1003 lines
119 KiB
Plaintext
1003 lines
119 KiB
Plaintext
Hands-on Time Series Analysis with Python
|
||
From Basics to Bleeding Edge Techniques — B V Vishwas Ashish Patel
|
||
|
||
Hands-on Time Series Analysis with Python
|
||
From Basics to Bleeding Edge Techniques
|
||
B V Vishwas Ashish Patel
|
||
|
||
Hands-on Time Series Analysis with Python: From Basics to Bleeding Edge Techniques
|
||
|
||
B V Vishwas Infosys Bengaluru, India
|
||
|
||
Ashish Patel Cygnet Infotech Pvt Ltd Ahmedabad, India
|
||
|
||
ISBN-13 (pbk): 978-1-4842-5991-7 ISBN-13 (electronic): 978-1-4842-5992-4 https://doi.org/10.1007/978-1-4842-5992-4
|
||
Copyright © 2020 by B V Vishwas and Ashish Patel
|
||
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed.
|
||
Trademarked names, logos, and images may appear in this Book. Rather than use a trademark symbol with every occurrence of a trademarked name, logo, or image we use the names, logos, and images only in an editorial fashion and to the benefit of the trademark owner, with no intention of infringement of the trademark.
|
||
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.
|
||
While the advice and information in this Book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.
|
||
Managing Director, Apress Media LLC: Welmoed Spahr Acquisitions Editor: Celestin Suresh John Development Editor: James Markham Coordinating Editor: Aditee Mirashi
|
||
Cover designed by eStudioCalamar
|
||
Cover image designed by Pixabay
|
||
Distributed to the book trade worldwide by Springer Science+Business Media New York, 233 Spring Street, 6th Floor, New York, NY 10013. Phone 1-800-SPRINGER, fax (201) 348-4505, e-mail orders-ny@springer-sbm.com, or visit www.springeronline.com. Apress Media, LLC is a California LLC and the sole member (owner) is Springer Science + Business Media Finance Inc (SSBM Finance Inc). SSBM Finance Inc is a Delaware corporation.
|
||
For information on translations, please e-mail booktranslations@springernature.com; for reprint, paperback, or audio rights, please e-mail bookpermissions@springernature.com.
|
||
Apress titles may be purchased in bulk for academic, corporate, or promotional use. eBook versions and licenses are also available for most titles. For more information, reference our Print and eBook Bulk Sales web page at www.apress.com/bulk-sales.
|
||
Any source code or other supplementary material referenced by the author in this Book is available to readers on GitHub via the Book’s product page, located at www.apress.com/ 978-1-4842-5991-7. For more detailed information, please visit www.apress.com/source-code.
|
||
Printed on acid-free paper
|
||
|
||
Table of Contents
|
||
About the Authors<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>xi About the Technical Reviewer<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>xiii Acknowledgments<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> xv Introduction<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> xvii
|
||
Chapter 1: T ime-Series Characteristics<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>1 Types of Data<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>2 Time-Series Data<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>2 Cross-Section Data<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>4 Panel Data/Longitudinal Data<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>4 Trend<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 6 Detecting Trend Using a Hodrick-Prescott Filter<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>6 Detrending a Time Series<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>7 Seasonality<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 11 Multiple Box Plots<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>12 Autocorrelation Plot<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>13 Deseasoning of Time-Series Data<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>14 Seasonal Decomposition<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>15 Cyclic Variations<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>16 Detecting Cyclical Variations<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>17 Errors, Unexpected Variations, and Residuals<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>18 Decomposing a Time Series into Its Components<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>18 Summary<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 21
|
||
iii
|
||
|
||
Table of Contents
|
||
Chapter 2: D ata Wrangling and Preparation for Time Series<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>23 Loading Data into Pandas<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>24 Loading Data Using CSV<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>24 Loading Data Using Excel<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>25 Loading Data Using JSON<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>25 Loading Data from a URL<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>26 Exploring Pandasql and Pandas Side by Side<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>27 Selecting the Top Five Records<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>27 Applying a Filter<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>28 Distinct (Unique)<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>29 IN<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 30 NOT IN<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>32 Ascending Data Order<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>33 Descending Data Order<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>34 Aggregation<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 35 GROUP BY<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>36 GROUP BY with Aggregation<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>38 Join (Merge)<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>39 INNER JOIN<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>41 LEFT JOIN<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>43 RIGHT JOIN<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>46 OUTER JOIN<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>48 Summary of the DataFrame<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>51 Resampling<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 52 Resampling by Month<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>53 Resampling by Quarter<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>53 Resampling by Year<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>53
|
||
iv
|
||
|
||
Table of Contents
|
||
Resampling by Week<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>54 Resampling on a Semimonthly Basis<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>55 Windowing Function<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>55 Rolling Window<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>56 Expanding Window<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>57 Exponentially Weighted Moving Window<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>57 Shifting<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 58 Handling Missing Data<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>60 BFILL<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 62 FFILL<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 62 FILLNA<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 63 INTERPOLATE<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 64 Summary<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 64
|
||
Chapter 3: S moothing Methods<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>65 Introduction to Simple Exponential Smoothing<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>66 Simple Exponential Smoothing in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>68 Introduction to Double Exponential Smoothing<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>76 Double Exponential Smoothing in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>78 Introduction to Triple Exponential Smoothing<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>86 Triple Exponential Smoothing in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>87 Summary<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 97
|
||
Chapter 4: R egression Extension Techniques for Time-Series Data<08><><EFBFBD>99 Types of Stationary Behavior in a Time Series<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>99 Making Data Stationery<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>101 Using Plots<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>101 Using Summary Statistics<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>101 Using Statistics Unit Root Tests<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>102
|
||
v
|
||
|
||
Table of Contents
|
||
Interpreting the P-value<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>104 Augmented Dickey-Fuller Test<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>104 Kwiatkowski-Phillips-Schmidt-Shin Test<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>105 Using Stationary Data Techniques<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>106
|
||
Differencing<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 106 Random Walk<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>107 First-Order Differencing (Trend Differencing)<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>108 Second-Order Differencing (Trend Differencing)<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>109 Seasonal Differencing<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>110 Autoregressive Models<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>111 Moving Average<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>113 Autocorrelation and Partial Autocorrelation Functions<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>116 Introduction to ARMA<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>117 Autoregressive Model<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>118 Moving Average<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>119 Introduction to Autoregressive Integrated Moving Average<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>119 The Integration (I)<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>121 ARIMA in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>122 Introduction to Seasonal ARIMA<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>129 SARIMA in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>131 Introduction to SARIMAX<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>143 SARIMAX in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>144 Introduction to Vector Autoregression<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>154 VAR in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>155 Introduction to VARMA<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>172 VARMA in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>172 Summary<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 184
|
||
vi
|
||
|
||
Table of Contents
|
||
Chapter 5: B leeding-Edge Techniques<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>185 Introduction to Neural Networks<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>185 Perceptron<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 186 Activation Function<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>188 Binary Step Function<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>188 Linear Activation Function<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>189 Nonlinear Activation Function<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>190 Forward Propagation<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>195 Backward Propagation<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>196 Gradient Descent Optimization Algorithms<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>199 Learning Rate vs. Gradient Descent Optimizers<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>199 Recurrent Neural Networks<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>202 Feed-Forward Recurrent Neural Network<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>204 Backpropagation Through Time in RNN<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>206 Long Short-Term Memory<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>210 Step-by-Step Explanation of LSTM<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>212 Peephole LSTM<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>214 Peephole Convolutional LSTM<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>215 Gated Recurrent Units<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>216 Convolution Neural Networks<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>219 Generalized CNN Formula<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>222 One-Dimensional CNNs<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>223 Auto-encoders<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 224 Summary<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 226
|
||
vii
|
||
|
||
Table of Contents
|
||
Chapter 6: B leeding-Edge Techniques for Univariate Time Series<08><><EFBFBD><EFBFBD>227 Single-Step Data Preparation for Time-Series Forecasting<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>227 Horizon-Style Data Preparation for Time-Series Forecasting<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>229 LSTM Univariate Single-Step Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>230 LSTM Univariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>242 Bidirectional LSTM Univariate Single-Step Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>253 Bidirectional LSTM Univariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>262 GRU Univariate Single-Step Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>271 GRU Univariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>279 Auto-encoder LSTM Univariate Single-Step Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>288 Auto-encoder LSTM Univariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>297 CNN Univariate Single-Step Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>306 CNN Univariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>315 Summary<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 324
|
||
Chapter 7: B leeding-Edge Techniques for Multivariate Time Series<08><> 325 LSTM Multivariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>325 Bidirectional LSTM Multivariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>337 Auto-encoder LSTM Multivariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>346 GRU Multivariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>356 CNN Multivariate Horizon Style in Action<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>365 Summary<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 374
|
||
viii
|
||
|
||
Table of Contents
|
||
Chapter 8: Prophet<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>375 The Prophet Model<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>375 Implementing Prophet<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>376 Adding Log Transformation<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>381 Adding Built-in Country Holidays<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>386 Adding Exogenous variables using add_regressors(function)<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>389 Summary<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> 394
|
||
Index<08><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>395
|
||
ix
|
||
|
||
About the Authors
|
||
B V Vishwas is a Data Scientist, AI researcher and AI Consultant, Currently living in Bengaluru(INDIA). His highest qualification is Master of Technology in Software Engineering from Birla Institute of Technology & Science, Pilani, India and his primary focus and inspiration is Data Warehousing, Big Data, Data Science (Machine Learning, Deep Learning, Timeseries, Natural Language Processing, Reinforcement Learning, and Operation Research). He has over seven years of IT experience currently working at Infosys as Data Scientist & AI Consultant. He has also worked on Data Migration, Data Profiling, ETL & ELT, OWB, Python, PL/SQL, Unix Shell Scripting, Azure ML Studio, Azure Cognitive Services, and AWS.
|
||
Ashish Patel is a Senior Data Scientist, AI researcher, and AI Consultant with over seven years of experience in the field of AI, Currently living in Ahmedabad(INDIA). He has a Master of Engineering Degree from Gujarat Technological University and his keen interest and ambition to research in the following domains such as (Machine Learning, Deep Learning, Time series, Natural Language Processing, Reinforcement Learning, Audio Analytics, Signal Processing, Sensor
|
||
xi
|
||
|
||
About the Authors Technology, IoT, Computer Vision). He is currently working as Senior Data Scientist for Cynet infotech Pvt Ltd. He has published more than 15 + Research papers in the field of Data Science with Reputed Publications such as IEEE. He holds Rank 3 as a kernel master in Kaggle. Ashish has immense experience working on cross-domain projects involving a wide variety of data, platforms, and technologies.
