Programmers like to write code in functions and procedures. They like to do this for all kinds of reasons. for example, they can put them into libraries and re-use them in other programs and save time, and they are easier to write, de-bug, understand and maintain because they tend to do just one job so aren't complicated. If you are going to write code using self-contained functions and procedures, then you need to be able to pass data to them each time you want to run them.
From now on, when we talk about a function, we will assume we are also talking about a procedure.
A piece of data that you pass to a function (or procedure) is called a parameter. A function expects you to give it the exact number of data items it needs, and these are defined in a function's definition. You can pass as many parameters as you like to a function but they must match what the function said it needs. Let's look at an example. Type the following code in and get it working.
#Calculate the area of a triangle from the base and height
def area_bh(base, height): # This function has two parameters.
'''Calculates the area of a triange from the base and height. Requires two float arguments to work and returns a single float.'''
area = 0.5 * base * height
return area #This is the answer 'returned' from the function
myBase = float(input('Please enter the base >>> ')) #get the base
myHeight = float(input('Please enter the height >>> ')) #get the height
triangle = area_bh(myBase, myHeight) #call the function area_bh and pass it two arguments.
print('The area of a triangle with base',myBase,'and height',myHeight,'is',triangle,'units squared.')
The above code uses a function to calculate the area of a triangle. In the definition of the function, we used the line:
def area_bh(base, height):
This says that you must pass this function two parameters for it to work. The first parameter you pass will be assigned to base and the second one will be assigned to height. If you don't give the function exactly two parameters, you will get an error. Try adding an extra line to the end of the program:
print (area_bh(12, 45, 23))
What error messages did you get?
Passing parameters by reference to it (REF))
Some programming languages allow you to pass parameters by reference to the original one. If you pass a parameter from your main program to a function by reference, that means that you are actually passing a pointer to the function that points to the original parameter, the original piece of data. If you make any changes to the parameter inside the function, then the original parameter will be changed as well. Consider the following pseudo-code example:
myData = 20
function addNumber (byRef num as integer)
print (num + 1)
Firstly, we print myData, which prints 20. Then we call the function and we pass it 20 by reference. Inside the function, we add 1 to it and print 21. After we exit the function, we print myData again and we see that it also prints 21. That's because when we passed myData to the function, we passed it by reference. The a pointer to the data was given to num and any changes that happended to num were also made to myData.
Passing parameters by value (VAL)
Some programming languages allow you to pass parameters by value. If you pass a parameter from your main program to a function by value, that means that you are actually taking a copy of the parameter first, and then passing that copy to the function. If you make any changes to the copy of the parameter inside the function, it doesn't change the data held in the original parameter. consider the example we just looked at, but this time, we will pass the data by value.
myData = 20
function addNumber (byVal num as integer)
print (num + 1)
Firstly, we print myData, which prints 20. Then we call the function and we pass it 20 by value. Inside the function, we add 1 to it and print 21. After we exit the function, we print myData again and we see that it prints 20 again. That's because when we passed myData to the function, we passed it by value. We passed a copy of the data in myData to num so any changes that happended to num were now not made to myData.
Look at this example:
data1 = 5
data2 = 20
data3 = 30
function mult(byXXX num1 as integer, byXXX num2 as integer, byXXX num3 as integer)
num1 = num1 * num2 * num3
print (data1, data2, data3)
mult(data1, data2, data3)
print (data1, data2, data3)
a) What is the output when XXX is Ref?
b) What is the output when XXX is Val?
In the first example, we passed 5, 10 and 30 by reference. That means we are passing pointers to the original data items. Any changes we make to these parameters inside the function are also made to the original data items. Inside the function, we assigned num1 to 3000 (5 * 20 * 30). That means that data1 also got changed to 3000. The output is therefore 5, 20, 30 , 3000, 3000, 20, 30
In the second example example, we passed 5, 20 and 30 to the function by value. That means that any changes made inside the function do not change the original data items. Although num1 was changed inside the function, it did not change the original data items. The output was therefore 5, 20 and 30, 3000, 5, 20, 30.
Passing data by reference and value in Python and other programming languages
In some programming languages like BASIC, Pascal and C++, there are specific instructions that allow you to pass values to functions either by reference or by value. In other languages, it doesn't really apply. In Python, for example, the distinction between reference and value parameter passing doesn't apply. In Python, variables are just names for values and if you want to give a value two names, you can do that in Python. When you call a function in Python, another name for the same value is passed to the function. It's not the same as pass by value because a copy isn't made of the original data. Neither is it using 'pass by reference' because a pointer to the original data isn't being passed to the function.
For example, consider this Python code. Type it in and get it working, but before you do, can you predict the output?
eggs = [2, 3]
ham = 
If Python used pass by value, eggs would have been passed a copy of ham, which is . What ever happened to eggs, wouldn't change ham, so when you printed ham out, it should print .
If Python passed parameters by reference, it would have been passed a pointer to ham. Any changes to eggs would then also change ham. It should have printed out [2, 3].
In fact, eggs simply became a second name pointing at the same data ham is pointing to, like this:
When the value is mutated, that value is seen by ham. However, when the program doesn't mutate but rebinds eggs to a different value, ham is still pointing to the original data.
What gets printed out is [0, 1]. This is quite a difficult idea to get your head around to start with, but a visualiser can help here. Copy or type the code into www.pythontutor.com and run it. You will be able to see exactly how the Python program (and any other program you are working with) is working. There are similar issues with other languages. Java essential passes parameters by value using native and class types whereas Ruby will call what Java does 'call by reference'! This can all get very confusing and is beyond the scope of A Level Computer Science. If you stick to the classic definitions of passing by reference or value, and then take note of how the languages you are learning works, you won't go far wrong!