[Eeglablist] pvaf parameter for ICA components

[David Jin]

Hi All,
We have two questions about ICA with eeglab. In our current dataset, we are
finding that the percent variance accounted for (when using eeg_pvaf) is
negative for the majority of our components such that when summed, they
come out to a negative percent variance accounted for. Is this a normal
finding (and what is its significance)? Secondly, do the components with
the maximum percent variance accounted for (found via eeg_pvaf) correspond
to the same components with the maximum mean power of back-projection
(output when running pop_envtopo)?



Many thanks,
David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20181206/c4addcfe/attachment.html>