[FieldTrip] cluster based permutation tests - grandaverage - grad structure

[cornelia.quaedflieg at uni-hamburg.de]

Dear fieldtrip users, 
 
we are having troubles running the cluster based permutation tests for MEG data analysis. We tried to run ft_freqstatistics for previously calculated grandaverages over subjects in two conditions. However, we are not sure how to specify cfg.neighbours. The grandaverage datasets in the fieldtrip tutorial contain a gradiometer structure, which is probably used to prepare neighbouring channels. When we run ft_freqgrandaverages on our data, this grad structure disappears because it cannot be averaged over subjects. Thus, we were wondering if there is a way to keep the gradiometer information in the averaged data or to append an example grad structure later on?
 
We appreciate any suggestions on that matter. Thank you in advance!
 
Best

Dr Conny Quaedflieg & Hanna Stoffregen 
Hamburg University 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20170512/a20b664c/attachment-0001.html>