Conceived in the late 1980s as a teaching and scripting language, Python has since become an essential tool for many programmers, engineers, researchers, and data scientists across academia and industry. As an astronomer focused on building and promoting the free open tools for data-intensive science, I’ve found Python to be a near-perfect fit for the types of problems I face day to day, whether it’s extracting meaning from large astronomical datasets, scraping and munging data sources from the Web, or automating day-to-day research tasks.
The appeal of Python is in its simplicity and beauty, as well as the convenience of the large ecosystem of domain-specific tools that have been built on top of it. For example, most of the Python code in scientific computing and data science is built around a group of mature and useful packages:
NumPy provides efficient storage and computation for multidimensional data arrays.
SciPy contains a wide array of numerical tools such as numerical integration and interpolation.
Pandas provides a DataFrame object along with a powerful set of methods to manipulate, filter, group, and transform data.
Matplotlib provides a useful interface for creation of publication-quality plots and figures.
Scikit-Learn provides a uniform toolkit for applying common machine learning algorithms to data.
IPython/Jupyter provides an enhanced terminal and an interactive notebook environment that is useful for exploratory analysis, as well as creation of interactive, executable documents. For example, the manuscript for this report was composed entirely in Jupyter notebooks.
No less important are the numerous other tools and packages which accompany these: if there is a scientific or data analysis task you want to perform, chances are someone has written a package that will do it for you.
To tap into the power of this data science ecosystem, however, first requires familiarity with the Python language itself. I often encounter students and colleagues who have (sometimes extensive) backgrounds in computing in some language—MATLAB, IDL, R, Java, C++, etc.—and are looking for a brief but comprehensive tour of the Python language that respects their level of knowledge rather than starting from ground zero. This report seeks to fill that niche.
As such, this report in no way aims to be a comprehensive introduction to programming, or a full introduction to the Python language itself; if that is what you are looking for, you might check out one of the recommended references listed in “Resources for Further Learning”. Instead, this will provide a whirlwind tour of some of Python’s essential syntax and semantics, built-in data types and structures, function definitions, control flow statements, and other aspects of the language. My aim is that readers will walk away with a solid foundation from which to explore the data science stack just outlined.
Supplemental material (code examples, IPython notebooks, etc.) is available for download at https://github.com/jakevdp/WhirlwindTourOfPython/.
This book is here to help you get your job done. In general, if example code is offered with this book, you may use it in your programs and documentation. You do not need to contact us for permission unless you’re reproducing a significant portion of the code. For example, writing a program that uses several chunks of code from this book does not require permission. Selling or distributing a CD-ROM of examples from O’Reilly books does require permission. Answering a question by citing this book and quoting example code does not require permission. Incorporating a significant amount of example code from this book into your product’s documentation does require permission.
We appreciate, but do not require, attribution. An attribution usually includes the title, author, publisher, and ISBN. For example: “A Whirlwind Tour of Python by Jake VanderPlas (O’Reilly). Copyright 2016 O’Reilly Media, Inc., 978-1-491-96465-1.”
If you feel your use of code examples falls outside fair use or the permission given above, feel free to contact us at [email protected].
Installing Python and the suite of libraries that enable scientific computing is straightforward whether you use Windows, Linux, or Mac OS X. This section will outline some of the considerations when setting up your computer.
This report uses the syntax of Python 3, which contains language enhancements that are not compatible with the 2.x series of Python. Though Python 3.0 was first released in 2008, adoption has been relatively slow, particularly in the scientific and web development communities. This is primarily because it took some time for many of the essential packages and toolkits to be made compatible with the new language internals. Since early 2014, however, stable releases of the most important tools in the data science ecosystem have been fully compatible with both Python 2 and 3, and so this report will use the newer Python 3 syntax. Even though that is the case, the vast majority of code snippets in this report will also work without modification in Python 2: in cases where a Py2-incompatible syntax is used, I will make every effort to note it explicitly.
Though there are various ways to install Python, the one I would suggest—particularly if you wish to eventually use the data science tools mentioned earlier—is via the cross-platform Anaconda distribution. There are two flavors of the Anaconda distribution:
Miniconda gives you the Python
interpreter itself, along with a command-line tool called conda
which
operates as a cross-platform package manager geared toward Python
packages, similar in spirit to the apt
or yum
tools that Linux users
might be familiar with.
Anaconda includes both Python and
conda
, and additionally bundles a suite of other pre-installed
packages geared toward scientific computing.
Any of the packages included with Anaconda can also be installed manually on top of Miniconda; for this reason, I suggest starting with Miniconda.
To get started, download and install the Miniconda package—make sure to choose a version with Python 3—and then install the IPython notebook package:
[~]$ conda install ipython-notebook
For more information on conda
, including information about creating
and using conda
environments, refer to the Miniconda package documentation
linked at the above page.
Python aficionados are often quick to point out how “intuitive”,
“beautiful”, or “fun” Python is. While I tend to agree, I also recognize
that beauty, intuition, and fun often go hand in hand with familiarity,
and so for those familiar with other languages such florid sentiments
can come across as a bit smug. Nevertheless, I hope that if you give
Python a chance, you’ll see where such impressions might come from. And
if you really want to dig into the programming philosophy that drives
much of the coding practice of Python power users, a nice little Easter
egg exists in the Python interpreter—simply close your eyes, meditate
for a few minutes, and run import this
:
In
[
1
]:
import
this
The Zen of Python, by Tim Peters Beautiful is better than ugly. Explicit is better than implicit. Simple is better than complex. Complex is better than complicated. Flat is better than nested. Sparse is better than dense. Readability counts. Special cases aren't special enough to break the rules. Although practicality beats purity. Errors should never pass silently. Unless explicitly silenced. In the face of ambiguity, refuse the temptation to guess. There should be one--and preferably only one--obvious way to do it. Although that way may not be obvious at first unless you're Dutch. Now is better than never. Although never is often better than *right* now. If the implementation is hard to explain, it's a bad idea. If the implementation is easy to explain, it may be a good idea. Namespaces are one honking great idea--let's do more of those!
With that, let’s start our tour of the Python language.
Python is a flexible language, and there are several ways to use it depending on your particular task. One thing that distinguishes Python from other programming languages is that it is interpreted rather than compiled. This means that it is executed line by line, which allows programming to be interactive in a way that is not directly possible with compiled languages like Fortran, C, or Java. This section will describe four primary ways you can run Python code: the Python interpreter, the IPython interpreter, via self-contained scripts, or in the Jupyter notebook.
The most basic way to execute Python code is line by line within the
Python interpreter. The Python interpreter can be started by
installing the Python language (see the previous section) and typing
python
at the command prompt (look for the Terminal on Mac OS X and
Unix/Linux systems, or the Command Prompt application in Windows):
$ python Python 3.5.1 |Continuum Analytics, Inc.| (default, Dec 7... Type "help", "copyright", "credits" or "license" for more... >>>
With the interpreter running, you can begin to type and execute code snippets. Here we’ll use the interpreter as a simple calculator, performing calculations and assigning values to variables:
>>>
1
+
1
2
>>>
x
=
5
>>>
x
*
3
15
The interpreter makes it very convenient to try out small snippets of Python code and to experiment with short sequences of operations.
If you spend much time with the basic Python interpreter, you’ll find
that it lacks many of the features of a full-fledged interactive
development environment. An alternative interpreter called IPython
(for Interactive Python) is bundled with the Anaconda distribution, and
includes a host of convenient enhancements to the basic Python
interpreter. It can be started by typing ipython
at the command
prompt:
$ ipython Python 3.5.1 |Continuum Analytics, Inc.| (default, Dec 7... Type "copyright", "credits" or "license" for more information. IPython 4.0.0 -- An enhanced Interactive Python. ? -> Introduction and overview of IPython's features. %quickref -> Quick reference. help -> Python's own help system. object? -> Details about 'object', use 'object??' for extra... In [1]:
The main aesthetic difference between the Python interpreter and the
enhanced IPython interpreter lies in the command prompt: Python uses
>>>
by default, while IPython uses numbered commands (e.g., In [1]:
).
Regardless, we can execute code line by line just as we did before:
In
[
1
]:
1
+
1
Out
[
1
]:
2
In
[
2
]:
x
=
5
In
[
3
]:
x
*
3
Out
[
3
]:
15
Note that just as the input is numbered, the output of each command is numbered as well. IPython makes available a wide array of useful features; for some suggestions on where to read more, see “Resources for Further Learning”.
Running Python snippets line by line is useful in some cases, but for more complicated programs it is more convenient to save code to file, and execute it all at once. By convention, Python scripts are saved in files with a .py extension. For example, let’s create a script called test.py that contains the following:
# file: test.py
(
"Running test.py"
)
x
=
5
(
"Result is"
,
3
*
x
)
To run this file, we make sure it is in the current directory and type
python
filename
at the command prompt:
$ python test.py Running test.py Result is 15
For more complicated programs, creating self-contained scripts like this one is a must.
A useful hybrid of the interactive terminal and the self-contained script is the Jupyter notebook, a document format that allows executable code, formatted text, graphics, and even interactive features to be combined into a single document. Though the notebook began as a Python-only format, it has since been made compatible with a large number of programming languages, and is now an essential part of the Jupyter Project. The notebook is useful both as a development environment and as a means of sharing work via rich computational and data-driven narratives that mix together code, figures, data, and text.
Python was originally developed as a teaching language, but its ease of use and clean syntax have led it to be embraced by beginners and experts alike. The cleanliness of Python’s syntax has led some to call it “executable pseudocode”, and indeed my own experience has been that it is often much easier to read and understand a Python script than to read a similar script written in, say, C. Here we’ll begin to discuss the main features of Python’s syntax.
Syntax refers to the structure of the language (i.e., what constitutes a correctly formed program). For the time being, we won’t focus on the semantics—the meaning of the words and symbols within the syntax—but will return to this at a later point.
Consider the following code example:
In
[
1
]:
# set the midpoint
midpoint
=
5
# make two empty lists
lower
=
[];
upper
=
[]
# split the numbers into lower and upper
for
i
in
range
(
10
):
if
(
i
<
midpoint
):
lower
.
append
(
i
)
else
:
upper
.
append
(
i
)
(
"lower:"
,
lower
)
(
"upper:"
,
upper
)
lower: [0, 1, 2, 3, 4] upper: [5, 6, 7, 8, 9]
This script is a bit silly, but it compactly illustrates several of the important aspects of Python syntax. Let’s walk through it and discuss some of the syntactical features of Python.
#
The script starts with a comment:
# set the midpoint
Comments in Python are indicated by a pound sign (#
), and anything on
the line following the pound sign is ignored by the interpreter. This
means, for example, that you can have standalone comments like the one
just shown, as well as inline comments that follow a statement. For example:
x
+=
2
# shorthand for x = x + 2
Python does not have any syntax for multiline comments, such as the
/* ... */
syntax used in C and C++, though multiline strings are
often used as a replacement for multiline comments (more on this in
“String Manipulation and Regular Expressions”).
The next line in the script is
midpoint
=
5
This is an assignment operation, where we’ve created a variable named
midpoint
and assigned it the value 5
. Notice that the end of this
statement is simply marked by the end of the line. This is in contrast
to languages like C and C++, where every statement must end with a
semicolon (;
).
In Python, if you’d like a statement to continue to the next line, it is
possible to use the marker to indicate this:
In
[
2
]:
x
=
1
+
2
+
3
+
4
+
5
+
6
+
7
+
8
It is also possible to continue expressions on the next line within
parentheses, without using the marker:
In
[
3
]:
x
=
(
1
+
2
+
3
+
4
+
5
+
6
+
7
+
8
)
Most Python style guides recommend the second version of line
continuation (within parentheses) to the first (use of the
marker).
Sometimes it can be useful to put multiple statements on a single line. The next portion of the script is:
lower
=
[];
upper
=
[]
This shows the example of how the semicolon (;
) familiar in C can be
used optionally in Python to put two statements on a single line.
Functionally, this is entirely equivalent to writing:
lower
=
[]
upper
=
[]
Using a semicolon to put multiple statements on a single line is generally discouraged by most Python style guides, though occasionally it proves convenient.
Next, we get to the main block of code:
for
i
in
range
(
10
):
if
i
<
midpoint
:
lower
.
append
(
i
)
else
:
upper
.
append
(
i
)
This is a compound control-flow statement including a loop and a conditional—we’ll look at these types of statements in a moment. For now, consider that this demonstrates what is perhaps the most controversial feature of Python’s syntax: whitespace is meaningful!
In programming languages, a block of code is a set of statements that should be treated as a unit. In C, for example, code blocks are denoted by curly braces:
// C code
for
(
int
i
=
0
;
i
<
100
;
i
++
)
{
// curly braces indicate code block
total
+=
i
;
}
In Python, code blocks are denoted by indentation:
for
i
in
range
(
100
):
# indentation indicates code block
total
+=
i
In Python, indented code blocks are always preceded by a colon (:
) on
the previous line.
The use of indentation helps to enforce the uniform, readable style that many find appealing in Python code. But it might be confusing to the uninitiated; for example, the following two snippets will produce different results:
>>>
if
x
<
4
:
>>>
if
x
<
4
:
...
y
=
x
*
2
...
y
=
x
*
2
...
(
x
)
...
(
x
)
In the snippet on the left, print(x)
is in the indented block, and
will be executed only if x
is less than 4
. In the snippet on the
right, print(x)
is outside the block, and will be executed regardless
of the value of x
!
Python’s use of meaningful whitespace often is surprising to programmers who are accustomed to other languages, but in practice it can lead to much more consistent and readable code than languages that do not enforce indentation of code blocks. If you find Python’s use of whitespace disagreeable, I’d encourage you to give it a try: as I did, you may find that you come to appreciate it.
Finally, you should be aware that the amount of whitespace used for indenting code blocks is up to the user, as long as it is consistent throughout the script. By convention, most style guides recommend to indent code blocks by four spaces, and that is the convention we will follow in this report. Note that many text editors like Emacs and Vim contain Python modes that do four-space indentation automatically.
While the mantra of meaningful whitespace holds true for whitespace before lines (which indicate a code block), whitespace within lines of Python code does not matter. For example, all three of these expressions are equivalent:
In
[
4
]:
x
=
1
+
2
x
=
1
+
2
x
=
1
+
2
Abusing this flexibility can lead to issues with code readability—in fact, abusing whitespace is often one of the primary means of intentionally obfuscating code (which some people do for sport). Using whitespace effectively can lead to much more readable code, especially in cases where operators follow each other—compare the following two expressions for exponentiating by a negative number:
x
=
10
**-
2
to
x
=
10
**
-
2
I find the second version with spaces much more easily readable at a single glance. Most Python style guides recommend using a single space around binary operators, and no space around unary operators. We’ll discuss Python’s operators further in “Basic Python Semantics: Variables and Objects”.
