Hi Paul,

the first question would be: what do 3dPeriodogram and 1dFFT really compute – magnitude or power, as you say?
Do they compute magnitude on the y-axis, so that I have to square their results (^2) or (**2 in Python) to get power?

The description of 3dPeriodogram says:

The result is the squared magnitude of the FFT of w(k)*data(k),
  divided by P.  This division makes the result be the 'power',
  which is to say the data's sum-of-squares ('energy') per unit
  time (in units of 1/TR, not 1/sec) ascribed to each FFT bin.

But I lack the understanding and knowledge to understand if that is the same as what I would compute in scipy Periodogram. It seems not to be the case, since the 3dPeriodogram PLE (slope of the linear regression) – that is to say the PLE computed based on the frequency-domain created by 3dPeriodogram – is roughly half the size of the one that I compute in scipy Periodogram or scipy Welch.

In other words: the 3dPeriodogram power spectra seem to result in a different power distribution in the frequency-domain, even though the description of 3dPeriodogram states that the result is "power".

Edit: I think I got it. [docs.scipy.org]

Scipy states that:

Selects between computing the power spectral density (‘density’) where Pxx has
units of V**2/Hz and computing the power spectrum (‘spectrum’) where Pxx has units of V**2,
if x is measured in V and fs is measured in Hz. Defaults to ‘density’

That would mean that I have to take the square root of the scipy results in order to obtain compareable results with the output of 3dPeriodogram, correct?


Philipp



Edited 4 time(s). Last edit at 05/27/2022 04:09PM by Philipp.