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:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 |
#include <iostream> int getValueFromUser() { std::cout << "Enter an integer: "; int input{ 0 }; std::cin >> input; return input; } int main() { int num { getValueFromUser() }; std::cout << num << " doubled is: " << num * 2 << '\n'; return 0; } |
However, what if we wanted to put the output line into its own function as well? You might try something like this:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 |
#include <iostream> int getValueFromUser() // this function now returns an integer value { std::cout << "Enter an integer: "; int input{ 0 }; std::cin >> input; return input; // added return statement to return input back to the caller } // This function won't compile // The void return type means the function won't return a value to the caller void printDouble() { std::cout << num << " doubled is: " << num * 2 << '\n'; } int main() { int num { getValueFromUser() }; printDouble(); return 0; } |
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():
|
1 2 3 4 5 |
void printDouble() { int num{ 0 }; // we added this line std::cout << num << " doubled is: " << num * 2 << '\n'; } |
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:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 |
// This function takes no parameters // It does not rely on the caller for anything void doPrint() { std::cout << "In doPrint()" << '\n'; } // This function takes one integer parameter named x // The caller will supply the value of x void printValue(int x) { std::cout << x << '\n'; } // This function has two integer parameters, one named x, and one named y // The caller will supply the value of both x and y int add(int x, int y) { return x + y; } |
An argument is a value that is passed from the caller to the function when a function call is made:
|
1 2 3 |
doPrint(); // this call has no arguments printValue(6); // 6 is the argument passed to function printValue() add(2, 3); // 2 and 3 are the arguments passed to function add() |
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:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 |
#include <iostream> // This function has two integer parameters, one named x, and one named y // The values of x and y are passed in by the caller void printValues(int x, int y) { std::cout << x << '\n'; std::cout << y << '\n'; } int main() { printValues(6, 7); // This function call has two arguments, 6 and 7 return 0; } |
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:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 |
#include <iostream> int getValueFromUser() // this function now returns an integer value { std::cout << "Enter an integer: "; int input{ 0 }; std::cin >> input; return input; // added return statement to return input back to the caller } void printDouble(int value) { std::cout << value << " doubled is: " << value * 2 << '\n'; } int main() { int num { getValueFromUser() }; printDouble(num); return 0; } |
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:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 |
#include <iostream> int getValueFromUser() // this function now returns an integer value { std::cout << "Enter an integer: "; int input{ 0 }; std::cin >> input; return input; // added return statement to return input back to the caller } void printDouble(int value) { std::cout << value << " doubled is: " << value * 2 << '\n'; } int main() { printDouble(getValueFromUser()); return 0; } |
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:
|
1 |
someFunction(a(), b()); // a() or b() may be called first |
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:
|
1 2 3 |
int avar{ a() }; // a() will always be called first int bvar{ b() }; // b() will always be called second someFunction(avar, bvar); // it doesn't matter whether avar or bvar are copied first because they are just values |
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.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
#include <iostream> // add() takes two integers as parameters, and returns the result of their sum // The values of x and y are determined by the function that calls add() int add(int x, int y) { return x + y; } // main takes no parameters int main() { std::cout << add(4, 5) << '\n'; // Arguments 4 and 5 are passed to function add() return 0; } |
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:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 |
#include <iostream> int add(int x, int y) { return x + y; } int multiply(int z, int w) { return z * w; } int main() { std::cout << add(4, 5) << '\n'; // within add() x=4, y=5, so x+y=9 std::cout << add(1 + 2, 3 * 4) << '\n'; // within add() x=3, y=12, so x+y=15 int a = 5; std::cout << add(a, a) << '\n'; // evaluates (5 + 5) std::cout << add(1, multiply(2, 3)) << '\n'; // evaluates 1 + (2 * 3) std::cout << add(1, add(2, 3)) << '\n'; // evaluates 1 + (2 + 3) return 0; } |
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:
|
1 2 |
int a{ 5 }; std::cout << add(a, a) << '\n'; // evaluates (5 + 5) |
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:
|
1 |
std::cout << add(1, multiply(2, 3)) << '\n'; // evaluates 1 + (2 * 3) |
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.
|
1 |
std::cout << add(1, add(2, 3)) << '\n'; // evaluates 1 + (2 + 3) |
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?
|
1 2 3 4 5 6 7 8 9 10 |
void multiply(int x, int y) { return x * y; } int main() { std::cout << multiply(4, 5) << '\n'; return 0; } |
Question #2
What two things are wrong with this program fragment?
|
1 2 3 4 5 6 7 8 9 10 |
int multiply(int x, int y) { int product{ x * y }; } int main() { std::cout << multiply(4) << '\n'; return 0; } |
Question #3
What value does the following program print?
