[FieldTrip] Question connectivity analysis / ICA / MVAR

["Jörn M. Horschig"]

Hi Martin,

that error most likely occurs because the rank of your datamatrix is not 
full. This can e.g. be the case for EEG data where you include the 
reference channel (or common average rereferenced) or when you used 
ft_channelrepair or, in your case, when you removed an ICA component, 
therefore reducing the rank of your data.
I am not sure what the best solution for this is, because there are 
several, e.g. you could do your MVar analysis on ICA data directly, do 
it on a subset of channels so that the datamatrix is rank sufficient 
again or just refrain from doing the mvar analysis.

Best,
Jörn


On 4/22/2013 8:36 PM, Martin Winkels wrote:
> Hello Guys,
>
> I am trying to perform connectivity analysis in sensor domain 
> (fieldtrip and matlab are up to date). Within doing that I am facing a 
> small problem in the ICA analysis in combination with a MVAR analysis. 
> If I execute the following code, I get an error message as follows:
>
>
> Error using chol
> Matrix must be positive definite.
>
> Error in armorf (line 40)
> ap(:,:,1) = inv((chol(ap(:,:,1)/Nr*(Nl-1)))');
>
> Error in ft_mvaranalysis (line 390)
>             [ar, tmpnoisecov] = armorf(dat, numel(rpt{rlop}), 
> size(tmpdata.trial{1},2), cfg.order);
>
> Error in mu_firstexample (line 90)
> mdata           = ft_mvaranalysis(cfg, data_ma_ica);
>
>
> Can anybody help me with that issue?
>
>
> % artifact correction: ica component analysis
>
> cfg             = [];
> cfg.method      = 'runica';
> cfg.runica.pca   = 40;
> comp            = ft_componentanalysis(cfg, data_ma);
>
> cfg             = [];
> cfg.layout      = 'CTF275.lay';
> cfg.viewmode    = 'component';
> ft_databrowser(cfg, comp)
>
> %% artifact correction: ica component rejection
>
> cfg             = [];
> cfg.component   = [4]; % to be removed component(s)
> data_ma_ica     = ft_rejectcomponent(cfg, comp, data_ma);
>
> %% compute MVAR, spectral transfer and granger
>
> % mvar
> cfg             = [];
> cfg.order       = 5;
> cfg.toolbox     = 'bsmart';
>
> mdata           = ft_mvaranalysis(cfg, data_ma_ica);
>
>
> Thank you,
>
> -- 
> Martin Winkels
>
> B.Sc. / Ingenieur in
> Medizininformatik und Biomedizintechnik
>
> University of Münster
>
> phone: +49 177 6090603 <tel:%2B49%20177%206090603>
> mail: m_wink10 at uni-muenster.de <mailto:m_wink10 at uni-muenster.de>
>
>
>
>
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip


-- 
Jörn M. Horschig
PhD Student
Donders Institute for Brain, Cognition and Behaviour
Centre for Cognitive Neuroimaging
Radboud University Nijmegen
Neuronal Oscillations Group
FieldTrip Development Team

P.O. Box 9101
NL-6500 HB Nijmegen
The Netherlands

Contact:
E-Mail: jm.horschig at donders.ru.nl
Tel:    +31-(0)24-36-68493
Web: http://www.ru.nl/donders

Visiting address:
Trigon, room 2.30
Kapittelweg 29
NL-6525 EN Nijmegen
The Netherlands

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20130423/3b023504/attachment-0002.html>