In the following code snippet, we see two uses of parentheses. First, they can be used in the typical way to group statements or mathematical operations:
In
[
5
]:
2
*
(
3
+
4
)
Out [5]: 14
They can also be used to indicate that a function is being called. In
the next snippet, the print()
function is used to display the
contents of a variable (see the sidebar that follows). The function call is indicated
by a pair of opening and closing parentheses, with the arguments to the
function contained within:
In
[
6
]:
(
'first value:'
,
1
)
first value: 1
In
[
7
]:
(
'second value:'
,
2
)
second value: 2
Some functions can be called with no arguments at all, in which case the
opening and closing parentheses still must be used to indicate a function
evaluation. An example of this is the sort
method of lists:
In
[
8
]:
L
=
[
4
,
2
,
3
,
1
]
L
.
sort
()
(
L
)
[1, 2, 3, 4]
The ()
after sort
indicates that the function should be executed,
and is required even if no arguments are necessary.
This has been a very brief exploration of the essential features of Python syntax; its purpose is to give you a good frame of reference for when you’re reading the code in later sections. Several times we’ve mentioned Python “style guides,” which can help teams to write code in a consistent style. The most widely used style guide in Python is known as PEP8, and can be found at https://www.python.org/dev/peps/pep-0008/. As you begin to write more Python code, it would be useful to read through this! The style suggestions contain the wisdom of many Python gurus, and most suggestions go beyond simple pedantry: they are experience-based recommendations that can help avoid subtle mistakes and bugs in your code.
This section will begin to cover the basic semantics of the Python language. As opposed to the syntax covered in the previous section, the semantics of a language involve the meaning of the statements. As with our discussion of syntax, here we’ll preview a few of the essential semantic constructions in Python to give you a better frame of reference for understanding the code in the following sections.
This section will cover the semantics of variables and objects, which are the main ways you store, reference, and operate on data within a Python script.
Assigning variables in Python is as easy as putting a variable name to
the left of the equals sign (=
):
# assign 4 to the variable x
x
=
4
This may seem straightforward, but if you have the wrong mental model of what this operation does, the way Python works may seem confusing. We’ll briefly dig into that here.
In many programming languages, variables are best thought of as containers or buckets into which you put data. So in C, for example, when you write
// C code
int
x
=
4
;
you are essentially defining a “memory bucket” named x
, and putting
the value 4
into it. In Python, by contrast, variables are best
thought of not as containers but as pointers. So in Python, when you
write
x
=
4
you are essentially defining a pointer named x
that points to some
other bucket containing the value 4
. Note one consequence of this:
because Python variables just point to various objects, there is no need
to “declare” the variable, or even require the variable to always point
to information of the same type! This is the sense in which people say
Python is dynamically typed: variable names can point to objects of
any type. So in Python, you can do things like this:
In
[
1
]:
x
=
1
# x is an integer
x
=
'hello'
# now x is a string
x
=
[
1
,
2
,
3
]
# now x is a list
While users of statically typed languages might miss the type-safety that comes with declarations like those found in C,
int
x
=
4
;
this dynamic typing is one of the pieces that makes Python so quick to write and easy to read.
There is a consequence of this “variable as pointer” approach that you need to be aware of. If we have two variable names pointing to the same mutable object, then changing one will change the other as well! For example, let’s create and modify a list:
In
[
2
]:
x
=
[
1
,
2
,
3
]
y
=
x
We’ve created two variables x
and y
that both point to the same
object. Because of this, if we modify the list via one of its names,
we’ll see that the “other” list will be modified as well:
In
[
3
]:
(
y
)
[1, 2, 3]
In
[
4
]:
x
.
append
(
4
)
# append 4 to the list pointed to by x
(
y
)
# y's list is modified as well!
[1, 2, 3, 4]
This behavior might seem confusing if you’re wrongly thinking of variables as buckets that contain data. But if you’re correctly thinking of variables as pointers to objects, then this behavior makes sense.
Note also that if we use =
to assign another value to x
, this will
not affect the value of y
—assignment is simply a change of what
object the variable points to:
In
[
5
]:
x
=
'something else'
(
y
)
# y is unchanged
[1, 2, 3, 4]
Again, this makes perfect sense if you think of x
and y
as pointers,
and the =
operator as an operation that changes what the name
points to.
You might wonder whether this pointer idea makes arithmetic operations in Python difficult to track, but Python is set up so that this is not an issue. Numbers, strings, and other simple types are immutable: you can’t change their value—you can only change what values the variables point to. So, for example, it’s perfectly safe to do operations like the following:
In
[
6
]:
x
=
10
y
=
x
x
+=
5
# add 5 to x's value, and assign it to x
(
"x ="
,
x
)
(
"y ="
,
y
)
x = 15 y = 10
When we call x += 5
, we are not modifying the value of the 5
object
pointed to by x
, but rather we are changing the object to which x
points. For this reason, the value of y
is not affected by the
operation.
Python is an object-oriented programming language, and in Python everything is an object.
Let’s flesh out what this means. Earlier we saw that variables are simply pointers, and the variable names themselves have no attached type information. This leads some to claim erroneously that Python is a type-free language. But this is not the case! Consider the following:
In
[
7
]:
x
=
4
type
(
x
)
Out [7]: int
In
[
8
]:
x
=
'hello'
type
(
x
)
Out [8]: str
In
[
9
]:
x
=
3.14159
type
(
x
)
Out [9]: float
Python has types; however, the types are linked not to the variable names but to the objects themselves.
In object-oriented programming languages like Python, an object is an entity that contains data along with associated metadata and/or functionality. In Python, everything is an object, which means every entity has some metadata (called attributes) and associated functionality (called methods). These attributes and methods are accessed via the dot syntax.
For example, before we saw that lists have an append
method, which adds
an item to the list, and is accessed via the dot syntax (.
):
In
[
10
]:
L
=
[
1
,
2
,
3
]
L
.
append
(
100
)
(
L
)
[1, 2, 3, 100]
While it might be expected for compound objects like lists to have
attributes and methods, what is sometimes unexpected is that in Python
even simple types have attached attributes and methods. For example,
numerical types have a real
and imag
attribute that return the
real and imaginary part of the value, if viewed as a complex number:
In
[
11
]:
x
=
4.5
(
x
.
real
,
"+"
,
x
.
imag
,
'i'
)
4.5 + 0.0 i
Methods are like attributes, except they are functions that you can
call using a pair of opening and closing parentheses. For example, floating-point
numbers have a method called is_integer
that checks whether the value
is an integer:
In
[
12
]:
x
=
4.5
x
.
is_integer
()
Out [12]: False
In
[
13
]:
x
=
4.0
x
.
is_integer
()
Out [13]: True
When we say that everything in Python is an object, we really mean that
everything is an object—even the attributes and methods of objects
are themselves objects with their own type
information:
In
[
14
]:
type
(
x
.
is_integer
)
Out [14]: builtin_function_or_method
We’ll find that the everything-is-object design choice of Python allows for some very convenient language constructs.
In the previous section, we began to look at the semantics of Python variables and objects; here we’ll dig into the semantics of the various operators included in the language. By the end of this section, you’ll have the basic tools to begin comparing and operating on data in Python.
Python implements seven basic binary arithmetic operators, two of which can double as unary operators. They are summarized in the following table:
Operator | Name | Description |
---|---|---|
|
Addition |
Sum of |
|
Subtraction |
Difference of |
|
Multiplication |
Product of |
|
True division |
Quotient of |
|
Floor division |
Quotient of |
|
Modulus |
Remainder after division of |
|
Exponentiation |
|
|
Negation |
The negative of |
|
Unary plus |
|
These operators can be used and combined in intuitive ways, using standard parentheses to group operations. For example:
In
[
1
]:
# addition, subtraction, multiplication
(
4
+
8
)
*
(
6.5
-
3
)
Out [1]: 42.0
Floor division is true division with fractional parts truncated:
In
[
2
]:
# True division
(
11
/
2
)
5.5
In
[
3
]:
# Floor division
(
11
//
2
)
5
The floor division operator was added in Python 3; you should be aware
if working in Python 2 that the standard division operator (/
) acts
like floor division for integers and like true division for floating-point numbers.
Finally, I’ll mention that an eighth arithmetic operator was added in Python 3.5: the a @ b
operator, which is meant to
indicate the matrix product of a
and b
, for use in various linear
algebra packages.
In addition to the standard numerical operations, Python includes operators to perform bitwise logical operations on integers. These are much less commonly used than the standard arithmetic operations, but it’s useful to know that they exist. The six bitwise operators are summarized in the following table:
Operator | Name | Description |
---|---|---|
|
Bitwise AND |
Bits defined in both |
|
Bitwise OR |
Bits defined in |
|
Bitwise XOR |
Bits defined in |
|
Bit shift left |
Shift bits of |
|
Bit shift right |
Shift bits of |
|
Bitwise NOT |
Bitwise negation of |
These bitwise operators only make sense in terms of the binary
representation of numbers, which you can see using the built-in bin
function:
In
[
4
]:
bin
(
10
)
Out [4]: '0b1010'
The result is prefixed with 0b
, which indicates a binary
representation. The rest of the digits indicate that the number 10 is
expressed as the sum:
Similarly, we can write:
In
[
5
]:
bin
(
4
)
Out [5]: '0b100'
Now, using bitwise OR, we can find the number which combines the bits of 4 and 10:
In
[
6
]:
4
|
10
Out [6]: 14
In
[
7
]:
bin
(
4
|
10
)
Out [7]: '0b1110'
These bitwise operators are not as immediately useful as the standard
arithmetic operators, but it’s helpful to see them at least once to
understand what class of operation they perform. In particular, users
from other languages are sometimes tempted to use XOR (i.e., a ^ b
)
when they really mean exponentiation (i.e., a ** b
).
We’ve seen that variables can be assigned with the =
operator, and
the values stored for later use. For example:
In
[
8
]:
a
=
24
(
a
)
24
We can use these variables in expressions with any of the operators
mentioned earlier. For example, to add 2 to a
we write:
In
[
9
]:
a
+
2
Out [9]: 26
We might want to update the variable a
with this new value; in this
case, we could combine the addition and the assignment and write a = a + 2
. Because this type of combined operation and assignment is
so common, Python includes built-in update operators for all of the
arithmetic operations:
In
[
10
]:
a
+=
2
# equivalent to a = a + 2
(
a
)
26
There is an augmented assignment operator corresponding to each of the binary operators listed earlier; in brief, they are:
|
|
|
|
|
|
|
|
|
|
|
|
Each one is equivalent to the corresponding operation followed by
assignment: that is, for any operator #
, the expression a #= b
is
equivalent to a = a # b
, with a slight catch. For mutable objects like
lists, arrays, or DataFrames, these augmented assignment operations are
actually subtly different than their more verbose counterparts: they
modify the contents of the original object rather than creating a new
object to store the result.
Another type of operation that can be very useful is comparison of
different values. For this, Python implements standard comparison
operators, which return Boolean values True
and False
. The
comparison operations are listed in the following table:
Operation | Description |
---|---|
|
|
|
|
|
|
|
|
|
|
|
|
These comparison operators can be combined with the arithmetic and bitwise operators to express a virtually limitless range of tests for the numbers. For example, we can check if a number is odd by checking that the modulus with 2 returns 1:
In
[
11
]:
# 25 is odd
25
%
2
==
1
Out [11]: True
In
[
12
]:
# 66 is odd
66
%
2
==
1
Out [12]: False
We can string together multiple comparisons to check more complicated relationships:
In
[
13
]:
# check if a is between 15 and 30
a
=
25
15
<
a
<
30
Out [13]: True
And, just to make your head hurt a bit, take a look at this comparison:
In
[
14
]:
-
1
==
~
0
Out [14]: True
Recall that ~
is the bit-flip operator, and evidently when you flip
all the bits of zero you end up with –1. If you’re curious as
to why this is, look up the two’s complement integer encoding scheme,
which is what Python uses to encode signed integers, and think about
happens when you start flipping all the bits of integers encoded this
way.
When working with Boolean values, Python provides operators to combine
the values using the standard concepts of “and”, “or”, and “not”.
Predictably, these operators are expressed using the words and
, or
,
and not
:
In
[
15
]:
x
=
4
(
x
<
6
)
and
(
x
>
2
)
Out [15]: True
In
[
16
]:
(
x
>
10
)
or
(
x
%
2
==
0
)
Out [16]: True
In
[
17
]:
not
(
x
<
6
)
Out [17]: False
Boolean algebra aficionados might notice that the XOR operator is not included; this can of course be constructed in several ways from a compound statement of the other operators. Otherwise, a clever trick you can use for XOR of Boolean values is the following:
In
[
18
]:
# (x > 1) xor (x < 10)
(
x
>
1
)
!=
(
x
<
10
)
Out [18]: False
These sorts of Boolean operations will become extremely useful when we begin discussing control flow statements such as conditionals and loops.
One sometimes confusing thing about the language is when to use Boolean
operators (and
, or
, not
), and when to use bitwise operations (&
,
|
, ~
). The answer lies in their names: Boolean operators should be
used when you want to compute Boolean values (i.e., truth or falsehood)
of entire statements. Bitwise operations should be used when you want
to operate on individual bits or components of the objects in
question.
Like and
, or
, and not
, Python also contains prose-like operators
to check for identity and membership. They are the following:
Operator | Description |
---|---|
|
True if |
|
True if |
|
True if |
|
True if |
The identity operators, is
and is not
, check for object
identity. Object identity is different than equality, as we can see
here:
In
[
19
]:
a
=
[
1
,
2
,
3
]
b
=
[
1
,
2
,
3
]
In
[
20
]:
a
==
b
Out [20]: True
In
[
21
]:
a
is
b
Out [21]: False
In
[
22
]:
a
is
not
b
Out [22]: True
What do identical objects look like? Here is an example:
In
[
23
]:
a
=
[
1
,
2
,
3
]
b
=
a
a
is
b
Out [23]: True
The difference between the two cases here is that in the first, a
and
b
point to different objects, while in the second they point to the
same object. As we saw in the previous section, Python variables are
pointers. The is
operator checks whether the two variables are
pointing to the same container (object), rather than referring to what
the container contains. With this in mind, in most cases that a beginner
is tempted to use is
, what they really mean is ==
.
Membership operators check for membership within compound objects. So, for example, we can write:
In
[
24
]:
1
in
[
1
,
2
,
3
]
Out [24]: True
In
[
25
]:
2
not
in
[
1
,
2
,
3
]
Out [25]: False
These membership operations are an example of what makes Python so easy to use compared to lower-level languages such as C. In C, membership would generally be determined by manually constructing a loop over the list and checking for equality of each value. In Python, you just type what you want to know, in a manner reminiscent of straightforward English prose.
When discussing Python variables and objects, we mentioned the fact that all Python objects have type information attached. Here we’ll briefly walk through the built-in simple types offered by Python. We say “simple types” to contrast with several compound types, which will be discussed in the following section.
Python’s simple types are summarized in Table 1-1.
Type | Example | Description |
---|---|---|
|
|
Integers (i.e., whole numbers) |
|
|
Floating-point numbers (i.e., real numbers) |
|
|
Complex numbers (i.e., numbers with a real and imaginary part) |
|
|
Boolean: True/False values |
|
|
String: characters or text |
|
|
Special object indicating nulls |
We’ll take a quick look at each of these in turn.
