<div dir="ltr"><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Dear Fieldtrip community, </span></p><div><br></div><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">my name is Stefan Dvoretskii and I am working in the PainLabMunich on a pain-related data preprocessing pipeline for EEG. Currently, I am analyzing connectivity in resting-state EEG data. In this e-mail, I would like to suggest an alternative, more memory-efficient algorithm for the dwPLI computation.</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">
During the analysis, we have faced memory problems when trying to compute the debiased weight phase lag index (dwPLI) on source-level data. We tried to compute a connectivity matrix from band passed ‘virtual channel data’ (2020 voxels x 500 samples x 299 trials) after spatial filtering with this code:</span></p><br><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">[Snippet 1]</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">fois = 8:0.5:13; </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% Alpha band with a freq. resolution of 0.5</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">connMatrix = zeros(2020,2020,length(fois));</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,255);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">for</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> idx = 1:length(fois)</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% Fourier components</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> cfg = [];</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> cfg.method = </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'mtmfft'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">;</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> Cfg.output = </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'fourier'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">;</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> cfg.keeptrials = </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'yes'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">;</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> cfg.foi = fois(idx);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> cfg.tapsmofrq = 1;</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> virtFreq = ft_reqanalysis(cfg,virtChan_data);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% Connectivity</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> cfg = [];</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> cfg.method = </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'wpli_debiased'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">;</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> source_conn = ft_connectiviyanalysis(cfg,virtFreq);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> connMatrix(:,:,idx) = source_conn.([</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'wpli_debiased'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">,</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'spctrm'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">]);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,255);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">end</span></p><br><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Matlab was just being killed by the system attempting to take too much memory, without any error message. By debugging, I figured that in our case the problem was a check of the data in the function ft_connectivityanalysis (line 251):</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">data = ft_checkdata(data,</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'datatype'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">,{</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'freqmvar'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(160,32,240);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">'freq'</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">});</span></p><br><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">The cross-spectral density was computed resulting in an array of 2020*2020*300 complex positions for one recording, totaling to 10GB of working memory. It was too much for the current working machine.</span></p><br><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">I have addressed this memory problem by creating a function combining the ft_connectivityanalysis() and ft_connectivity_wpli() that avoids the memory-costly precomputation of the CSD and computes the cross-spectrum from the Fourier spectrum. It processes one frequency of interest at a go [Snippet 2]:</span></p><br><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">[Snippet 2]</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">%% Compute dwPLI using only Fourier spectrum to save space</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% input: </span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% - data: virtual channels frequency decomposition for one frequency of interest (e.g. 8.5 Hz). Output of ft_frequencyanalysis; includes Tapers x Channels Fourier spectrum as a field.</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,255);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">function</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> connMatrix = conn_dwpli_fourier(data)</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> nchan = size(data.label,1);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> nrpt = size(data.cumtapcnt,1);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> sumtapcnt = [0;cumsum(data.cumtapcnt(:,1))];</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% see ft_connectivity_wpli.m</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> outsum = complex(zeros(nchan, nchan));</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> outsumW = zeros(nchan, nchan);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> outssq = complex(zeros(nchan, nchan));</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% alternatively: loop along channel combinations. reduces the csdTrial size</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% (converts to the float), but increases the amount of operations. good for</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% cluster/peer network jobs.</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,255);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">for</span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> p=1:nrpt</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> indx=(sumtapcnt(p)+1):sumtapcnt(p+1);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> fourierTrial = transpose(data.fourierspctrm(indx,:));</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> csdTrial = (fourierTrial*fourierTrial')./length(indx);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(60,118,61);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">% imaginary part (phase info) only</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> input = imag(csdTrial);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> outsum = outsum + input;</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> outsumW = outsumW + abs(input);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> outssq = outssq + (input.^2);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> </span><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,255);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">end</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"> connMatrix = (outsum.^2 - outssq)./(outsumW.^2 - outssq);</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:"Courier New";color:rgb(0,0,255);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">end</span></p><br><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">This function [Snippet 2] is to be used in place of ft_connectivityanalysis in [Snippet 1]. We attach an example of virtChan_data for reproducibility: </span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><a href="https://www.dropbox.com/s/cmrv6zc4rsawfvh/virtFreq_example_one_foi.mat?dl=0" target="_blank">https://www.dropbox.com/s/cmrv6zc4rsawfvh/virtFreq_example_one_foi.mat?dl=0</a><br></span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">We hope to help people facing similar problems and would be looking forward to eventual suggestions of integrating this algorithm into fieldtrip codebase. </span></p><br><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Best regards,</span></p><p dir="ltr" style="line-height:1.2;margin-top:0pt;margin-bottom:0pt"><span style="font-size:11pt;font-family:Calibri,sans-serif;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Stefan Dvoretskii</span></p></div>