Search

2.3 — Introduction to function parameters and arguments

In the previous lesson, we learned that we could have a function return a value back to the function’s caller. We used that to create a modular getValueFromUser function that we used in this program:

However, what if we wanted to put the output line into its own function as well? You might try something like this:

This won’t compile, because function printDouble doesn’t know what identifier num is. You might try defining num as a variable inside function printDouble():

While this addresses the compiler error and makes the program compile-able, the program still doesn’t work correctly (it always prints “0 doubled is: 0”). The core of the problem here is that function printDouble doesn’t have a way to access the value the user entered.

We need some way to pass the value of variable num to function printDouble so that printDouble can use that value in the function body.

Function parameters and arguments

In many cases, it is useful to be able to pass information to a function being called, so that the function has data to work with. For example, if we wanted to write a function to add two numbers, we need some way to tell the function which two numbers to add when we call it. Otherwise, how would the function know what to add? We do that via function parameters and arguments.

A function parameter is a variable used in a function. Function parameters work almost identically to variables defined inside the function, but with one difference: they are always initialized with a value provided by the caller of the function.

Function parameters are defined in the function declaration by placing them in between the parenthesis after the function identifier, with multiple parameters being separated by commas.

Here’s some examples of functions with different numbers of parameters:

An argument is a value that is passed from the caller to the function when a function call is made:

Note that multiple arguments are also separated by commas.

How parameters and arguments work together

When a function is called, all of the parameters of the function are created as variables, and the value of each of the arguments is copied into the matching parameter. This process is called pass by value.

For example:

When function printValues is called with arguments 6 and 7, printValues‘s parameter x is created and assigned the value of 6, and printValues‘s parameter y is created and assigned the value of 7.

This results in the output:

6
7

Note that the number of arguments must generally match the number of function parameters, or the compiler will throw an error. The argument passed to a function can be any valid expression (as the argument is essentially just an initializer for the parameter, and initializers can be any valid expression).

Fixing our challenge program

We now have the tool we need to fix the program we presented at the top of the lesson:

In this program, variable num is first initialized with the value entered by the user. Then, function printDouble is called, and the value of argument num is copied into the value parameter of function printDouble. Function printDouble then uses the value of parameter value.

Using return values as arguments

In the above problem, we can see that variable num is only used once, to transport the return value of function getValueFromUser to the argument of the call to function printDouble.

We can simplify the above example slightly as follows:

Now, we’re using the return value of function getValueFromUser directly as an argument to function printDouble!

Although this program is more concise (and makes it clear that the value read by the user will be used for nothing else), you may also find this “compact syntax” a bit hard to read. If you’re more comfortable sticking with the version that uses the variable instead, that’s fine.

A warning about function argument order of evaluation

The C++ specification does not define whether arguments are matched with parameters in left to right order or right to left order. When copying values, is of no consequence. However, if the arguments are function calls, then this can be problematic:

If the architectures evaluates left to right, a() will be called before b(). If the architecture evaluates right to left, b() will be called before a(). This may or not be of consequence, depending on what a() and b() do.

If it is important that one argument evaluate first, you should explicitly define the order of execution, like so:

Warning

The C++ specification does not define whether function calls evaluate arguments left to right or right to left. Take care not to make function calls where argument order matters.

How parameters and return values work together

By using both parameters and a return value, we can create functions that take data as input, do some calculation with it, and return the value to the caller.

Here is an example of a very simple function that adds two numbers together and returns the result to the caller.

Execution starts at the top of main. When add(4, 5) is evaluated, function add is called, with parameter x being initialized with value 4, and parameter y being initialized with value 5.

The return statement in function add evaluates x + y to produce the value 9, which is then returned back to main. This value of 9 is then sent to std::cout to be printed on the console.

Output:

9

In pictorial format:

More examples

Let’s take a look at some more function calls:

This program produces the output:

9
15
10
7
6

The first statement is straightforward.

In the second statement, the arguments are expressions that get evaluated before being passed. In this case, 1 + 2 evaluates to 3, so 3 is copied to parameter x. 3 * 4 evaluates to 12, so 12 is copied to parameter y. add(3, 12) resolves to 15.

The next pair of statements is relatively easy as well:

In this case, add() is called where the value of a is copied into both parameters x and y. Since a has value 5, add(a, a) = add(5, 5), which resolves to value 10.

Let’s take a look at the first tricky statement in the bunch:

When the function add is executed, the program needs to determine what the values for parameters x and y are. x is simple since we just passed it the integer 1. To get a value for parameter y, it needs to evaluate multiply(2, 3) first. The program calls multiply and initializes z = 2 and w = 3, so multiply(2, 3) returns the integer value 6. That return value of 6 can now be used to initialize the y parameter of the add function. add(1, 6) returns the integer 7, which is then passed to std::cout for printing.

Put less verbosely:
add(1, multiply(2, 3)) evaluates to add(1, 6) evaluates to 7

The following statement looks tricky because one of the parameters given to add is another call to add.

But this case works exactly the same as the prior case. add(2, 3) resolves first, resulting in the return value of 5. Now it can resolve add(1, 5), which evaluates to the value 6, which is passed to std::cout for printing.