The most basic numerical type is the integer. Any number without a decimal point is an integer:
In
[
1
]:
x
=
1
type
(
x
)
Out [1]: int
Python integers are actually quite a bit more sophisticated than
integers in languages like C
. C integers are fixed-precision, and usually
overflow at some value (often near 231 or
263, depending on your system). Python integers are
variable-precision, so you can do computations that would overflow in
other languages:
In
[
2
]:
2
**
200
Out [2]: 1606938044258990275541962092341162602522202993782792835301376
Another convenient feature of Python integers is that by default, division upcasts to floating-point type:
In
[
3
]:
5
/
2
Out [3]: 2.5
Note that this upcasting is a feature of Python 3; in Python 2, like in many statically typed languages such as C, integer division truncates any decimal and always returns an integer:
# Python 2 behavior
>>>
5
/
2
2
To recover this behavior in Python 3, you can use the floor-division operator:
In
[
4
]:
5
//
2
Out [4]: 2
Finally, note that although Python 2.x had both an int
and long
type, Python 3 combines the behavior of these two into a single int
type.
The floating-point type can store fractional numbers. They can be defined either in standard decimal notation, or in exponential notation:
In
[
5
]:
x
=
0.000005
y
=
5e-6
(
x
==
y
)
True
In
[
6
]:
x
=
1400000.00
y
=
1.4e6
(
x
==
y
)
True
In the exponential notation, the e
or E
can be read “…times ten to
the…”, so that 1.4e6
is interpreted as
1.4 × 106.
An integer can be explicitly converted to a float with the float
constructor:
In
[
7
]:
float
(
1
)
Out [7]: 1.0
One thing to be aware of with floating-point arithmetic is that its precision is limited, which can cause equality tests to be unstable. For example:
In
[
8
]:
0.1
+
0.2
==
0.3
Out [8]: False
Why is this the case? It turns out that it is not a behavior unique to Python, but is due to the fixed-precision format of the binary floating-point storage used by most, if not all, scientific computing platforms. All programming languages using floating-point numbers store them in a fixed number of bits, and this leads some numbers to be represented only approximately. We can see this by printing the three values to high precision:
In
[
9
]:
(
"0.1 = {0:.17f}"
.
format
(
0.1
))
(
"0.2 = {0:.17f}"
.
format
(
0.2
))
(
"0.3 = {0:.17f}"
.
format
(
0.3
))
0.1 = 0.10000000000000001 0.2 = 0.20000000000000001 0.3 = 0.29999999999999999
We’re accustomed to thinking of numbers in decimal (base-10) notation, so that each fraction must be expressed as a sum of powers of 10:
In the familiar base-10 representation, we represent this in the familiar decimal expression: 0.125.
Computers usually store values in binary notation, so that each number is expressed as a sum of powers of 2:
In a base-2 representation, we can write this 0.0012, where the subscript 2 indicates binary notation. The value 0.125 = 0.0012 happens to be one number which both binary and decimal notation can represent in a finite number of digits.
In the familiar base-10 representation of numbers, you are probably familiar with numbers that can’t be expressed in a finite number of digits. For example, dividing 1 by 3 gives, in standard decimal notation:
The 3s go on forever: that is, to truly represent this quotient, the number of required digits is infinite!
Similarly, there are numbers for which binary representations require an infinite number of digits. For example:
Just as decimal notation requires an infinite number of digits to perfectly represent 1/3, binary notation requires an infinite number of digits to represent 1/10. Python internally truncates these representations at 52 bits beyond the first nonzero bit on most systems.
This rounding error for floating-point values is a necessary evil of working with floating-point numbers. The best way to deal with it is to always keep in mind that floating-point arithmetic is approximate, and never rely on exact equality tests with floating-point values.
Complex numbers are numbers with real and imaginary (floating-point) parts. We’ve seen integers and real numbers before; we can use these to construct a complex number:
In
[
10
]:
complex
(
1
,
2
)
Out [10]: (1+2j)
Alternatively, we can use the j
suffix in expressions to indicate
the imaginary part:
In
[
11
]:
1
+
2j
Out [11]: (1+2j)
Complex numbers have a variety of interesting attributes and methods, which we’ll briefly demonstrate here:
In
[
12
]:
c
=
3
+
4j
In
[
13
]:
c
.
real
# real part
Out [13]: 3.0
In
[
14
]:
c
.
imag
# imaginary part
Out [14]: 4.0
In
[
15
]:
c
.
conjugate
()
# complex conjugate
Out [15]: (3-4j)
In
[
16
]:
abs
(
c
)
# magnitude--that is, sqrt(c.real ** 2 + c.imag ** 2)
Out [16]: 5.0
Strings in Python are created with single or double quotes:
In
[
17
]:
message
=
"what do you like?"
response
=
'spam'
Python has many extremely useful string functions and methods; here are a few of them:
In
[
18
]:
# length of string
len
(
response
)
Out [18]: 4
In
[
19
]:
# Make uppercase. See also str.lower()
response
.
upper
()
Out [19]: 'SPAM'
In
[
20
]:
# Capitalize. See also str.title()
message
.
capitalize
()
Out [20]: 'What do you like?'
In
[
21
]:
# concatenation with +
message
+
response
Out [21]: 'what do you like?spam'
In
[
22
]:
# multiplication is multiple concatenation
5
*
response
Out [22]: 'spamspamspamspamspam'
In
[
23
]:
# Access individual characters (zero-based indexing)
message
[
0
]
Out [23]: 'w'
For more discussion of indexing in Python, see “Lists”.
Python includes a special type, the NoneType
, which has only a single
possible value: None
. For example:
In
[
24
]:
type
(
None
)
Out [24]: NoneType
You’ll see None
used in many places, but perhaps most commonly it is
used as the default return value of a function. For example, the
print()
function in Python 3 does not return anything, but we can
still catch its value:
In
[
25
]:
return_value
=
(
'abc'
)
abc
In
[
26
]:
(
return_value
)
None
Likewise, any function in Python with no return value is, in reality,
returning None
.
The Boolean type is a simple type with two possible values: True
and
False
, and is returned by comparison operators discussed previously:
In
[
27
]:
result
=
(
4
<
5
)
result
Out [27]: True
In
[
28
]:
type
(
result
)
Out [28]: bool
Keep in mind that the Boolean values are case-sensitive: unlike some
other languages, True
and False
must be capitalized!
In
[
29
]:
(
True
,
False
)
True False
Booleans can also be constructed using the bool()
object constructor:
values of any other type can be converted to Boolean via predictable
rules. For example, any numeric type is False if equal to zero, and True
otherwise:
In
[
30
]:
bool
(
2014
)
Out [30]: True
In
[
31
]:
bool
(
0
)
Out [31]: False
In
[
32
]:
bool
(
3.1415
)
Out [32]: True
The Boolean conversion of None
is always False:
In
[
33
]:
bool
(
None
)
Out [33]: False
For strings, bool(s)
is False for empty strings and True otherwise:
In
[
34
]:
bool
(
""
)
Out [34]: False
In
[
35
]:
bool
(
"abc"
)
Out [35]: True
For sequences, which we’ll see in the next section, the Boolean representation is False for empty sequences and True for any other sequences:
In
[
36
]:
bool
([
1
,
2
,
3
])
Out [36]: True
In
[
37
]:
bool
([])
Out [37]: False
We have seen Python’s simple types: int
, float
, complex
, bool
,
str
, and so on. Python also has several built-in compound types, which act
as containers for other types. These compound types are:
Type Name | Example | Description |
---|---|---|
|
|
Ordered collection |
|
|
Immutable ordered collection |
|
|
Unordered (key,value) mapping |
|
|
Unordered collection of unique values |
As you can see, round, square, and curly brackets have distinct meanings when it comes to the type of collection produced. We’ll take a quick tour of these data structures here.
Lists are the basic ordered and mutable data collection type in Python. They can be defined with comma-separated values between square brackets; here is a list of the first several prime numbers:
In
[
1
]:
L
=
[
2
,
3
,
5
,
7
]
Lists have a number of useful properties and methods available to them. Here we’ll take a quick look at some of the more common and useful ones:
In
[
2
]:
# Length of a list
len
(
L
)
Out [2]: 4
In
[
3
]:
# Append a value to the end
L
.
append
(
11
)
L
Out [3]: [2, 3, 5, 7, 11]
In
[
4
]:
# Addition concatenates lists
L
+
[
13
,
17
,
19
]
Out [4]: [2, 3, 5, 7, 11, 13, 17, 19]
In
[
5
]:
# sort() method sorts in-place
L
=
[
2
,
5
,
1
,
6
,
3
,
4
]
L
.
sort
()
L
Out [5]: [1, 2, 3, 4, 5, 6]
In addition, there are many more built-in list methods; they are well-covered in Python’s online documentation.
While we’ve been demonstrating lists containing values of a single type, one of the powerful features of Python’s compound objects is that they can contain objects of any type, or even a mix of types. For example:
In
[
6
]:
L
=
[
1
,
'two'
,
3.14
,
[
0
,
3
,
5
]]
This flexibility is a consequence of Python’s dynamic type system. Creating such a mixed sequence in a statically typed language like C can be much more of a headache! We see that lists can even contain other lists as elements. Such type flexibility is an essential piece of what makes Python code relatively quick and easy to write.
So far we’ve been considering manipulations of lists as a whole; another essential piece is the accessing of individual elements. This is done in Python via indexing and slicing, which we’ll explore next.
Python provides access to elements in compound types through indexing for single elements, and slicing for multiple elements. As we’ll see, both are indicated by a square-bracket syntax. Suppose we return to our list of the first several primes:
In
[
7
]:
L
=
[
2
,
3
,
5
,
7
,
11
]
Python uses zero-based indexing, so we can access the first and second element in using the following syntax:
In
[
8
]:
L
[
0
]
Out [8]: 2
In
[
9
]:
L
[
1
]
Out [9]: 3
Elements at the end of the list can be accessed with negative numbers, starting from -1:
In
[
10
]:
L
[
-
1
]
Out [10]: 11
In
[
12
]:
L
[
-
2
]
Out [12]: 7
You can visualize this indexing scheme this way:
Here values in the list are represented by large numbers in the squares;
list indices are represented by small numbers above and below. In this
case, L[2]
returns 5
, because that is the next value at index 2
.
Where indexing is a means of fetching a single value from the list, slicing is a means of accessing multiple values in sublists. It uses a colon to indicate the start point (inclusive) and end point (non-inclusive) of the subarray. For example, to get the first three elements of the list, we can write it as follows:
In
[
12
]:
L
[
0
:
3
]
Out [12]: [2, 3, 5]
Notice where 0
and 3
lie in the preceding diagram, and how the slice
takes just the values between the indices. If we leave out the first
index, 0
is assumed, so we can equivalently write the following:
In
[
13
]:
L
[:
3
]
Out [13]: [2, 3, 5]
Similarly, if we leave out the last index, it defaults to the length of the list. Thus, the last three elements can be accessed as follows:
In
[
14
]:
L
[
-
3
:]
Out [14]: [5, 7, 11]
Finally, it is possible to specify a third integer that represents the step size; for example, to select every second element of the list, we can write:
In
[
15
]:
L
[::
2
]
# equivalent to L[0:len(L):2]
Out [15]: [2, 5, 11]
A particularly useful version of this is to specify a negative step, which will reverse the array:
In
[
16
]:
L
[::
-
1
]
Out [16]: [11, 7, 5, 3, 2]
Both indexing and slicing can be used to set elements as well as access them. The syntax is as you would expect:
In
[
17
]:
L
[
0
]
=
100
(
L
)
[100, 3, 5, 7, 11]
In
[
18
]:
L
[
1
:
3
]
=
[
55
,
56
]
(
L
)
[100, 55, 56, 7, 11]
A very similar slicing syntax is also used in many data science–oriented packages, including NumPy and Pandas (mentioned in the introduction).
Now that we have seen Python lists and how to access elements in ordered compound types, let’s take a look at the other three standard compound data types mentioned earlier.
Tuples are in many ways similar to lists, but they are defined with parentheses rather than square brackets:
In
[
19
]:
t
=
(
1
,
2
,
3
)
They can also be defined without any brackets at all:
In
[
20
]:
t
=
1
,
2
,
3
(
t
)
(1, 2, 3)
Like the lists discussed before, tuples have a length, and individual elements can be extracted using square-bracket indexing:
In
[
21
]:
len
(
t
)
Out [21]: 3
In
[
22
]:
t
[
0
]
Out [22]: 1
The main distinguishing feature of tuples is that they are immutable: this means that once they are created, their size and contents cannot be changed:
In
[
23
]:
t
[
1
]
=
4
--------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-23-141c76cb54a2> in <module>() ----> 1 t[1] = 4 TypeError: 'tuple' object does not support item assignment
In
[
24
]:
t
.
append
(
4
)
--------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-24-e8bd1632f9dd> in <module>() ----> 1 t.append(4) AttributeError: 'tuple' object has no attribute 'append'
Tuples are often used in a Python program; a particularly common case is
in functions that have multiple return values. For example, the
as_integer_ratio()
method of floating-point objects returns a
numerator and a denominator; this dual return value comes in the form of
a tuple:
In
[
25
]:
x
=
0.125
x
.
as_integer_ratio
()
Out [25]: (1, 8)
These multiple return values can be individually assigned as follows:
In
[
26
]:
numerator
,
denominator
=
x
.
as_integer_ratio
()
(
numerator
/
denominator
)
0.125
The indexing and slicing logic covered earlier for lists works for tuples as well, along with a host of other methods. Refer to the Data Structures documentation for a more complete list of these.
Dictionaries are extremely flexible mappings of keys to values, and form
the basis of much of Python’s internal implementation. They can be
created via a comma-separated list of key:value
pairs within curly
braces:
In
[
27
]:
numbers
=
{
'one'
:
1
,
'two'
:
2
,
'three'
:
3
}
Items are accessed and set via the indexing syntax used for lists and tuples, except here the index is not a zero-based order but valid key in the dictionary:
In
[
28
]:
# Access a value via the key
numbers
[
'two'
]
Out [28]: 2
New items can be added to the dictionary using indexing as well:
In
[
29
]:
# Set a new key/value pair
numbers
[
'ninety'
]
=
90
(
numbers
)
{'three': 3, 'ninety': 90, 'two': 2, 'one': 1}
Keep in mind that dictionaries do not maintain any sense of order for the input parameters; this is by design. This lack of ordering allows dictionaries to be implemented very efficiently, so that random element access is very fast, regardless of the size of the dictionary (if you’re curious how this works, read about the concept of a hash table). The Python documentation has a complete list of the methods available for dictionaries.