|
||
xii
|
||
|
||
About the Technical Reviewer
|
||
Over his nearly three-decade career, Alexey Panchekha, PhD, CFA, has spent 10 years in academia, where he focused on nonlinear and dynamic processes; 10 years in the technology industry, where he specialized in program design and development; and eight years in financial services. In the latter arena, he specialized in applying mathematical techniques and technology to risk management and alpha generation. For example, Panchekha was involved in the equity derivative trading technology platform at Goldman Sachs, and he led the creation of the multi-asset multigeographies portfolio risk management system at Bloomberg. He also served as the head of research at Markov Process International, a leader in portfolio attribution and analytics. Most recently, Panchekha cofounded Turing Technology Associates, Inc., with Vadim Fishman. Turing is a technology and intellectual property company that sits at the intersection of mathematics, machine learning, and innovation. Its solutions typically service the financial technology (fintech) industry. Turing primarily focuses on enabling technology that supports the burgeoning ensemble active management (EAM). Prior to Turing, Panchekha was managing director at Incapital, and head of research at F-Squared Investments, where he designed innovative volatility-based risk-sensitive investment strategies. He is fluent in multiple computer programming languages and software and database programs and is certified in deep learning software. He earned a PhD from Kharkiv Polytechnic University with studies in physics and mathematics as well as an MS in physics. Panchekha is a CFA charterholder.
|
||
xiii
|
||
|
||
Acknowledgments
|
||
This being my first book, I found transforming idea and real-world experience into its current shape to be a strenuous task. I am grateful to almighty God for blessing and guiding me in all endeavors. I would like to thank my parents (Vijay Kumar and Rathnamma), brother (Shreyas), other family, and friends for helping me sail though the sea of life.
|
||
—B V Vishwas First and foremost, praises and thanks to God, the almighty, for His showers of blessings throughout the book-writing process. I would like to express my deep and sincere gratitude to my parents (Dinesh Kumar N Patel and Javnika ben D Patel), my sister (Nisha Patel), and my family and friends (Shailesh Patel, Sanket Patel, Nikit Patel, Mansi Patel, Khushboo Shah) for their support and valuable prayers.
|
||
—Ashish Patel Special thanks to Celestin Suresh John, Aditee Mirashi, James Markham, Alexey Panchekha, and the Apress team for bringing this book to life.
|
||
xv
|
||
|
||
Introduction
|
||
This book explains the concepts of time series from traditional to bleeding-edge techniques with full-fledged examples.
|
||
The book begins by covering time-series fundamentals and their characteristics, Structure & Components of time series data, preprocessing, and ways of crafting features through data wrangling. Next, it covers traditional time-series techniques such as the smoothing methods ARMA, ARIMA, SARIMA, SARIMAX, VAR, and VARMA using trending frameworks such as Statsmodels and Pmdarima.
|
||
Further covers how to leverage advanced deep learning-based techniques such as ANN, CNN, RNN, LSTM, GRU, and Autoencoder to solve time-series problems using Tensorflow. It concludes by explaining how to use the popular framework fbprophet for modeling time-series analysis.
|
||
After completion of the book, the reader will have thorough knowledge of concepts and techniques to solve time-series problems. All the code presented in this book is available in Jupyter Notebooks; this allows readers to do hands-on experiments and enhance them in exciting ways.
|
||
xvii
|
||
|
||
CHAPTER 1
|
||
|
||
Time-Series Characteristics
|
||
A time series is a collection of data points that are stored with respect to their time. Mathematical and statistical analysis performed on this kind of data to find hidden patterns and meaningful insight is called time- series analysis. Time-series modeling techniques are used to understand past patterns from the data and try to forecast future horizons. These techniques and methodologies have been evolving for decades.
|
||
Observations with continuous timestamps and target variables are sometimes framed as straightforward regression problems by decomposing dates into minutes, hours, days, weeks, months, years, and so on, which is not the right way to handle such data because the results obtained are poor. In this chapter, you will learn the right approach for handling time-series data.
|
||
There are different kinds of data, such as structured, semistructured, and unstructured, and each type should be handled in its own way to gain maximum insight. In this book, we are going to be looking at time-series data that is structured in manner such as data from the stock market, weather, birth rates, traffic, bike-sharing apps, etc.
|
||
This chapter is a gentle introduction to the types of time-series data, its components, and ways to decompose it.
|
||
|
||
© B V Vishwas and Ashish Patel 2020
|
||
|
||
1
|
||
|
||
B V Vishwas and A. Patel, Hands-on Time Series Analysis with Python,
|
||
|
||
https://doi.org/10.1007/978-1-4842-5992-4_1
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
T ypes of Data
|
||
Time-series analysis is a statistical technique that measures a sequential set of data points. This is a standard measure in terms of time that comes in three types, as shown in Figure 1-1.
|
||
Data Types
|
||
|
||
Time-Series Data
|
||
|
||
Cross-Section Data
|
||
|
||
Figure 1-1. Types of data
|
||
|
||
Panel/Longitudinal Data
|
||
|
||
T ime-Series Data
|
||
A time series contains data points that increase, decrease, or otherwise change in chronological order over a period. A time series that incorporates the records of a single feature or variable is called a univariate time series. If the records incorporate more than one feature or variable, the series is called a multivariate time series. In addition, a time series can be designated in two ways: continuous or discrete.
|
||
In a continuous time series, data observation is carried out continuously throughout the period, as with earthquake seismograph magnitude data, speech data, etc. Figure 1-2 illustrates earthquake data measured continuously from 1975 to 2015.
|
||
|
||
2
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
EARTHQUAKE SEISMOGRAPHIC MAGNITUDE 1975 to 2015
|
||
6.60 6.40 6.20 6.00 5.80 5.60 5.40 5.20 5.00
|
||
|
||
MAGNITUDE
|
||
|
||
YEARS
|
||
Figure 1-2. Forty years of earthquake seismograph magnitude data
|
||
|
||
Figure 1-3 Illustrates temperature behavior in India over a century and clearly shows that temperature is increasing monotonically.
|
||
|
||
AVG TEMPRATURE
|
||
|
||
27 26.5
|
||
26 25.5
|
||
25 24.5
|
||
24 23.5
|
||
23 1880
|
||
|
||
1900
|
||
|
||
1920
|
||
|
||
1940
|
||
|
||
1960
|
||
|
||
1980
|
||
|
||
YEARS
|
||
|
||
2000
|
||
|
||
2020
|
||
|
||
2040
|
||
|
||
Figure 1-3. India’s temperature data from 1901 to 2017
|
||
|
||
In a discrete time series, data observation is carried out at a specific time or equally spaced, as with temperature increases or decreases, exchange rates of currencies, air pressure data, etc. Figure 1-2 illustrates the analysis of the average temperature of India from 1901 to 2017, which either increases or decreases with time. This data behavior is discrete.
|
||
|
||
3
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
C ross-Section Data
|
||
Cross-section data is data gathered at a specific point of time for several subjects such as closing prices of a particular group of stocks on a specific date, opinion polls of elections, obesity level in population, etc. Crosssection studies are utilized in many research areas such as medical, economics, psychology, etc. For instance, high blood pressure is one of the significant risk factors for cause of death in India according to a 2017 WHO report. WHO has carried out the study of several risk factors (considered various subjects), which reflects cross-section survey data. Figure 1-4 illustrates the cross-section data.
|
||
|
||
1800000 1600000 1400000 1200000 1000000
|
||
800000 600000 400000 200000
|
||
0
|
||
|
||
NUMBERS OF DEATH BY RISK FACTOR IN INDIA WHO 2017 STUDY REPORT
|
||
1544920.203
|
||
|
||
DEATH COUNT Unsafe water source…
|
||
Poor sanitation… No access to…
|
||
Indoor air pollution… Non-exclusive… Discontinued…
|
||
Child wasting (deaths) Child stunting (deaths)
|
||
Low birth weight… Secondhand smoke… Alcohol use (deaths)
|
||
Drug use (deaths) Diet low in fruits…
|
||
Diet low in… Unsafe sex (deaths) Low physical activity…
|
||
High blood sugar… High cholesterol… Obesity (deaths) High blood pressure… Smoking (deaths)
|
||
Iron deficiency… Zinc deficiency… Vitamin-A deficiency… Low bone mineral…
|
||
Air pollution… Outdoor air pollution…
|
||
|
||
RISK FACTOR
|
||
Figure 1-4. Number of deaths by risk factor in India
|
||
Panel Data/Longitudinal Data
|
||
Panel data/longitudinal data contains observations of multiple occurrences collected over various durations of time for the same individuals. It is data that is determined periodically by the number of observations in cross-sectional data units such as individuals, companies,
|
||
4
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
|
||
or government agencies. Table 1-1 provides examples of data available for multiple people over the course of a few years where the data gathered comprises income, age, and sex.
|
||
|
||
Table 1-1. Example of Panel Data
|
||
|
||
Panel Data A
|
||
|
||
Name
|
||
|
||
Year
|
||
|
||
Allen
|
||
|
||
2016
|
||
|
||
Allen
|
||
|
||
2017
|
||
|
||
Allen
|
||
|
||
2018
|
||
|
||
Malissa
|
||
|
||
2016
|
||
|
||
Malissa
|
||
|
||
2017
|
||
|
||
Malissa
|
||
|
||
2018
|
||
|
||
Panel Data B
|
||
|
||
Name
|
||
|
||
Year
|
||
|
||
Malissa
|
||
|
||
2016
|
||
|
||
Malissa
|
||
|
||
2017
|
||
|
||
Allen
|
||
|
||
2016
|
||
|
||
Allen
|
||
|
||
2017
|
||
|
||
Allen
|
||
|
||
2018
|
||
|
||
Alicia
|
||
|
||
2017
|
||
|
||
Income
|
||
|
||
Age
|
||
|
||
42145
|
||
|
||
24
|
||
|
||
47797
|
||
|
||
21
|
||
|
||
41391
|
||
|
||
23
|
||
|
||
41100
|
||
|
||
22
|
||
|
||
25800
|
||
|
||
23
|
||
|
||
34508
|
||
|
||
22
|
||
|
||
Income
|
||
|
||
Age
|
||
|
||
42688
|
||
|
||
27
|
||
|
||
21219
|
||
|
||
25
|
||
|
||
46340
|
||
|
||
26
|
||
|
||
22715
|
||
|
||
22
|
||
|
||
34653
|
||
|
||
21
|
||
|
||
31553
|
||
|
||
29
|
||
|
||
Sex Female Female Female Male Male Male
|
||
Sex Female Female Male Male Male Female
|
||
|
||
In Table 1-1, datasets A and B (with the attributes income, age, and sex) gathered throughout the years are for different people. Dataset A is a depiction of two people, Allen and Malissa, who were subject to observation over three years (2016, 2017, 2018); this is known as balanced panel data. Dataset B is called unbalanced panel data because data does not exist for every individual every year.
