NxN spectral granger producing odd results

[moon94]

I am doing spectral granger causality analysis from the process GUI and I am obtaining odd results, such as the first figure.

When I did the analysis previously I obtained an expected result (second figure). Note that the values are no longer between 0 and 1.

Note that the two figures have a different number of channels, but this does not matter.

This is what I'm inputting into the process GUI. This has not changed and I am very confused as to why the results are so different now.

It seems to be fine for certain data but not for others. To my knowledge, there isn't anything wrong with the other datasets.

[MartinC]

Hi,

Can you describe what you find "odd" about your results? If the issue is with the range of values, that might be because you have changed the parameters of your displayed colormap. You can double click on the colorbar of any figure to reset it to its default values.

Martin

[moon94]

Oh, perhaps that is the problem. How do I reset it to default values?

[EDIT]; if you're refering to option "restore defaults" then the result is the same

[EDIT2]: what's odd about the result is that almost every other value is 0, and the values are not between 0 and 1. It's as though the algorithm is only working on a couple of the channels. This is evident when checking the power spectrum.

[MartinC]

Yes, I was referring to the restore defaults option.

The Spectral Granger causality computation hasn't been changed recently so I don't think this is a bug but rather a question about interpreting its results.
@hossein27en Can you help answering this question?

Thanks,
Martin

[moon94]

Here is the difference in power spectrums

(what is presumably the erroneous one)