The fourth basic collection is the set, which contains unordered collections of unique items. They are defined much like lists and tuples, except they use the curly brackets of dictionaries:
In
[
30
]:
primes
=
{
2
,
3
,
5
,
7
}
odds
=
{
1
,
3
,
5
,
7
,
9
}
If you’re familiar with the mathematics of sets, you’ll be familiar with operations like the union, intersection, difference, symmetric difference, and others. Python’s sets have all of these operations built in via methods or operators. For each, we’ll show the two equivalent methods:
In
[
31
]:
# union: items appearing in either
primes
|
odds
# with an operator
primes
.
union
(
odds
)
# equivalently with a method
Out [31]: {1, 2, 3, 5, 7, 9}
In
[
32
]:
# intersection: items appearing in both
primes
&
odds
# with an operator
primes
.
intersection
(
odds
)
# equivalently with a method
Out [32]: {3, 5, 7}
In
[
33
]:
# difference: items in primes but not in odds
primes
-
odds
# with an operator
primes
.
difference
(
odds
)
# equivalently with a method
Out [33]: {2}
In
[
34
]:
# symmetric difference: items appearing in only one set
primes
^
odds
# with an operator
primes
.
symmetric_difference
(
odds
)
# equivalently with a method
Out [34]: {1, 2, 9}
Many more set methods and operations are available. You’ve probably already guessed what I’ll say next: refer to Python’s online documentation for a complete reference.
Python contains several other data structures that you might find
useful; these can generally be found in the built-in collections
module. The collections
module is fully documented in Python’s online
documentation, and you can read more about the various objects available
there.
In particular, I’ve found the following very useful on occasion:
collections.namedtuple
Like a tuple, but each value has a name
collections.defaultdict
Like a dictionary, but unspecified keys have a user-specified default value
collections.OrderedDict
Like a dictionary, but the order of keys is maintained
Once you’ve seen the standard built-in collection types, the use of these extended functionalities is very intuitive, and I’d suggest reading about their use.
Control flow is where the rubber really meets the road in programming. Without it, a program is simply a list of statements that are sequentially executed. With control flow, you can execute certain code blocks conditionally and/or repeatedly: these basic building blocks can be combined to create surprisingly sophisticated programs!
Here we’ll cover conditional statements (including if
, elif
,
and else
) and loop statements (including for
and while
, and the
accompanying break
, continue
, and pass
).
Conditional statements, often referred to as if-then statements, allow the programmer to execute certain pieces of code depending on some Boolean condition. A basic example of a Python conditional statement is this:
In
[
1
]:
x
=
-
15
if
x
==
0
:
(
x
,
"is zero"
)
elif
x
>
0
:
(
x
,
"is positive"
)
elif
x
<
0
:
(
x
,
"is negative"
)
else
:
(
x
,
"is unlike anything I've ever seen..."
)
-15 is negative
Note especially the use of colons (:
) and whitespace to denote
separate blocks of code.
Python adopts the if
and else
often used in other languages; its
more unique keyword is elif
, a contraction of “else if”. In these
conditional clauses, elif
and else
blocks are optional;
additionally, you can optionally include as few or as many elif
statements as you would like.
Loops in Python are a way to repeatedly execute some code statement. So,
for example, if we’d like to print each of the items in a list, we can
use a for
loop:
In
[
2
]:
for
N
in
[
2
,
3
,
5
,
7
]:
(
N
,
end
=
' '
)
# print all on same line
2 3 5 7
Notice the simplicity of the for
loop: we specify the variable we want
to use, the sequence we want to loop over, and use the in
operator
to link them together in an intuitive and readable way. More precisely,
the object to the right of the in
can be any Python iterator. An
iterator can be thought of as a generalized sequence, and we’ll discuss
them in “Iterators”.
For example, one of the most commonly used iterators in Python is the
range
object, which generates a sequence of numbers:
In
[
3
]:
for
i
in
range
(
10
):
(
i
,
end
=
' '
)
0 1 2 3 4 5 6 7 8 9
Note that the range starts at zero by default, and that by convention the top of the range is not included in the output. Range objects can also have more complicated values:
In
[
4
]:
# range from 5 to 10
list
(
range
(
5
,
10
))
Out [4]: [5, 6, 7, 8, 9]
In
[
5
]:
# range from 0 to 10 by 2
list
(
range
(
0
,
10
,
2
))
Out [5]: [0, 2, 4, 6, 8]
You might notice that the meaning of range
arguments is very similar to
the slicing syntax that we covered in “Lists”.
Note that the behavior of range()
is one of the differences between
Python 2 and Python 3: in Python 2, range()
produces a list, while in
Python 3, range()
produces an iterable object.
The other type of loop in Python is a while
loop, which iterates until
some condition is met:
In
[
6
]:
i
=
0
while
i
<
10
:
(
i
,
end
=
' '
)
i
+=
1
0 1 2 3 4 5 6 7 8 9
The argument of the while
loop is evaluated as a Boolean statement,
and the loop is executed until the statement evaluates to False.
There are two useful statements that can be used within loops to fine-tune how they are executed:
The break
statement breaks out of the loop entirely
The continue
statement skips the remainder of the current loop, and
goes to the next iteration
These can be used in both for
and while
loops.
Here is an example of using continue
to print a string of even
numbers. In this case, the result could be accomplished just as well with
an if-else
statement, but sometimes the continue
statement can be a
more convenient way to express the idea you have in mind:
In
[
7
]:
for
n
in
range
(
20
):
# check if n is even
if
n
%
2
==
0
:
continue
(
n
,
end
=
' '
)
1 3 5 7 9 11 13 15 17 19
Here is an example of a break
statement used for a less trivial task.
This loop will fill a list with all Fibonacci numbers up to a certain
value:
In
[
8
]:
a
,
b
=
0
,
1
amax
=
100
L
=
[]
while
True
:
(
a
,
b
)
=
(
b
,
a
+
b
)
if
a
>
amax
:
break
L
.
append
(
a
)
(
L
)
[1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89]
Notice that we use a while True
loop, which will loop forever unless
we have a break statement!
One rarely used pattern available in Python is the else
statement as
part of a for
or while
loop. We discussed the else
block earlier: it
executes if all the if
and elif
statements evaluate to False
. The
loop-else
is perhaps one of the more confusingly named statements in
Python; I prefer to think of it as a nobreak
statement: that is, the
else
block is executed only if the loop ends naturally, without
encountering a break
statement.
As an example of where this might be useful, consider the following (non-optimized) implementation of the Sieve of Eratosthenes, a well-known algorithm for finding prime numbers:
In
[
9
]:
L
=
[]
nmax
=
30
for
n
in
range
(
2
,
nmax
):
for
factor
in
L
:
if
n
%
factor
==
0
:
break
else
:
# no break
L
.
append
(
n
)
(
L
)
[2, 3, 5, 7, 11, 13, 17, 19, 23, 29]
The else
statement only executes if none of the factors divide the
given number. The else
statement works similarly with the while
loop.
So far, our scripts have been simple, single-use code blocks. One way to
organize our Python code and to make it more readable and reusable is to
factor-out useful pieces into reusable functions. Here we’ll cover two
ways of creating functions: the def
statement, useful for any type of
function, and the lambda
statement, useful for creating short
anonymous functions.
Functions are groups of code that have a name and can be called using
parentheses. We’ve seen functions before. For example, print
in Python
3 is a function:
In
[
1
]:
(
'abc'
)
abc
Here print
is the function name, and 'abc'
is the function’s
argument.
In addition to arguments, there are keyword arguments that are
specified by name. One available keyword argument for the print()
function (in Python 3) is sep
, which tells what character or
characters should be used to separate multiple items:
In
[
2
]:
(
1
,
2
,
3
)
1 2 3
In
[
3
]:
(
1
,
2
,
3
,
sep
=
'--'
)
1--2--3
When non-keyword arguments are used together with keyword arguments, the keyword arguments must come at the end.
Functions become even more useful when we begin to define our own,
organizing functionality to be used in multiple places. In Python,
functions are defined with the def
statement. For example, we can
encapsulate a version of our Fibonacci sequence code from the previous
section as follows:
In
[
4
]:
def
fibonacci
(
N
):
L
=
[]
a
,
b
=
0
,
1
while
len
(
L
)
<
N
:
a
,
b
=
b
,
a
+
b
L
.
append
(
a
)
return
L
Now we have a function named fibonacci
which takes a single argument
N
, does something with this argument, and return
s a value; in this
case, a list of the first N
Fibonacci numbers:
In
[
5
]:
fibonacci
(
10
)
Out [5]: [1, 1, 2, 3, 5, 8, 13, 21, 34, 55]
If you’re familiar with strongly typed languages like C
, you’ll immediately
notice that there is no type information associated with the function
inputs or outputs. Python functions can return any Python object, simple
or compound, which means constructs that may be difficult in other
languages are straightforward in Python.
For example, multiple return values are simply put in a tuple, which is indicated by commas:
In
[
6
]:
def
real_imag_conj
(
val
):
return
val
.
real
,
val
.
imag
,
val
.
conjugate
()
r
,
i
,
c
=
real_imag_conj
(
3
+
4j
)
(
r
,
i
,
c
)
3.0 4.0 (3-4j)
Often when defining a function, there are certain values that we want
the function to use most of the time, but we’d also like to give the
user some flexibility. In this case, we can use default values for
arguments. Consider the fibonacci
function from before. What if we
would like the user to be able to play with the starting values? We
could do that as follows:
In
[
7
]:
def
fibonacci
(
N
,
a
=
0
,
b
=
1
):
L
=
[]
while
len
(
L
)
<
N
:
a
,
b
=
b
,
a
+
b
L
.
append
(
a
)
return
L
With a single argument, the result of the function call is identical to before:
In
[
8
]:
fibonacci
(
10
)
Out [8]: [1, 1, 2, 3, 5, 8, 13, 21, 34, 55]
But now we can use the function to explore new things, such as the effect of new starting values:
In
[
9
]:
fibonacci
(
10
,
0
,
2
)
Out [9]: [2, 2, 4, 6, 10, 16, 26, 42, 68, 110]
The values can also be specified by name if desired, in which case the order of the named values does not matter:
In
[
10
]:
fibonacci
(
10
,
b
=
3
,
a
=
1
)
Out [10]: [3, 4, 7, 11, 18, 29, 47, 76, 123, 199]
Sometimes you might wish to write a function in which you don’t
initially know how many arguments the user will pass. In this case, you
can use the special form *args
and **kwargs
to catch all arguments
that are passed. Here is an example:
In
[
11
]:
def
catch_all
(
*
args
,
**
kwargs
):
(
"args ="
,
args
)
(
"kwargs = "
,
kwargs
)
In
[
12
]:
catch_all
(
1
,
2
,
3
,
a
=
4
,
b
=
5
)
args = (1, 2, 3) kwargs = {'a': 4, 'b': 5}
In
[
13
]:
catch_all
(
'a'
,
keyword
=
2
)
args = ('a',) kwargs = {'keyword': 2}
Here it is not the names args
and kwargs
that are important, but
the *
characters preceding them. args
and kwargs
are just the
variable names often used by convention, short for “arguments” and
“keyword arguments”. The operative difference is the asterisk
characters: a single *
before a variable means “expand this as a
sequence”, while a double **
before a variable means “expand this as a
dictionary”. In fact, this syntax can be used not only with the function
definition, but with the function call as well!
In
[
14
]:
inputs
=
(
1
,
2
,
3
)
keywords
=
{
'pi'
:
3.14
}
catch_all
(
*
inputs
,
**
keywords
)
args = (1, 2, 3) kwargs = {'pi': 3.14}
Earlier we quickly covered the most common way of defining functions, the
def
statement. You’ll likely come across another way of defining
short, one-off functions with the lambda
statement. It looks something
like this:
In
[
15
]:
add
=
lambda
x
,
y
:
x
+
y
add
(
1
,
2
)
Out [15]: 3
This lambda function is roughly equivalent to
In
[
16
]:
def
add
(
x
,
y
):
return
x
+
y
So why would you ever want to use such a thing? Primarily, it comes down to the fact that everything is an object in Python, even functions themselves! That means that functions can be passed as arguments to functions.
As an example of this, suppose we have some data stored in a list of dictionaries:
In
[
17
]:
data
=
[{
'first'
:
'Guido'
,
'last'
:
'Van Rossum'
,
'YOB'
:
1956
},
{
'first'
:
'Grace'
,
'last'
:
'Hopper'
,
'YOB'
:
1906
},
{
'first'
:
'Alan'
,
'last'
:
'Turing'
,
'YOB'
:
1912
}]
Now suppose we want to sort this data. Python has a sorted
function
that does this:
In
[
18
]:
sorted
([
2
,
4
,
3
,
5
,
1
,
6
])
Out [18]: [1, 2, 3, 4, 5, 6]
But dictionaries are not orderable: we need a way to tell the function
how to sort our data. We can do this by specifying the key
function,
a function which given an item returns the sorting key for that item:
In
[
19
]:
# sort alphabetically by first name
sorted
(
data
,
key
=
lambda
item
:
item
[
'first'
])
Out [19]: [{'YOB': 1912, 'first': 'Alan', 'last': 'Turing'}, {'YOB': 1906, 'first': 'Grace', 'last': 'Hopper'}, {'YOB': 1956, 'first': 'Guido', 'last': 'Van Rossum'}]
In
[
20
]:
# sort by year of birth
sorted
(
data
,
key
=
lambda
item
:
item
[
'YOB'
])
Out [20]: [{'YOB': 1906, 'first': 'Grace', 'last': 'Hopper'}, {'YOB': 1912, 'first': 'Alan', 'last': 'Turing'}, {'YOB': 1956, 'first': 'Guido', 'last': 'Van Rossum'}]
While these key functions could certainly be created by the normal,
def
syntax, the lambda
syntax is convenient for such short one-off
functions like these.
No matter your skill as a programmer, you will eventually make a coding mistake. Such mistakes come in three basic flavors:
Errors where the code is not valid Python (generally easy to fix)
Errors where syntactically valid code fails to execute, perhaps due to invalid user input (sometimes easy to fix)
Errors in logic: code executes without a problem, but the result is not what you expect (often very difficult to identify and fix)
Here we’re going to focus on how to deal cleanly with runtime errors. As we’ll see, Python handles runtime errors via its exception handling framework.
If you’ve done any coding in Python, you’ve likely come across runtime errors. They can happen in a lot of ways.
For example, if you try to reference an undefined variable:
In
[
1
]:
(
Q
)
--------------------------------------------------------- NameError Traceback (most recent call last) <ipython-input-3-e796bdcf24ff> in <module>() ----> 1 print(Q) NameError: name 'Q' is not defined
Or if you try an operation that’s not defined:
In
[
2
]:
1
+
'abc'
--------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-4-aab9e8ede4f7> in <module>() ----> 1 1 + 'abc' TypeError: unsupported operand type(s) for +: 'int' and 'str'
Or you might be trying to compute a mathematically ill-defined result:
In
[
3
]:
2
/
0
--------------------------------------------------------- ZeroDivisionError Traceback (most recent call last) <ipython-input-5-ae0c5d243292> in <module>() ----> 1 2 / 0 ZeroDivisionError: division by zero
Or maybe you’re trying to access a sequence element that doesn’t exist:
In
[
4
]:
L
=
[
1
,
2
,
3
]
L
[
1000
]
--------------------------------------------------------- IndexError Traceback (most recent call last) <ipython-input-6-06b6eb1b8957> in <module>() 1 L = [1, 2, 3] ----> 2 L[1000] IndexError: list index out of range
Note that in each case, Python is kind enough to not simply indicate that an error happened, but to spit out a meaningful exception that includes information about what exactly went wrong, along with the exact line of code where the error happened. Having access to meaningful errors like this is immensely useful when trying to trace the root of problems in your code.