|
||
5
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Trend
|
||
A trend is a pattern that is observed over a period of time and represents the mean rate of change with respect to time. A trend usually shows the tendency of the data to increase/uptrend or decrease/downtrend during the long run. It is not always necessary that the increase or decrease is in the same direction throughout the given period of time. A trend line is also drawn using candlestick charts.
|
||
For example, you may have heard about an increase or decrease in different market commodities such as gold, silver, stock prices, gas, diesel, etc., or about the rate of interest for banks or home loans increasing or decreasing. These are all commodity market conditions, which may either increase or decrease over time, that show a trend in data.
|
||
Detecting Trend Using a Hodrick-Prescott Filter
|
||
The Hodrick-Prescott (HP) filter has become a benchmark for getting rid of trend movements in data. This method is broadly employed for econometric methods in applied macroeconomics research. The technique is nonparametric and is used to dissolve a time series into a trend; it is a cyclical component unaided by economic theory or prior trend specification. Like all nonparametric methods, the HP filter is contingent significantly on a tuning parameter that controls the degree of smoothing. This method is broadly employed in applied macroeconomics utilized in central banks, international economics agencies, industry, and government.
|
||
With the following example code, you can see how the EXINUS stock changes over a period of time:
|
||
import pandas as pd from statsmodels.tsa.filters.hp_filter import hpfilter df = pd.read_excel(r'\Data\India_Exchange_Rate_Dataset.xls',\
|
||
6
|
||
|
||
Chapter 1 Time-Series Characteristics index_col=0,parse_dates=True) EXINUS_cycle,EXINUS_trend = hpfilter(df['EXINUS'], lamb=1600) EXINUS_trend.plot(figsize=(15,6)).autoscale(axis='x',tight=True)
|
||
Figure 1-5 shows an upward trend over the period.
|
||
Figure 1-5. EXINUS stock showing an upward trend
|
||
Detrending a Time Series
|
||
Detrending is the process of removing a trend from time-series data, or it mentions a change in the mean over time. It is continuously increasing or decreasing over the duration of time. Identification, modeling, and even removing trend data from time-series datasets can be beneficial. The following are methods to detrend time-series data:
|
||
• Pandas differencing • SciPy signal • HP filter
|
||
7
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Detrending Using Pandas Differencing
|
||
The Pandas library has a built-in function to calculate the difference in a dataset. This diff() function is used both for series and for DataFrames. It can provide a period value to shift in order to form the difference. The following code is an example of Pandas differencing.
|
||
• Warning is a built-in module of Python that handles the warning messages.
|
||
• Pyplot is a submodule of Matplotlib that is used to design the graphical representation of the data.
|
||
import pandas as pd import matplotlib.pyplot as plt import warnings warnings.filterwarnings("ignore") df = pd.read_excel(r'\Data\India_Exchange_Rate_Dataset.xls',\ index_col=0,parse_dates=True) diff = df.EXINUS.diff() plt.figure(figsize=(15,6)) plt.plot(diff) plt.title('Detrending using Differencing', fontsize=16) plt.xlabel('Year') plt.ylabel('EXINUS exchange rate') plt.show()
|
||
Figure 1-6 shows the data without a trend by using Pandas differencing.
|
||
8
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Figure 1-6. Trend removal using differencing
|
||
Detrending Using a SciPy Signal
|
||
A signal is another form of time-series data. Every signal either increases or decreases in a different order. Using the SciPy library, this can be removing the linear trend from the signal data. The following code shows an example of SciPy detrending.
|
||
• Signal.detrend is a submodule of SciPy that is used to remove a linear trend along an axis from data.
|
||
import pandas as pd import matplotlib.pyplot as plt from scipy import signal import warnings warnings.filterwarnings("ignore") df = pd.read_excel(r'\Data\India_Exchange_Rate_Dataset.xls',\ index_col=0,parse_dates=True) detrended = signal.detrend(df.EXINUS.values) plt.figure(figsize=(15,6)) plt.plot(detrended) plt.xlabel('EXINUS')
|
||
9
|
||
|
||
Chapter 1 Time-Series Characteristics plt.ylabel('Frequency') plt.title('Detrending using Scipy Signal', fontsize=16) plt.show()
|
||
Figure 1-7 shows the detrended data using SciPy.
|
||
Figure 1-7. Removing a linear trend in a signal using SciPy
|
||
Detrend Using an HP Filter
|
||
An HP filter is also used to detrend a time series and smooth the data. It’s used for removing short-term fluctuations. The following code shows an example of HP filter detrending.
|
||
• Hpfilter is a submodule of Statmodels that is used to remove a smooth trend. import pandas as pd import matplotlib.pyplot as plt from statsmodels.tsa.filters.hp_filter import hpfilter import warnings warnings.filterwarnings("ignore") df = pd.read_excel(r'\Data\India_Exchange_Rate_Dataset.xls',\ index_col=0,parse_dates=True)
|
||
10
|
||
|
||
Chapter 1 Time-Series Characteristics EXINUS_cycle,EXINUS_trend = hpfilter(df['EXINUS'], lamb=1600) df['trend'] = EXINUS_trend . detrended = df.EXINUS - df['trend'] plt.figure(figsize=(15,6)) plt.plot(detrended) plt.title('Detrending using HP Filter', fontsize=16) plt.xlabel('Year') plt.ylabel('EXINUS exchange rate') plt.show() Figure 1-8 shows the data after removing a smooth trend.
|
||
Figure 1-8. Trend removal using an HP filter
|
||
S easonality
|
||
Seasonality is a periodical fluctuation where the same pattern occurs at a regular interval of time. It is a characteristic of economics, weather, and stock market time-series data; less often, it’s observed in scientific data. In other industries, many phenomena are characterized by periodically recurring seasonal effects. For example, retail sales tend to increase during Christmas and decrease afterward.
|
||
11
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
The following methods can be used to detect seasonality:
|
||
• Multiple box plots
|
||
• Autocorrelation plots
|
||
M ultiple Box Plots
|
||
A box plot is an essential graph to depict data spread out over a range. It is a standard approach to showing the minimum, first quartile, middle, third quartile, and maximum. The following code shows an example of detecting seasonality with the help of multiple box plots. See Figure 1-9.
|
||
• Seaborn is a graphical representation package similar to Matplotlib.
|
||
import pandas as pd import seaborn as sns import matplotlib.pyplot as plt import warnings warnings.filterwarnings("ignore") df=pd.read_excel(r'\Data\India_Exchange_Rate_Dataset. xls',\ parse_dates=True) df['month'] = df['observation_date'].dt.strftime('%b') df['year'] = [d.year for d in df.observation_date] df['month'] = [d.strftime('%b') for d in df.observation_date] years = df['year'].unique() plt.figure(figsize=(15,6)) sns.boxplot(x='month', y='EXINUS', data=df).set_ title("Multi Month-wise Box Plot") plt.show()
|
||
12
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Figure 1-9. Multiple-box plot to identify seasonality
|
||
A utocorrelation Plot
|
||
Autocorrelation is used to check randomness in data. It helps to identify types of data where the period is not known. For instance, for the monthly data, if there is a regular seasonal effect, we would hope to see massive peak lags after every 12 months. Figure 1-10 demonstrates an example of detecting seasonality with the help of an autocorrelation plot. from pandas.plotting import autocorrelation_plot import pandas as pd import matplotlib.pyplot as plt df = pd.read_excel(r'\Data\India_Exchange_Rate_Dataset.xls',\ index_col=0,parse_dates=True) plt.rcParams.update({'figure.figsize':(15,6), 'figure. dpi':220}) autocorrelation_plot(df.EXINUS.tolist())
|
||
13
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Figure 1-10. Autocorrelation plot to identify seasonality Note Sometimes identifying seasonality is not easy; in that case, we need to evaluate other plots such as sequence or seasonal subseries plots.
|
||
Deseasoning of Time-Series Data
|
||
Deseasoning means to remove seasonality from time-series data. It is stripped of the pattern of seasonal effect to deseasonalize the impact. Time-series data contains four main components.
|
||
• Level means the average value of the time-series data. • Trend means an increasing or decreasing value in
|
||
time-series data. • Seasonality means repeating the pattern of a cycle in
|
||
the time-series data. • Noise means random variance in time-series data. 14
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Note An additive model is when time-series data combines these four components for linear trend and seasonality, and a multiplicative model is when components are multiplied to gather for nonlinear trends and seasonality.
|
||
S easonal Decomposition
|
||
Decomposition is the process of understanding generalizations and problems related to time-series forecasting. We can leverage seasonal decomposition to remove seasonality from data and check the data only with the trend, cyclic, and irregular variations. Figure 1-11 illustrates data without seasonality.
|
||
import pandas as pd import matplotlib.pyplot as plt from statsmodels.tsa.seasonal import seasonal_decompose import warnings warnings.filterwarnings("ignore") df = pd.read_excel(r'\Data\India_Exchange_Rate_Dataset.xls', index_col=0,parse_dates=True) result_mul = seasonal_decompose(df['EXINUS'], model='multiplicative', extrapolate_trend='freq') deseason = df['EXINUS'] - result_mul.seasonal
|
||
plt.figure(figsize=(15,6)) plt.plot(deseason) plt.title('Deseasoning using seasonal_decompose', fontsize=16) plt.xlabel('Year') plt.ylabel('EXINUS exchange rate') plt.show()
|
||
15
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Figure 1-11. Deseasoning using seasonal_decompose from Statsmodels
|
||
Cyclic Variations
|
||
Cyclical components are fluctuations around a long trend observed every few units of time; this behavior is less frequent compared to seasonality. It is a recurrent process in a time series. In the field of business/economics, the following are three distinct types of cyclic variations examples:
|
||
• Prosperity: As we know, when organizations prosper, prices go up, but the benefits also increase. On the other hand, prosperity also causes over-development, challenges in transportation, increments in wage rate, insufficiency in labor, high rates of returns, deficiency of cash in the market and price concessions, etc., leading to depression
|
||
• Depression: As we know, when there is cynicism in exchange and enterprises, processing plants close down, organizations fall flat, joblessness spreads, and the wages and costs are low.
