[FieldTrip] Single-trial time-frequency connectivity

[daniel.strahnen at uni-ulm.de]

Dear FieldTrip-Community,

 

Is it possible to calculate time-frequency coherence on a single trial?

When using the configuration below it works perfectly when including all
trials, but if I specify cfg_conn.trials = 1 then I receive an error.

I went through the code and found that data.cumtapcnt does not have the
usual dimensions (ntrials x nfreq) but now is changed to (nfreq x 1).

Since I don't want to mess up the FieldTrip-code I wanted to ask if there is
a solution to this problem?

 

        cfg              = [];

        cfg.output       = 'fourier';

        cfg.channel      = channels;

        cfg.method       = 'wavelet';

        cfg.foi          = 1:1:48;

        cfg.t_ftimwin    = 20./cfg.foi;

        cfg.toi          = 1:0.01:12;

        cfg.width      = 3;

        cfg.keeptrials = 'yes';

        freq_data = ft_freqanalysis(cfg, data);

 

        cfg_conn = [];

        cfg_conn.method = 'coh';

        cfg_conn.trials = 1;

        coh_str = ft_connectivityanalysis(cfg_conn, freq_data);        

 

Error:

Index in position 1 exceeds array bounds (must not

exceed 1).

 

Error in ft_checkdata>fixcsd (line 966)

        tmpdat =

        transpose(data.fourierspctrm(indx,:,m,k));

 

Error in ft_checkdata (line 693)

    data = fixcsd(data, cmbrepresentation,

    channelcmb);

 

Error in univariate2bivariate (line 63)

        data = ft_checkdata(data,

        'cmbrepresentation', 'full');

 

Error in ft_connectivityanalysis (line 398)

          [data, powindx, hasrpt] =

          univariate2bivariate(data,

          'fourierspctrm', 'crsspctrm', dtype,

          'cmb', cfg.channelcmb, 'keeprpt',

          normrpt);

 

 

 

Thank you and best regards

 

Daniel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20201207/bd10df5d/attachment.htm>