The main tool Python gives you for handling runtime exceptions is the
try
…except
clause. Its basic structure is this:
In
[
5
]:
try
:
(
"this gets executed first"
)
except
:
(
"this gets executed only if there is an error"
)
this gets executed first
Note that the second block here did not get executed: this is because
the first block did not return an error. Let’s put a problematic
statement in the try
block and see what happens:
In
[
6
]:
try
:
(
"let's try something:"
)
x
=
1
/
0
# ZeroDivisionError
except
:
(
"something bad happened!"
)
let's try something: something bad happened!
Here we see that when the error was raised in the try
statement (in
this case, a ZeroDivisionError
), the error was caught, and the except
statement was executed.
One way this is often used is to check user input within a function or another piece of code. For example, we might wish to have a function that catches zero-division and returns some other value, perhaps a suitably large number like 10100:
In
[
7
]:
def
safe_divide
(
a
,
b
):
try
:
return
a
/
b
except
:
return
1E100
In
[
8
]:
safe_divide
(
1
,
2
)
Out [8]: 0.5
In
[
9
]:
safe_divide
(
2
,
0
)
Out [9]: 1e+100
There is a subtle problem with this code, though: what happens when another type of exception comes up? For example, this is probably not what we intended:
In
[
10
]:
safe_divide
(
1
,
'2'
)
Out [10]: 1e+100
Dividing an integer and a string raises a TypeError
, which our
over-zealous code caught and assumed was a ZeroDivisionError
! For this
reason, it’s nearly always a better idea to catch exceptions
explicitly:
In
[
11
]:
def
safe_divide
(
a
,
b
):
try
:
return
a
/
b
except
ZeroDivisionError
:
return
1E100
In
[
12
]:
safe_divide
(
1
,
0
)
Out [12]: 1e+100
In
[
13
]:
safe_divide
(
1
,
'2'
)
--------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-15-2331af6a0acf> in <module>() ----> 1 safe_divide(1, '2') <ipython-input-13-10b5f0163af8> in safe_divide(a, b) 1 def safe_divide(a, b): 2 try: ----> 3 return a / b 4 except ZeroDivisionError: 5 return 1E100 TypeError: unsupported operand type(s) for /: 'int' and 'str'
We’re now catching zero-division errors only, and letting all other errors pass through unmodified.
We’ve seen how valuable it is to have informative exceptions when using parts of the Python language. It’s equally valuable to make use of informative exceptions within the code you write, so that users of your code (foremost yourself!) can figure out what caused their errors.
The way you raise your own exceptions is with the raise
statement. For
example:
In
[
14
]:
raise
RuntimeError
(
"my error message"
)
--------------------------------------------------------- RuntimeError Traceback (most recent call last) <ipython-input-16-c6a4c1ed2f34> in <module>() ----> 1 raise RuntimeError("my error message") RuntimeError: my error message
As an example of where this might be useful, let’s return to the
fibonacci
function that we defined previously:
In
[
15
]:
def
fibonacci
(
N
):
L
=
[]
a
,
b
=
0
,
1
while
len
(
L
)
<
N
:
a
,
b
=
b
,
a
+
b
L
.
append
(
a
)
return
L
One potential problem here is that the input value could be negative.
This will not currently cause any error in our function, but we might
want to let the user know that a negative N
is not supported. Errors
stemming from invalid parameter values, by convention, lead to a
ValueError
being raised:
In
[
16
]:
def
fibonacci
(
N
):
if
N
<
0
:
raise
ValueError
(
"N must be non-negative"
)
L
=
[]
a
,
b
=
0
,
1
while
len
(
L
)
<
N
:
a
,
b
=
b
,
a
+
b
L
.
append
(
a
)
return
L
In
[
17
]:
fibonacci
(
10
)
Out [17]: [1, 1, 2, 3, 5, 8, 13, 21, 34, 55]
In
[
18
]:
fibonacci
(
-
10
)
--------------------------------------------------------- RuntimeError Traceback (most recent call last) <ipython-input-20-3d291499cfa7> in <module>() ----> 1 fibonacci(-10) <ipython-input-18-01d0cf168d63> in fibonacci(N) 1 def fibonacci(N): 2 if N < 0: ----> 3 raise ValueError("N must be non-negative") 4 L = [] 5 a, b = 0, 1 ValueError: N must be non-negative
Now the user knows exactly why the input is invalid, and could even use
a try
…except
block to handle it!
In
[
19
]:
N
=
-
10
try
:
(
"trying this..."
)
(
fibonacci
(
N
))
except
ValueError
:
(
"Bad value: need to do something else"
)
trying this... Bad value: need to do something else
Briefly, I want to mention here some other concepts you might run into. I’ll not go into detail on these concepts and how and why to use them, but instead simply show you the syntax so you can explore more on your own.
Sometimes in a try
…except
statement, you would like to be able to
work with the error message itself. This can be done with the as
keyword:
In
[
20
]:
try
:
x
=
1
/
0
except
ZeroDivisionError
as
err
:
(
"Error class is: "
,
type
(
err
))
(
"Error message is:"
,
err
)
Error class is: <class 'ZeroDivisionError'> Error message is: division by zero
With this pattern, you can further customize the exception handling of your function.
In addition to built-in exceptions, it is possible to define custom
exceptions through class inheritance. For instance, if you want a
special kind of ValueError
, you can do this:
In
[
21
]:
class
MySpecialError
(
ValueError
):
pass
raise
MySpecialError
(
"here's the message"
)
--------------------------------------------------------- MySpecialError Traceback (most recent call last) <ipython-input-23-92c36e04a9d0> in <module>() 2 pass 3 ----> 4 raise MySpecialError("here's the message") MySpecialError: here's the message
This would allow you to use a try
…except
block that only catches
this type of error:
In
[
22
]:
try
:
(
"do something"
)
raise
MySpecialError
(
"[informative error message here]"
)
except
MySpecialError
:
(
"do something else"
)
do something do something else
You might find this useful as you develop more customized code.
In addition to try
and except
, you can use the else
and finally
keywords to further tune your code’s handling of exceptions. The basic
structure is this:
In
[
23
]:
try
:
(
"try something here"
)
except
:
(
"this happens only if it fails"
)
else
:
(
"this happens only if it succeeds"
)
finally
:
(
"this happens no matter what"
)
try something here this happens only if it succeeds this happens no matter what
The utility of else
here is clear, but what’s the point of finally
?
Well, the finally
clause really is executed no matter what: I
usually see it used to do some sort of cleanup after an operation
completes.
Often an important piece of data analysis is repeating a similar
calculation, over and over, in an automated fashion. For example, you
may have a table of names that you’d like to split into first and
last, or perhaps of dates that you’d like to convert to some standard
format. One of Python’s answers to this is the iterator syntax. We’ve
seen this already with the range
iterator:
In
[
1
]:
for
i
in
range
(
10
):
(
i
,
end
=
' '
)
0 1 2 3 4 5 6 7 8 9
Here we’re going to dig a bit deeper. It turns out that in Python 3,
range
is not a list, but is something called an iterator, and
learning how it works is key to understanding a wide class of very
useful Python functionality.
Iterators are perhaps most easily understood in the concrete case of iterating through a list. Consider the following:
In
[
2
]:
for
value
in
[
2
,
4
,
6
,
8
,
10
]:
# do some operation
(
value
+
1
,
end
=
' '
)
3 5 7 9 11
The familiar for x in y
syntax allows us to repeat some operation
for each value in the list. The fact that the syntax of the code is so
close to its English description (for [each] value in [the] list) is
just one of the syntactic choices that makes Python such an intuitive
language to learn and use.
But the face-value behavior is not what’s really happening. When you
write something like for val in L
, the Python interpreter checks
whether it has an iterator interface, which you can check yourself
with the built-in iter
function:
In
[
3
]:
iter
([
2
,
4
,
6
,
8
,
10
])
Out [3]: <list_iterator at 0x104722400>
It is this iterator object that provides the functionality required by
the for
loop. The iter
object is a container that gives you access
to the next object for as long as it’s valid, which can be seen with the
built-in function next
:
In
[
4
]:
I
=
iter
([
2
,
4
,
6
,
8
,
10
])
In
[
5
]:
(
next
(
I
))
2
In
[
6
]:
(
next
(
I
))
4
In
[
7
]:
(
next
(
I
))
6
What is the purpose of this level of indirection? Well, it turns out this is incredibly useful, because it allows Python to treat things as lists that are not actually lists.
Perhaps the most common example of this indirect iteration is the
range()
function in Python 3 (named xrange()
in Python 2), which
returns not a list, but a special range()
object:
In
[
8
]:
range
(
10
)
Out [8]: range(0, 10)
range
, like a list, exposes an iterator:
In
[
9
]:
iter
(
range
(
10
))
Out [9]: <range_iterator at 0x1045a1810>
So Python knows to treat it as if it’s a list:
In
[
10
]:
for
i
in
range
(
10
):
(
i
,
end
=
' '
)
0 1 2 3 4 5 6 7 8 9
The benefit of the iterator indirection is that the full list is never
explicitly created! We can see this by doing a range calculation that
would overwhelm our system memory if we actually instantiated it (note
that in Python 2, range
creates a list, so running the following will
not lead to good things!):
In
[
11
]:
N
=
10
**
12
for
i
in
range
(
N
):
if
i
>=
10
:
break
(
i
,
end
=
', '
)
0, 1, 2, 3, 4, 5, 6, 7, 8, 9,
If range
were to actually create that list of one trillion values, it
would occupy tens of terabytes of machine memory: a waste, given the
fact that we’re ignoring all but the first 10 values!
In fact, there’s no reason that iterators ever have to end at all!
Python’s itertools
library contains a count
function that acts as
an infinite range:
In
[
12
]:
from
itertools
import
count
for
i
in
count
():
if
i
>=
10
:
break
(
i
,
end
=
', '
)
0, 1, 2, 3, 4, 5, 6, 7, 8, 9,
Had we not thrown in a loop break here, it would go on happily counting
until the process is manually interrupted or killed (using, for example,
ctrl-C
).
This iterator syntax is used nearly universally in Python built-in types as well as the more data science–specific object we’ll explore in later sections. Here we’ll cover some of the more useful iterators in the Python language.
Often you need to iterate not only the values in an array, but also keep track of the index. You might be tempted to do things this way:
In
[
13
]:
L
=
[
2
,
4
,
6
,
8
,
10
]
for
i
in
range
(
len
(
L
)):
(
i
,
L
[
i
])
0 2 1 4 2 6 3 8 4 10
Although this does work, Python provides a cleaner syntax using the enumerate
iterator:
In
[
14
]:
for
i
,
val
in
enumerate
(
L
):
(
i
,
val
)
0 2 1 4 2 6 3 8 4 10
This is the more “Pythonic” way to enumerate the indices and values in a list.
Other times, you may have multiple lists that you want to iterate over
simultaneously. You could certainly iterate over the index as in the
non-Pythonic example we looked at previously, but it is better to use the zip
iterator,
which zips together iterables:
In
[
15
]:
L
=
[
2
,
4
,
6
,
8
,
10
]
R
=
[
3
,
6
,
9
,
12
,
15
]
for
lval
,
rval
in
zip
(
L
,
R
):
(
lval
,
rval
)
2 3 4 6 6 9 8 12 10 15
Any number of iterables can be zipped together, and if they are
different lengths, the shortest will determine the length of the zip
.
The map
iterator takes a function and applies it to the values in an
iterator:
In
[
16
]:
# find the first 10 square numbers
square
=
lambda
x
:
x
**
2
for
val
in
map
(
square
,
range
(
10
)):
(
val
,
end
=
' '
)
0 1 4 9 16 25 36 49 64 81
The filter
iterator looks similar, except it only passes through
values for which the filter function evaluates to True:
In
[
17
]:
# find values up to 10 for which x % 2 is zero
is_even
=
lambda
x
:
x
%
2
==
0
for
val
in
filter
(
is_even
,
range
(
10
)):
(
val
,
end
=
' '
)
0 2 4 6 8
The map
and filter
functions, along with the reduce
function
(which lives in Python’s functools
module) are fundamental components
of the functional programming style, which, while not a dominant
programming style in the Python world, has its outspoken proponents (see,
for example, the pytoolz library).
We saw in “*args and **kwargs: Flexible Arguments” that *args
and **kwargs
can be used to pass
sequences and dictionaries to functions. It turns out that the *args
syntax works not just with sequences, but with any iterator:
In
[
18
]:
(
*
range
(
10
))
0 1 2 3 4 5 6 7 8 9
So, for example, we can get tricky and compress the map
example from before
into the following:
In
[
19
]:
(
*
map
(
lambda
x
:
x
**
2
,
range
(
10
)))
0 1 4 9 16 25 36 49 64 81
Using this trick lets us answer the age-old question that comes up in
Python learners’ forums: why is there no unzip()
function that does
the opposite of zip()
? If you lock yourself in a dark closet and think
about it for a while, you might realize that the opposite of zip()
is… zip()
! The key is that zip()
can zip together any number of
iterators or sequences. Observe:
In
[
20
]:
L1
=
(
1
,
2
,
3
,
4
)
L2
=
(
'a'
,
'b'
,
'c'
,
'd'
)
In
[
21
]:
z
=
zip
(
L1
,
L2
)
(
*
z
)
(1, 'a') (2, 'b') (3, 'c') (4, 'd')
In
[
22
]:
z
=
zip
(
L1
,
L2
)
new_L1
,
new_L2
=
zip
(
*
z
)
(
new_L1
,
new_L2
)
(1, 2, 3, 4) ('a', 'b', 'c', 'd')
Ponder this for a while. If you understand why it works, you’ll have come a long way in understanding Python iterators!
We briefly looked at the infinite range
iterator,
itertools.count
, earlier. The itertools
module contains a whole host of
useful iterators; it’s well worth your while to explore the module
to see what’s available. As an example, consider the
itertools.permutations
function, which iterates over all permutations
of a sequence:
In
[
23
]:
from
itertools
import
permutations
p
=
permutations
(
range
(
3
))
(
*
p
)
(0, 1, 2) (0, 2, 1) (1, 0, 2) (1, 2, 0) (2, 0, 1) (2, 1, 0)
Similarly, the itertools.combinations
function iterates over all
unique combinations of N
values within a list:
In
[
24
]:
from
itertools
import
combinations
c
=
combinations
(
range
(
4
),
2
)
(
*
c
)
(0, 1) (0, 2) (0, 3) (1, 2) (1, 3) (2, 3)
Somewhat related is the product
iterator, which iterates over all sets
of pairs between two or more iterables:
In
[
25
]:
from
itertools
import
product
p
=
product
(
'ab'
,
range
(
3
))
(
*
p
)
('a', 0) ('a', 1) ('a', 2) ('b', 0) ('b', 1) ('b', 2)
Many more useful iterators exist in itertools
: the full list can be
found, along with some examples, in Python’s
online
documentation.
