McNicholas Paul D. - Data Science with Julia

Data Science with Julia Data Science with Julia By Paul D. McNicholas and Peter A. Tait CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 2019 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed on acid-free paper Version Date: 20191119 International Standard Book Number-13: 978-1-138-49998-0 (Paperback) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained.

If any copyright material has not been acknowledged, please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users.

For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: McNicholas, Paul D., author. | Tait, Peter A., author. Title: Data science with Julia / Paul D. Tait. Tait.

Description: Boca Raton : Taylor & Francis, CRC Press, 2018. | Includes bibliographical references and index. Identifiers: LCCN 2018025237 | ISBN 9781138499980 (pbk.) Subjects: LCSH: Julia (Computer program language) | Data structures (Computer science) Classification: LCC QA76.73.J85 M37 2018 | DDC 005.7/3--dc23 LC record available at https://lccn.loc.gov/2018025237 Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.comand the CRC Press Web site at
http://www.crcpress.comFor Oscar, who tries and tries.PDMTo my son, Xavier, Gettin after it does pay off.PATContents Foreword The 21st century will probably be the century of the data revolution. Our numerical world is creating masses of data every day and the volume of generated data is increasing more and more (the number of produced numerical data is doubling every two years according to the most recent estimates). In such a context, data science is nowadays an unavoidable field for anyone interested in exploiting data. People may be interested in either understanding a phenomenon or in predicting the future behavior of this phenomenon.

To this end, it is important to have significant knowledge of both the rationale (the theory) behind data science techniques and their practical use on real-world data. Indeed, data science is a mix of data, statistical/machine learning methods and software. Software is actually the link between data and data science techniques. It allows the practitioner to load the data and apply techniques on it for analysis. It is therefore important to master at least one of the data science languages. R and Python are probably the two most popular languages for data science. R and Python are probably the two most popular languages for data science.

On the one hand, R has been made by statisticiansmostly for statisticians! It is, however, an excellent tool for data science since the most recent statistical learning techniques are provided on the R platform (named CRAN). Using R is probably the best way to be directly connected to current research in statistics and data science through the packages provided by researchers. Python is, on the other hand, an actual computer science language (with all appropriate formal aspects) for which some advanced libraries for data science exist. In this context, the Julia language has the great advantage to permit users to interact with both R and Python (but also C, Fortran, etc.), within a software language designed for efficient and parallel numerical computing while keeping a high level of human readability. Professor Paul McNicholas and Peter Tait propose in this book to learn both fundamental aspects of data science: theory and application. First, the book will provide you with the significant elements to understand the mathematical aspects behind the most used data science techniques.

The book will also allow you to discover advanced recent techniques, such as probabilistic principal components analysis (PPCA), mixtures of PPCAs, and gradient boosting. In addition, the book will ask you to dive into the Julia language such that you directly apply the learned techniques on concrete examples. This is, in my opinion, the most efficient way to learn such an applied science. In addition, the focus made by this book on the Julia language is a great choice because of the numerous qualities of this language regarding data science practice. These include ease of learning for people familiar with R or Python, nice syntax, easy code debugging, the speed of the compiled language, and code reuse. Both authors have extensive experience in data science.

Professor Paul McNicholas is Canada Research Chair in Computational Statistics at McMaster University and Director of the MacDATA Institute of the same university. In his already prolific career, McNicholas has made important contributions to statistical learning. More precisely, his research is mainly focused on model-based learning with high-dimensional and skew-distributed data. He is also a researcher deeply involved in the spreading of research products through his numerous contributions to the R software with packages. Peter Tait is currently a Ph.D. student but, before returning to academia, he had a professional life dedicated to data science in industry.

His strong knowledge of the needs of industry regarding data science problems was really an asset for the book. This book is a great way to both start learning data science through the promising Julia language and to become an efficient data scientist. Professor Charles Bouveyron Professor of Statistics INRIA Chair in Data Science Universit Cte dAzur Nice, France Preface This is a book for people who want to learn about the Julia language with a view to using it for data science. Some effort has gone into making this book suitable for someone who has familiarity with the R software and wants to learn about Julia. However, prior knowledge of R is not a requirement. While this book is not intended as a textbook for a course, some may find it a useful book to follow for a course that introduces statistics or data science students to Julia.