Hello, and welcome to Chapter Four,

Functions, in the book Python for
Informatics.

As always, these slides and this audio and

this video are copyright Creative
Commons Attribution.

Now we are to the point, you know, Chapter
Four, we're sort of well into the class.

So I figure I should introduce myself a
little bit, let you know a little bit.

As I said before, I think in the beginning,

we're tape, I'm taping this in a

wonderful building at the University
of Michigan called the North Quad.

It's a relatively new building, 
it's got some

residential sections and some academic
sections and some

classrooms, and one of the classrooms that

I typically teach in is actually 
2255 North Quad.

It's a really beautiful room

with great ways for people to interact,
and so sometimes

I'm teaching, you know, little tiny
Dr. Chuck down here.

With a smile on the face.

And sometimes my students are
taking me on, taking my

classes on campus and sometimes students
are watching me through lecture.

And so this building is really 
beautiful and if

you ever get a chance to come to Ann Arbor

and look at it, maybe walk through it, 
it's really, it's really quite nice.

One of the things I like about it is

that I think it's really highly 
inspired by Harry Potter.

The kind of, of course, Oxford and
Cambridge are

the real inspiration for Harry Potter, 
but our, our cafeteria,

for example, it kind of looks 
like the four tables

in Hogwarts and you can kind of 
imagine a snowy owl

flying around and a Sorting Hat at the,
at the front sorting people.

And so the nickname, the nickname for the
place is Quadwarts, because it's North Quad

Quadwarts, that's like Hogwarts and
North Quad kind of jammed together.

And of course, given that we sort of think

of ourselves a little bit as Harry Potter,

people when they first come in September
often sort of decide

to sort themselves and a few years back

when we first started the building
the students

decided that I did not get to 
be in Gryffindor.

As a matter of fact, it's probably 
time for me to to

show you who I am, and who 
I've been sorted to be.

So the students decided that 
I couldn't be in Griffindor.

That I had to be in Slitherin,

and that's because of
my name, Charles Severance.

And sever is snake.

What's even cooler, of course,

is given that I teach Python, 
Slitherin's house is

a snake, right?
So it makes a lot of sense, I even have

this really fancy Slitherin teacup that I
use to drink tea during lectures.

Sometimes I drink coffee, and

sometimes I drink tea.
Oh wow, this thing itches.

So let me just get rid of it.

If I had any hair, that would 
mess my hair up.

So let me get rid of this for the 
rest of the lecture.

So there I am, okay.
Enough of that, back to,

back to Dr. Chuck.
[CLAP]

So, with that sort of brief, brief
interlude, the topic of

the actual topic of this lecture is
Functions. And so storing and

reusing is basically an idea that

we will often have a series of 
steps that we will

want to use over and over in a program,
increasingly complex.

The things that we'll use in this lecture
are kind of silly because

I have to keep them short so the slides
don't get too long.

But a good example of, you know, the kind
of work is maybe I am going to use

Google's geocoding service and I am
going to send some unstructured

data and get a GPS coordinate back.

And that's a service that I want 
to call and it would

maybe be about this much lines of, this
many lines of code.

And I'm going to want to do that 
all over the place.

So do I want to put this many lines
of code 40 places in my program?

Or do I want to put it one place and

then call it in the various places 
that I need it?

And so that's why I call it the store and
the reuse function.

So if we take a look at the 
simple syntax here.

These things are called functions and in some
languages it's called subprograms,

but we call them functions in Python.
And the keyword

that we're really going to focus on is 
def, which stands for

define, and what happens here is 
when Python sees this def keyword

it actually doesn't run the code.

It says, oh, you're going to make a
function and you're

going to kind of turn on a recorder and
start recording this code.

So it has a colon at the end of it.

So it has an indented block afterwards.

And so the indented block becomes
recorded.

So instead of running the code, 
like if we just

put print hello and print fun, 
it would run it.

Instead it says, hey don't run it right
now, name it hello.

We give it a name.

It's kind of like a variable.
We choose the name.

We've chosen hello as the name of this.

Define it as hello, have it have these two

lines of Python in it, and we'll 
use it later, okay?

And so that's the function
definition.

That's the store phase.