If you read enough Python code, you’ll eventually come across the terse and efficient construction known as a list comprehension. This is one feature of Python I expect you will fall in love with if you’ve not used it before; it looks something like this:
In
[
1
]:
[
i
for
i
in
range
(
20
)
if
i
%
3
>
0
]
Out [1]: [1, 2, 4, 5, 7, 8, 10, 11, 13, 14, 16, 17, 19]
The result of this is a list of numbers that excludes multiples of 3. While this example may seem a bit confusing at first, as familiarity with Python grows, reading and writing list comprehensions will become second nature.
List comprehensions are simply a way to compress a list-building
for
loop into a single short, readable line. For example, here is a loop
that constructs a list of the first 12 square integers:
In
[
2
]:
L
=
[]
for
n
in
range
(
12
):
L
.
append
(
n
**
2
)
L
Out [2]: [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121]
The list comprehension equivalent of this is the following:
In
[
3
]:
[
n
**
2
for
n
in
range
(
12
)]
Out [3]: [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121]
As with many Python statements, you can almost read off the meaning of
this statement in plain English: “construct a list consisting of the
square of n
for each n
up to 12”.
This basic syntax, then, is [expr for var in
iterable]
, where expr
is any valid expression, var
is a
variable name, and iterable
is any iterable Python object.
Sometimes you want to build a list not just from one value, but from
two. To do this, simply add another for
expression in the
comprehension:
In
[
4
]:
[(
i
,
j
)
for
i
in
range
(
2
)
for
j
in
range
(
3
)]
Out [4]: [(0, 0), (0, 1), (0, 2), (1, 0), (1, 1), (1, 2)]
Notice that the second for
expression acts as the interior index,
varying the fastest in the resulting list. This type of construction can
be extended to three, four, or more iterators within the comprehension, though
at some point code readability will suffer!
You can further control the iteration by adding a conditional to the end of the expression. In the first example of the section, we iterated over all numbers from 1 to 20, but left out multiples of 3. Look at this again, and notice the construction:
In
[
5
]:
[
val
for
val
in
range
(
20
)
if
val
%
3
>
0
]
Out [5]: [1, 2, 4, 5, 7, 8, 10, 11, 13, 14, 16, 17, 19]
The expression (i % 3 > 0)
evaluates to True
unless val
is
divisible by 3. Again, the English language meaning can be immediately
read off: “Construct a list of values for each value up to 20, but only
if the value is not divisible by 3”. Once you are comfortable with it, this is
much easier to write—and to understand at a glance—than the
equivalent loop syntax:
In
[
6
]:
L
=
[]
for
val
in
range
(
20
):
if
val
%
3
:
L
.
append
(
val
)
L
Out [6]: [1, 2, 4, 5, 7, 8, 10, 11, 13, 14, 16, 17, 19]
If you’ve programmed in C, you might be familiar with the single-line
conditional enabled by the ?
operator:
int
absval
=
(
val
<
0
)
?
-
val
:
val
Python has something very similar to this, which is most often used
within list comprehensions, lambda
functions, and other places where a
simple expression is desired:
In
[
7
]:
val
=
-
10
val
if
val
>=
0
else
-
val
Out [7]: 10
We see that this simply duplicates the functionality of the built-in
abs()
function, but the construction lets you do some really
interesting things within list comprehensions. This is getting pretty
complicated now, but you could do something like this:
In
[
8
]:
[
val
if
val
%
2
else
-
val
for
val
in
range
(
20
)
if
val
%
3
]
Out [8]: [1, -2, -4, 5, 7, -8, -10, 11, 13, -14, -16, 17, 19]
Note the line break within the list comprehension before the for
expression: this is valid in Python, and is often a nice way to break-up
long list comprehensions for greater readability. Look this over: what
we’re doing is constructing a list, leaving out multiples of 3, and
negating all multiples of 2.
Once you understand the dynamics of list comprehensions, it’s straightforward to move on to other types of comprehensions. The syntax is largely the same; the only difference is the type of bracket you use.
For example, with curly braces you can create a set
with a set
comprehension:
In
[
9
]:
{
n
**
2
for
n
in
range
(
12
)}
Out [9]: {0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121}
Recall that a set
is a collection that contains no duplicates. The
set comprehension respects this rule, and eliminates any duplicate
entries:
In
[
10
]:
{
a
%
3
for
a
in
range
(
1000
)}
Out [10]: {0, 1, 2}
With a slight tweak, you can add a colon (:
) to create a dict
comprehension:
In
[
11
]:
{
n
:
n
**
2
for
n
in
range
(
6
)}
Out [11]: {0: 0, 1: 1, 2: 4, 3: 9, 4: 16, 5: 25}
Finally, if you use parentheses rather than square brackets, you get what’s called a generator expression:
In
[
12
]:
(
n
**
2
for
n
in
range
(
12
))
Out [12]: <generator object <genexpr> at 0x1027a5a50>
A generator expression is essentially a list comprehension in which elements are generated as needed rather than all at once, and the simplicity here belies the power of this language feature: we’ll explore this more next.
Here we’ll take a deeper dive into Python generators, including generator expressions and generator functions.
The difference between list comprehensions and generator expressions is sometimes confusing; here we’ll quickly outline the differences between them.
This is a representative list comprehension:
In
[
1
]:
[
n
**
2
for
n
in
range
(
12
)]
Out [1]: [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121]
While this is a representative generator expression:
In
[
2
]:
(
n
**
2
for
n
in
range
(
12
))
Out [2]: <generator object <genexpr> at 0x104a60518>
Notice that printing the generator expression does not print the
contents; one way to print the contents of a generator expression is to
pass it to the list
constructor:
In
[
3
]:
G
=
(
n
**
2
for
n
in
range
(
12
))
list
(
G
)
Out [3]: [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121]
When you create a list, you are actually building a collection of values, and there is some memory cost associated with that. When you create a generator, you are not building a collection of values, but a recipe for producing those values. Both expose the same iterator interface, as we can see here:
In
[
4
]:
L
=
[
n
**
2
for
n
in
range
(
12
)]
for
val
in
L
:
(
val
,
end
=
' '
)
0 1 4 9 16 25 36 49 64 81 100 121
In
[
5
]:
G
=
(
n
**
2
for
n
in
range
(
12
))
for
val
in
G
:
(
val
,
end
=
' '
)
0 1 4 9 16 25 36 49 64 81 100 121
The difference is that a generator expression does not actually compute the values until they are needed. This not only leads to memory efficiency, but to computational efficiency as well! This also means that while the size of a list is limited by available memory, the size of a generator expression is unlimited!
An example of an infinite generator expression can be created using the count
iterator defined in itertools
:
In
[
6
]:
from
itertools
import
count
count
()
Out [6]: count(0)
In
[
7
]:
for
i
in
count
():
(
i
,
end
=
' '
)
if
i
>=
10
:
break
0 1 2 3 4 5 6 7 8 9 10
The count
iterator will go on happily counting forever until you tell
it to stop; this makes it convenient to create generators that will
also go on forever:
In
[
8
]:
factors
=
[
2
,
3
,
5
,
7
]
G
=
(
i
for
i
in
count
()
if
all
(
i
%
n
>
0
for
n
in
factors
))
for
val
in
G
:
(
val
,
end
=
' '
)
if
val
>
40
:
break
1 11 13 17 19 23 29 31 37 41
You might see what we’re getting at here: if we were to expand the list of factors appropriately, what we would have the beginnings of is a prime number generator, using the Sieve of Eratosthenes algorithm. We’ll explore this more momentarily.
This is one of those potential gotchas of generator expressions. With a list, we can straightforwardly do this:
In
[
9
]:
L
=
[
n
**
2
for
n
in
range
(
12
)]
for
val
in
L
:
(
val
,
end
=
' '
)
()
for
val
in
L
:
(
val
,
end
=
' '
)
0 1 4 9 16 25 36 49 64 81 100 121 0 1 4 9 16 25 36 49 64 81 100 121
A generator expression, on the other hand, is used up after one iteration:
In
[
10
]:
G
=
(
n
**
2
for
n
in
range
(
12
))
list
(
G
)
Out [10]: [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121]
In
[
11
]:
list
(
G
)
Out [11]: []
This can be very useful because it means iteration can be stopped and started:
In
[
12
]:
G
=
(
n
**
2
for
n
in
range
(
12
))
for
n
in
G
:
(
n
,
end
=
' '
)
if
n
>
30
:
break
(
"
doing something in between"
)
for
n
in
G
:
(
n
,
end
=
' '
)
0 1 4 9 16 25 36 doing something in between 49 64 81 100 121
One place I’ve found this useful is when working with collections of data files on disk; it means that you can quite easily analyze them in batches, letting the generator keep track of which ones you have yet to see.
We saw in the previous section that list comprehensions are best used to
create relatively simple lists, while using a normal for
loop can be
better in more complicated situations. The same is true of generator
expressions: we can make more complicated generators using generator
functions, which make use of the yield
statement.
Here we have two ways of constructing the same list:
In
[
13
]:
L1
=
[
n
**
2
for
n
in
range
(
12
)]
L2
=
[]
for
n
in
range
(
12
):
L2
.
append
(
n
**
2
)
(
L1
)
(
L2
)
[0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121] [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121]
Similarly, here we have two ways of constructing equivalent generators:
In
[
14
]:
G1
=
(
n
**
2
for
n
in
range
(
12
))
def
gen
():
for
n
in
range
(
12
):
yield
n
**
2
G2
=
gen
()
(
*
G1
)
(
*
G2
)
0 1 4 9 16 25 36 49 64 81 100 121 0 1 4 9 16 25 36 49 64 81 100 121
A generator function is a function that, rather than using return
to
return a value once, uses yield
to yield a (potentially infinite)
sequence of values. Just as in generator expressions, the state of the
generator is preserved between partial iterations, but if we want a
fresh copy of the generator we can simply call the function again.
Here I’ll show my favorite example of a generator function: a function to generate an unbounded series of prime numbers. A classic algorithm for this is the Sieve of Eratosthenes, which works something like this:
In
[
15
]:
# Generate a list of candidates
L
=
[
n
for
n
in
range
(
2
,
40
)]
(
L
)
[2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]
In
[
16
]:
# Remove all multiples of the first value
L
=
[
n
for
n
in
L
if
n
==
L
[
0
]
or
n
%
L
[
0
]
>
0
]
(
L
)
[2, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31, 33, 35, 37, 39]
In
[
17
]:
# Remove all multiples of the second value
L
=
[
n
for
n
in
L
if
n
==
L
[
1
]
or
n
%
L
[
1
]
>
0
]
(
L
)
[2, 3, 5, 7, 11, 13, 17, 19, 23, 25, 29, 31, 35, 37]
In
[
18
]:
# Remove all multiples of the third value
L
=
[
n
for
n
in
L
if
n
==
L
[
2
]
or
n
%
L
[
2
]
>
0
]
(
L
)
[2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37]
If we repeat this procedure enough times on a large enough list, we can generate as many primes as we wish.
Let’s encapsulate this logic in a generator function:
In
[
19
]:
def
gen_primes
(
N
):
"""Generate primes up to N"""
primes
=
set
()
for
n
in
range
(
2
,
N
):
if
all
(
n
%
p
>
0
for
p
in
primes
):
primes
.
add
(
n
)
yield
n
(
*
gen_primes
(
70
))
2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67
That’s all there is to it! While this is certainly not the most computationally efficient implementation of the Sieve of Eratosthenes, it illustrates how convenient the generator function syntax can be for building more complicated sequences.
One feature of Python that makes it useful for a wide range of tasks is the fact that it comes “batteries included”—that is, the Python standard library contains useful tools for a wide range of tasks. On top of this, there is a broad ecosystem of third-party tools and packages that offer more specialized functionality. Here we’ll take a look at importing standard library modules, tools for installing third-party modules, and a description of how you can make your own modules.
For loading built-in and third-party modules, Python provides the
import
statement. There are a few ways to use the statement, which we
will mention briefly here, from most recommended to least recommended.
Explicit import of a module preserves the module’s content in a
namespace. The namespace is then used to refer to its contents with a
.
between them. For example, here we’ll import the built-in math
module and compute the sine of pi:
In
[
1
]:
import
math
math
.
cos
(
math
.
pi
)
Out [1]: -1.0
For longer module names, it’s not convenient to use the full module name
each time you access some element. For this reason, we’ll commonly use
the import ... as ...
pattern to create a shorter alias for the
namespace. For example, the NumPy (Numerical Python) package, a
popular third-party package useful for data science, is by convention
imported under the alias np
:
In
[
2
]:
import
numpy
as
np
np
.
cos
(
np
.
pi
)
Out [2]: -1.0
Sometimes rather than importing the module namespace, you would just
like to import a few particular items from the module. This can be done
with the from ... import ...
pattern. For example, we can import
just the cos
function and the pi
constant from the math
module:
In
[
3
]:
from
math
import
cos
,
pi
cos
(
pi
)
Out [3]: -1.0
Finally, it is sometimes useful to import the entirety of the module
contents into the local namespace. This can be done with the
from ... import *
pattern:
In
[
4
]:
from
math
import
*
sin
(
pi
)
**
2
+
cos
(
pi
)
**
2
Out [4]: 1.0
This pattern should be used sparingly, if at all. The problem is that such imports can sometimes overwrite function names that you do not intend to overwrite, and the implicitness of the statement makes it difficult to determine what has changed.
For example, Python has a built-in sum
function that can be used for
various operations:
In
[
5
]:
help
(
sum
)
Help on built-in function sum in module builtins: sum(...) sum(iterable[, start]) -> value Return the sum of an iterable of numbers (NOT strings) plus the value of parameter 'start' (which defaults to 0). When the iterable is empty, return start.
We can use this to compute the sum of a sequence, starting with a
certain value (here, we’ll start with -1
):
In
[
6
]:
sum
(
range
(
5
),
-
1
)
Out [6]: 9
Now observe what happens if we make the exact same function call after
importing *
from numpy
:
In
[
7
]:
from
numpy
import
*
In
[
8
]:
sum
(
range
(
5
),
-
1
)
Out [8]: 10
The result is off by one! The reason for this is that the import *
statement replaces the built-in sum
function with the numpy.sum
function, which has a different call signature: in the former, we’re
summing range(5)
starting at -1
; in the latter, we’re summing
range(5)
along the last axis (indicated by -1
). This is the type of
situation that may arise if care is not taken when using
import *
—for this reason, it is best to avoid this unless you know
exactly what you are doing.
Python’s standard library contains many useful built-in modules, which
you can read about fully in Python’s
documentation. Any of these can be imported with the import
statement, and then explored using the help function discussed in
the previous section. Here is an extremely incomplete list of some of the modules
you might wish to explore and learn about:
|
Tools for interfacing with the operating system, including navigating file directory structures and executing shell commands |
|
Mathematical functions and operations on real and complex numbers |
|
Tools for constructing and interacting with iterators and generators |
|
Tools that assist with functional programming |
|
Tools for generating pseudorandom numbers |
|
Tools for object persistence: saving objects to and loading objects from disk |
|
Tools for reading JSON-formatted and CSV-formatted files |
|
Tools for doing HTTP and other web requests |
You can find information on these, and many more, in the Python standard library documentation: https://docs.python.org/3/library/.
