Spectral analysis of unevenly spaced or discontinuous data using the “normperiod” code

Stella Pytharouli, Stathis Stiros

Research output: Contribution to journalArticle

36 Citations (Scopus)

Abstract

''Normperiod'', a Fortran code permitting spectral analysis of non-equidistant data in various fields of engineering and natural sciences is presented. This code is based on the Lomb normalized periodogram and has two basic merits; first, it permits analysis of discontinuous or non-equidistant time series without prior interpolations or their fragmentation, as is the case with FFT, and hence noisy results or loss of information is avoided. Second, it is very simple, can be easily modified, can be used in the analysis of both very short and very long time series and permits the determination of the statistical significance of obtained results. The tolerance of this code is demonstrated on the basis of several examples, in which obtained dominant frequencies are compared with independently known values.
LanguageEnglish
Pages190-196
Number of pages7
JournalComputers and Structures
Volume86
Issue number1-2
DOIs
Publication statusPublished - Jan 2008

Fingerprint

Spectral Analysis
Spectrum analysis
Time series
Natural sciences
Fast Fourier transforms
Interpolation
Periodogram
Statistical Significance
Fragmentation
Tolerance
Interpolate
Engineering

Keywords

  • spectral analysis
  • lomb normalized periodogram
  • unevenly spaced data
  • least-squares spectrum
  • time series
  • geodetic data

Cite this

@article{c98353e94a9b4535a341f7d9c11a9d85,
title = "Spectral analysis of unevenly spaced or discontinuous data using the “normperiod” code",
abstract = "''Normperiod'', a Fortran code permitting spectral analysis of non-equidistant data in various fields of engineering and natural sciences is presented. This code is based on the Lomb normalized periodogram and has two basic merits; first, it permits analysis of discontinuous or non-equidistant time series without prior interpolations or their fragmentation, as is the case with FFT, and hence noisy results or loss of information is avoided. Second, it is very simple, can be easily modified, can be used in the analysis of both very short and very long time series and permits the determination of the statistical significance of obtained results. The tolerance of this code is demonstrated on the basis of several examples, in which obtained dominant frequencies are compared with independently known values.",
keywords = "spectral analysis, lomb normalized periodogram, unevenly spaced data, least-squares spectrum, time series, geodetic data",
author = "Stella Pytharouli and Stathis Stiros",
year = "2008",
month = "1",
doi = "10.1016/j.compstruc.2007.02.022",
language = "English",
volume = "86",
pages = "190--196",
journal = "Computers and Structures",
issn = "0045-7949",
number = "1-2",

}

Spectral analysis of unevenly spaced or discontinuous data using the “normperiod” code. / Pytharouli, Stella; Stiros, Stathis.

In: Computers and Structures, Vol. 86, No. 1-2, 01.2008, p. 190-196.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Spectral analysis of unevenly spaced or discontinuous data using the “normperiod” code

AU - Pytharouli, Stella

AU - Stiros, Stathis

PY - 2008/1

Y1 - 2008/1

N2 - ''Normperiod'', a Fortran code permitting spectral analysis of non-equidistant data in various fields of engineering and natural sciences is presented. This code is based on the Lomb normalized periodogram and has two basic merits; first, it permits analysis of discontinuous or non-equidistant time series without prior interpolations or their fragmentation, as is the case with FFT, and hence noisy results or loss of information is avoided. Second, it is very simple, can be easily modified, can be used in the analysis of both very short and very long time series and permits the determination of the statistical significance of obtained results. The tolerance of this code is demonstrated on the basis of several examples, in which obtained dominant frequencies are compared with independently known values.

AB - ''Normperiod'', a Fortran code permitting spectral analysis of non-equidistant data in various fields of engineering and natural sciences is presented. This code is based on the Lomb normalized periodogram and has two basic merits; first, it permits analysis of discontinuous or non-equidistant time series without prior interpolations or their fragmentation, as is the case with FFT, and hence noisy results or loss of information is avoided. Second, it is very simple, can be easily modified, can be used in the analysis of both very short and very long time series and permits the determination of the statistical significance of obtained results. The tolerance of this code is demonstrated on the basis of several examples, in which obtained dominant frequencies are compared with independently known values.

KW - spectral analysis

KW - lomb normalized periodogram

KW - unevenly spaced data

KW - least-squares spectrum

KW - time series

KW - geodetic data

U2 - 10.1016/j.compstruc.2007.02.022

DO - 10.1016/j.compstruc.2007.02.022

M3 - Article

VL - 86

SP - 190

EP - 196

JO - Computers and Structures

T2 - Computers and Structures

JF - Computers and Structures

SN - 0045-7949

IS - 1-2

ER -