Data Types

Just like any other tool, understanding what kind of data you are dealing with at any given time means understanding what it can do and where it can go.

Think about vehicles. A pickup truck has a flatbed that is good for moving things, whereas a minivan is better for ferrying a group of people around. A bicycle might be a good way to get to places close to home, but you be in trouble if you tried to take a bicycle out sailing when what you need is a boat.

So it is with data. Every piece of data has a type and that type dictates things like what methods and attributes it has, what functions it can be passed to, and what other types of data it can play nicely with.

Table of Contents

Part 1: Classes and instances

In modern Python the terms type and class are more or less synonymous.

A class is both like a category and a blueprint for data. It defines the characteristics and behavior of values in a particular category. Those individual values are called instances of a particular class, sometimes referred to as objects.

As an analogy, you can think of a Ford Mustang as a class of cars that have doors = 2, seats = 4, and a body_color that varies per car; and have the ability to drive, turn and stop. Individual Mustang cars parked in a garage or driving down the street are instances of the Mustang class.

For example:

• int is a class

• 1 and 10 are instances of the int class

• str is a class

• "hello" and "goodbye" are instances of the str class

The built-in standard types like int, str, and bool are all lowercase. Most classes defined elsewhere use CamelCasing, like Path from the pathlib library.

Part 1.1: Creating instances

I’ve said that classes are like blueprints. That is in part because classes know how to create their own instances.

Classes are callable just like functions and methods. When called, classes return an empty instance of that class. This is called instantiation.

Lets look at what happens when we call some of the built in type classes that we are already familiar with.

Listing 112 Python shell

>>> float()
0.0
>>> int()
0
>>> str()
''
>>> list()
[]

Part 1.2: Type conversion

Classes may also take arguments. For example, most of the built in type classes can be used to convert a value to its type. This is called typecasting.

Listing 113 Python shell

>>> str(5.0)
'5.0'

>>> float(5)
5.0

>>> int(5.25)
5

>>> list("hello")
['h', 'e', 'l', 'l', 'o']

Not all types can be converted to all other types though.

Listing 114 Python shell

>>> dict(5)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-29-2e3c53b6b812> in <module>
----> 1 dict(5)

TypeError: 'int' object is not iterable

Sometimes you may need to convert to another type first to get the desired results. For example, lets say we want a list with all of the digits in 32562. An int cannot be converted to a list. Instead, convert it to a str first, then convert each of the digits back to an int.

Listing 115 Python shell

>>> list(32562)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-46-947a2aee1ed3> in <module>
----> 1 list(32562)

TypeError: 'int' object is not iterable

>>> str(32562)
'32562'

>>> list(str(32562))
['3', '2', '5', '6', '2']

>>> digits = list(str(32562))
>>> for i, num in enumerate(digits):
        digits[i] = int(digits[i])

>>> digits
[3, 2, 5, 6, 2]

Part 1.3: Exercise

Exercise 43 (classes)

Do each of the following in a Python shell.

1. Use the type classes to create new empty objects for each of the following:

   • bool

   • int

   • float

   • str

   • dict

   • list

   • tuple

2. Convert data types of one type to another. Here are some to start with, but feel free to come up with your own combinations. Note: not all of these can be directly converted, so some exceptions are expected.

   • "5" to int

   • 5 to str

   • 5 to bool

   • 5 to float

   • 5.0 to int

   • 5.25 to int

   • "5.25" to int

   • 5 to list

   • [1, 2, 3] to tuple

   • (1, 2, 3) to dict

   • {'a': '1', 'b': '2', 'c': '3'} to list

Solution to Exercise 43 (classes)

Listing 116 #1

>>> bool()
False

>>> int()
0

>>> float()
0.0

>>> str()
''

>>> dict()
{}

>>> list()
[]

>>> tuple()
()

Listing 117 #2

>>> int("5")
5

>>> str(5)
'5'

>>> bool(5)
True

>>> float(5)
5.0

>>> int(5.0)
5

>>> int(5.25)
5

>>> int("5.25")
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-63-12a4bd2b0343> in <module>
----> 1 int("5.25")

ValueError: invalid literal for int() with base 10: '5.25'

>>> int(float("5.25"))
5

>>> list(5)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-64-0c7f5cd48ec1> in <module>
----> 1 list(5)

TypeError: 'int' object is not iterable

>>> tuple([1, 2, 3])
(1, 2, 3)

>>> dict((1, 2, 3))
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-66-237d7a195025> in <module>
----> 1 dict((1, 2, 3))

TypeError: cannot convert dictionary update sequence element #0 to a sequence

>>> list({'a': '1', 'b': '2', 'c': '3'})
['a', 'b', 'c']

Part 2: Finding out the type

Part 2.1: Type checking functions

We can find out the type of any value by using the type() function.

Listing 118 Python shell

>>> type(5)
<class 'int'>

>>> type(2.5)
<class 'float'>

>>> type("hello")
<class 'str'>

You can even find out the type of a type.

Listing 119 Python shell

>>> type(int)
<class 'type'>

>>> type(str)
<class 'type'>

>>> type(float)
<class 'type'>

You can check if a value is a particular type by using the isinstance() function. The first argument is the value you want to check, and the second argument is a type, or tuple of types, that you want to check it against.

Listing 120 Python shell

>>> isinstance("5", int)
False

>>> isinstance(5.0, float)
True

>>> isinstance(5, (int, float))
True

Part 2.2: Exercise

Exercise 44 (classes)

Do each of the following in a Python shell.

1. Use the type() function on the following values, then come up with some of your own.

   • False

   • None

   • "1"

   • 5.0

   • [1, 2, 3]

2. Use the isinstance() function to check the following, then come up with some of your own.

   • is "5" a str

   • is 5 an int

   • is [1, 2, 3] a list or a tuple

   • is 0 a bool

Solution to Exercise 44 (classes)

Listing 121 #1

>>> type(False)
bool

>>> type(None)
NoneType

>>> type("1")
str

>>> type(5.0)
float

>>> type([1, 2, 3])
list

Listing 122 #2

>>> isinstance("5", str)
True

>>> isinstance(5, int)
True

>>> isinstance([1, 2, 3], (list, tuple))
True

>>> isinstance(0, bool)
False

Summary

• A type or class is the classification or category of a value, as well as a blueprint or template for a particular kind of data. They have the type type.

• Individual values are called instances or objects

• When you call a class it returns an empty instance of that type, which is called instantiation.

• You can use the type() and isinstance() functions to find out the type of a value.

• You can often call a class with an argument to convert it to that type, which is called typecasting.

• Builtin types are lowercased, most other types are CamelCased.

Reference

Glossary

Data Types

type

data type

class

The classification of a value which tells Python what operations can be performed on it. Some examples include str, int, list, and dict.
When called the type or class will return a new empty object of that type. For example str() returns "".

typecasting

Converting from one type to another.

instance

An individual value of a particular type. A realized version of a class.

instantiation

instantiate

When a new object is created.