That is it's sort of like, it doesn't
really run those lines.

It sort of makes a variable called hello

that actually contains Python code

rather than containing like 12 or a 
string or something like that

that we've worked with before.

So this is the store part, and then the 
reuse part is, we then have extended Python.

We now can call our bit of code. So we say,
hello, hello name is what we came up with,

parentheses, and then that says,
remember that code that I

put in there under the name hello?

Run it now and, so, so, so if I

start looking at that and then 
it just continues.

So let me kind of clear this 
and start over again.

And, so if I watch what Python does from
the beginning as

it reads here and it goes oh, you're
defining a function named hello.

Great, I will sort of remember, remember.

I got that remembered for you, let's
continue on.

Oh, hello!

You want me to run that stuff that

you just got done storing under 
the name hello.

So then it kind of goes and runs it and
out comes Hello Fun.

Then after that it runs to the print 
and out comes print Zip.

And we say, you know what?

I want to reuse that again.

I stored it once, I can reuse it as many
times as I want.

And now, hello, and then these 
two lines of code run a second time.

So we stored them once, gave them 
a name, and then

ran them twice, in the context of 
wherever it is we wanted.

Now this is not sort of a profound

a profound reason to use it in this.

I'm just trying to give you 
the notion that there's a

way to store and name code that 
then you can retrieve later.

That's really what's going on here.

There's two kind of functions 
inside of Python, and

we've actually been using them almost from
the very first lecture.

And that is there are built-in functions
that Python provides

to us like float, raw_input, int, 
those kinds of functions.

Those are just part of Python but 
we call them as functions.

The difference is we don't write them.

And then there's user-defined functions,
functions that we write.

Functions that create functionality 
that we want to make use of.

Like encapsulating the ability to compute
pay for time and a half for overtime.

And so, we name these things and we

treat them as new reserved words that
we've created.

They're kind of an extension 
to the language

as it were.

So when we're coming along, we define a
function with the def keyword, right?

The def keyword

is a reserved word. It's one of the many
reserved words back in Chapter One

that we talked about, and it indicates to
Python the beginning of a function.

We define it, and then when we call it,
which is called invoking,

it's like, we're building it,
and then we're invoking it.

And you can build it once, and then invoke
it many, many times.

So for example, here is a built-in
function called max that

finds the largest character, the sort

of lexicographically largest character,
in a string.

And so it's like okay, tell me 
the maximum character.

And so max is not some code that we've

written but we are invoking a 
function here and we're

passing in an argument to that.

So the argument is this stuff 
in between the parentheses.

So the max function can find the maximum
of many different things.

At this moment we want it to find the
maximum of that particular string.

The highest character

in that particular string.
So, this is a left, a right-hand side of

an assignment statement too.
So that has to be evaluated to a value.

So it goes into the function,

does whatever things the function 
wants to do,

and then the function 
gives us back a value

that becomes the value for 
max parentheses Hello world.

And that value in this case is 
the letter w, okay?

Because the letter w was 
decided to be

the highest letter and that's 
what max gives us back.

And then when we're done with that,
then that w ends up being assigned.

The assignment statement completes.

And so you can think of the 
function evaluation

as happening as part of the right-hand
side expression calculation.

There could a plus here and other stuff

and it's just, at some point, 
a big expression.

And this one, it's a simple expression
with just one function call.

Now if we look at this, there is some
code somewhere.

Somebody wrote some code.

It's part of Python.

You didn't write it.
There's a max function somewhere.

And you can think of a function 
as having some input.

It's kind of like a program, that's why

some languages call these things
subprograms.

Because they have an input, they do some
kind of useful works, whatever

that useful work happens to be, and then
they produce some kind of an output.

Right? So Hello world

is the input, a string, the
arguments, the thing we're passing in.

Hello world is what's being
passed in to the function.

The function is running and then something
comes back and is sent back.

So it has input, processing, and output.

Input, processing, and output. 
So that's how a function.

Some stored code, whether we 
wrote it or not,

they work the same when we call

functions, right?

So you could think of this as, somewhere
inside of the Python library

is some code that maybe has a little 
def in there, and the name,

they named the function max, and it 
takes a single parameter.

