1.12.

Defining Functions Problem Solving with Algorithms and

6/11/17,
Dat... 2:47 PM

1.12. Defining Functions

The earlier example of procedural abstraction called upon a Python f unction called sqrt f rom the math
module to compute the square root. In general, w e can hide the details of any computation by def ining a
f unction. A f unction def inition requires a name, a group of parameters, and a body. It may also explicitly
return a value. For example, the simple f unction def ined below returns the square of the value you
pass into it.

>>> def square(n):

... return n**2
...
>>> square(3)
9
>>> square(square(3))
81
>>>

The syntax f or this f unction def inition includes the name, square , and a parenthesized list of f ormal
parameters. For this f unction, n is the only f ormal parameter, w hich suggests that square needs only
one piece of data to do its w ork. The details, hidden inside the box, simply compute the result of n**2
and return it. We can invoke or call the square f unction by asking the Python environment to evaluate it,
passing an actual parameter value, in this case, 3 . Note that the call to square returns an integer that
can in turn be passed to another invocation.

We could implement our ow n square root f unction by using a w ell-know n technique called New tons
Method. New tons Method f or approximating square roots perf orms an iterative computation that
1 n
converges on the correct value. The equation newguess = 2 (oldguess + oldguess ) takes a value
n and repeatedly guesses the square root by making each newguess the oldguess in the subsequent
n
iteration. The initial guess used here is 2 . Listing 1 show s a f unction def inition that accepts a value n
and returns the square root of n af ter making 20 guesses. Again, the details of New tons Method are
hidden inside the f unction def inition and the user does not have to know anything about the
implementation to use the f unction f or its intended purpose. Listing 1 also show s the use of the #
character as a comment marker. Any characters that f ollow the # on a line are ignored.

Lis ting 1

def squareroot(n):
root = n/2 #initial guess will be 1/2 of n
for k in range(20):
root = (1/2)*(root + (n / root))

return root

1 of 2
1.12. Defining Functions Problem Solving with Algorithms and
6/11/17,
Dat... 2:47 PM

>>>squareroot(9)
3.0
>>>squareroot(4563)
67.549981495186216
>>>

Se lf Che ck
Heres a self check that really covers everything so f ar. You may have heard of the inf inite
monkey theorem? The theorem states that a monkey hitting keys at random on a typew riter
keyboard f or an inf inite amount of time w ill almost surely type a given text, such as the
complete w orks of William Shakespeare. Well, suppose w e replace a monkey w ith a Python
f unction. How long do you think it w ould take f or a Python f unction to generate just one
sentence of Shakespeare? The sentence w ell shoot f or is: methinks it is like a w easel

Youre not going to w ant to run this one in the brow ser, so f ire up your f avorite Python IDE. The
w ay w ell simulate this is to w rite a f unction that generates a string that is 27 characters long
by choosing random letters f rom the 26 letters in the alphabet plus the space. Well w rite
another f unction that w ill score each generated string by comparing the randomly generated
string to the goal.

A third f unction w ill repeatedly call generate and score, then if 100% of the letters are correct
w e are done. If the letters are not correct then w e w ill generate a w hole new string.To make it
easier to f ollow your programs progress this third f unction should print out the best string
generated so f ar and its score every 1000 tries.

Se lf Che ck Challe nge

See if you can improve upon the program in the self check by keeping letters that are correct
and only modif ying one character in the best string so f ar. This is a type of algorithm in the
class of hill climbing algorithms, that is w e only keep the result if it is better than the previous
one.

2 of 2