Notes5

WEEK
FIVE
Writing your own functions
Well done for making it to the final week of the challenge!
This week we have got some really interesting questions lined up for you.
You have been calling functions for several weeks now, either builtin functions like raw_input, functions from
modules like random.shuffle, or methods like reverse for lists.
Now that your programs are becoming more complicated it is time for us to show you how to write your own
functions. This will allow you to make your programs simpler, and therefore easier to understand and debug.
1 Functions
A function is an independent, named chunk of code that performs a specific operation. It can be run or called
by referring to it by name with any information called arguments it needs to do its job. By now you will have
had lots of experience calling builtin functions like raw_input or int. Both raw_input and int take a single
argument. For raw_input this is the message to display to the user, e.g. ’Enter a number: ’, and for int
it is the value to be converted into an integer, e.g. ’10’.
Different languages (and textbooks) sometimes have different terms for functions. They may also be called
routines, subroutines, subprograms, procedures, messages or methods. Many languages differentiate procedures
from functions by saying that functions return a value but procedures don’t. Since languages like C/C++ and Java
have a void or empty return type (equivalent to None in Python) the distinction no longer makes sense.
Using functions in Python is easy — so easy that we have been using them without really bothering to properly
introduce function calls. On the whole, defining your own functions in Python is straightforward. However, there
are a number of tricks that can catch you out occasionally. We mention some of those things below.
The clever thing about functions is that the same bit of code can be called again and again, rather than having to
write it out multiple times. If we need to make a change to that code we only have to do it in one place rather than
each copy. Also, we can use functions to hide complicated code so we don’t have to think about it so carefully
when we use it.
2 Function Calls
You have already an old hand at calling functions, but let’s recap quickly to make sure you understand it all.
To call a function you need to how many arguments the function is expecting and what type of values they are.
For instance, the abs builtin function takes one argument which must be an integer, float (or complex number).
Hopefully, if the function’s purpose is well defined it will be obvious what information the function needs and
why. That is, unless you have a number, then it is meaningless to ask about its absolute value!
The examples below show what happens when you give abs the wrong number or type of arguments:
c National Computer Science School 2005-2009 1

NCSS Challenge (Beginners) WEEK FIVE
>>> abs(1, 3)
Traceback (most recent call last):
File "<pyshell#0>", line 1, in -toplevel-
abs(1, 3)
TypeError: abs() takes exactly one argument (2 given)
>>> abs()
abs()
TypeError: abs() takes exactly one argument (0 given)
>>> abs(’a’)
abs(’a’)
TypeError: bad operand type for abs()
A function call consists of the function name, which may need to be preceded (or qualified by the module name,
followed by the arguments surrounded by parentheses. A common mistake is to forget to put the brackets when
the function takes no arguments.
>>> abs
<built-in function abs>
Unfortunately, abs without the parentheses is not an error since abs is a variable which holds the actual code
that performs the absolute value. Typing abs, is asking to see the contents of the variable, not call the function.
3 Creating your own functions
Functions in Python are created using a definition statement. The def creates a function from the indented
function body and assigns it to the name of the function.
>>> def hello(name):
... print ’Hello’, name
...
>>> hello(’James’)
Hello James
A definition statement consists of the keyword def, followed by the function name (here hello), and then a list
of zero or more arguments in parentheses (here just one, name). The parentheses must be present even when the
function takes no arguments. The def line must end with a colon (like control statements). The code block of
the function must be indented as in control statements.
The def statement is executed in the same order as other statements which means your program must reach the
def statement before it tries to call the function:
hello(’James’)
def hello(name):
print ’Hello’, name
When you run this, Python spits out the following error:
File ‘‘example.py’’, line 3, in ?
hello(’James’)
NameError: name ’hello’ is not defined

This is because Python attempts to run the call to hello before the function itself actually exists, which is why
the name hello is undeﬁned.
Here is a function that performs factorial, that is, for a number n it calculates n x n - 1 x ... x 1, so 5 factorial,
written 5!, is 5 x 4 x 3 x 2 x 1 = 120. The Python version is:
>>> def fact(n):
... result = 1
... while n > 1:
... result *= n
... n = n - 1
... return result
...
>>>
Running this code with 3 calls to fact gives:
>>> print fact(5), fact(10), fact(20)
120 3628800 2432902008176640000
As you can see those factorial numbers get very large very quickly!
4 return Statements
As we have seen, some functions (such as fact above) need to give the result of their execution back to the
caller. This is done using a return statement.
>>> def add(x, y):
... return x + y
...
>>> r = add(5, 10)
>>> r
15
A return statement immediately terminates the running function and control is passed back to the caller, that
is, the line of code that called the function continues. If return is given an argument, then it is returned to the
caller and can be stored in a variable, used in an expression, or discarded.
If return is not given a value, then no value is returned to the caller (None is returned instead). This is the same
as a function which has ﬁnished executing and does not have a return statement. The hello function above is
an example without a return. Because a return statement terminates the execution immediately, it is often used to
exit from a function before the last line of code.
>>> def hi():
... print ’hello’
... return
... print ’there’
...
>>> hi()
hello
In this example, print ’there’ will never get called because the return statement, will cause the function to
exit before the second print is reached.
As in the above examples, when a function needs to return an argument, the return statement is followed by
a single expression which is evaluated and returned to the caller. A return statement can appear anywhere,
including inside if, for and while statements, in which case the function immediately returns.