And it does some blah, blah, blah, blah,

loopy blah, blah stuff, whatever 
max wants to do.

Whatever we need max to do based on

the specifications that max is 
supposed to support.

But somewhere

there is code inside of Python

that actually represents the 
function definition.

It's a built-in function because it 
comes with Python

and we didn't have to do 
anything to add it.

So some common built-in functions 
that we have

been using all along, good examples 
are the float, which

takes as input anything and

returns you a floating
point number version of that.

Type, which takes a parameter 
of a variable or a constant

and says, what is the type of this.

Float, again converting.

Type, again, and float.

So these are all things that, we've been
calling functions all along.

And it passes the input value 
into the function, the

function runs and then gives
us back a return value

which then participates in the rest of the
expression on the right-hand side.

You can think of it's pausing the calculation

on the right-hand side, calling the function,

getting the result of the function back,

and then continuing the evaluation of the
right-hand side.

Then coming up with whatever value and
then printing that value out.

Okay?
Another thing that we've done

is we've done string conversions.
Right?

So we've converted, in this case, 
a string to an integer.

And asked what type it is.

We've converted a string to an integer.
So, int converts its argument,

whatever that happens to be, 
into an integer.

So that's just some of the built-in

functions that we have talked about
so far.

Now, this becomes more interesting

when we can make our own, own functions.

Oops, there goes my teabag right in the
middle of the thing.

Got to take the teabag out, I think it's,
whoa, hang on, be right back.

Teabag, okay, there's my tea.
So, so we want to make a new function.

Like I said in the other example,
we use the def keyword,

the def keyword here, and then we have
some indented bit.

We create a name for it and then 
have some parentheses.

These parentheses will later tell 
the inputs

that we're going to pass in, but this

function has no input, so we just go 
parenthesis, parenthesis and then the

all-important colon character which 
indicates the beginning of an

indented block of Python, that then is, 
this detects the function.

So, it's important to remember

that while this is executing
when Python first

looks at this, it doesn't run 
these lines of code.

It just remembers them, and names them
print lyrics.

So it doesn't cause any printout, it just
causes Python to remember.

I probably said that a few too many times.
So, so here is

a difficult problem, and I'll let you
think about

it for a while.

I want you to kind of mentally go through
and execute this code.

And ask what, ask yourself what the output
of this program would produce.

How many lines?

How many lines of output would this
program produce?

So, how many of you said three?

How many of you said five?
Well, the right answer

is actually three. You see five 
print statements, two, three,

four, five, but two of the print statements
are sitting inside of this.

And we never called, we never invoked a
function down here, okay?

So, this one,

let's clear this.
This one prints, these two get skipped,

this one prints, and this one prints.

So that that's why there are three
statements that print.

There is stored, but we never used, a
function called print lyrics.

And it's got two statements in it, but we
never used it.

So the output of this is 
Hello Yo 7, and that's because we

never actually invoked it.

We had to say print lyrics parentheses
or whatever to cause it to call this.

Okay? That's just to emphasize that as it

looks at it, it does not execute 
these lines.

So once we've defined a function, once we
have given it a name, given it

code that is a part of it, then we can invoke
it or call it as many times as we like.

So now,

our little example works a little better
if we actually call our function.

Python really doesn't care if you 
don't call your function.

It's like you told me to 
make one, I made one.

You didn't use it.
There you go.

But if you look at this one now, 
so here we go, x equals 5.

print Hello, out comes Hello.

Define.

Nothing happens here.
Nothing happens here.

It's just remembering.

Okay? Then it says, print Yo.

Then it calls the function print lyrics,
which sort of stops us here,

runs these two lines of code.
So out comes that and that.

Then it sort of finishes this and it comes
back, x equals x plus 2, then it prints x.

That must mean that x is 7, 
and so out that comes.

And so, again, it's on the 
first time through.

No, go back, go back, go back.

On the first time through, it 
doesn't print.

But then when it hits this,

it prints.

You could say print lyrics 
several more times and

it would run this as many times as it did,

and it needed to, as many times as you
want, and it would make output for you.

So you can invoke this is the definition,
let's clear this.

This is the definition.
This is the call or invoke.

So we are invoking the function, 
we're calling the function,

we're causing the function to execute.