|
||
16
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
• Accessibility: This causes idealness of money, accessibility of cash at a low interest, an increase in demand for goods or money at a low interest rate, an increase in popular merchandise and ventures described by the circumstance of recuperation that at last prompts for prosperity or boom.
|
||
Detecting Cyclical Variations
|
||
The following code shows how to decompose time-series data and visualize only a cyclic pattern:
|
||
from statsmodels.tsa.filters.hp_filter import hpfilter import pandas as pd import matplotlib.pyplot as plt import warnings warnings.filterwarnings("ignore") %matplotlib inline df = pd.read_excel(r'\Data\India_Exchange_Rate_Dataset.xls', index_col=0,parse_dates=True) EXINUS_cycle,EXINUS_trend = hpfilter(df['EXINUS'], lamb=1600) df['cycle'] =EXINUS_cycle df['trend'] =EXINUS_trend df[['cycle']].plot(figsize=(15,6)).autoscale(axis='x',tight=True) plt.title('Extracting Cyclic Variations', fontsize=16) plt.xlabel('Year') plt.ylabel('EXINUS exchange rate') plt.show()
|
||
Figure 1-12 illustrates isolated cyclic behavior from other data.
|
||
17
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Figure 1-12. Extracting cyclic variations using an HP filter
|
||
Errors, Unexpected Variations, and Residuals
|
||
When trend and cyclical variations are removed from time-series data, the patterns left behind that cannot be explained are called errors, unexpected variations, or residuals. Various methods are available to check for irregular variations such as probability theory, moving averages, and autoregressive time-series methods. If we can find any cyclic variation in data, it is considered to be part of the residuals. These variations that occur due to unexpected circumstances are called unexpected variations or unpredictable errors.
|
||
Decomposing a Time Series into Its Components
|
||
Decomposition is a method used to isolate the time-series data into different elements such as trends, seasonality, cyclic variance, and residuals. We can leverage seasonal decomposition from a stats model
|
||
18
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
to decompose the data into its constituent parts, considering series as additive or multiplicative.
|
||
• Trends(T(t)) means an increase or decrease in the value of ts data.
|
||
• Seasonality(S[t]) means repeating a short-term cycle of ts data.
|
||
• Cyclic variations(c[t]) means a fluctuation in long trends of ts data.
|
||
• Residuals(e[t]) means an irregular variation of ts data. The additive model works with linear trends of time-series data such as changes constantly over time. The additive model formula is as follows:
|
||
Y[t] = T[t] + S[t] + c[t] + e[t] The multiplicative model works with a nonlinear type of data such as quadric or exponential. The multiplicative model formula is as follows:
|
||
Y[t] = T[t] * S[t] * c[t] * e[t] The following example depicts the additive and multiplicative models:
|
||
from statsmodels.tsa.seasonal import seasonal_decompose import pandas as pd df = pd.read_excel(r'Data\India_Exchange_Rate_Dataset.xls', index_col=0,parse_dates=True)
|
||
result = seasonal_decompose(df['EXINUS'], model='add') result.plot()
|
||
result = seasonal_decompose(df['EXINUS'], model='mul') result.plot()
|
||
Figure 1-13 and Figure 1-14 illustrate two methods of time-series decomposition.
|
||
19
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Figure 1-13. Decomposing time-series data using an additive model
|
||
Figure 1-14. Decomposing time-series data using a multiplicative model 20
|
||
|
||
Chapter 1 Time-Series Characteristics
|
||
Summary
|
||
In this chapter, we discussed what time-series data is and glanced at various ways of isolating time-series components such as trend, seasonality, cyclic variations, and errors using multiple techniques. In the next chapter, you will learn how to perform data wrangling for time-series data.
|
||
21
|
||
|
||
CHAPTER 2
|
||
Data Wrangling and Preparation for Time Series
|
||
The expression “Garbage in, garbage out” certainly applies when it comes to data science.
|
||
Data wrangling is the art of transformation and mapping raw data into valuable information with the help of preprocessing strategies. It aims to provide rich data that can be utilized to gain maximum insights. It comprises operations such as loading, imputing, applying transformations, treating outliers, cleaning, integrating, dealing with inconsistency, reducing dimensionality, and engineering features. Data wrangling is an integral and main part of machine learning modeling. Real-world data is undoubtedly messy, and it is not reasonable to use data directly for modeling without performing some wrangling.
|
||
This chapter introduces how to read multiple data file formats with the help of Pandas. It covers how to do data selection, transformation, and cleansing as well as how to derive basic statistics with help of Pandas and Pandasql.
|
||
|
||
© B V Vishwas and Ashish Patel 2020
|
||
|
||
23
|
||
|
||
B V Vishwas and A. Patel, Hands-on Time Series Analysis with Python,
|
||
|
||
https://doi.org/10.1007/978-1-4842-5992-4_2
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Notes The functions and methods introduced in this chapter have a lot more functionalities than covered here. You can explore them in more detail at the official Pandas and Pandasql sites.
|
||
Loading Data into Pandas
|
||
Pandas is the most notable framework for data wrangling. It includes data manipulation and analysis tools intended to make data analysis rapid and convenient. In the real world, data is generated via various tools and devices; hence, it comes in different formats such as CSV, Excel, and JSON. In addition, sometimes data needs to be read from a URL. All data comprises several records and variables. In this section, you will learn how various data formats are handled with Pandas.
|
||
Loading Data Using CSV
|
||
This dataset depicts the number of female births over time. Pandas has a built-in function to read CSV files. The following code imports the dataset: import pandas as pd df = pd.read_csv(r'\Data\daily-total-female-births-CA.csv') df.head(5)
|
||
The following are the top five records from the CSV files:
|
||
24
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series If files have different separators, use sep = ‘ ’ and fill in the separator, as in (;, |,\t,‘ ’).
|
||
Loading Data Using Excel
|
||
The Istanbul Stock Exchange dataset comprises the returns of the Istanbul Stock Exchange along with seven other international indices (SP, DAX, FTSE, NIKKEI, BOVESPA, MSCE_EU, MSCI_EM) from June 5, 2009, to February 22, 2011. The data is organized by workday. This data is available in Excel format, and Pandas has a built-in function to read the Excel file. The following code imports the dataset: import pandas as pd dfExcel = pd.read_excel(r'\Data\istambul_stock_exchange.xlsx', sheet_name = 'Data') dfExcel.head(5)
|
||
The following are the top five records from the Excel file:
|
||
Loading Data Using JSON
|
||
This example dataset is in JSON format. The following code imports the dataset: import pandas as pd dfJson = pd.read_json(r'\Data\test.json') dfJson.head(5)
|
||
25
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series The following are the records from the JSON:
|
||
Loading Data from a URL
|
||
The Abalone dataset comprises 4,177 observations and 9 variables. It is not in any file structure; instead, we can display it as text at the specific URL. The following code imports the dataset: import pandas as pd dfURL = pd.read_csv(r'https://archive.ics.uci.edu/ml/ machine-learning-databases/abalone/abalone.data', names =['Sex', 'Length','Diameter', 'Height','Whole weight', 'Shucked weight','Viscera weight', 'Shell weight', 'Rings']) dfURL.head(5)
|
||
The following are the records from the URL:
|
||
26
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Exploring Pandasql and Pandas Side by Side
|
||
Pandasql is a Python framework for running SQL queries on Pandas DataFrames. It has plenty of essential attributes and a similar purpose as the sqldf framework in R. It helps us to query Pandas DataFrames using SQL syntax. It provides a SQL interface to perform data wrangling on a Pandas DataFrame. Pandasql helps analysts and data engineers who are masters of SQL to transition into Python (Pandas) smoothly.
|
||
Use pip install pandasql to install the library. In this chapter, we will explore ways to use both Pandas and Pandasql for solving different problems, such as selecting data, transforming data, and getting a basic summary of the data.
|
||
Selecting the Top Five Records
|
||
With the help of the pandas.head() function, we can fetch the first N records from the dataset. We can do the same operation with the help of Pandassql. The example illustrates how to do it with Pandas and Pandasql.
|
||
Here is how to do it with Pandas: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp.head(5)
|
||
27
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series Here is how to do it with Pandasql:
|
||
from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select * from dfpsql limit 5 """ sqldf(Query_string, globals())
|
||
The main function used in Pandasql is sqldf. It accepts two parameters: a SQL query string and a set of session/environment variables (locals() or globals()).
|
||
Applying a Filter
|
||
Data filtering is a significant part of data preprocessing. With filtering, we can choose a smaller partition of the dataset and use that subset for viewing and munging; we need specific criteria or a rule to filter the data. This is also known as subsetting data or drill-down data. The following example illustrates how to apply a filter with Pandas and Pandasql.
|
||
Here is how to do it with Pandas: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp[(dfp['Age'] >=30) & (dfp['Age'] <=45)]
|
||
28
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Here is how to do it with Pandasql: from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select * from dfpsql where age>=30 and age<=45 """ sqldf(Query_string, globals())
|
||
Distinct (Unique)
|
||
Several duplicate records exist in the dataset. If we want to select the number of unique values for the specific variable, then we can use the unique() function of Pandas. The following example illustrates how to do this with Pandas and Pandasql.
|
||
Here is how to do it with Pandas:
|
||
29
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp['ID'].unique()
|
||
Here is how to do it with Pandasql: from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select distinct ID from dfpsql;""" sqldf(Query_string, globals())
|
||
IN
|
||
Data filtering is a process of extracting essential data from a dataset by using some condition. There are several methods to do filtering on a dataset. Sometimes we want to investigate whether the data has been associated with a particular DataFrame or Series. In such an event, we can 30
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series use Pandas’ isin() function, which checks whether values are present in the sequence. The procedure can also be carried out in Pandasql. The following example illustrates how to do it with Pandas and Pandasql.
|
||
Here is how to do it with Pandas: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp[dfp.Age.isin([20,30,40])]
|
||
Here is how to do it with Pandasql: from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select * from dfpsql where Age in (20,30,40);""" sqldf(Query_string, globals())
|
||
31
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
NOT IN
|
||
The NOT IN operation is used for a similar purpose as explained earlier. If we want to check whether a value is not part of a sequence, we can use the tilde (~) symbol to perform a NOT IN operation.
