- Understand some of the benefits of splitting a program into multiple functions.
- Learn how to create our own functions.
- Understand the purpose of parameters and how to write functions that use them.
- Understand the purpose of return values and how to use them.
- Learn what "scope" is and how it affects variables in functions.
10.1 What is a Function?
We have been calling functions in Python ever since the very first "Hello World!" program:
Here we are calling the
Functions can have parameters, which are the things between the parenthesis. Parameters allow us to pass data to a function to control how it works. The print function prints out each of its parameters.
Consider the following function call:
math.sqrt function takes a number as a parameter. The job it has to do is to calculate the square root of the parameter. Unlike the print function, the sqrt function returns a value. We use parameters to give information to a function, and return values let functions give information back to us. So here, 144.0 is the parameter, and the return value will be 12.
Not all functions return values. If we try to print the return value from the
So parameters allow us to pass information to functions. Some functions take no parameters, some take one, and some take more. Likewise some functions return a value back and some do not.
10.2 Why Write Functions?
In addition to calling functions that already exist, we can create our own. There are many reasons to do this. The first is to split our code up into more manageable pieces. Just like books are divided into paragraphs and chapters, programs can be divided up into functions to make them easier to understand and write.
Another reason is to decrease repetitive code. Imagine you were writing a Program that reads in the size of a rectangle so it can compute the area and perimeter. It needs the width and height to do this, and both numbers should be greater than zero. We should make sure that's the case, so our code to read these values might look like this:
# read the width width = int(input("Width: ")) while width <= 0: print("Please enter a positive number.") width = int(input("Width: ")) # read the height height = int(input("Height: ")) while height <= 0: print("Please enter a positive number.") height = int(input("Height: "))
This code is pretty repetitive. The input line, while condition and loop bodies are all basically the same. The only things different are the variable name being used and the input prompt.
Having repetitive code in programs is not ideal. For one, we have to write more code. A basic tenet of programming is to not do more work than you have to. Another, even bigger, reason is that now we have to maintain the code in two places. If we want to, say, make it so that it doesn't crash the program when a letter is accidentally entered, then would have to make that change in two places. As we write larger programs that take longer to debug, having repetitive code like this is a bad idea.
If we were to write a function to read in a positive number, then we can just call upon it whenever we need to, and know that it already does this job. That might look like this:
We pass in the prompts, because that was the one thing that's really different. Of course we have to create the
readPositive function for this to work. So we will see how to make functions next.
10.3 Writing a Function
So now that we know we want to write functions, we need to know how to do so. Functions are created using the following syntax:
def starts a function, it stands for define. Function names follow the same rules as variable names: any letter or underscore, followed by any numbers of letters, numbers and underscores. Parameters are specified the same as in a function call: as a list separated by commas inside of parenthesis. Then the colon signifies the start of the function. Each line of code in the function must be indented over, as in a loop or if statement.
As an example, we can define a function that prints out a greeting:
Here the function's name is
greet, and it takes no parameters. The body of the function is the code that will be run when the function gets called. This function doesn't return anything either.
If we run this program, nothing actually happens. Defining a function does not run the code inside it. It just says what would happen should the function ever actually be called. So in order for this program to do something, we would need to call the function after defining it, like this:
Calling the functions we make works the same as calling a functions that come with Python. Now the program will first define the function, and then actually call it, so the print statement will run and we should see this:
Functions must be defined before they are called. So this program will only work if the definition of greet is before the call to it. If we flip them around, Python would give us an error message.
Suppose we want to personalize the greet function so that it greets people by name. To do this, the function will need to know the name of the person it is supposed to greet. To send information into a function, we use parameters. In this case, we'll add a parameter to the function that for the person's name:
Now this function takes one parameter called
name. Inside the function, we can refer to
name when we need that information. In this case, we just print it out. Notice that the function isn't setting the
name variable anywhere. The function just assumes it has been set to something already.
