[FieldTrip] Concerning powcorr_orth method in ft_connectivityanalysis

[Schoffelen, J.M. (Jan Mathijs)]

Dear Sean,

Thanks for catching this. The original line 852 does not make sense at all, indeed.
May I suggest to replace this with data.fourierspctrm(:,:,i).’ (note the ‘.’)?
Can you please submit this as a pull request on github? http://www.fieldtriptoolbox.org/development/git


Many thanks,

Jan-Mathijs



On 14 Nov 2018, at 21:26, SEAN TANABE <stanabe at wisc.edu<mailto:stanabe at wisc.edu>> wrote:

Dear all,
I am trying to use the powcorr_orth method in ft_connectivityanalysis and noticed the results always produce a random connectivity pattern. A similar issue has been raised a year ago here;

https://mailman.science.ru.nl/pipermail/fieldtrip/2017-July/011695.html

I found after editing a line from ft_connectivityanalysis the results look more reasonable, and wanted to confirm if this editing is correct with the fieldtrip developers.
In fieldtrip-20170820 ft_connectivityanalysis I noticed a line which reshapes the fourierspctrm before inputting to ft_connectivity_powcorr_orth (line 852 in fieldtrip-20181113).

        dat = reshape(data.fourierspctrm(:,:,i)', nrpttap, nchan).';
        datout{i} = ft_connectivity_powcorr_ortho(dat, optarg{:});

This seems to not fit the description of ft_connectivity_powcorr_orth, which requires "dat" to have nchan rows.
I changed this to

        dat       = data.fourierspctrm(:,:,i)';
        datout{i} = ft_connectivity_powcorr_ortho(dat, optarg{:});

The above gave natural looking blobs of connectivity in the topology.

Thank you.
Sean



_______________________________________________
fieldtrip mailing list
https://mailman.science.ru.nl/mailman/listinfo/fieldtrip
https://doi.org/10.1371/journal.pcbi.1002202

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20181115/5ad18edb/attachment-0002.html>