|
||
Here is the example with Pandas: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp[~dfp.Age.isin([20,30,40])]
|
||
Here is the example with Pandasql: from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select * from dfpsql where Age not in(20,30,40);""" sqldf(Query_string, globals())
|
||
32
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Ascending Data Order
|
||
ORDER BY sorts the result-set in ascending or descending order based on the value selected. Pandas has a sort_value() function that can use different sorting algorithms, such as quicksort, mergesort, and heapsort. The default value is ascending order, whose Boolean value is True. Except for that axis-wise, we do perform sorting such as 0 for index and 1 for columns. SQL has an ORDER BY clause to perform a similar sorting operation.
|
||
Here is the example using Pandas: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp.sort_values(['Age','Service_time'], ascending= True)
|
||
Here is the example using Pandasql: ORDER BY in SQLite by default is ascending. from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select * from dfpsql order by Age,Service_ time;""" sqldf(Query_string, globals())
|
||
33
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Descending Data Order
|
||
As mentioned, the sort_value() function can sort results in ascending or descending order. It needs the parameter ascending = False to sort values in descending order. You can also update some other parameters, as explained for ORDER BY ascending.
|
||
Here is the example with Pandas: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp.sort_values(['Age','Service_time'], ascending= False)
|
||
34
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series You can sort results in descending order by using a desc keyword next to the column names, as shown here: from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select * from dfpsql order by Age Desc,Service_time Desc;""" sqldf(Query_string, globals())
|
||
Aggregation
|
||
Aggregation is the process of mining data so the data can be investigated, collected, and presented in a summarized manner. It allows you to perform a calculation on single or multiple vectors, usually accompanied by the group by() functions in Pandas. It performs numerous operations such as splitting, applying, and combining. This example illustrates the aggregation operation in Pandas and Pandasql.
|
||
For the following example, we are leveraging .agg in Pandas to achieve the required result: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp.agg({'Transportation_expense': ['count','min', 'max', 'mean']})
|
||
35
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
There is also an option to aggregate directly on a column, as shown here: dfp.Transportation_expense.min()
|
||
Here is the example with Pandasql: from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select count(Transportation_expense) as count, min(Transportation_expense) as min, max(Transportation_expense) as max, avg(Transportation_expense) as mean from dfp;""" sqldf(Query_string, globals())
|
||
GROUP BY
|
||
We can use GROUP BY to arrange identical data into groups based on the aggregation function used. In other words, it groups rows that have similar values into summary rows. We perform this operation with the groupby() function in Pandas. SQL has its GROUP BY clause to perform a similar operation. The example illustrates the GROUP BY operation in Pandas and Pandasql.
|
||
Here is the example with Pandas: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp.groupby('ID')['Service_time'].sum() 36
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Here is the example with Pandasql: from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') Query_string = """ select ID , sum(Service_time) as Sum_ Service_time from dfp group by ID;""" sqldf(Query_string, globals())
|
||
37
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
GROUP BY with Aggregation
|
||
The groupby() function provides a group of similar data based on a selected feature with that data; we can perform different aggregation operations with the .agg() function on other features in Pandas. In SQL, we use GROUP BY to apply aggregate functions on groups of data returned from a query. FILTER is a modifier used with an aggregate function to bound the values used in an aggregation.
|
||
Here is the example with Pandas: import pandas as pd dfp = pd.read_excel(r'\Data\Absenteeism_at_work.xls') dfp.groupby('Reason_for_absence').agg({'Age': ['mean','min','max']})
|
||
Here is the example with Pandasql: from pandasql import sqldf dfpsql = pd.read_excel(r'\Data\Absenteeism_at_work.xls') 38
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series Query_string = """ select Reason_for_absence , avg(Age) as mean, min(Age) as min, max(Age) as max from dfp group by Reason_for_absence;""" sqldf(Query_string, globals())
|
||
Join (Merge)
|
||
The terms join and merge mean the same thing in Pandas, Python, and other languages like SQL and R. In reality, their fundamental operation is equivalent, but their way of executing an operation is different. A join in Pandas utilizes an index to consolidate two data sources, while a merge looks for overlapping columns to merge. In Pandas, the merge() and join() functions are used. In SQL, the MERGE and JOIN operators perform the same operation as in Pandas. This example illustrates the aggregation operation in Pandas and Pandasql.
|
||
We’ve created Sample Employee and Department tables to illustrate the join examples. import pandas as pd data1 = {
|
||
'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'],
|
||
39
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series 'Designation': ['Manager', 'Research
|
||
Engineer', 'Research Engineer', 'VP', 'Delivery Manager'], 'Date_of_joining': ['01-Jan-2000', '23-sep-2006', '11-Jan-2012','21-Jan-1991',
|
||
'12-Jan-1990']} Emp_df = pd.DataFrame(data1, columns = ['Empid', 'Name', 'Designation','Date_of_
|
||
joining']) print(Emp_df.head(5))
|
||
import pandas as pddata2 = { 'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research', 'Research', 'Management', 'Delivery'], 'Total_Experience': [18, 10, 10, 28, 22]}
|
||
Dept_df = pd.DataFrame(data2, columns = ['Empid', 'Deptartment', 'Total_Experience']) print(Dept_df.head(5))
|
||
40
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
INNER JOIN
|
||
An inner merge/inner join keeps rows where the merge value contains an “on” value in both the DataFrames. It selects the records that have matching values (joining columns) in one or more tables.
|
||
Here is the example with Pandas: import pandas as pd data1 = {
|
||
'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'], 'Designation': ['Manager', 'Research
|
||
Engineer', 'Research Engineer', 'VP', 'Delivery Manager'], 'Date_of_joining': ['01-Jan-2000', '23-sep-2006', '11-Jan-2012','21-Jan-1991',
|
||
'12-Jan-1990']}
|
||
41
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series Emp_df = pd.DataFrame(data1, columns = ['Empid', 'Name', 'Designation','Date_of_joining']) data2 = {
|
||
'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research',
|
||
'Research', 'Management', 'Delivery'], 'Total_Experience': [18, 10, 10, 28, 22]} Dept_df = pd.DataFrame(data2, columns = ['Empid',
|
||
'Deptartment', 'Total_Experience']) pd.merge(Emp_df, Dept_df, left_on='Empid',right_on='Empid', how='inner')
|
||
Here is the example with Pandasql: import pandas as pd data1 = {
|
||
'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'],
|
||
'Designation': ['Manager', 'Research Engineer', 'Research Engineer', 'VP', 'Delivery Manager'],
|
||
'Date_of_joining': ['01-Jan-2000', '23-sep-2006', '11-Jan-2012','21-Jan-1991',
|
||
'12-Jan-1990']} Emp_df = pd.DataFrame(data1, columns = ['Empid',
|
||
'Name', 'Designation','Date_of_joining'])
|
||
42
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series data2 = {
|
||
'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research',
|
||
'Research', 'Management', 'Delivery'], 'Total_Experience': [18, 10, 10, 28, 22]} Dept_df = pd.DataFrame(data2, columns = ['Empid',
|
||
'Deptartment', 'Total_Experience']) Query_string = """ select * from Emp_df a INNER JOIN Dept_df b ON a.Empid = b.Empid;""" sqldf(Query_string, globals())
|
||
LEFT JOIN
|
||
A left merge or left join keeps a row on the left DataFrame when the missing value finds the “on” variable in the right DataFrame and adds the empty or NaN value in that result. All the records from the left table and selected matching records based on the joining condition from one or more tables.
|
||
43
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series Here is the example with Pandas:
|
||
import pandas as pd data1 = {
|
||
'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'], 'Designation': ['Manager', 'Research
|
||
Engineer', 'Research Engineer', 'VP', 'Delivery Manager'], 'Date_of_joining': ['01-Jan-2000', '23-sep-2006', '11-Jan-2012','21-Jan-1991',
|
||
'12-Jan-1990']} Emp_df = pd.DataFrame(data1, columns = ['Empid', 'Name', 'Designation','Date_of_
|
||
joining']) data2 = {
|
||
'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research',
|
||
'Research', 'Management', 'Delivery'], 'Total_Experience': [18, 10, 10, 28, 22]} Dept_df = pd.DataFrame(data2, columns = ['Empid',
|
||
'Deptartment', 'Total_Experience']) pd.merge(Emp_df, Dept_df,left_on='Empid',right_on='Empid', how='left')
|
||
44
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series Here is the example with Pandasql: import pandas as pd data1 = {
|
||
'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'],
|
||
'Designation': ['Manager', 'Research Engineer', 'Research Engineer', 'VP', 'Delivery Manager'],
|
||
'Date_of_joining': ['01-Jan-2000', '23-sep-2006', '11-Jan-2012','21-Jan-1991',
|
||
'12-Jan-1990']} Emp_df = pd.DataFrame(data1, columns = ['Empid', 'Name', 'Designation','Date_of_
|
||
joining']) data2 = {
|
||
'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research',
|
||
'Research', 'Management', 'Delivery'], 'Total_Experience': [18, 10, 10, 28, 22]} Dept_df = pd.DataFrame(data2, columns = ['Empid',
|
||
'Deptartment', 'Total_Experience']) Query_string = """ select * from Emp_df a LEFT JOIN Dept_df b ON a.Empid = b.Empid;""" sqldf(Query_string, globals())
|
||
45
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
RIGHT JOIN
|
||
A right merge or right join keeps every row on the right DataFrame. Where the missing values find the “on” variable in the left columns, you add the empty/ NaN values to the result. All the records from the left table and selected matching records based on the joining condition from one or more tables.
|
||
Here is the example with Pandas: import pandas as pd data1 = {
|
||
'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'], 'Designation': ['Manager', 'Research
|
||
Engineer', 'Research Engineer', 'VP', 'Delivery Manager'], 'Date_of_joining': ['01-Jan-2000', '23-sep-2006', '11-Jan-2012','21-Jan-1991',
|
||
'12-Jan-1990']} Emp_df = pd.DataFrame(data1, columns = ['Empid', 'Name', 'Designation','Date_of_
|
||
joining']) data2 = {
|
||
'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research',
|
||
'Research', 'Management', 'Delivery'], 'Total_Experience': [18, 10, 10, 28, 22]}
|
||
46
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series Dept_df = pd.DataFrame(data2, columns = ['Empid', 'Deptartment', 'Total_Experience']) pd.merge(Emp_df, Dept_df,left_on='Empid',right_on='Empid', how='right')
|
||
Here is the example with Pandasql: import pandas as pd data1 = {
|
||
'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'], 'Designation': ['Manager', 'Research
|
||
Engineer', 'Research Engineer', 'VP', 'Delivery Manager'], 'Date_of_joining': ['01-Jan-2000', '23-sep-2006', '11-Jan-2012','21-Jan-1991',
|
||
'12-Jan-1990']} Emp_df = pd.DataFrame(data1, columns = ['Empid', 'Name', 'Designation','Date_of_
|
||
joining']) data2 = {
|
||
'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research',
|
||
'Research', 'Management', 'Delivery'],
|
||
47
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series 'Total_Experience': [18, 10, 10, 28, 22]}
|
||
Dept_df = pd.DataFrame(data2, columns = ['Empid', 'Deptartment', 'Total_Experience'])
|
||
Query_string = """ select a.Empid,Name,Designation,Date_of_ joining,Deptartment,Total_Experience from Dept_df a LEFT JOIN Emp_df b ON a.Empid = b.Empid;""" sqldf(Query_string, globals())
|
||
OUTER JOIN
|
||
An outer merge/full outer join returns all the rows from left and right DataFrames only where it matches or returns NaNs or empty values. It returns all the records from both the tables based on the joining condition regardless of whether they match or not. Not matching ones will be nulls.