When we call the function, we pass in the actual value that should fill in for the parameter. For example, we add three calls to this function in the program below:
This program outputs the following:
Hello Alice how are you? Hello Bob how are you? Hello Claire how are you?
Parameters let us pass information into the function to change the way it works. The function above just prints out its name variable, but that will be different based on what was passed in.
We have to get the number of parameters right, or Python won't be able to call the function. If we pass the
greet function zero parameters, or more than one parameter, then Python will complain:
>>> greet() Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: greet() takes exactly 1 argument (0 given) >>> greet("Too", "Many") Traceback (most recent call last): File "<stdin>", line 1, in <module> TypeError: greet() takes exactly 1 positional argument (2 given)
As another example, let's write a function that will print a countdown from a starting point. Sometimes we will want the function to print out a countdown starting at 10 and going down to 1. Sometimes we will want to start at just 5 instead.
Because sometimes we want the starting point to be 10 and sometimes we want it to be 5, that has to be a parameter. We can then use that parameter, which we will call
start, to control the for loop that will do the counting down. The code for this function is below:
def countdown(start): for i in range(start, 0, -1): print(i) print("Done!") countdown(10) countdown(5)
We are calling the
countdown function with two different values for the start parameter. The output of this program appears below:
10 9 8 7 6 5 4 3 2 1 Done! 5 4 3 2 1 Done!
10.5 Return Values
Now that we have seen how to pass information to functions using parameters, we will look at how to pass information back from functions using return values.
A return value is a value that is produced by a function call. We have seen functions that produce return values such as
math.sqrt. We can return values from our own functions using the
return statement. This statement takes the value we want to return. The value we return can be any value in a Python program such as a constant, variable or expression.
For example, let's say we want to write a function to calculate the area of a rectangle. To do this, the information we need is the width and height of the rectangle. Because we need to know this to do our task, these should be parameters into the function. The result of our work will be the area, so this is what we should return back. The function could look like this:
The return statement sends back the value of the
area variable which was computed. When we call this function, we'll need to save the value it gives us into a variable. If we don't than the result it gives us will be lost (just like we have to store the value that
input gives us into a variable). The following bit of code calls the function above and then prints the result of it:
We can also take the output of a function and pass it directly as a parameter to another function. For instance, we can print the area of a rectangle to the screen directly by calling it from inside of print:
Here the rectangleArea function is called with parameters of 5 and 7. This function is going to return the value 35 which is then passed directly as a parameter to the print function. The upshot is that the 35 will be printed directly to the screen, without being stored in a variable first. We have actually been doing this same thing with the
input function for a while in code like this:
input function's return value (a string) is passed directly to the
int function, which then returns us the integer version of that string.
A function can also have multiple
return statements for different cases. For example, we can write a function to convert a number grade into a letter grade:
def numberToLetter(grade): if grade >= 92: return "A" elif grade >= 89: return "A-" elif grade >= 87: return "B+" elif grade >= 82: return "B" elif grade >= 79: return "B-" elif grade >= 77: return "C+" elif grade >= 72: return "C" elif grade >= 69: return "C-" elif grade >= 67: return "D+" elif grade >= 60: return "D" else: return "F"
Here we have a
return statement for each condition in the function. When Python executes this function, the first condition that is true will have its return statement executed.
While we can have multiple
return statements, we can only ever execute one of them for each function. As soon as we reach the
return statement, we leave the function and don't do anything else. For example, this program has a statement after the
def rectangleArea(width, height): area = width * height return area print("Hello!") # will never be printed
This will never print the "Hello!" message, because Python leaves the function as soon as the
return is done. No statements after that will be executed.