One of the things that makes Python useful, especially within the world
of data science, is its ecosystem of third-party modules. These can be
imported just as the built-in modules, but first the modules must
be installed on your system. The standard registry for such modules is
the Python Package Index (PyPI for short), found on the Web at
http://pypi.python.org/. For convenience, Python comes with a program
called pip
(a recursive acronym meaning “pip installs packages”), which
will automatically fetch packages released and listed on PyPI (if you
use Python version 2, pip
must be installed separately). For example,
if you’d like to install the supersmoother
package that I wrote, all
that is required is to type the following at the command line:
$ pip install supersmoother
The source code for the package will be automatically downloaded from the PyPI repository, and the package installed in the standard Python path (assuming you have permission to do so on the computer you’re using).
For more information about PyPI and the pip
installer, refer to the
documentation at http://pypi.python.org/.
One place where the Python language really shines is in the manipulation of strings. This section will cover some of Python’s built-in string methods and formatting operations, before moving on to a quick guide to the extremely useful subject of regular expressions. Such string manipulation pattens come up often in the context of data science work, and is one big perk of Python in this context.
Strings in Python can be defined using either single or double quotations (they are functionally equivalent):
In
[
1
]:
x
=
'a string'
y
=
"a string"
x
==
y
Out [1]: True
In addition, it is possible to define multiline strings using a triple-quote syntax:
In
[
2
]:
multiline
=
"""
one
two
three
"""
With this, let’s take a quick tour of some of Python’s string manipulation tools.
For basic manipulation of strings, Python’s built-in string methods can be extremely convenient. If you have a background working in C or another low-level language, you will likely find the simplicity of Python’s methods extremely refreshing. We introduced Python’s string type and a few of these methods earlier; here we’ll dive a bit deeper.
Python makes it quite easy to adjust the case of a string. Here we’ll
look at the upper()
, lower()
, capitalize()
, title()
, and
swapcase()
methods, using the following messy string as an example:
In
[
3
]:
fox
=
"tHe qUICk bROWn fOx."
To convert the entire string into uppercase or lowercase, you can use
the upper()
or lower()
methods respectively:
In
[
4
]:
fox
.
upper
()
Out [4]: 'THE QUICK BROWN FOX.'
In
[
5
]:
fox
.
lower
()
Out [5]: 'the quick brown fox.'
A common formatting need is to capitalize just the first letter of each
word, or perhaps the first letter of each sentence. This can be done
with the title()
and capitalize()
methods:
In
[
6
]:
fox
.
title
()
Out [6]: 'The Quick Brown Fox.'
In
[
7
]:
fox
.
capitalize
()
Out [7]: 'The quick brown fox.'
The cases can be swapped using the swapcase()
method:
In
[
8
]:
fox
.
swapcase
()
Out [8]: 'ThE QuicK BrowN FoX.'
Another common need is to remove spaces (or other characters) from the
beginning or end of the string. The basic method of removing characters
is the strip()
method, which strips whitespace from the beginning and
end of the line:
In
[
9
]:
line
=
' this is the content '
line
.
strip
()
Out [9]: 'this is the content'
To remove just space to the right or left, use rstrip()
or lstrip()
, respectively:
In
[
10
]:
line
.
rstrip
()
Out [10]: ' this is the content'
In
[
11
]:
line
.
lstrip
()
Out [11]: 'this is the content '
To remove characters other than spaces, you can pass the desired
character to the strip()
method:
In
[
12
]:
num
=
"000000000000435"
num
.
strip
(
'0'
)
Out [12]: '435'
The opposite of this operation, adding spaces or other characters, can
be accomplished using the center()
, ljust()
, and rjust()
methods.
For example, we can use the center()
method to center a given string
within a given number of spaces:
In
[
13
]:
line
=
"this is the content"
line
.
center
(
30
)
Out [13]: ' this is the content '
Similarly, ljust()
and rjust()
will left-justify or right-justify
the string within spaces of a given length:
In
[
14
]:
line
.
ljust
(
30
)
Out [14]: 'this is the content '
In
[
15
]:
line
.
rjust
(
30
)
Out [15]: ' this is the content'
All these methods additionally accept any character which will be used to fill the space. For example:
In
[
16
]:
'435'
.
rjust
(
10
,
'0'
)
Out [16]: '0000000435'
Because zero-filling is such a common need, Python also provides
zfill()
, which is a special method to right-pad a string with zeros:
In
[
17
]:
'435'
.
zfill
(
10
)
Out [17]: '0000000435'
If you want to find occurrences of a certain character in a string, the
find()
/rfind()
, index()
/rindex()
, and replace()
methods are
the best built-in methods.
find()
and index()
are very similar, in that they search for the
first occurrence of a character or substring within a string, and return
the index of the substring:
In
[
18
]:
line
=
'the quick brown fox jumped over a lazy dog'
line
.
find
(
'fox'
)
Out [18]: 16
In
[
19
]:
line
.
index
(
'fox'
)
Out [19]: 16
The only difference between find()
and index()
is their behavior
when the search string is not found; find()
returns -1
, while
index()
raises a ValueError
:
In
[
20
]:
line
.
find
(
'bear'
)
Out [20]: -1
In
[
21
]:
line
.
index
(
'bear'
)
--------------------------------------------------------- ValueError Traceback (most recent call last) <ipython-input-21-4cbe6ee9b0eb> in <module>() ----> 1 line.index('bear') ValueError: substring not found
The related rfind()
and rindex()
work similarly, except they search
for the first occurrence from the end rather than the beginning of the
string:
In
[
22
]:
line
.
rfind
(
'a'
)
Out [22]: 35
For the special case of checking for a substring at the beginning or end
of a string, Python provides the startswith()
and endswith()
methods:
In
[
23
]:
line
.
endswith
(
'dog'
)
Out [23]: True
In
[
24
]:
line
.
startswith
(
'fox'
)
Out [24]: False
To go one step further and replace a given substring with a new string,
you can use the replace()
method. Here, let’s replace 'brown'
with
'red'
:
In
[
25
]:
line
.
replace
(
'brown'
,
'red'
)
Out [25]: 'the quick red fox jumped over a lazy dog'
The replace()
function returns a new string, and will replace all
occurrences of the input:
In
[
26
]:
line
.
replace
(
'o'
,
'--'
)
Out [26]: 'the quick br--wn f--x jumped --ver a lazy d--g'
For a more flexible approach to this replace()
functionality, see the
discussion of regular expressions in “Flexible Pattern Matching with Regular Expressions”.
If you would like to find a substring and then split the string based
on its location, the partition()
and/or split()
methods are what
you’re looking for. Both will return a sequence of substrings.
The partition()
method returns a tuple with three elements: the
substring before the first instance of the split-point, the split-point
itself, and the substring after:
In
[
27
]:
line
.
partition
(
'fox'
)
Out [27]: ('the quick brown ', 'fox', ' jumped over a lazy dog')
The rpartition()
method is similar, but searches from the right of the
string.
The split()
method is perhaps more useful; it finds all instances of
the split-point and returns the substrings in between. The default is to
split on any whitespace, returning a list of the individual words in a
string:
In
[
28
]:
line
.
split
()
Out [28]: ['the', 'quick', 'brown', 'fox', 'jumped', 'over', 'a', 'lazy', 'dog']
A related method is splitlines()
, which splits on newline characters.
Let’s do this with a haiku popularly attributed to the 17th-century
poet Matsuo Bashō:
In
[
29
]:
haiku
=
"""matsushima-ya
aah matsushima-ya
matsushima-ya"""
haiku
.
splitlines
()
['matsushima-ya', 'aah matsushima-ya', 'matsushima-ya']
Note that if you would like to undo a split()
, you can use the
join()
method, which returns a string built from a split-point and an
iterable:
In
[
30
]:
'--'
.
join
([
'1'
,
'2'
,
'3'
])
Out [30]: '1--2--3'
A common pattern is to use the special character
(newline) to
join together lines that have been previously split, and recover the
input:
In
[
31
]:
(
"
"
.
join
([
'matsushima-ya'
,
'aah matsushima-ya'
,
'matsushima-ya'
]))
matsushima-ya aah matsushima-ya matsushima-ya
In the preceding methods, we have learned how to extract values from strings,
and to manipulate strings themselves into desired formats. Another use
of string methods is to manipulate string representations of values of
other types. Of course, string representations can always be found using
the str()
function; for example:
In
[
32
]:
pi
=
3.14159
str
(
pi
)
Out [32]: '3.14159'
For more complicated formats, you might be tempted to use string arithmetic as outlined in “Basic Python Semantics: Operators”:
In
[
33
]:
"The value of pi is "
+
str
(
pi
)
Out [33]: 'The value of pi is 3.14159'
A more flexible way to do this is to use format strings, which are strings with special markers (noted by curly braces) into which string-formatted values will be inserted. Here is a basic example:
In
[
34
]:
"The value of pi is {}"
.
format
(
pi
)
Out [34]: 'The value of pi is 3.14159'
Inside the {}
marker you can also include information on exactly
what you would like to appear there. If you include a number, it will
refer to the index of the argument to insert:
In
[
35
]:
"""First letter: {0}. Last letter: {1}."""
.
format
(
'A'
,
'Z'
)
Out [35]: 'First letter: A. Last letter: Z.'
If you include a string, it will refer to the key of any keyword argument:
In
[
36
]:
"""First: {first}. Last: {last}."""
.
format
(
last
=
'Z'
,
first
=
'A'
)
Out [36]: 'First: A. Last: Z.'
Finally, for numerical inputs, you can include format codes that control how the value is converted to a string. For example, to print a number as a floating point with three digits after the decimal point, you can use the following:
In
[
37
]:
"pi = {0:.3f}"
.
format
(
pi
)
Out [37]: 'pi = 3.142'
As before, here the 0
refers to the index of the value to be
inserted. The :
marks that format codes will follow. The .3f
encodes the desired precision: three digits beyond the decimal point,
floating-point format.
This style of format specification is very flexible, and the examples here barely scratch the surface of the formatting options available. For more information on the syntax of these format strings, see the “Format Specification” section of Python’s online documentation.
The methods of Python’s str
type give you a powerful set of tools for
formatting, splitting, and manipulating string data. But even more
powerful tools are available in Python’s built-in regular expression
module. Regular expressions are a huge topic; there are entire books written on the topic (including Jeffrey E.F. Friedl’s Mastering Regular Expressions, 3rd Edition), so it will be hard to do justice within just a single
subsection.
My goal here is to give you an idea of the types of problems that might be addressed using regular expressions, as well as a basic idea of how to use them in Python. I’ll suggest some references for learning more in “Resources for Further Learning”.
Fundamentally, regular expressions are a means of flexible pattern
matching in strings. If you frequently use the command line, you are
probably familiar with this type of flexible matching with the *
character, which acts as a wildcard. For example, we can list all the
IPython notebooks (i.e., files with extension .ipynb) with “Python”
in their filename by using the *
wildcard to match any characters in
between:
In
[
38
]:
!
ls
*
Python
*.
ipynb
01-How-to-Run-Python-Code.ipynb 02-Basic-Python-Syntax.ipynb
Regular expressions generalize this “wildcard” idea to a wide range of
flexible string-matching syntaxes. The Python interface to regular
expressions is contained in the built-in re
module; as a simple
example, let’s use it to duplicate the functionality of the string
split()
method:
In
[
39
]:
import
re
regex
=
re
.
compile
(
's+'
)
regex
.
split
(
line
)
Out [39]: ['the', 'quick', 'brown', 'fox', 'jumped', 'over', 'a', 'lazy', 'dog']
Here we’ve first compiled a regular expression, then used it to
split a string. Just as Python’s split()
method returns a list of
all substrings between whitespace, the regular expression split()
method returns a list of all substrings between matches to the input
pattern.
In this case, the input is s+
: s
is a special character that
matches any whitespace (space, tab, newline, etc.), and the +
is a
character that indicates one or more of the entity preceding it.
Thus, the regular expression matches any substring consisting of one or
more spaces.
The split()
method here is basically a convenience routine built upon
this pattern matching behavior; more fundamental is the match()
method, which will tell you whether the beginning of a string matches
the pattern:
In
[
40
]:
for
s
in
[
" "
,
"abc "
,
" abc"
]:
if
regex
.
match
(
s
):
(
repr
(
s
),
"matches"
)
else
:
(
repr
(
s
),
"does not match"
)
' ' matches 'abc ' does not match ' abc' matches
Like split()
, there are similar convenience routines to find the first
match (like str.index()
or str.find()
) or to find and replace (like
str.replace()
). We’ll again use the line from before:
In
[
41
]:
line
=
'the quick brown fox jumped over a lazy dog'
With this, we can see that the regex.search()
method operates a lot
like str.index()
or str.find()
:
In
[
42
]:
line
.
index
(
'fox'
)
Out [42]: 16
In
[
43
]:
regex
=
re
.
compile
(
'fox'
)
match
=
regex
.
search
(
line
)
match
.
start
()
Out [43]: 16
Similarly, the regex.sub()
method operates much like str.replace()
:
In
[
44
]:
line
.
replace
(
'fox'
,
'BEAR'
)
Out [44]: 'the quick brown BEAR jumped over a lazy dog'
In
[
45
]:
regex
.
sub
(
'BEAR'
,
line
)
Out [45]: 'the quick brown BEAR jumped over a lazy dog'
With a bit of thought, other native string operations can also be cast as regular expressions.
But, you might ask, why would you want to use the more complicated and verbose syntax of regular expressions rather than the more intuitive and simple string methods? The advantage is that regular expressions offer far more flexibility.
Here we’ll consider a more complicated example: the common task of matching email addresses. I’ll start by simply writing a (somewhat indecipherable) regular expression, and then walk through what is going on. Here it goes:
In
[
46
]:
=
re
.
compile
(
'w+@w+.[a-z]{3}'
)
Using this, if we’re given a line from a document, we can quickly extract things that look like email addresses:
In
[
47
]:
text
=
"To email Guido, try [email protected]
or the older address [email protected]."
.
findall
(
text
)
Out [47]: ['[email protected]', '[email protected]']
(Note that these addresses are entirely made up; there are probably better ways to get in touch with Guido).
We can do further operations, like replacing these email addresses with another string, perhaps to hide addresses in the output:
In
[
48
]:
.
sub
(
'[email protected]'
,
text
)
Out [48]: 'To email Guido, try [email protected] or the older address [email protected].'
Finally, note that if you really want to match any email address, the preceding regular expression is far too simple. For example, it only allows addresses made of alphanumeric characters that end in one of several common domain suffixes. So, for example, the period used here means that we only find part of the address:
In
[
49
]:
.
findall
(
'[email protected]'
)
Out [49]: ['[email protected]']
This goes to show how unforgiving regular expressions can be if you’re not careful! If you search around online, you can find some suggestions for regular expressions that will match all valid emails, but beware: they are much more involved than the simple expression used here!
The syntax of regular expressions is much too large a topic for this short section. Still, a bit of familiarity can go a long way: I will walk through some of the basic constructs here, and then list some more complete resources from which you can learn more. My hope is that the following quick primer will enable you to use these resources effectively.