|
||
Here is the example with Pandas: import pandas as pd data1 = {
|
||
'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'], 48
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series 'Designation': ['Manager', 'Research
|
||
Engineer', 'Research Engineer', 'VP', 'Delivery Manager'], 'Date_of_joining': ['01-Jan-2000', '23-sep-2006',
|
||
'11-Jan-2012','21-Jan-1991', '12-Jan-1990']}
|
||
Emp_df = pd.DataFrame(data1, columns = ['Empid', 'Name', 'Designation','Date_of_
|
||
joining']) data2 = {
|
||
'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research',
|
||
'Research', 'Management', 'Delivery'], 'Total_Experience': [18, 10, 10, 28, 22]} Dept_df = pd.DataFrame(data2, columns = ['Empid',
|
||
'Deptartment', 'Total_Experience']) pd.merge(Emp_df, Dept_df,left_on='Empid',right_on='Empid', how='outer')
|
||
Here is the example with Pandasql: import pandas as pd data1 = {
|
||
49
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series 'Empid': [1011, 1012, 1013, 1014, 1015], 'Name': ['John', 'Rahul', 'Rick', 'Morty', 'Tim'],
|
||
'Designation': ['Manager', 'Research Engineer', 'Research Engineer', 'VP', 'Delivery Manager'],
|
||
'Date_of_joining': ['01-Jan-2000', '23-sep-2006', '11-Jan-2012','21-Jan-1991',
|
||
'12-Jan-1990']} Emp_df = pd.DataFrame(data1, columns = ['Empid', 'Name', 'Designation','Date_of_
|
||
joining']) data2 = {
|
||
'Empid': [1011, 1017, 1013, 1019, 1015], 'Deptartment': ['Management', 'Research',
|
||
'Research', 'Management', 'Delivery'], 'Total_Experience': [18, 10, 10, 28, 22]} Dept_df = pd.DataFrame(data2, columns = ['Empid',
|
||
'Deptartment', 'Total_Experience']) # OUTER join in not currenlty supported in SQL lite Query_string = """ select * from Emp_df a left OUTER JOIN Dept_ df b ON a.Empid = b.Empid;""" sqldf(Query_string, globals())
|
||
50
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Summary of the DataFrame
|
||
Descriptive statistics is a method used to depict data in a meaningful way to represent either the entire population or a sample. It has two sections: the measure of central tendency, which is used to measure the summary of a sample and population (e.g., mean, median, and mode), and the measure of variability, which is used to measure the spread or dispersion in the dataset (e.g., range, interquartile, variance, and standard deviation). In Pandas, it provides a built-in function called describe(). The following example illustrates the detailed result in Pandas.
|
||
In Pandas, the describe() function returns a number of descriptive statistics values (e.g., mean, standard deviation, min, median, max, first quartile, third quartile). import pandas as pd dfsumm = pd.read_csv(r'https://archive.ics.uci.edu/ml/machinelearning-databases/abalone/abalone.data', names =['Sex', 'Length','Diameter', 'Height','Whole
|
||
weight', 'Shucked weight','Viscera weight', 'Shell weight', 'Rings']) dfsumm.head(5)
|
||
51
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series dfsumm.describe()
|
||
Resampling
|
||
Resampling is a method to convert the time-based observed data into a different interval. In other words, it is used to change the time frequency into another time-frequency format. For instance, say we want to change monthly data into a year-wise format or upsample week data into hours. We would perform either upsample or downsample operations. For that, every data object must have a DateTime-like index(Datetimeindex, PeriodIdnex, TimedeltaIndex). The following example illustrates the result of resampling operations in Pandas: import pandas as pd df = pd.read_csv(r'\Data\daily-total-female-births-CA. csv',index_col =0, parse_dates=['date']) df.head(5)
|
||
52
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Resampling by Month
|
||
Here is how to resample by month: df.births.resample('M').mean()
|
||
Resampling by Quarter
|
||
Here is how to resample by quarter: df.births.resample('Q').mean()
|
||
Resampling by Year
|
||
Here is how to resample by year: df.births.resample('Y').mean()
|
||
53
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Resampling by Week
|
||
Here is how to resample by week: df.births.resample('W').mean()
|
||
54
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Resampling on a Semimonthly Basis
|
||
Here is how to resample on a semimonthly basis: df.births.resample('SM').mean()
|
||
Windowing Function
|
||
A windowing function is used to calculate the number of operations, such as rolling count, rolling sum, rolling mean, rolling median, rolling variance, rolling standard deviation, rolling min, rolling max, rolling correlation, rolling covariance, rolling skewness, rolling kurtosis, rolling quantile, etc. In addition, we can perform some other operations such as expanding window and exponential weighted moving window. The following example illustrates the result of windowing operations in Pandas: import pandas as pd dfExcelwin = pd.read_excel(r'\Data\istambul_stock_exchange. xlsx', sheet_name = 'Data',index_col =0,parse_dates=['date']) dfExcelwin.head(5)
|
||
55
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Rolling Window
|
||
The rolling window feature supports the following methods: count, sum, mean, median, var, std, min, max, corr, cov, skew, kurt, quantile, sum, and aggregate. import pandas as pd dfExcelwin = pd.read_excel(r'\Data\istambul_stock_exchange. xlsx', sheet_name = 'Data',index_col =0,parse_dates=['date']) dfExcelwin.rolling(window=4).mean().head(10)
|
||
56
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Expanding Window
|
||
Expanding window supports the following methods: count, sum, mean, median, var, std, min, max, corr, cov, skew, kurt, quantile, sum, aggregate, and quantile. import pandas as pd dfExcelwin = pd.read_excel(r'\Data\istambul_stock_exchange. xlsx', sheet_name = 'Data',index_col =0,parse_dates=['date']) dfExcelwin.expanding(min_periods=4).mean().head(10)
|
||
Exponentially Weighted Moving Window
|
||
Expanding window supports the following methods: mean, std, var, corr, and cov. import pandas as pd dfExcelwin = pd.read_excel(r'\Data\istambul_stock_exchange. xlsx', sheet_name = 'Data',index_col =0,parse_dates=['date']) dfExcelwin.ewm(com=0.5).mean().head(10)
|
||
57
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
Shifting
|
||
Shifting is also known as “lag” and moves a value back or forward in time. Pandas has a df.shift() function to shift an index by the desired number of periods with an optional time frequency. The following example illustrates the shifting operations result in Pandas: import pandas as pd dfshift = pd.read_excel(r'\Data\istambul stock exchange.xlsx', sheet_name = 'Data',index_col =0, parse_dates=['date']) dfshift.head(5)
|
||
dfshift.shift(periods=3).head(7)
|
||
58
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series dfshift.shift(periods=-1).head(7) dfshift.shift(periods=3, axis =1).head(7)
|
||
59
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series dfshift.shift(periods=3,fill_value=0).head(7)
|
||
Handling Missing Data
|
||
A missing value or missing data occurs when no data value is found at a respective number of places in either a DataFrame or dataset. A missing value presents many problems in the dataset. It reduces the statistical power of data, and the lost data can increase the bias in the dataset. So, it is essential to handle this missing value to maintain the characteristics of the data. Pandas has a number of methods to handle the missing values such as bfill(), ffill(), interpolate(), and fillna(). The following example illustrates the missing value–handling operations result in Pandas: import pandas as pd dfmiss = pd.read_csv(r'\Data\daily-total-female-births-CA-with_ nulls.csv',index_col =0, parse_dates=['date']) dfmiss
|
||
60
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
• Isnull() is a function to detect missing values for an array-like object. dfmiss.isnull().sum()
|
||
61
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
BFILL
|
||
The backward method fills the missing values in the dataset and uses the next valid observation to fill the gap. dfmiss.bfill()
|
||
FFILL
|
||
The forward fill method propagates the last valid observation forward to the next valid one. dfmiss.ffill()
|
||
62
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
FILLNA
|
||
This replaces NA/NaN values using a constant value. dfmiss.fillna(10)
|
||
63
|
||
|
||
Chapter 2 Data Wrangling and Preparation for Time Series
|
||
INTERPOLATE
|
||
This method interpolates values linearly in a forward direction. dfmiss.interpolate(method='linear',limit_direction='forward')
|
||
Summary
|
||
In this chapter, we discussed different techniques for performing data wrangling with the help of Pandas and Pandasql. In the next chapter, you will learn some simple, double, and triple exponential smoothing techniques.
|
||
64
|
||
|
||
CHAPTER 3
|
||
|
||
Smoothing Methods
|
||
|
||
Smoothing is a statistical method we can use to create an approximation function to remove irregularities in data and attempt to capture significant patterns. Robert Goodell Brown was the father of exponential smoothing, and in 1956 he published “Exponential Smoothing for Predicting Demand” (https://industrydocuments.library.ucsf.edu/tobacco/ docs/#id=jzlc0130). In 1957, professor Charles C. Holt was working at CMU on forecasting and published a paper called “Forecasting Seasonals and Trends by Exponentially Weighted Moving Averages” that discussed double exponential smoothing. Three years, in 1960, a student of Holt’s, Peter R. Winters, developed an algorithm by uniting seasonality and published “Forecasting Sales by Exponentially Weighted Moving Averages” (https://pubsonline.informs.org/doi/abs/10.1287/mnsc.6.3.324).
|
||
The smoothing technique is a family of time-series forecasting algorithms, which utilizes the weighted averages of a previous observation to predict or forecast a new value. The main idea of this technique is to overweight recent values in a time series. It is utilized for short-term forecasting models. This technique is more efficient when time-series data is moving slowly over time. It harmonizes errors, trends, and seasonal components into computing smoothening parameters. These components are pooled either additively or multiplicatively.
