Friday, 28 June, 2002
Sunday, 2 June, 2002

Simple Types in Python

In which I attempt to explain value types

Ok, it’s been over a year since the first step of my Python Babysteps Tutorial. It’s about time to dig a little deeper.

In this step, we will take our first look at variables and types in the Python language. We will use variables for storing and retrieving information. We will also tiptoe into the wild and woolly world of types, which computer languages rely on for tasks such as telling the difference between numbers and words.

You’re not expected to be some kind of expert to work through this tutorial. I only expect you to know how to use your development environment (such as IDLE) to work within the Python shell and create your own Python scripts. If this sounds unfamiliar to you, you’re in luck — I happen to cover exactly that in the Babysteps tutorial.

For the rest of us non-experts, let’s take the first steps into real, live programming!


Your program needs something to work with. You need names and addresses for your mailing list program, you need debits and credits for your accounting program, and you need weapons and critters for your fast-paced arcade game.

How are we supposed to do that? Well, that’s where variables come in. What’s a variable? In my attempts to uncover the secrets of computer science, I found something describing them as:

named locations in memory used to store a value.

Ummm … that’s all fine and good, but if you’ve got a small brain like me, you’ll spend too much time trying to remember what those words mean to actually use them in a program. We need a simpler definition that we can rattle off in casual conversation. How about this?

A variable is something you want your program to remember.

It’s to the point and uses words in a way I can understand. I like that. It also kinda describes the situation in Python, where nearly anything can be used as a variable. Numbers, text, code, files, objects, other programs… the list is longer than we care to imagine this early in our studies.

To create a variable, make a name and assign a value to it using the equals = character. To get the value, use the variable name in a statement.

>>> number = 5
>>> print number

You can change the value of a variable any time by using =.

>>> print number
>>> number = 7
>>> print number

Identifiers aka Variable Names

Python gives us a lot of freedom in creating variable names, also known as identifiers. As long as they start with a letter or underscore (_), and are followed by letters, numbers, or underscores, you’re generally good to go. Case is important, so number, NUMBER, and Number are three different names. The biggest restriction is that you can’t use an identifier that is already used by Python. This means no variables named print, for example.

Here is the complete list of identifiers that Python has reserved for its own purposes::

and       del       for       is        raise
assert    elif      from      lambda    return
break     else      global    not       try
class     except    if        or        while
continue  exec      import    pass      yield
def       finally   in        print

Here are a few guidelines to follow when deciding on a variable name:

  • Identifiers starting with _ are treated differently by Python, so avoid them until you know what they’re for.
  • Use descriptive names rather than abbreviations or inside jokes. When you are naming a variable that holds the radius of a circle, it is usually better to use radius than r or halfway_there
  • Use a name that indicates what the variable will be used for. radius is much better than fnord for describing the radius of a circle. Plus, “fnord” breaks the “no inside jokes” guideline.
  • Find a balance between names that are too long or too short.
    • Too short would be n
    • Too long might be name_of_my_favorite_customer_in_walla_walla_washington
    • Just right might be name or customer_name
  • It is common practice to use all upper case letters for identifiers that describe constants — variables which will not be changing their values. Since pi will always have the value 3.1415926 — or so — you would use an identifier of PI for this variable.
  • It’s okay to bend the guidelines in favor of common terms. If you are writing code to figure out the distance between two points, then x1, y1, x2, and y2 are perfectly sensible identifiers.

There are two popular approaches to devising longer names (the ones that consist of two or more words pasted together). In the first, the two words are separated by underscores where the spaces would be. In this tradition, “customer name” would be written as customer_name. The other school uses capitalization to show separation, and would write the same “customer name” as CustomerName, or maybe customerName.

I use underscores in my own code. I don’t really care which one you use, but stay consistent. Try not to follow customer_name with CustomerAddress. And whatever you do, please don’t mix the two in the same variable. Identifiers like Customer_Name will only serve to aggravate me and hasten the approach of carpal tunnel syndrome for you. I am so emphatic about this silly little issue because I am saying this as somebody who has to read code written by other people.

The sad truth is that you are going to encounter plenty of otherwise great code that is downright mean about variable names. It’s just something we have to live with.

Literals aka Plain Old Values

Let’s throw in another fancy technical term, since it’ll help us understand types in the next section. Literal is a fancy name for a plain old value. In the statement number = 5, 5 is the literal. Most often, you’ll see literals being used in assigning a value to your variables. There are different ways to write literals, depending on what type of variable you are assigning to.


I’ve used the word “type” a few times already, but haven’t explained what a type is. That’s because I’ve been stalling. It’s a broad concept and the best definition I could think of is circular:

Type describes the type of variable you are using.

See what I mean? Let’s try a different approach.

You and I know that there’s a difference between numbers and words. Imagine the following exchange between you and a random stranger who we’ll call “Bob”:

Hi there! What’s your name?
Uhh … Brian
What’s your name plus 5?
Oh look, there’s my bus!

Right away, you realized that Bob is a raving lunatic — or maybe a Zen master — because "Brian" + 5 just doesn’t work. You’re not supposed to combine words and numbers like that. It isn’t polite.

Types are used by programming languages to recognize when programmers are starting to sound like a raving lunatic. Every type has rules for what you can do with them. You can do numbery things to numbers (add, subtract, find the square root of, etcetera). You can do stringy things to strings of text (search, capitalize, concatenate, and so on). Capitalizing numbers is a no-no, as is finding the square root of your name. When you learn how to create your own types later on, you will also be writing the rules for how your new types can be used.