10.6 A couple more examples
Another benefit of functions is they sometimes make code easier to read. We have seen that we can check if a number is even or odd by using the modulus operator. To refresh your memory, we can write a program to tell the user if their age is an even number like this:
age = int(input("What is your age? ")) if age % 2 == 0: print("Your age is even") else: print("Your age is odd")
What's going on is that the modulus operator divides the age variable by 2, and checks the remainder. If it's 0, then the age is evenly divided by 2, so it must be even. How this code works is not exactly clear. To help make this easier to read and understand, we can write functions called
isOdd to perform these calculations:
def isEven(num): if num % 2 == 0: return True else: return False def isOdd(num): if num % 2 == 1: return True else: return False
These functions both take a number as a parameter. They then do the modulus trick to determine if the number is even or odd and then return a boolean True or False value. Now we can call them whenever we want to know if a number is even or odd, instead of having to write out the modulus expression directly.
As another example, let's say we want to write our own version of Python's
max function? Given a list of numbers, it should return the biggest item out of it.
This function will take the list as a parameter. The return value should be the biggest item found in the list. We can do this with the following function:
Here we loop through the list and keep track of the biggest item that we have seen so far. Once we've gone through the whole thing, we return the biggest.
Finally, let's write the code for the
readPositive function we talked about earlier. To do this, we will make the prompt for the user a parameter, then read values until one is positive, and then use a
return to send it back to the user. This might look like this:
def readPositive(prompt): value = int(input(prompt)) while value <= 0: print("Please enter a positive number.") value = int(input(prompt)) return value
We could even make a more general version of this function which gets a number as input from the user between any lower and upper limits. The limits could be passed into the function along with the prompt.
Scope refers to the parts of code that can access things like variables. Before we used functions, we could access any variable any place in our program, after it was created. But when we create variables inside of a function, they can only be accessed inside of that function. For example, the
hello function below makes a variable called
message. We are allowed to access that variable inside the
hello function, but not outside of it:
def hello(): message = "Hello!" # this is OK because we're inside the function print(message) # this is not allowed print(message)
This program produces an error because, in the last print statement, we are not inside of the hello function. Therefore the variable
message cannot be accessed. Even if the function hello() is called, and
message is created, we still cannot access
message outside of the function.
def hello(): message = "Hello!" print(message) # call the function hello() # still not OK and will cause an error print(message)
The scope of the
message variable is only inside the hello function because that's where it was made. If we made
message outside of the function, this would be OK:
# message is not in a function, so its scope is the whole program message = "Hello!" def hello(): # this variable CAN be called here print(message) # this is OK too print(message)
message is set up outside of the function and can be accessed inside of functions in the program.
Variables created outside of a function are called global and variables inside of a function are called local. So the scope of a global variable is the whole program, and the scope of a local variable is just the function it's in.
The reason Python works this way is to try to keep programs more organized. If our program has a bunch of functions in it, that all make several variables, things would become messy if all of those variables could be accessed anywhere in the program. By keeping the scope of a variable just to the function it was made in, things stay organized. If you want to use some value from a function in other parts of your program, you need to use a
return to send it back.
10.8 Designing Programs with Functions
When writing programs that are long and complicated, it's a good idea to break the program up into functions. We start the program by thinking about what pieces we need. We then make functions for each of the pieces and write them one by one. This is sometimes called a "divide and conquer" approach because we divide the program into parts and "conquer" them one by one.
In general each function should do one specific job. If a piece of code does more than one thing, you should consider splitting it up into multiple functions.
- Functions allow us to split programs into independent parts. This makes programs easier to write and test, and keeps code more organized.
- Parameters allow us to pass information into functions, which allows us to customize the way that they work. When we call the function, we supply it with the values it needs.
- Return values allow functions to pass information back to the code that called them. To use the return value, you normally store it into a variable.
- When a variable is created inside of a function, it can only be used inside of that function. We say the scope of that variable is inside that function which is called local. When a variable is created outside of a function, its scope is global which means it can be used anywhere.
- When working on a big program, functions allow us to split the work into smaller units and make solving a big problem more approachable.
Copyright © 2019–2023 Ian Finlayson
Licensed under a Creative Commons Attribution 4.0 International License.