|
||
This chapter introduces the math behind simple, double, and triple exponential smoothing and how to leverage them efficiently to solve time- series problems.
|
||
|
||
© B V Vishwas and Ashish Patel 2020
|
||
|
||
65
|
||
|
||
B V Vishwas and A. Patel, Hands-on Time Series Analysis with Python,
|
||
|
||
https://doi.org/10.1007/978-1-4842-5992-4_3
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Introduction to Simple Exponential Smoothing
|
||
Simple exponential smoothing (SES) is one of the minimal models of the exponential smoothing algorithm. Such techniques are employed for a univariate observation (data), which has no clear trend or seasonal patterns. It involves a parameter called alpha, which is the smoothing parameter. The core idea is to employ a weighted moving average, including exponentially decreasing weights, which designates higher weight to the most recent observation. The concept behind SES is forecasting future values by using a weighted average of all the previous values in the series. This method can be used to predict series that do not have trends or seasonality.
|
||
In this section, let’s look at the high-level math behind simple exponential smoothing.
|
||
Assume that a series has the following:
|
||
• Level (Lt)
|
||
• No trends
|
||
• No seasonality
|
||
• Noise (somewhat constant)
|
||
SES is the forecast estimate level at a most recent point in time.
|
||
Ft+k = Lt
|
||
Let’s estimate and update the level (Lt). Here is the level updating equation:
|
||
Lt = αYt + (1-α) Lt-1 We are taking the level at time Lt and updating the previous level, Lt-1, by integrating information from most recent data point, Yt. We can see that it is a weighted average, where α and 1-α are the weights.
|
||
66
|
||
|
||
Chapter 3 Smoothing Methods
|
||
The smoothing constant equals α. The 0≤α≤1 alpha value lies between 0 and 1. The algorithm learns a new level from the new data that it sees. One of the possible ways of initializing the whole system is to assign L1 equal to the first record in the series.
|
||
F1 = L1 = Y1
|
||
Why is it called exponential smoothing? Here is the level update equation:
|
||
Lt = αYt + (1-α) Lt-1 Substitute Lt with its formula, as shown here:
|
||
Lt = αYt + (1-α) [(αYt-1 +(1-α)Lt-2] = = αYt+ α(1-α) Yt-1+(1-α)2 Lt-2 = …
|
||
= αYt+ α(1-α) Yt-1+ α (1-α)2 Yt-2 + … We can see that we are ending up with an average of all the values in a series, but they have weights that are decaying exponentially into the past. Hence, this is called exponential smoothing when α = 1 and past values do not influence the forecast. It’s called under-smoothing when α = 0 and past values have equal weight in the average. In this case, we do not give more weight to recent information, which is called over-smoothing. Here is the effect of alpha:
|
||
|
||
α
|
||
|
||
α (1- α)
|
||
|
||
α (1- α)2
|
||
|
||
α (1- α)3
|
||
|
||
0.9
|
||
|
||
0.089
|
||
|
||
0.0089
|
||
|
||
0.00089
|
||
|
||
0.5
|
||
|
||
0.25
|
||
|
||
0.125
|
||
|
||
0.0625
|
||
|
||
0.1
|
||
|
||
0.09
|
||
|
||
0.081
|
||
|
||
0.0729
|
||
|
||
67
|
||
|
||
Chapter 3 Smoothing Methods In all cases, decay as we go into the past with alpha is fast, whereas
|
||
with a smaller alpha decay it is slower.
|
||
Simple Exponential Smoothing in Action
|
||
In the previous section, you learned about the math behind simple exponential smoothing. Now let’s implement it on a time-series dataset.
|
||
Import the required libraries and load the CSV data, which is Facebook stock data from 2014 to 2019.
|
||
• NumPy is the core library for scientific computing in Python. It provides a high-performance multidimensional array object and tools for working with these arrays.
|
||
• sklearn is a machine learning library for the Python programming language.
|
||
from statsmodels.tsa.api import SimpleExpSmoothing import pandas as pd import numpy as np from sklearn import metrics df = pd.read_csv(r'\Data\FB.csv')
|
||
Figure 3-1 provides a sneak peek at the data. Print(df)
|
||
68
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Figure 3-1. First ten rows of dataset Modeling will be done only for the Close column in the dataset. Make
|
||
a copy of the data to perform the train/test split, where the train is used for training, and after getting satisfactory results, we will evaluate the results on the test data.
|
||
The train will have all the data expected for the last 30 days, and the test contains only the last 30 days to evaluate against predictions. X = df['Close'] test = X.iloc[-30:] train = X.iloc[:-30]
|
||
Let’s look at the simple exponential smoothing algorithm’s hyperparameters for reference.
|
||
The SimpleExpSmoothing parameter specifies the following: smoothing_level (float, optional): This is the smoothing_level value of the simple exponential smoothing. If the value is set, then this value will be used as the value. 69
|
||
|
||
Chapter 3 Smoothing Methods
|
||
optimized (bool): This specifies whether the values that have not been set earlier should be optimized automatically.
|
||
Create a function that has all the required evaluation metrics, which will give us the results in one go. This function helps us understand how far off our forecasts are against the actuals.
|
||
• Mean squared error (MSE) tells you how close a regression line is to a set of points. It does this by taking the distances from the points to the regression line (these distances are the errors) and squaring them. The closer to zero the error is, the better the model.
|
||
• Mean absolute error (MAE) measures the average magnitude of the errors in a set of predictions, without considering their direction. The closer to zero the error is, the better the model.
|
||
• Root mean square error (RMSE) is a quadratic scoring rule that also measures the average magnitude of the error. It’s the square root of the average of squared differences between prediction and actual observation. The closer to zero the error is, the better the model.
|
||
• Mean absolute percentage error (MAPE) is a statistical measure of how accurate a forecast system is. It is a measure in terms of percentage. It is mostly used for time-series forecasting. The closer to zero the error is, the better the model.
|
||
• R-squared determines the proportion of variance in the dependent variable that can be explained by the independent variable. The closer to zero the error is, the better the model.
|
||
70
|
||
|
||
Chapter 3 Smoothing Methods
|
||
MAPE and RMSE are notable statistical measures used to check the accuracy of the forecasting model.
|
||
def timeseries_evaluation_metrics_func(y_true, y_pred):
|
||
def mean_absolute_percentage_error(y_true, y_pred): y_true, y_pred = np.array(y_true), np.array(y_pred) return np.mean(np.abs((y_true - y_pred) / y_true)) * 100
|
||
print('Evaluation metric results:-') print(f'MSE is : {metrics.mean_squared_error(y_true,
|
||
y_pred)}') print(f'MAE is : {metrics.mean_absolute_error(y_true,
|
||
y_pred)}') print(f'RMSE is : {np.sqrt(metrics.mean_squared_error
|
||
(y_true, y_pred))}') print(f'MAPE is : {mean_absolute_percentage_error(y_true,
|
||
y_pred)}') print(f'R2 is : {metrics.r2_score(y_true, y_pred)}',end='
|
||
\n\n')
|
||
The following snippet is used to find the best smoothing parameter, which ranges from 0 to 1. The smoothing level model will be fit, and its results will be captured in temp_df. The following results show that 1.0 gives the least RMSE.
|
||
The SimpleExpSmoothing parameter was imported from the Statsmodels package. Here we are using it to train the data. The training data is converted into a NumPy array. See Figure 3-2 for the results.
|
||
resu = [] temp_df = pd.DataFrame() for i in [0, 0.10, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, 0.90,1]:
|
||
71
|
||
|
||
Chapter 3 Smoothing Methods print(f'Fitting for smoothing level= {i}') fit_v = SimpleExpSmoothing(np.asarray(train)).fit(i) fcst_pred_v= fit_v.forecast(30) timeseries_evaluation_metrics_func(test,fcst_pred_v)
|
||
rmse = np.sqrt(metrics.mean_squared_error(test, fcst_ pred_v)) df3 = {'smoothing parameter':i, 'RMSE': rmse} temp_df = temp_df.append(df3, ignore_index=True)
|
||
temp_df.sort_values(by=['RMSE']).head(3)
|
||
Figure 3-2. Optimal parameter search results From the previous search, the least RMSE was achieved with
|
||
smoothing_level equal to 0.1. Let’s use the same value and train the model. 72
|
||
|
||
Chapter 3 Smoothing Methods fitSES = SimpleExpSmoothing(np.asarray(train)).fit(smoothing_ level = 0.1,optimized= False) fcst_gs_pred = fitSES.forecast(30) timeseries_evaluation_metrics_func(test,fcst_gs_pred)
|
||
Let’s check if SimpleExpSmoothing is able to find the best parameters by itself by adjusting a few settings. See Figure 3-3.
|
||
• optimized = True estimates the model parameters by maximizing the log likelihood.
|
||
• use_brute= True searches for good starting values using a brute-force (grid) optimizer.
|
||
fitSESauto = SimpleExpSmoothing(np.asarray(train)).fit( optimized= True, use_brute = True) fcst_auto_pred = fitSESauto.forecast(30) timeseries_evaluation_metrics_func(test,fcst_auto_pred)
|
||
fitSESauto.summary()
|
||
73
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Figure 3-3. Summary of best parameter selected Now that we have the results, let’s set the index similar to the test
|
||
index, such that plots align properly. See Figure 3-4. df_fcst_gs_pred = pd.DataFrame(fcst_gs_pred, columns=['Close_ grid_Search']) df_fcst_gs_pred["new_index"] = range(1229, 1259) df_fcst_gs_pred = df_fcst_gs_pred.set_index("new_index") df_fcst_auto_pred = pd.DataFrame(fcst_auto_pred, columns=['Close_auto_search']) df_fcst_auto_pred["new_index"] = range(1229, 1259) df_fcst_auto_pred = df_fcst_auto_pred.set_index("new_index") import matplotlib.pyplot as plt get_ipython().run_line_magic('matplotlib', 'inline') plt.rcParams["figure.figsize"] = [16,9] 74
|
||
|
||
Chapter 3 Smoothing Methods plt.plot( train, label='Train') plt.plot(test, label='Test') plt.plot(fcst_auto_pred, label='Simple Exponential Smoothing using optimized =True') plt.plot(fcst_gs_pred, label='Simple Exponential Smoothing using custom grid search') plt.legend(loc='best') plt.show()
|
||
Figure 3-4. Line plot depicting results comparision between different configurations of simple exponential smoothing
|
||
We can clearly see that the simple exponential smoothing is not performing well as the stock market data, which will have trends and seasonality. Our basic model will not be able to capture these details.
