Stationary regime in time series.

AI Thread Summary
To determine if a time series is in a stationary regime, one can analyze whether its statistical properties remain constant over time. While no specific mathematical tool is strictly necessary, statistical analysis may be required if noise affects the data. Using a fast Fourier transform can help identify signals within white noise, provided there is a sufficient data set. If the signal has a broad spectrum, alternative methods may be more appropriate. Understanding the nature of the noise is also crucial for accurate analysis.
Horaci Castellini
Hi all.

Anyone can say to me as I can know if a time serie is in stationary
regime?. I.E. What mathematical tool I must use to find out this when
the time series is empirical?

Thantks Horacio.
 
Physics news on Phys.org
Stationary regime usually means that nothing changes with time.
No mathematical tool should be needed to check if a time series is stationary.
However, if noise perturbs the experimental data, then a statistical analysis would be needed as well as a knowledge of the nose source. (!)

For a white noise, I think I would simply use the fast Fourier transform with a sufficient amount of data. Then, it should be possible to check if something comes out of the noise. This is assuming that the "signal" does not have a broad spectrum, otherwise other methods could be more suitable.
 
Back
Top