[FieldTrip] TFR Cluster permutation on Single-sensor

[Tzvetan Popov]

Hi Tahereh,
Re: first- this shouldn’t happen unless you have really only freqs 1-10Hz
Re: second- since you have only one channel you shouldn’t specify ‘neighbourdist’, but cfg.neighbours = []. In that way clustering is done along the freq and time dimensions the adjacency of which is defined by the time freq resolution.

best
tzvetan


> Hi fieldtripers,
> 
> I am trying to do a cluster permutation on Time-frequency data of a single channel (of 64 channel EEG) in a between subject design based on the example provided here. Two weird things happen. 
> First, although I set cfg.frequency = 'all' for stats, but it always returns results for only ~1-10Hz instead of 1:15 Hz I have in my freq data (to be precise : stat.cfg.frequency =[0.9995    9.9951]).
> Second, based on the explanation in this paper, I assume cfg.neighbourdist in single-sensor test somehow shows the number of adjacent points, but when I change this number the clusters don't change at all. 
> Here is my code, I would appreciate if anyone could tell me what is happening.
> 
> Thanks,
> Tahereh   
> 
> % Time frequency analysis
> cfg = [];
> cfg.output     = 'pow';
> cfg.method     = 'mtmconvol';
> cfg.taper      = 'hanning';
> f1 = 1;f2 = 15; fstep = 0.1;
> cfg.foi        =f1:fstep:f2;
> t1 = -1; t2 = 1;
> cfg.toi        = t1:0.02:t2;
> n_cycles = 5;
> cfg.t_ftimwin  = n_cycles ./cfg.foi; 
> cfg.keeptrials ='no';
> 
> [freqdata] = ft_freqanalysis(cfg, data);
> 
> %store freq data for each condition/subject
>  TFDesc(Condition,subject) =  {ft_freqdescriptives(cfg,freqdata)}
> 
> %Comparing Condition1 and Condition2
> cfg = [];
> Condition1_avg= ft_freqgrandaverage(cfg, TFDesc{cond1,: });
> Condition2_avg =  ft_freqgrandaverage(cfg,  TFDesc{cond2,: });
> 
> % make a dummy structure with the difference between Condition 1 & 2
> avgDiff = Condition1_avg;
> avgDiff.powspctrm = Condition1_avg.powspctrm-Condition2_avg.powspctrm;
> 
> cfg = [];
> cfg.channel          =  'FP2'
> cfg.frequency        = 'all'
> cfg.neighbourdist    = 4;
> cfg.latency          = 'all'
> cfg.avgovertime      = 'no';
> cfg.avgoverfreq      ='no';
> cfg.avgoverchan      = 'no';
> 
> cfg.clusteralpha     = 0.05;
> cfg.statistic        = 'ft_statfun_depsamplesT';
> cfg.numrandomization = 500;
> cfg.correctm         =   'cluster';
> cfg.method           = 'montecarlo';
> cfg.design           = [
>     1 2 3 4 5 6 7 8 9  1 2 3 4 5 6 7  8 9     % subject number
>     1 1 1 1 1 1 1 1 1   2 2 2 2 2 2 2 2 2   ];  % condition number
> cfg.uvar = 1;                                   % "subject" is unit of observation
> cfg.ivar = 2;                                   % "condition" is the independent variable
> cfg
> stat = ft_freqstatistics(cfg, TFDesc{cond1,selectedSubj}, TFDesc{cond2,selectedSubj});
> 
> 
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20150605/c2b98c35/attachment-0002.html>