|
||
In the next section, we will learn the math behind double exponential smoothing and also how to use it efficiently.
|
||
75
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Introduction to Double Exponential Smoothing
|
||
As you saw in the previous section, simple exponential smoothing does not function well when the data has trends. In those cases, we can use double exponential smoothing. This is a more reliable method for handling data that consumes trends without seasonality than compared to other methods. This method adds a time trend equation in the formulation. Two different weights, or smoothing parameters, are used to update these two components at a time.
|
||
Holt’s exponential smoothing is also sometimes called double exponential smoothing. The main idea here is to use SES and advance it to capture the trend component.
|
||
Assume that a series has the following: • Level • Trend s • No seasonality • Noise
|
||
Forecast = estimate level + trend at the most recent time point.
|
||
Ft+k= lt + kTt I is the most recent level. Tt is the most recent time point. K is how many steps into the future we are trying to forecast. We will have two updated equations, level and trends, and we learn both from the data. Here is the updated equation for level:
|
||
Lt = αYt + (1-α) (Lt-1 + Tt-1)
|
||
76
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Lt-1 adjusts the previous level. Tt-1 adds the trend. (Lt-1 + Tt-1) adjusts the previous level by adding a trend. Here is the updated equation for trends:
|
||
Tt = β (Lt-Lt-1) + (1-β) Tt-1 Take the trend in the previous time period, Tt-1, and update it by showing the difference in most recent level estimates: Lt-Lt-1. This setup allows the trend to vary over time adaptively. β is controlling the speed of adjusting the trend. If the trend changes quickly in the series, we want it to learn faster, so we have this:
|
||
α >0 <1 β>0 <1 We can adjust Holt’s exponential smoothing forecast formula to capture either an additive or multiplicative trend. Here is the additive trend: Forecast = most recent estimate level + estimate trend. Ft+k =Lt + KTt Here is the multiplicative trend: Forecast = most recent estimate level + estimate trend. Ft+k =Lt x ( Tt )K Dampening means decreasing future time trends on a straight line (no trend). The predictions generated by Holt’s linear method show a constant trend (increasing or decreasing) in the future.
|
||
77
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Double Exponential Smoothing in Action
|
||
In the previous section, you learned about the math behind double exponential smoothing. Now let’s implement it on a time-series dataset.
|
||
Import the required libraries and load the CSV data, which is Facebook stock data from 2014 to 2019. import pandas as pd import numpy as np from sklearn import metrics from timeit import default_timer as timer from statsmodels.tsa.api import ExponentialSmoothing, SimpleExpSmoothing, Holt df = pd.read_csv(r'\Data\FB.csv')
|
||
Let’s take a sneak peek at the data (see Figure 3-5). Print(df)
|
||
Figure 3-5. First ten rows of loaded dataset
|
||
78
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Modeling will be done only for the Close column in the dataset. Let’s perform a test train split.
|
||
The train will have all the data except the last 30 days. The test contains only the last 30 days to evaluate against predictions.
|
||
train = df.Close[0:-30] test = df.Close[-30:]
|
||
Create a function that has all the required evaluation metrics, which will give us results in one go.
|
||
def timeseries_evaluation_metrics_func(y_true, y_pred):
|
||
def mean_absolute_percentage_error(y_true, y_pred): y_true, y_pred = np.array(y_true), np.array(y_pred) return np.mean(np.abs((y_true - y_pred) / y_true)) * 100
|
||
print('Evaluation metric results:-') print(f'MSE is : {metrics.mean_squared_error(y_true,
|
||
y_pred)}') print(f'MAE is : {metrics.mean_absolute_error(y_true,
|
||
y_pred)}') print(f'RMSE is : {np.sqrt(metrics.mean_squared_error
|
||
(y_true, y_pred))}') print(f'MAPE is : {mean_absolute_percentage_error(y_true,
|
||
y_pred)}') print(f'R2 is : {metrics.r2_score(y_true, y_pred)}',end='
|
||
\n\n')
|
||
The following snippet is a simple grid search. First we are defining a range of parameters for smoothing_level, smoothing_slope, damping_ slope, and damped, and then we are trying to find the best possible combination that yields at least RMSE and R2.
|
||
79
|
||
|
||
Chapter 3 Smoothing Methods
|
||
from sklearn.model_selection import ParameterGrid
|
||
param_grid = {'smoothing_level': [0.10, 0.20,.30,.40,.50,.60, .70,.80,.90], 'smoothing_slope':[0.10, 0.20,.30, .40,.50,.60,.70,.80,.90],
|
||
'damping_slope': [0.10, 0.20,.30,.40,.50,.60,.70, .80,.90],'damped' : [True, False]}
|
||
pg = list(ParameterGrid(param_grid))
|
||
Let’s discuss the double explanation smoothing algorithm’s hyperparameters for reference.
|
||
Here are the Holt parameters:
|
||
endog (array-like): Time series
|
||
exponential (bool, optional): Type of trend component
|
||
damped (bool, optional): Whether the trend component should be damped
|
||
Here are the fit parameters:
|
||
smoothing_level (float, optional): This is the alpha value of the simple exponential smoothing. If the value is set, then this setting will be used as the value.
|
||
smoothing_slope (float, optional): This is the beta value of the Holt trend method. If the value is set, then this setting will be used as the value.
|
||
damping_slope (float, optional): This is the phi value of the damped method. If the value is set, then this setting will be used as the value.
|
||
Optimized (bool, optional). This specifies whether the values that have not been set be optimized automatically.
|
||
80
|
||
|
||
Chapter 3 Smoothing Methods
|
||
df_results_moni = pd.DataFrame(columns=['smoothing_level', 'smoothing_slope', 'damping_slope','damped','RMSE','r2']) start = timer()
|
||
for a,b in enumerate(pg): smoothing_level = b.get('smoothing_level') smoothing_slope = b.get('smoothing_slope') damping_slope = b.get('damping_slope') damped = b.get('damped')
|
||
print(smoothing_level, smoothing_slope, damping_ slope,damped)
|
||
fit1 = Holt(train,damped =damped ).fit(smoothing_ level=smoothing_level, smoothing_slope=smoothing_slope, damping_slope = damping_slope, optimized=False) #fit1.summary z = fit1.forecast(30) print(z) df_pred = pd.DataFrame(z, columns=['Forecasted_result'])
|
||
RMSE = np.sqrt(metrics.mean_squared_error(test, df_ pred.Forecasted_result)) r2 = metrics.r2_score(test, df_pred.Forecasted_result)
|
||
print( f' RMSE is {np.sqrt(metrics.mean_squared_error(test, df_pred.Forecasted_result))}')
|
||
df_results_moni = df_results_moni.append({'smoothing_level' :smoothing_level, 'smoothing_slope':smoothing_slope, 'damping_slope' :damping_slope,'damped':damped,'RMSE': RMSE,'r2':r2}, ignore_index=True)
|
||
end = timer() print(f' Total time taken to complete grid search in seconds: {(end - start)}')
|
||
print(f' Below mentioned parameter gives least RMSE and r2') df_results_moni.sort_values(by=['RMSE','r2']).head(1)
|
||
81
|
||
|
||
Chapter 3 Smoothing Methods See Figure 3-6 and Figure 3-7.
|
||
Figure 3-6. Best parameter selected from custom grid search fit1 = Holt(train,damped =False ).fit(smoothing_level=0.9, smoothing_slope=0.6, damping_slope = 0.1, optimized=False) Forecast_custom_pred = fit1.forecast(30) fit1.summary()
|
||
Figure 3-7. Custom grid search model summary 82
|
||
|
||
Chapter 3 Smoothing Methods Now let’s fit the double exponential smoothing model custom grid search results and evaluate the results. See Figure 3-8. timeseries_evaluation_metrics_func(test,Forecast_custom_pred)
|
||
Figure 3-8. Results after using custom grid search parameters Let’s check whether the double exponential smoothing is able to find
|
||
the best parameters by itself by adjusting a few settings, as follows: • optimized = True estimates model parameters by maximizing the log likelihood. • use_brute= True searches for good starting values using a brute-force (grid) optimizer.
|
||
fitESAUTO = Holt(train).fit(optimized= True, use_brute = True) fitESAUTO.summary()
|
||
See Figure 3-9, Figure 3-10, and Figure 3-11.
|
||
83
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Figure 3-9. Automated grid search model summary fitESAUTOpred = fitESAUTO.forecast(30) timeseries_evaluation_metrics_func(test,fitESAUTOpred)
|
||
Figure 3-10. Results after using automated grid search parameters import matplotlib.pyplot as plt get_ipython().run_line_magic('matplotlib', 'inline') plt.rcParams["figure.figsize"] = [16,9] 84
|
||
|
||
Chapter 3 Smoothing Methods plt.plot( train, label='Train') plt.plot(test, label='Test') plt.plot(fitESAUTOpred, label='Automated grid search') plt.plot(Forecast_custom_pred, label='Double Exponential Smoothing with custom grid search') plt.legend(loc='best') plt.show()
|
||
Figure 3-11. Line plot depicting results comparision between different configurations of double exponential smoothing
|
||
From the evaluation metrics and graph, we can see that the double exponential smoothing performed significantly better than simple exponential smoothing.
|
||
In the next section, we will learn the math behind triple exponential smoothing and how to use it efficiently.
|
||
85
|
||
|
||
Chapter 3 Smoothing Methods
|
||
Introduction to Triple Exponential Smoothing
|
||
Triple exponential smoothing is a forecasting method that enforces exponential smoothing three times. This method can be applied when the data consumes trends and seasonality over time. It includes all smoothing component equations such as trends and seasonality. Seasonality comprises two different types, such as additive and multiplicative, which is a similar operation in mathematics. The Winter method uses the idea of the Holt method and adds seasonality. Winter was a student of Holt and extended his approach by adding an additional equation to update seasonality. This triple exponential smoothing is also known as the Holt- Winters method.
|
||
Let’s assume that a series has the following: • Level • Trend • Seasonality • Noise
|
||
The forecasting equation can be adjusted to handle additive or multiplicative trends and seasonality.
|
||
This is the additive seasonality model: Forecast = estimate level + trend + seasonality at most recent time point.
|
||
Ft+k= lt + kTt + St+k-M This is the multiplicative seasonality model with the additive trend: Forecast = estimate (level × trend) × seasonality at most recent time point.
|
||
Ft+k= (lt + kTt ) St+k-M
|
||
86
|
||
|