Here we're just causing the function 
to be looked at and defined,

but not actually executed. 
Hope that's clear.

Now, when we pass data into a 
function, and functions

that don't take data are not as
useful as they could be.

There are plenty of things that do, times

that you build a function doesn't take
data.

But the most interesting functions 
are the ones

that you could hand them something to
work on and

they could do their work and then come
back with whatever.

So this max function is a good example of
this, one that's taking an argument.

We call the things in between the
parentheses

when we're invoking the function,

we call the things in between the
parentheses arguments, okay?

So that's passing into the function.

Feeding data into the function.
So we put arguments in between them.

So for example, here we have 
a little program.

That

that is, it's a function named greet, and
now we are

going to define this function and we're
going to say, you know what?

I would like to take a parameter, let's
take a parameter.

Let's have one parameter come in.

And we need kind of a placeholder for that

parameter, so within the function we're
going to use lang.

Now this isn't actually a real variable.

It's kind of like a, it's a placeholder
variable.

So this first parameter, whatever it is,

when it's called, is going to be lang.

And so if that first parameter is 
equal to es,

we're going to print Hola.

And else if it's equal to fr
we'll print Bonjour.

And otherwise, we'll print Hello.

So there's apparently three languages in
the world, Spanish, French, and English.

And if it's not Spanish or French, then it
must be English.

But, I, you have to keep this kind of
small, so my screen doesn't get too big.

So this is again just the definition and
if you type this

into the interactive thing it gives you
this dot dot dot prompt.

And so we now have this thing called greet

and now we've extended Python to add 
our own function to Python.

And now we can say greet en and so 
it runs this code

except that en is lang and so that comes,
and then it prints Hello.

So out comes Hello.
Now later we can say, oh,

I would to do a greeting, but this time
I'm going to pass es in as it.

So lang becomes, for this execution, es.
And then, so it prints out Hola.

And then the next execution, lang is fr.
So it executes this three times but

lang is different each time because we've
passed in different parameters each time.

So that's how we can kind of write 
general-purpose code

inside the function and then reuse that
general-purpose code in different ways.

Okay? It's a real powerful, powerful
mechanism

that makes functions far more useful.

Now, functions don't necessarily 
just have to do stuff.

A real powerful mechanism in a function is
what we call a return value.

So a function can take its 
arguments, do some work.

We've seen that.

And then it can return a value. And the key
to the return value is, when we call the

function, like we were calling max,

it gives us back some value like the 
little w, okay?

So here we're going to make a function
called greet that takes no parameters.

Doesn't take parameters, but it has 
another keyword.

It's another reserved word in Python

and whatever we put on this

return statement shows up as the
replacement in this expression.

So, whatever greet is, it runs greet and

then the return is kind of a 
residual value.

So if we say print greet, comma, Glenn.

it says Hello Glenn, because 
the return value

for the greet function 
is the string Hello.

And if we say greet Sally,

it says Hello Sally.

And so, and it's run the code twice and
the return function, return value has

been put in here instead. And so the Hello
came there and the Hello came there,

so we get the two lines.

So, return is a statement that both
terminates the

execution of the function and 
defines the value of

what will be replaced when 
the function call comes back,

in the line that the function was
called from.

So here is a little smarter version of
our greet function.

It's very similar, it's called greet
still, takes lang as a parameter.

And if the language is es, then 
it returns the string Hola.

If the language is French, 
it returns Bonjour.

Otherwise, it returns Hello.

So we're not actually doing 
the print, if you

go back on the other slides, 
we were printing.

But now we're just returning a string.
Okay?

And so now, I can call print greet, 
and pass en in,

so then that runs the code once, 
with lang equal to en.

And I get back Hello, and then
comma, Glenn.

Then I call it again and I pass es in.
And then

that time it returns, the return value
here becomes Hola, a string Hola.

Hola Sally. And then Michael, 
I'll pass in one more time.

Lang is now fr,

the string fr, and so it returns a
Bonjour and

so the, the residual that is here is
Bonjour, and so out comes Bonjour Michael.

So there is lot to this, right?

You're passing stuff in.

You have this kind of placeholder
variable.

And you have this return that sort of
appears where it was called from.

