Ctrl+K

Writing Functions

Writing Functions#

Introduction#

You can define your own function in R. This is particularly useful if you want to perform the same task with many different data sets. Your definition requires you to declare your function’s arguments and whether they have a set default value or not. Arguments cannot by default be forced to a certain mode or class, but you can check for them in the function an coerce them if necessary. Variables within your function are limited to only that function, and after it has run will simply disappear. If you want to store a result from a function, you must return it to the main program.

Defining a basic function#

Functions always have the same structure:

function_name <- function(argument, argument = default value) {
      statement or operations
      return(result)
}

You need to define a name for your function that you will later use to call it with. The curly brackets define where the function starts and ends. The return command returns the result back to the main program. Let’s have a look at a first example. We would like to write a function that converts temperature Fahrenheit to Celsius.

# defining function
f_to_c <- function(temp_F) {          #define a function and argument temp_F, no default arguments
  temp_C <- (temp_F - 32) * 5 / 9     #perform calculations using the argument
  return(temp_C)                      #return the result to the main program
}

# using the function
f_to_c(70) # = 21.1                   #using the name of the function to call it

The next function decrypts numbers into letters. We define two arguments: the necessary argument x and the optional argument offset which is set to 0 by default.

# defining a function
caesarDecrypt <- function(x,offset=0){
  new_x = x - offset - 1                  # Remove the offset and minus 1 for the next line
  new_x = new_x%%26 + 1                   # Find the modulo, add 1 to move from 0:25 to 1:26
  string = letters[new_x]                 # Translate numbers to letters
  return(string)                          # Return the answer
}
# Using it
x <- c(3,8,18,9,19)
caesarDecrypt(x)
x <- c(7,14,20,17,20,12,4)
caesarDecrypt(x, offset = 12)

You can create functions as complex as you like. For example, we can include an if statement or for loops. This next function only multiplies by 3 the input if it is an even number.

#multiply all even number by a certain factor
OnlyIfEven <- function(number, factor = 3){             #define function. mulitplication factor is set to 3 by default
  temp_res <- number %% 2                               #calculate modulo
  if (temp_res == 0){                                   #if modulo is zero then the number is even
  res <- number*factor                                  #mulitplication
  print(res)                                            #print result. Careful, this does not return the result!
  }
  else{print("error: expected even number")              #print error message if modulo not zero
}

#using the function
OnlyIfEven(4) #=12
OnlyIfEven(5) #error: expected even number

Organising Functions#

If you need to write many different functions for your data set it is recommended to keep them saved in separate files. This keeps your scripts nice and tidy. You can always call another file in your current script by using the source command.

source("path_to_your_file/filename.R")

Exercises#

  • Write a function to add up all numbers in a vector except for the highest

  • Write a function that deciphers letters into numbers. (hint: use the which function)

Next: Loading and installing packages

Contents