The Analysis of Time Series: An Introduction

When I was a graduate student at Stanford my advisor taught an elementary time series course out of Chatfield's book. It was either the first or the second edition. I was his teaching assistant. The book has been very successful and is now in its fifth edition. It covers most of the important topics concisely and in an intuitive manner. This book gives the student a feel for time series analysis and an appreciation for its applicability. It is not meant for someone who wants a rigorous treatment and a strong understanding of the theory. For that the text of Brockwell and Davis or Anderson or Brillinger or Priestley are more appropriate.

I have read the entire book, and I am quite satisfied with it. I think it is an excellent introduction for anyone with the requisite background in calculus, probability, and statistics.

This tidy book is a highly readable, introductory survey to the topic of modern time series analysis. It excels in its ability to focus on the more intuitive aspects of analysis and model identification. The discussion of both time- and frequency-domain approaches is reasonably balanced, and Kalman filtering is also introduced. While it touches on many modern aspects of time series analysis, it sometimes (intentionally) lacks important technical depth necessary for implementation.The author has done an admirable job at keeping the book manageably small. However, the reader is occasionally left wanting where interesting details are omitted because the author considered them "beyond the scope" of the book. For example, the preface mentions that several new topics are incorporated into the 5th edition (wavelets, for example), but the reader only finds a gratuitous single paragraph with references to complementary journal articles. In these few rare cases, the discussions are not intuitive enough for the reader to know whether it would be profitable to bother with further research at the professional journal level. Still, this title does well to reference the most important landmark works in the time series literature. Those performing remedial research may find it is easier - and more productive - to simply consult Chatfield's recommendations of important topical works before resorting to online or library literature searches.This text has been in print since 1975 with new editions arriving every 5 years or so (perhaps even a 6th edition is close, since the last edition is copyrighted 1996). I am usually suspicious of textbooks having increasingly larger numbers of editions because the continual re-writing implies some level of recurring insufficiency. However, the frequency of update is probably justified due to continuing advances in this field of study. As a result, this title is surprisingly current given its introductory status (although the 4th and 5th editions do not differ too much).For someone new to time series analysis, this may be one of the better places to start, especially for the price. Readers lacking in intuition or experience in time series analysis - especially non-statisticians - will certainly appreciate this introductory title. The more experienced analyst will also be well served by the author's expert perspectives - but to do practical work, this text will still likely need to be supplemented. The generous citation of additional literature will help the reader to know where to go next.

Realmente explica de manera clara, conceptos básicos en series de tiempo, por lo cual lo recomiendo para cualquier persona, así conocera el fantástico Mundo del análisis estocástico en el tiempo...