Practical Data Science Cookbook

396 Pages·2014·4.855 MB·English

by  Tony Ojeda, Sean Patrick Murphy, Benjamin Bengfort, Abhijit Dasgupta

Practical Data
Science Cookbook
89 hands-on recipes to help you complete real-world data
science projects in R and Python
Tony Ojeda
Sean Patrick Murphy
Benjamin Bengfort
Abhijit Dasgupta
BIRMINGHAM - MUMBAI
Practical Data Science Cookbook
Copyright © 2014 Packt Publishing
All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, without the prior written permission of the publisher,
except in the case of brief quotations embedded in critical articles or reviews.
Every effort has been made in the preparation of this book to ensure the accuracy of the
information presented. However, the information contained in this book is sold without
warranty, either express or implied. Neither the authors, nor Packt Publishing, and its dealers
and distributors will be held liable for any damages caused or alleged to be caused directly or
indirectly by this book.
Packt Publishing has endeavored to provide trademark information about all of the companies
and products mentioned in this book by the appropriate use of capitals. However, Packt
Publishing cannot guarantee the accuracy of this information.
First published: September 2014
Production reference: 1180914
Published by Packt Publishing Ltd.
Livery Place
35 Livery Street
Birmingham B3 2PB, UK.
ISBN 978-1-78398-024-6
www.packtpub.com
Cover image by Pratyush Mohanta ([e-mail])
Credits
Authors Project Coordinator
Tony Ojeda Priyanka Goel
Sean Patrick Murphy
Proofreaders
Benjamin Bengfort
Simran Bhogal
Abhijit Dasgupta
Maria Gould
Ameesha Green
Reviewers
Richard Heimann Paul Hindle
Sarah Kelley Kevin McGowan
Liang Shi Lucy Rowland
Will Voorhees
Indexers
Rekha Nair
Commissioning Editor
James Jones Priya Sane
Acquisition Editor Graphics
James Jones Abhinash Sahu
Content Development Editor Production Coordinator
Arvind Koul Adonia Jones
Technical Editors Cover Work
Pankaj Kadam Adonia Jones
Sebastian Rodrigues
Copy Editors
Insiya Morbiwala
Sayanee Mukherjee
Stuti Srivastava
About the Authors
Tony Ojeda is an accomplished data scientist and entrepreneur, with expertise in business
process optimization and over a decade of experience creating and implementing innovative
data products and solutions. He has a Master's degree in Finance from Florida International
University and an MBA with concentrations in Strategy and Entrepreneurship from DePaul
University. He is the founder of District Data Labs, a cofounder of Data Community DC, and is
actively involved in promoting data science education through both organizations.
First and foremost, I'd like to thank my coauthors for the tireless work they
put in to make this book something we can all be proud to say we wrote
together. I hope to work on many more projects and achieve many great
things with you in the future.
I'd like to thank our reviewers, specifically Will Voorhees and Sarah Kelley,
for reading every single chapter of the book and providing excellent
feedback on each one. This book owes much of its quality to their great
advice and suggestions.
I'd also like to thank my family and friends for their support and
encouragement in just about everything I do.
Last, but certainly not least, I'd like to thank my fiancée and partner in
life, Nikki, for her patience, understanding, and willingness to stick with
me throughout all my ambitious undertakings, this book being just one of
them. I wouldn't dare take risks and experiment with nearly as many things
professionally if my personal life was not the stable, loving, supportive
environment she provides.
Sean Patrick Murphy spent 15 years as a senior scientist at The Johns Hopkins
University Applied Physics Laboratory, where he focused on machine learning, modeling
and simulation, signal processing, and high performance computing in the Cloud. Now, he
acts as an advisor and data consultant for companies in SF, NY, and DC. He completed his
graduation from The Johns Hopkins University and his MBA from the University of Oxford. He
currently co-organizes the Data Innovation DC meetup and cofounded the Data Science MD
meetup. He is also a board member and cofounder of Data Community DC.
Benjamin Bengfort is an experienced data scientist and Python developer who has worked
in military, industry, and academia for the past 8 years. He is currently pursuing his PhD in
Computer Science at the University of Maryland, College Park, doing research in Metacognition
and Natural Language Processing. He holds a Master's degree in Computer Science from North
Dakota State University, where he taught undergraduate Computer Science courses. He is
also an adjunct faculty member at Georgetown University, where he teaches Data Science and
Analytics. Benjamin has been involved in two data science start-ups in the DC region: leveraging
large-scale machine learning and Big Data techniques across a variety of applications. He has a
deep appreciation for the combination of models and data for entrepreneurial effect, and he is
currently building one of these start-ups into a more mature organization.
I'd like to thank Will Voorhees for his tireless support in everything I've
been doing, even agreeing to review my technical writing. He made my
chapters understandable, and I'm thankful that he reads what I write. It's
been essential to my career and sanity to have a classmate, a colleague,
and a friend like him. I'd also like to thank my coauthors, Tony and Sean,
for working their butts off to make this book happen; it was a spectacular
effort on their part. I'd also like to thank Sarah Kelley for her input and
fresh take on the material; so far, she's gone on many adventures with us,
and I'm looking forward to the time when I get to review her books! Finally,
I'd especially like to thank my wife, Jaci, who puts up with a lot, especially
when I bite off more than I can chew and end up working late into the night.
Without her, I wouldn't be writing anything at all. She is an inspiration, and
one of the writers in my family, she is the one who students will be reading,
even a hundred years from now.
Abhijit Dasgupta is a data consultant working in the greater DC-Maryland-Virginia area,