>>> def day(name):
... if name in [’Saturday’, ’Sunday’]:
... return ’weekend’
... elif name in [’Monday’, ’Tuesday’, ’Wednesday’, ’Thursday’, ’Friday’]:
... return ’weekday’
... else:
... return ’not a day’
>>> day(’Monday’)
’weekday’
>>> day(’Sunday’)
’weekend’
In this function there are three separate return statements that correspond to the three different conditions that
this function handles.
5 Arguments
Communicating information to a function can occur in two ways: Either the function declares a number of
arguments (or parameters) which must be passed to it when the function is called, or the function uses variables
that are defined in the main program.
You have probably guessed from the examples above how arguments can be used because it’s kind of hard to
do interesting examples without them. That’s not surprising since arguments provide the information functions
need to do their work.
Adding arguments to a function is as simple as adding variable names between the parentheses in the def
statement. Then when the function is called, the value you want that argument to have in that call will need
to be added between the parentheses too. The names of the variables inside (the function definition) and outside
(in the function call) need not be the same. For this week’s problems it is relatively easy since our checking code
will tell you if you have defined your functions incorrectly.
Arguments are the trickiest part of creating functions because they don’t quite behave how you might expect. The
biggest surprise is that arguments (and variables) that are declared as part of a function are not visible outside
that function (in the main program or other functions):
>>> def hello(name):
... print ’Hello’, name
...
>>> hello(’James’)
Hello James
>>> name
NameError: name ’name’ is not defined
Another complication is that you can’t change values that are passed into a function:
>>> def addone(x):
... x += 1
...
You might expect that if we call this function on a variable, say y, then the value of y would be incremented by
one after the function call:
>>> y = 5
>>> addone(y)
>>> y
5
>>>

Outside the function nothing has changed, but x inside the function would have been incremented by one. This is
because Python functions make a copy of their arguments (called pass by value) rather than access the original
variables.
So in the case above, a new local variable x has been creating containing a copy of the contents of the variable y
that was passed in as an argument during the function call. When we modify x inside the function we are only
modifying that local variable.
What this means is that if we want to modify a variable we need to pass it in to the function and then return it,
like this:
>>> def addone(x):
... return x + 1
...
>>> y = 5
>>> y = addone(5)
>>> y
6
>>>
There is one final complication with passing by value, and that is that Python data structures are stored as
references. We don’t want to go into the details of what that means now, but the gotcha is that they can be
modified (but not assigned) inside a function call:
>>> def appendone(x):
... x.append(1)
...
>>> y = []
>>> appendone(y)
>>> y
[1]
>>> appendone(y)
>>> y
[1, 1]
>>>
6 Testing functions in IDLE
In the examples above we have developed our functions in the Python shell rather than in a separate file IDLE.
However, you normally develop programs in separate Python files and then run them. So how do you do that for
your own functions?
Say you have just written your own function in a separate file in IDLE and you want to test it. For example:
def greet(name):
return ’Hello ’ + name
You can test this in the interactive interpreter in the same way by pressing F5 (or selecting Run Module from
the Run menu). But when you do this nothing seems to happen because you have no code apart from the def
statement in your file, and the def statement doesn’t produce any output on the Python shell:
>>> ================================ RESTART ================================
>>>
What you need to do then is call your function manually:
>>> greet(’James’)
’Hello James’
>>>

7 More dictionary methods
Sometimes you’ll need to either retrieve the value for a particular key from the dictionary, or use a default value
if the key doesn’t exist. This typically takes 3 lines of Python:
>>> val = ’default’
>>> if key in d:
... val = d[key]
>>>
However, getting a key-value pair or default is so common, dictionaries provide the get method to simplify this:
>>> val = d.get(key, ’default’)
The ﬁrst argument to get is the key. If the key is in the dictionary then the corresponding value is returned,
otherwise the second argument to get (here ’default’) is returned.
If you want to copy the key-value pairs from one dictionary to another, Python provides the update method:
>>> d1 = {’a’: 1, ’b’: 5}
>>> d2 = {’a’: 4, ’c’: 6}
>>> d1.update(d2)
>>> d1
{’a’: 4, ’c’: 6, ’b’: 5}
Notice that not only does the new key ’c’ have the value 5 from d2, but the value for ’a’ has also been updated
to the value from d2.
8 Miscellaneous string tricks
To make a couple of the questions a bit easier, here are a last couple of miscellaneous tricks:
Firstly, the string module has some useful constants:
>>> import string
>>> string.lowercase
’abcdefghijklmnopqrstuvwxyz’
>>> string.punctuation
’!"#$%&’()*+,-./:;<=>?%%[]^_‘{|}~’
>>>
If you need to strip off all punctuation, you can now use:
>>> x = "’Hello!’"
>>> x.strip(string.punctuation)
’Hello’
>>>

Notes5

Recommended

More Related Content

What's hot (19)

Viewers also liked (7)

Similar to Notes5 (20)

More from hccit (20)

Recently uploaded (20)

Notes5