This goes in, does its work, it comes back,
and there's sort of the

residual value that sits here.

You don't have to have a return in a
function, but if you want to

do something with the value, then you have
to have a return in the function.

We call the functions that produce values

fruitful, and the other ones 
are called void.

[LAUGH] So, that's a good name for them.

So, to review sort of this, arguments,

parameters, and results, if we 
look at max,

the original thing, where it's looking for
the largest, lexicographically

largest letter, it looks Hello world is
the argument that's passed in.

We have this sort of formal parameter here
called inp which is not really a

variable, it just happens to refer to
whatever is the

first argument in any particular call.

And then it does its little thing and
runs loops and does all these things and

at some point it returns w, so that the
thing that comes out when the function

quits that becomes the replacement value
here is a lowercase w string.

And then that is the w that 
goes over in the big.

So the return is what defines what comes
back here.

Because you think of this as, it's looking
at this. It suspends for the moment.

It runs this code.

It's holding.

It's holding itself here.

It's running this code and then 
it comes back to here.

Okay?

And the return value is what defines
coming back.

So, of course you can have more than one
parameter and they are in order.

So here we have an a and a b.

These, the name of these things doesn't
really matter.

They're just relevant inside of the
function definition.

So we are going to add 
two numbers together

by taking a + b, and then 
returning the sum.

The added variable is just kind of local
to this function.

And now we can say, you know, 
addtwo, 3, comma, 5.

And then this will come back as 8, 
and then 8 will get assigned

into x, and so that will 
print out 8.

And so you can have as many of 
these as you want

and the order matters and there is a
one-to-one correspondence.

3 goes to a and 5 goes to b

when the thing is called.

And then the return value again 
comes back.

Okay?
So that's, sort of, arguments.

And like I said, not all functions 
have to return values.

We call them void functions when they 
don't return anything.

It's totally fine for that to be the case.

So at this point you might be thinking
to yourself, okay, great, well I still

don't quite get why to use functions.

And in reality, in the first 10, 11
chapters of this book, other than

using lots of functions, we're not
really going to spend a lot

of time making functions because most of
our programs are going to kind of be

that long and we're not going to 
do a lot reuse in the program.

And there'll be a time when your 
programs become complex enough.

You'll be like, oh, thank heaven for
functions.

I think it's premature to say you 
must use functions, even though

there are some exercises that just say,
hey, do this with a function.

Just so you kind of get the 
understanding of a function.

You will find soon enough,

as your programs grow, you'll go like, oh,

I keep doing this same thing 
over and over again.

Let me pull it up into a function and pass

a parameter in, have a return value, and
away you go.

Or, you might find that you're moving from

one program to another and you have this
common thing

that you want to do, so you make yourself
a library that you drag along.

And we will do lots of libraries.

The book in the second half does 
lots and lots of library

stuff, doing things like parsing XML and
this, that, and the other thing.

So, so don't feel like you need to use
functions on every

assignment, because they're a natural
thing when a program gets big enough.

So, so just kind of understand them on a

mechanical level, but

it'll come to you at the right time

when it's time to start building 
your own functions.

So in this class, we kind of, you know,
talked about functions.

Just got you started.

Talked about parameters, talked about
built-in functions,

talked about return values, the 
store and reuse pattern.

So, the problems at the end of the

chapter for this particular chapter 
are relatively straightforward,

in that, like I said, we don't have
a real strong

need to do functions yet in this class
because the programs aren't large enough.

But I just said okay, take, take one

of your previous assignments and 
refactor the code

so that at the top there is a def 
computepay, and you put like the if

and whatever in here, and then later on

you do your code and then you call 
computepay.

So you took code that you already had, you

move it up into a function, and 
make a function.

And I've also online got sort of a

sample of this, because it's, it's 
a little complex.

And so you should be able to find, on
Python Learn or on the course site,

you should be able to find a good example,

because I really want you to 
sort of get this.

Like I said, there will 
come a time

when functions will make the most 
sense to you.

But coming up next, of course, is
Chapter Five, and that's loops.

And loops are going to rock the house,
And so we really, that's our

fourth major pattern is loops and 
and I'm looking forward to it.

So we'll we'll see you at the next lecture.