Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/67382 
Year of Publication: 
2012
Series/Report no.: 
FZID Discussion Paper No. 60-2012
Publisher: 
Universität Hohenheim, Forschungszentrum Innovation und Dienstleistung (FZID), Stuttgart
Abstract: 
Spectral analysis is one of the most important areas of time series econometrics. The use of spectral measures is widespread in different science fields such as economics, physics, engineering, geology. The SPECTRAN toolbox has been developed to facilitate the application of spectral concepts to univariate as well as to multivariate series. It offers a variety of frequency-domain techniques and supports the statistical inference. It also provides convenient tools for the examination of the results, e.g.functions for writing the output to a file or functions specially designed for plotting the estimated spectral measures. The key feature of SPECTRAN is the user-friendliness embodied in, e.g., the central function spectran which performs the whole analysis with default settings, but also gives the user the possibility to adjust them. This document sets out the most relevant spectral concepts and their implementation in SPECTRAN. Finally, three examples shall illustrate the application of different toolbox function to macroeconomic data.
Subjects: 
univariate spectral analysis
multivariate spectral analysis
Matlab
JEL: 
C18
C22
C32
C87
Persistent Identifier of the first edition: 
Document Type: 
Working Paper

Files in This Item:
File
Size
542.05 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.