If you build a regular expression on a simple string of characters or digits, it will match that exact string:
In
[
50
]:
regex
=
re
.
compile
(
'ion'
)
regex
.
findall
(
'Great Expectations'
)
Out [50]: ['ion']
While simple letters or numbers are direct matches, there are a handful of characters that have special meanings within regular expressions. They are:
. ^ $ * + ? { } [ ] | ( )
We will discuss the meaning of some of these momentarily. In the meantime, you should know that if you’d like to match any of these characters directly, you can escape them with a backslash:
In
[
51
]:
regex
=
re
.
compile
(
r
'$'
)
regex
.
findall
(
"the cost is $20"
)
Out [51]: ['$']
The r
preface in r'$'
indicates a raw string; in standard Python
strings, the backslash is used to indicate special characters. For
example, a tab is indicated by
:
In
[
52
]:
(
'a
b
c'
)
a b c
Such substitutions are not made in a raw string:
In
[
53
]:
(
r
'a b c'
)
a b c
For this reason, whenever you use backslashes in a regular expression, it is good practice to use a raw string.
Just as the character within regular expressions can escape
special characters, turning them into normal characters, it can also be
used to give normal characters special meaning. These special characters
match specified groups of characters, and we’ve seen them before. In the
email address regexp from before, we used the character
w
, which is a
special marker matching any alphanumeric character. Similarly, in the
simple split()
example, we also saw s
, a special marker indicating
any whitespace character.
Putting these together, we can create a regular expression that will match any two letters/digits with whitespace between them:
In
[
54
]:
regex
=
re
.
compile
(
r
'wsw'
)
regex
.
findall
(
'the fox is 9 years old'
)
Out [54]: ['e f', 'x i', 's 9', 's o']
This example begins to hint at the power and flexibility of regular expressions.
The following table lists a few of these characters that are commonly useful:
Character | Description |
---|---|
|
Match any digit |
|
Match any non-digit |
|
Match any whitespace |
|
Match any non-whitespace |
|
Match any alphanumeric char |
|
Match any non-alphanumeric char |
This is not a comprehensive list or description; for more details, see Python’s regular expression syntax documentation.
If the built-in character groups aren’t specific enough for you, you can use square brackets to specify any set of characters you’re interested in. For example, the following will match any lowercase vowel:
In
[
55
]:
regex
=
re
.
compile
(
'[aeiou]'
)
regex
.
split
(
'consequential'
)
Out [55]: ['c', 'ns', 'q', '', 'nt', '', 'l']
Similarly, you can use a dash to specify a range: for example, [a-z]
will match any lowercase letter, and [1-3]
will match any of 1
,
2
, or 3
. For instance, you may need to extract from a document specific numerical
codes that consist of a capital letter followed by a
digit. You could do this as follows:
In
[
56
]:
regex
=
re
.
compile
(
'[A-Z][0-9]'
)
regex
.
findall
(
'1043879, G2, H6'
)
Out [56]: ['G2', 'H6']
If you would like to match a string with, say, three alphanumeric
characters in a row, it is possible to write, for example, www
. Because
this is such a common need, there is a specific syntax to match
repetitions—curly braces with a number:
In
[
57
]:
regex
=
re
.
compile
(
r
'w{3}'
)
regex
.
findall
(
'The quick brown fox'
)
Out [57]: ['The', 'qui', 'bro', 'fox']
There are also markers available to match any number of repetitions—for
example, the +
character will match one or more repetitions of
what precedes it:
In
[
58
]:
regex
=
re
.
compile
(
r
'w+'
)
regex
.
findall
(
'The quick brown fox'
)
Out [58]: ['The', 'quick', 'brown', 'fox']
The following is a table of the repetition markers available for use in regular expressions:
Character | Description | Example |
---|---|---|
|
Match zero or one repetitions of preceding |
|
|
Match zero or more repetitions of preceding |
|
|
match one or more repetitions of preceding |
|
|
Match |
|
|
Match between |
|
With these basics in mind, let’s return to our email address matcher:
In
[
59
]:
=
re
.
compile
(
r
'w+@w+.[a-z]{3}'
)
We can now understand what this means: we want one or more alphanumeric
characters (w+
) followed by the at sign (@
), followed by one
or more alphanumeric characters (w+
), followed by a period (.
—note the need for a backslash escape), followed by exactly three
lowercase letters.
If we want to now modify this so that the Obama email address matches, we can do so using the square-bracket notation:
In
[
60
]:
email2
=
re
.
compile
(
r
'[w.]+@w+.[a-z]{3}'
)
email2
.
findall
(
'[email protected]'
)
Out [60]: ['[email protected]']
We have changed w+
to [w.]+
, so we will match any alphanumeric
character or a period. With this more flexible expression, we can
match a wider range of email addresses (though still not all—can you
identify other shortcomings of this expression?).
For compound regular expressions like our email matcher, we often want to extract their components rather than the full match. This can be done using parentheses to group the results:
In
[
61
]:
email3
=
re
.
compile
(
r
'([w.]+)@(w+).([a-z]{3})'
)
In
[
62
]:
text
=
"To email Guido, try [email protected]"
"or the older address [email protected]."
email3
.
findall
(
text
)
Out [62]: [('guido', 'python', 'org'), ('guido', 'google', 'com')]
As we see, this grouping actually extracts a list of the sub-components of the email address.
We can go a bit further and name the extracted components using the
(?P<name> )
syntax, in which case the groups can be extracted as a
Python dictionary:
In
[
63
]:
email4
=
re
.
compile
(
r
'(?P<user>[w.]+)@(?P<domain>w+)'
'.(?P<suffix>[a-z]{3})'
)
match
=
email4
.
match
(
'[email protected]'
)
match
.
groupdict
()
Out [63]: {'domain': 'python', 'suffix': 'org', 'user': 'guido'}
Combining these ideas (as well as some of the powerful regexp syntax that we have not covered here) allows you to flexibly and quickly extract information from strings in Python.
The preceding discussion is just a quick (and far from complete) treatment of this large topic. If you’d like to learn more, I recommend the following resources:
re
package documentationI find that I promptly forget how to use regular expressions just about every time I use them. Now that I have the basics down, I’ve found this page to be an incredibly valuable resource to recall what each specific character or sequence means within a regular expression.
A more narrative approach to regular expressions in Python.
This is a 500+ page book on the subject. If you want a really complete treatment of this topic, this is the resource for you.
For some examples of string manipulation and regular expressions in action at a larger scale, see “Pandas: Labeled Column-Oriented Data”, where we look at applying these sorts of expressions across tables of string data within the Pandas package.
If you would like to spring from here and go farther in using Python for scientific computing or data science, there are a few packages that will make your life much easier. This section will introduce and preview several of the more important ones, and give you an idea of the types of applications they are designed for. If you’re using the Anaconda or Miniconda environment suggested at the beginning of this report, you can install the relevant packages with the following command:
$ conda install numpy scipy pandas matplotlib scikit-learn
Let’s take a brief look at each of these in turn.
NumPy provides an efficient way to store and manipulate multidimensional dense arrays in Python. The important features of NumPy are:
It provides an ndarray
structure, which allows efficient
storage and manipulation of vectors, matrices, and higher-dimensional
datasets.
It provides a readable and efficient syntax for operating on this data, from simple element-wise arithmetic to more complicated linear algebraic operations.
In the simplest case, NumPy arrays look a lot like Python lists. For
example, here is an array containing the range of numbers 1 to 9
(compare this with Python’s built-in range()
):
In
[
1
]:
import
numpy
as
np
x
=
np
.
arange
(
1
,
10
)
x
Out [1]: array([1, 2, 3, 4, 5, 6, 7, 8, 9])
NumPy’s arrays offer both efficient storage of data, as well as
efficient element-wise operations on the data. For example, to square
each element of the array, we can apply the **
operator to the array
directly:
In
[
2
]:
x
**
2
Out [2]: array([ 1, 4, 9, 16, 25, 36, 49, 64, 81])
Compare this with the much more verbose Python-style list comprehension for the same result:
In
[
3
]:
[
val
**
2
for
val
in
range
(
1
,
10
)]
Out [3]: [1, 4, 9, 16, 25, 36, 49, 64, 81]
Unlike Python lists (which are limited to one dimension), NumPy arrays can
be multidimensional. For example, here we will reshape our x
array
into a 3x3 array:
In
[
4
]:
M
=
x
.
reshape
((
3
,
3
))
M
Out [4]: array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])
A two-dimensional array is one representation of a matrix, and NumPy
knows how to efficiently do typical matrix operations. For example, you
can compute the transpose using .T
:
In
[
5
]:
M
.
T
Out [5]: array([[1, 4, 7], [2, 5, 8], [3, 6, 9]])
or a matrix-vector product using np.dot
:
In
[
6
]:
np
.
dot
(
M
,
[
5
,
6
,
7
])
Out [6]: array([ 38, 92, 146])
and even more sophisticated operations like eigenvalue decomposition:
In
[
7
]:
np
.
linalg
.
eigvals
(
M
)
Out [7]: array([ 1.61168440e+01, -1.11684397e+00, -1.30367773e-15])
Such linear algebraic manipulation underpins much of modern data analysis, particularly when it comes to the fields of machine learning and data mining.
For more information on NumPy, see “Resources for Further Learning”.
Pandas is a much newer package than NumPy, and is in fact built on top of it. What Pandas provides is a labeled interface to multidimensional data, in the form of a DataFrame object that will feel very familiar to users of R and related languages. DataFrames in Pandas look something like this:
In
[
8
]:
import
pandas
as
pd
df
=
pd
.
DataFrame
({
'label'
:
[
'A'
,
'B'
,
'C'
,
'A'
,
'B'
,
'C'
],
'value'
:
[
1
,
2
,
3
,
4
,
5
,
6
]})
df
Out [8]: label value 0 A 1 1 B 2 2 C 3 3 A 4 4 B 5 5 C 6
The Pandas interface allows you to do things like select columns by name:
In
[
9
]:
df
[
'label'
]
Out [9]: 0 A 1 B 2 C 3 A 4 B 5 C Name: label, dtype: object
Apply string operations across string entries:
In
[
10
]:
df
[
'label'
]
.
str
.
lower
()
Out [10]: 0 a 1 b 2 c 3 a 4 b 5 c Name: label, dtype: object
Apply aggregates across numerical entries:
In
[
11
]:
df
[
'value'
]
.
sum
()
Out [11]: 21
And, perhaps most importantly, do efficient database-style joins and groupings:
In
[
12
]:
df
.
groupby
(
'label'
)
.
sum
()
Out [12]: value label A 5 B 7 C 9
Here in one line we have computed the sum of all objects sharing the same label, something that is much more verbose (and much less efficient) using tools provided in NumPy and core Python.
For more information on using Pandas, see the resources listed in “Resources for Further Learning”.
Matplotlib is currently the most popular scientific visualization packages in Python. Even proponents admit that its interface is sometimes overly verbose, but it is a powerful library for creating a large range of plots.
To use Matplotlib, we can start by enabling the notebook mode (for use
in the Jupyter notebook) and then importing the package as plt
:
In
[
13
]:
# run this if using Jupyter notebook
%
matplotlib
notebook
In
[
14
]:
import
matplotlib.pyplot
as
plt
plt
.
style
.
use
(
'ggplot'
)
# make graphs in the style of R's ggplot
Now let’s create some data (as NumPy arrays, of course) and plot the results:
In
[
15
]:
x
=
np
.
linspace
(
0
,
10
)
# range of values from 0 to 10
y
=
np
.
sin
(
x
)
# sine of these values
plt
.
plot
(
x
,
y
);
# plot as a line
If you run this code live, you will see an interactive plot that lets you pan, zoom, and scroll to explore the data.
This is the simplest example of a Matplotlib plot; for ideas on the wide range of plot types available, see Matplotlib’s online gallery as well as other references listed in “Resources for Further Learning”.
SciPy is a collection of scientific functionality that is built on NumPy. The package began as a set of Python wrappers to well-known Fortran libraries for numerical computing, and has grown from there. The package is arranged as a set of submodules, each implementing some class of numerical algorithms. Here is an incomplete sample of some of the more important ones for data science:
|
Fast Fourier transforms |
|
Numerical integration |
|
Numerical interpolation |
|
Linear algebra routines |
|
Numerical optimization of functions |
|
Sparse matrix storage and linear algebra |
|
Statistical analysis routines |
For example, let’s take a look at interpolating a smooth curve between some data:
In
[
16
]:
from
scipy
import
interpolate
# choose eight points between 0 and 10
x
=
np
.
linspace
(
0
,
10
,
8
)
y
=
np
.
sin
(
x
)
# create a cubic interpolation function
func
=
interpolate
.
interp1d
(
x
,
y
,
kind
=
'cubic'
)
# interpolate on a grid of 1,000 points
x_interp
=
np
.
linspace
(
0
,
10
,
1000
)
y_interp
=
func
(
x_interp
)
# plot the results
plt
.
figure
()
# new figure
plt
.
plot
(
x
,
y
,
'o'
)
plt
.
plot
(
x_interp
,
y_interp
);
What we see is a smooth interpolation between the points.
Built on top of these tools are a host of other data science packages, including general tools like Scikit-Learn for machine learning, Scikit-Image for image analysis, and StatsModels for statistical modeling, as well as more domain-specific packages like AstroPy for astronomy and astrophysics, NiPy for neuro-imaging, and many, many more.
No matter what type of scientific, numerical, or statistical problem you are facing, it’s likely there is a Python package out there that can help you solve it.
This concludes our whirlwind tour of the Python language. My hope is that if you read this far, you have an idea of the essential syntax, semantics, operations, and functionality offered by the Python language, as well as some idea of the range of tools and code constructs that you can explore further.
I have tried to cover the pieces and patterns in the Python language that will be most useful to a data scientist using Python, but this has by no means been a complete introduction. If you’d like to go deeper in understanding the Python language itself and how to use it effectively, here are a handful of resources I’d recommend:
This is an excellent O’Reilly book that explores best practices and idioms for Python, including getting the most out of the standard library.
This is a free online book that provides a ground-up introduction to the Python language.
This book follows a “learn by trying” approach, and deliberately emphasizes developing what may be the most useful skill a programmer can learn: Googling things you don’t understand.
This 700-page monster is well written, and covers virtually everything there is to know about the Python language and its built-in libraries. For a more application-focused Python walk-through, see Beazley’s Python Cookbook.
To dig more into Python tools for data science and scientific computing, I recommend the following books:
This book starts precisely where this report leaves off, and provides a comprehensive guide to the essential tools in Python’s data science stack, from data munging and manipulation to machine learning.
This book is applicable to people far beyond the world of physics research. It is a step-by-step, ground-up introduction to scientific computing, including an excellent introduction to many of the tools mentioned in this report.
This book covers the Pandas library in detail, as well as giving useful information on some of the other tools that enable it.
Finally, for an even broader look at what’s out there, I recommend the following:
O’Reilly features a number of excellent books on Python itself and specialized topics in the Python world.
The PyCon, SciPy, and PyData conferences draw thousands of attendees each year, and archive the bulk of their programs each year as free online videos. These have turned into an incredible set of resources for learning about Python itself, Python packages, and related topics. Search online for videos of both talks and tutorials: the former tend to be shorter, covering new packages or fresh looks at old ones. The tutorials tend to be several hours, covering the use of the tools mentioned here as well as others.
18.218.89.173