An exception is a special type in Python that we’ll be seeing a lot of as we learn. Think back to that conversation with Bob. As soon as he asked you to add 5 to your name, a little red flag went up in your brain. This red flag told you that something was very wrong with dear old Bob. Exceptions are the little red flags that Python uses to tell you that something is very wrong. There are a lot of ways that things can go wrong, so there are a lot of different types of exception that we will see. Let’s see what Python does when we play the role of Bob.

>>> name = "Brian"
>>> print name + 5
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: cannot concatenate 'str' and 'int' objects

All of the “File” gibberish is called a traceback, showing where things went wrong. Yours will probably look a little different, but the idea is the same.

The last line is the exception itself. It tells us what kind of problem was found — exceptions.TypeError — and describes the specific issue — cannot concatenate 'str' and 'int' objects.

We won’t really understand what to do with an exception for a while, but the basic idea is clear. Python uses exceptions to warn us that we are heading towards the path of madness.


Nothing illustrates the idea of types as clearly as the number. Most of us know what a number is, and can understand the sorts of things we do with numbers: add them, subtract them, ignore them when they’re written on a bill, and so on. We know how to compare numbers to determine which is greater. Things get a little murky when we start talking about the String type or the HttpServer type, but we know what numbers are. Lucky for us, so does Python.

To make things easier “under the hood”, Python has four different categories of numbers:

Whole numbers, like 1, 2, and 186212.
long integers
Really big integers, like 6209389143
floating point numbers
Decimal numbers, like 1.0, 2.2, or 3.1415926
imaginary numbers
I can’t imagine what these would be. A little more seriously, these are the complex numbers that higher-math types play with all the time. Something to do with the square root of -1, I think.

Amazingly enough, Python takes care of telling one from the other, and we don’t need to worry about it most of the time. Just feed it a literal, and it’ll try to do the right thing. Each of these types do have their own rules, though, and these rules will bite us if we try to treat Python numbers exactly the same as the numbers in our checkbook.

Numeric Literals: How To Describe Numbers

>>> 1
>>> 35
>>> 34456432
>>> 355556499871154247854

What? Where did that L come from? It turns out that computers are faster at dealing with integers if they can fit them into a small chunk of memory. That “small chunk” is still more than enough room for most of the numbers that you’ll deal with in your day-to-day programming. Until you start converting parsecs to inches, 2,147,483,647 should be more than enough room.

When you hand it a really large literal like 355556499871154247854, Python notices right away that it won’t fit as a regular integer, and automatically makes it a long integer. There are times when you know in advance that your number will end up being very large. Specifying long integer type is very very easy. Just paste the letter L to the end of your literal.

>>> 6209389143L

While it is technically okay to use a lowercase l, it is much harder to read when scanning program code. 6209389143l looks too much like 62093891431. It’s a good idea to always use the upper case L for specifiying long integers.

The next type of number is floating point. You can either write floating point literals in the familiar decimal notation, or you can use something more like scientific notation.

>>> 1.0
>>> 1.234e+02
>>> 1.1

This is where things start to bite us if we’re not paying attention. 1.0 made sense, but what in the world went wrong with our other two numbers?

It’s like this, see. Computers are basically made up of switches. On and off. 1 and 0. That’s all a computer knows. It takes many layers of programming to translate your value into something that a computer can understand, and then more layers to turn it back into something you can understand. With integers, it’s relatively easy. It’s a simple matter for the computer to handle whole numbers. Floating point, however, is much more fluid. It takes a lot of work for the computer to translate a floating point value into a series of 1 and 0 switches. To save time and memory, it fudges the number a little bit. For most programs, it’s not a problem. It certainly hasn’t been an issue for me in any program I’ve written. There are also some excellent programming libraries out there when you do hit that particular wall. Everyday python does an excellent job of protecting us from all the chaos when printing a variable:

>>> value = 1.1
>>> print value

Internally, it’s still 1.1000000000000001 or so, but Python realizes that human readers are not going to be interested in the extra 14 zeroes.

Now, before you go storming off to Perl or C, you should learn the dark secret: this is not limited to Python. It’s something to do with computers in general, so this will come up no matter what language you use. The classier ones like Python just hide it from you when they can.

Our last kind of number is the imaginary number. My own math skills haven’t progressed far enough to present anything truly useful, but I know how to write an imaginary literal.

I think I just melted my brain by putting those two words together.

An imaginary … uh … literal … thing … consists of a floating point number (the real part), a + character, and another floating point number followed by a J (the imaginary part). Here, just look at my example, and you math people can figure it out yourself.

>>> number = 1.1 + 9.87J
>>> print number

I guess imaginary numbers don’t make it into everyday math that much, but they do show up in my checkbook a lot. Wrong kind of “imaginary”, I guess.

Now that I’ve made a complete fool of myself, let’s move along quickly to look at some things we can do to numbers.

Numeric Operations: Some Things You Can Do With Numbers

The Python shell does make a handy calculator. You can do all of the handy four-function operations, plus we get a nifty exponent operator and modulus operator at no extra charge.

>>> print 2 + 2
>>> print 10 - 2
>>> print 11 * 3
>>> print 27 / 3
>>> print 2 ** 3
>>> print 20 / 3
>>> print 20 % 3