This is a general question on multitaper spectral analysis of signals and stationarity

I work with time series which are non stationary. In this regard I have two questions

  1. Is weak sense stationarity of second order sufficient to apply multitaper spectrum to compute valid estimates from contiguous time series objects or strick sense stationarity is required.

  2. Multitaper works with signals that are periodic, quasi periodic, nearly periodic, and semi periodic signals however I find hard to diferenciate between periodic and the rest as to be able to see if the series I am working with belong to any of the mentioned groups.

Your help will be welcomed

  • $\begingroup$ If you find your own answer to be good, why don't you accept it? $\endgroup$ – Marcus Müller Jul 31 '16 at 8:08

The answers to the questions are

  1. Yes WSS is sufficient.
    • Quasiperiodic

    enter image description here

    • Semiperiodic

    enter image description here

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.