content/blog/2008/08/beauty-in-computer-science-recursion.html @ ee49211d75fb

{% extends "_post.html" %}

{% hyde
    title: "Beauty in Computer Science"
    snip: "Why I love what I do."
    created: 2008-08-29 15:30:38
    flattr: true
%}


{% block article %}

When I went to college, I majored in Computer Science. I haven't really
written anything about this part of my life yet, so I figured this might be a
good time to start.

I decided to major in CS when I was in high school. I learned to program on my
own and enjoyed the challenge of it. I also knew that programming jobs are
fairly common and pay well enough that I wouldn't have to worrying about
paying my rent. I wasn't artistic in high school, and CS seemed like a
logical, sterile field that I felt I could do well in.

During the years I spent at RIT I changed quite a bit. I think the most
important, or at least the most obvious, change has been my becoming more
artistic. Between dancing, playing bass and photography I've developed a
creative side that I never thought I could. Has this led me away from Computer
Science at all?

No. While I became a programmer for practical reasons, I stayed one for
entirely different reasons. Computer Science (and math in general) is
beautiful. It took me years to slowly realize this, but now that I have I see
that it's more beautiful than any dance, photograph of music I've ever
encountered. This post is the first in a series of posts that I hope can
communicate why I find CS beautiful, or at least point in the right direction.

A Quick Primer on "Functions" for Non-Programmers
---------

I know that most (if not all) of the people that read my website don't
program. I'm going to try to avoid using extremely technical, computer sciency
terms in these posts, but there are at least a few basic things that I need to
explain. The first is what a function is.

Think of a function as a set of instructions. A recipe is a decent example.
Imagine you have a piece of paper with the following written on it:

1. Heat a frying pan.
2. Crack two eggs into a bowl.
3. Mix the eggs.
4. Pour the mixture into the hot frying pan.
5. Stir the mixture until it is solid.
6. Take the mixture out of the pan.

That's the simplest example of a function: a list of instructions that you
follow in the order you get them. The next idea is a "parameter." That recipe
only told us how to make food for one person, wouldn't it be nice to know how
to make enough for two or more? Imagine the paper now says this:

1. Heat a frying pan.
2. Crack NUMBER_OF_PEOPLE * 2 eggs into a bowl.
3. Mix the eggs.
4. Pour the mixture into the hot frying pan.
5. Stir the mixture until it is solid.
6. Take the mixture out of the pan.

Now we still have a single piece of paper, but by just substituting in the
number of people we're cooking for we have instructions for making any amount
of food. If we have three people, NUMBER_OF_PEOPLE * 2 eggs becomes 3 * 2
eggs, or 6 eggs.

The last important concept is "calling." Once you have one function, you can
use it in other functions, like so:

1. Put a plate on the table.
2. Refer to the other piece of paper and do what it says, for 1 person.
3. Put the result of that on the plate.
4. Eat it.

See how we did a few things, then referred to the other piece of paper instead
of repeating ourselves? This lets you make small tasks and put them together
to form bigger ones. That's enough of a primer for now; if something's not
clear please find me on [Twitter][twsl] and let me know.

[twsl]: {{links.twsl}}

What's Recursion?
---------------

Recursion is a term that means, basically, a function calls (or refers to)
itself. This concept can be hard to grasp, but once you do it slowly turns
into something breathtakingly beautiful.

Here's a simple example: imagine you're 10 meters away from a doorway and you
want to walk out of it. A function that could help you do this might be: "Walk
11 meters." That's pretty simple.

What if we want to be able to walk out of a doorway that's any number of
meters ahead of us? We could say: "Walk DISTANCE + 1 meters." This works, but
requires that you know how many meters away the door is before you even start
walking. What if we don't?

How about we just do a little at a time and see how it goes? "Walk 2 meters.
If you're out of the doorway, stop. Otherwise, repeat this function." If the
doorway is 3 meters in front of us, we'll walk 2, check if we're out (we're
not), and repeat. We'll walk 2, check if we're out (we are), and stop. This
function is recursive; it refers back to itself.

Why is it beautiful?
--------------

The beauty of recursion, for me, is that you can build infinitely large or
complex tasks or structures using one or two tiny, simple parts. I'll give an
example of this because I think it's the easiest way for me to show it.

Imagine that you only know three things: how to add 1 to a number, how to
subtract 1 from a number, and what 0 means. Now imagine that you want to be
able to add any two (positive) numbers (integers) together. How could you do
this? Here's a function:

    :::text
    function add(x, y):
        if y = 0:
            return x
        otherwise:
            return add( x+1, y-1 )

Don't let the new notation scare you; it's the same kind of function as
always. In English, it says "This function is named 'add' and needs two
parameters (x and y). Step one is to see if y is 0. If it is, the result is x.
If it's not, then the result is whatever you get from performing this function
with the following parameters: x+1, y-1."

It might not be immediately obvious that this will actually perform the
addition we're used to, so let's try it. First, let's try adding 1 + 0. In
that case, we check if y is 0. It is, so the result is x, or 1. So far, so
good.

Let's try 2 + 1. Is y equal to 0? No, so we need to call add with some new
parameters. 2+1 is 3, 1-0 is 0, so the result is now whatever we get from
add(3, 0). We start following the instructions of add. Does y equal 0? Yes, so
the result is 3, which is what we expect.

One more for good measure: 9 + 2. First we check if y is 0. It's not, so the
result is then add(9+1, 2-1), or add(10,1). Start over. Is y zero? Nope, so
the result is now add(11, 0). Start over. Is y zero? Yes, so the result is x,
or 11.

So What?
--------

This will work for any x and any y (as long as they're both positive, negative
numbers make it just a tiny bit more complicated). This might not seem
impressive, but think about what we've done. We've created something that can
add any two numbers together. There are an infinite amount of numbers. This
tiny function that we've made from the simplest of parts can generate more
results than there are people on this planet. Or atoms in the universe. To me,
this is amazing. Not "magic trick" amazing or "miracle" amazing but "I can
understand and create something that can describe more knowledge than the
whole of humanity couldn't hope to describe in a thousand lifetimes" amazing.

Computer Science (and math in general) is full of this kind of beauty. I've
tried to find parallels in my other interests; the closest I've found is the
photograph [Pale Blue Dot][]. It's a photo taken by the Voyager 1 space probe
showing an immense amount of space with the tiniest blue speck in the middle,
which is Earth.

When you view the photograph, it's a mostly black field with a little fleck of
blue pigment. Not terribly complicated or interesting, until you realize that
that blue fleck is a representation of the planet that billions of people have
lived and died on. A smidgen of blue ink, in the right context, represents
every place a human has ever called "home."

Our addition function might not seems nearly as nifty as this, but it actually
represents more than even this photo ever can. There are a finite number of
people on Earth, but the addition function can add more numbers that that.
Even if you count every second in the life of every person that has ever lived
on our planet, the addition function can create more numbers than that.

This awes me more than any myth, photograph, story, song, legend or dance ever
has. It's the reason I stayed a Computer Scientist.

[Pale Blue Dot]: http://en.wikipedia.org/wiki/Pale_Blue_Dot

{% endblock %}