with several years of experience in biomedical consulting, business analytics, bioinformatics,
and bioengineering consulting. He has a PhD in Biostatistics from the University of
Washington and over 40 collaborative peer-reviewed manuscripts, with strong interests in
bridging the statistics/machine-learning divide. He is always on the lookout for interesting and
challenging projects, and is an enthusiastic speaker and discussant on new and better ways
to look at and analyze data. He is a member of Data Community DC and a founding member
and co-organizer of Statistical Programming DC (formerly, R Users DC).
About the Reviewers
Richard Heimann is a technical fellow and Chief Data Scientist at L-3 National Security
Solutions (NSS) (NYSE:LLL), and is also an EMC-certified data scientist with concentrations in
spatial statistics, data mining, and Big Data. Richard also leads the data science team at the
L-3 Data Tactics Business Unit. L-3 NSS and L-3 Data Tactics are both premier Big Data and
analytics service providers based in Washington DC and serve customers globally.
Richard is an adjunct professor at the University of Maryland, Baltimore County, where he
teaches Spatial Analysis and Statistical Reasoning. Additionally, he is an instructor at George
Mason University, teaching Human Terrain Analysis; he is also a selection committee member
for the 2014-2015 AAAS Big Data and Analytics Fellowship Program and member of the
WashingtonExec Big Data Council.
Richard has recently published a book titled Social Media Mining with R, Packt Publishing.
He recently supported DARPA, DHS, the US Army, and the Pentagon with analytical support.
Sarah Kelley is a junior Python developer and aspiring data scientist. She currently works
at a start-up in Bethesda, Maryland, where she spends most of her time on data ingestion
and wrangling. Sarah holds a Master's degree in Education from Seattle University. She is a
self-taught programmer who became interested in the field through her desire to inspire her
students to pursue careers in Mathematics, Science, and technology.
Liang Shi received his PhD in Computer Science and a Master's degree in Statistics from
the University of Georgia in 2008 and 2006, respectively. His PhD study is on Machine
Learning and AI, mainly solving surrogate model-assisted optimization problems. After
graduation, he joined the Data Mining Research team at McAfee; his job was to detect
network threats through machine-learning approaches based on Big Data and cloud
computing platforms. He later joined Microsoft as a software engineer, and continued his
security research and development leveraged by machine-learning algorithms, basically
for online advertisement fraud detection on very large, real-time data scales. In 2012, he
rejoined McAfee (Intel) as a senior researcher, conducting network threat research, again
with the help of machine-learning and cloud computing techniques. Early this year, he joined
Pivotal as a senior data scientist; his work is mainly on data scientist projects with clients of
popular companies, mainly for IT and security data analytics. He is very familiar with statistical
and machine-learning modeling and theories, and he is proficient with many programming
languages and analytical tools. He has several journal- and conference-proceeding
publications, and he also published a book chapter.
Will Voorhees is a software developer with experience in all sorts of interesting things from
mobile app development and natural language processing to infrastructure security. After
teaching English in Austria and bootstrapping an education technology start-up, he moved
to the West Coast, joined a big tech company, and is now happily working on infrastructure
security software used by thousands of developers.
In his free time, Will enjoys reviewing technical books, watching movies, and convincing his
dog that she's a good girl, yes she is.
www.PacktPub.com
Support files, eBooks, discount offers, and more
You might want to visit www.PacktPub.com for support files and downloads related to
your book.
Did you know that Packt offers eBook versions of every book published, with PDF and ePub
files available? You can upgrade to the eBook version at www.PacktPub.com and as a print
book customer, you are entitled to a discount on the eBook copy. Get in touch with us at
[e-mail] for more details.
At www.PacktPub.com, you can also read a collection of free technical articles, sign up
for a range of free newsletters and receive exclusive discounts and offers on Packt books
and eBooks.
TM
http://PacktLib.PacktPub.com
Do you need instant solutions to your IT questions? PacktLib is Packt's online digital book
library. Here, you can access, read and search across Packt's entire library of books.
Why Subscribe?
f Fully searchable across every book published by Packt
f Copy and paste, print and bookmark content
f On demand and accessible via web browser
Free Access for Packt account holders
If you have an account with Packt at www.PacktPub.com, you can use this to access
PacktLib today and view nine entirely free books. Simply use your login credentials for
immediate access.
Table of Contents
Preface 1
Chapter 1: Preparing Your Data Science Environment 7
Introduction 7
Understanding the data science pipeline 9
Installing R on Windows, Mac OS X, and Linux 11
Installing libraries in R and RStudio 14
Installing Python on Linux and Mac OS X 17
Installing Python on Windows 18
Installing the Python data stack on Mac OS X and Linux 21
Installing extra Python packages 24
Installing and using virtualenv 26
Chapter 2: Driving Visual Analysis with Automobile Data (R) 31
Introduction 31
Acquiring automobile fuel efficiency data 32
Preparing R for your first project 34
Importing automobile fuel efficiency data into R 35
Exploring and describing fuel efficiency data 38
Analyzing automobile fuel efficiency over time 43
Investigating the makes and models of automobiles 54
Chapter 3: Simulating American Football Data (R) 59
Introduction 59
Acquiring and cleaning football data 61
Analyzing and understanding football data 65
Constructing indexes to measure offensive and defensive strength 74
Simulating a single game with outcomes decided by calculations 77
Simulating multiple games with outcomes decided by calculations 81