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
#include <iostream> int add(int x, int y, int z) { return x + y + z; } int multiply(int x, int y) { return x * y; } int main() { std::cout << multiply(add(1, 2, 3), 4) << '\n'; return 0; } |
Question #4
Write a function called doubleNumber() that takes one integer parameter and returns twice the value passed in.
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.
 2.4 -- Introduction to local scope |
 Index |
 2.2 -- Function return values |
It seems that the code snippet above :
won't compile properly because the variable name is the same as the function name, right?
Giving global variables the same name as functions will probably result in an error.
Giving local variables the same name as functions will probably result in an compiler warning.
Best to avoid giving similar names.
I am also getting the same type of error. Why do we have that snippet in the tutorial then..??
==================================================================================
pushb@pushb-VirtualBox:~/working_dir/learncpp(master)$ g++ -o 2_3_2 2_3_2.cpp
2_3_2.cpp: In function ‘int num()’:
2_3_2.cpp:5:9: error: ‘int num()’ redeclared as different kind of symbol
int num(){
^
2_3_2.cpp:4:5: note: previous declaration ‘int num’
int num;
^~~
2_3_2.cpp: In function ‘int main()’:
2_3_2.cpp:9:12: error: ‘num’ cannot be used as a function
num = num();
^
pushb@pushb-VirtualBox:~/working_dir/learncpp(master)$ cat 2_3_2.cpp
pushb@pushb-VirtualBox:~/working_dir/learncpp(master)$
Oops. I updated the snippet to remove the conflict. Thanks for pointing this out.
First of all. Best site to learn cpp. Your doing a great job.
This is my code for the last exercise.
As you can see in my doubleNumber() func i declare a result var to store the result of the calculation. Is there any major difference if i only use return x * 2? (Seen on a larger scale it might be speed of the program but on a smaller scale like this it shouldn't give me any disadvantages, right?
* Line 5: Initialize your variables with brace initializers.
* Missing printed trailing line feed at the end of the program.
If you
You'll have to
Now you have duplicate code and a duplicate calculation. That's bad style and bad performance.
Hey guys, I'm trying to explain the pass by value process in my own words to put into anki. Would you say the following description of the process is accurate?
When a function call is made all of the parameters of the called function are created as variables, and the value of each of the arguments from the caller is copied into the matching parameters of the callee.
Correct
I noticed and followed the original code in the examples, and when answering the quiz question, I choose different variable names for defining the user's input. But I decided to try naming the variable 'x' across all functions and their calls, in the hopes of limiting the amount of variable names I declared, and it worked! Is 'x' being redefined each time I declare it, staying in the same memory location? Is this a bad practice? I realize that we can assign the same value to multiple variables, but I feel like this is wasteful. I can see situations where defining different variable names is desired, but for the scope of this program's intended use, maintaining the same variable name makes more sense, please correct/coach my idea, and a HUGE thank you for your series...
The next section answers my question! Answer to my question is: Variable 'x' is defined locally, and every function has it's own variables, which are separate and can have the same name as other variables in other functions, including main(). So although 'x' is the variable name in each function, it still has it's own memory location and they aren't related in any way.
Hi Alex!
From "A warning about function argument order of evaluation" onward, you're using copy initialization.
Updated. Thanks for pointing these out.
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.
Quite right on both counts. Thanks for pointing these out.
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?)
The extraneous "can" has been sacked. Thanks!
Hello,
The first & second program in this tutorial cause compiler error also on the line no. 14. It should be -
instead of -
And fixed. Thanks!
Thank you for great tutorials!
under more examples you have:
There shouldn't be a 6 between the 9 and 15
Fixed. Thanks!
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
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:
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
so, I ended up doing this, and it works. Why does using Y instead of x work? Sorry if this is an odd question
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.
Thanks! Clears that up!
You dont even have to have the
instead you could just type:
and it should work!
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. :)
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.
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?
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.
Thank you Sir. :)
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;
}
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
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!