Spectral Analysis for Univariate Time Series

Spectral analysis is a statistical theory that interprets time series data from various fields using running examples. It covers nonparametric and parametric techniques, emphasizing the multitaper method. The book provides exercises and examples with accompanying R language code for readers to test their understanding.

Format: Hardback
Length: 780 pages
Publication date: 19 March 2020
Publisher: Cambridge University Press

Spectral analysis is a widely employed technique for interpreting time series collected across various domains. This comprehensive book delves into the statistical foundations of spectral analysis, equipping data analysts with the necessary skills to translate theory into practical applications. Throughout the text, real-world time series from oceanography, metrology, atmospheric science, and other fields are utilized, facilitating a clear comparison of different methods in addressing specific research questions. The authors present a unified approach to quantifying the bandwidth of nonparametric spectral estimates, emphasizing the multitaper method, both in its original formulation with Slepian tapers and in a popular alternative using sinusoidal tapers. An extensive set of exercises enables readers to test their understanding of the theory and practical analysis, while supplementary materials, including example time series and R language code, are readily accessible on the book's website. This resource serves as a valuable guide for both novice and experienced practitioners in the field of spectral analysis.

Weight: 1450g
Dimension: 184 x 114 x 20 (mm)
ISBN-13: 9781107028142