Less verbosely:
add(1, add(2, 3)) evaluates to add(1, 5) => evaluates to 6

Conclusion

Function parameters and return values are the key mechanisms by which functions can be written in a reusable way, as it allows us to write functions that can perform tasks and return retrieved or calculated results back to the caller without knowing what the specific inputs or outputs are ahead of time.

Quiz time

Question #1

What’s wrong with this program fragment?

Show Solution

Question #2

What two things are wrong with this program fragment?

Show Solution

Question #3

What value does the following program print?

Show Solution

Question #4

Write a function called doubleNumber() that takes one integer parameter and returns twice the value passed in.

Show Solution

Question #5

5) Write a complete program that reads an integer from the user, doubles it using the doubleNumber() function you wrote for quiz answer 4, and then prints the doubled value out to the console.

Show Solution


2.4 -- Introduction to local scope
Index
2.2 -- Function return values

432 comments to 2.3 — Introduction to function parameters and arguments

  • Hi Alex!

    From "A warning about function argument order of evaluation" onward, you're using copy initialization.

  • Louis Cloete

    Under the heading "More examples," just below the paragraph just below the code example after "Let’s take a look at the first tricky statement in the bunch:" the second to last sentence should read: "That return value of 6 can now be used to *initialize* the y parameter of the add function."

    Also, in Quiz Question #5, int x should be initialized, even if it is assigned in the next line, as per the suggestion given earlier.

  • elFlaco

    Seems this sentence above is not formed correctly? -
    "If it is important that one argument evaluate first, you can should explicitly define the order of execution, like so:"
    (you can should?)

  • yuvi

    Hello,
    The first & second program in this tutorial cause compiler error also on the line no. 14. It should be -

    instead of -

  • abaj

    under more examples you have:

    There shouldn't be a 6 between the 9 and 15

  • Dear Teacher,
    Please let me say you that it is worthy to add in lesson that function with parameters is executed even when arguments are unused. For example:

    With regards and friendship
    Georges Theodosou

    • Alex

      It would be awfully confusing if a function call were ignored simply because a parameter wasn't used. As you've rightfully discovered, the use or non-use of parameters has no bearing on whether the function is called or not.

      Here's another interesting tidbit -- if a function parameter isn't used in the function, it actually doesn't need a name. So in the above, you could have written:

  • John

    Hi Alex, I just want to say thank you so much for these tutorials so far. I took two courses for C++ at my college but I really never understood them. You explain C++ so well and I'm actually starting to understand things rather than just memorizing them. I will keep doing these tutorials and expand on my knowledge! Thank you so much again! :D

  • Nathan

    so, I ended up doing this, and it works.  Why does using Y instead of x work? Sorry if this is an odd question

    • Alex

      The parameter name in function doubleNumber() is independent from the local variable name in function main(). You could rename "y" to whatever you want and it will still work.

      The only association between the two is that during the function call to doubleNumber(x), variable x is evaluated to get its value, and then that value is copied into the function parameter.

    • You dont even have to have the

      instead you could just type:

      and it should work!

  • Devvrat Singh

    Hi Alex! This website is in my list of my top 5 favourite websites of all time. The dedication you have (it's been a very long time)... it's just awesome! :)
    But I had a question. If I have to take input from the user, for example, I can very well do it inside a function. What, then, are parameters/arguments useful for?
    PS: Are you on any social network? I would love to know you better. :)

    • Alex

      Parameters/arguments are useful for passing data to a function that needs to perform a calculation or perform a task. For example, if you had a list of names and you wanted to sort them, writing a sort function makes a lot of sense. But where would that function get the list of names from? The answer is from the caller, passed as an argument to the function.

      • Devvrat Singh

        Thanks a lot for replying! :)
        But instead of passing values to the function, couldn't we just have called the function, and taken the input values inside the function. And, if so, isn't the need for parameters eliminated?

        • Alex

          Sure, if you wanted the user to always be the one to input values. But it's better to keep your input gathering and calculation functions separate -- that way your calculation functions can work with user input or numbers provided by the caller. If your calculation function asks for user input, then it can only be used that one way.

  • GSP

    Been loving this site so far.  Here's the code I made for this quiz.  Please let me know where I can improve so far.

    #include <iostream>

    int doubleNumber()
    {
        std::cout << "Input number to double: ";
        int x;
        std::cin >> x;
        std::cout << "Answer: ";
        return x * 2;
    }
    int main()
    {
        std::cout << doubleNumber();
        return 0;
    }

  • Ayush agarwal

    Hi Alex!
    I tried to write the program as asked in quiz 5 just the way u showed in solution but when I try to compile it , it gives me an error message C2220. "No object file generated" . Can u please tell me what is causing the problem and how to fix it

    • Alex

      This sounds like a compiler-specific issue that I probably won't be able to diagnose. I suggest using Google search and typing in your compiler's name and the error message together. Undoubtedly someone else has run into this issue and there will be suggestions you can try to fix it.

    • I ran into this same issue using VS and the default Microsoft compiler. Seems it's related to the "Treat warnings as errors" compiler flag. You probably are having some warning that is triggering the error because of that flag, so yous should either solve the warning(s) or disable the "Treat warning as errors" flag. Hope this helps!

Leave a Comment

Put all code inside code tags: [code]your code here[/code]