically takes in input, does something to the input and sends the result as an output. Not all functions require inputs. R does require a function to have an output though.

We have already been using functions! Every time you see something like some_function() that is a function. The input goes in the parentheses some_function(4) would do something with the number for return the result.

For a simple example lets use the function to generate random numbers from a uniform distribution. A uniform distribution just means any number in the specified range is equally likely. Rolling a die, for instance, is a random uniform distribution from 1 to 6. In R the function is runif(). One difference between a die and runif() is that runif() generates random floating numbers. runif() requires one parameter for input, how many numbers to generate.


> runif(1)[1]
[1] 0.1975651

# Generates 10 random numbers and returns a vector.
> runif(10)
 [1] 0.6910957 0.1875335 0.5158921 0.4604096 0.6535119 0.8049887 0.7999748 0.8750682
 [9] 0.1394821 0.4831887

Many functions also have optional parameters that allow you to change how the function behaves. We can change runif() to return numbers between 0 and 10.


> runif(10, min = 0, max = 10)
 [1] 7.219652 7.746382 6.813922 5.046449 8.404445 3.504893 6.244533 1.103338 7.530735
[10] 5.119951

We can use the result of one function as the input for another function. Here we calculate the mean of the results from runif() using the function mean().


> mean(runif(10))
[1] 0.3963005

> mean(runif(10, min = 0, max = 10))
[1] 4.584673

To find out more about a function you can use the help() function. The documentation will be displayed in the lower right pane of R Studio. Typically you will get a Description, Usage, a list of Arguments with explanations, more Details, relevant References and Examples.


> help("runif")

> help("help")  # If you want to get kinda meta about it.

You can also create your own functions. This is very useful if you are repeating code. Functions can also help make your code more readable. Creating a function is pretty much like creating any other object. Below are four ways to create a function to convert miles into kilometers.


miles_to_km <- function(miles) { 
  km <- miles * 1.60934
  return(km)    
}


# Return is optional. R returns the last statement in a function.
miles_to_km2 <- function(miles) { 
  km <- miles * 1.60934
  km     
}

# This shortens the function. Usually shorter is better
miles_to_km3 <- function(miles) { 
  return(miles * 1.60934)
}

# A little more refractoring
miles_to_km4 <- function(miles) { 
  miles * 1.60934
}


# A more general converter
converter <- function(num, conversion  = 1.60934) {
  num * conversion
}

Try It Out

  1. Investigate the floor() function. See what it does and how to use it.
  2. Write a function that models a six sided die using runif(). (HINT: Die tosses are integers! HINT: R typically applies a function to each element of a vector.)
  3. Write a function to return the average roll from N number of tosses.
  4. Check your function by using the mean() function. What is the expected average roll of a six sided die.
  5. Advanced: Edit your function to model S sided dice. Make six sides the default.