[R] significance of spectal peak with spectrum()

Dieter Menne dieter.menne at menne-biomed.de
Fri Sep 16 09:41:26 CEST 2005

Sebastian Leuzinger <Sebastian.Leuzinger <at> unibas.ch> writes:

> Hello, has anybody got a simple recepie to test the significance level of the 
> peaks after using spectrum() ?
Having worked in circadian rhythmic consultancy for a few years, I know how 
popular this question is and how difficult it is to tell people that there is 
no simple answer. It may be possible to find simple answers for astrophysical 
signals, where almost all background is white noise and long signals are 
available, but in biology the null-hypothesis is so badly defined that checking 
the pre-conditions is the most important part of the job.

See also http://finzi.psych.upenn.edu/R/Rhelp02a/archive/33423.html.

This said, you may try package GeneTS where you can find some methods.


More information about the R-help mailing list