Help needed script frequency

Dear Community

I have created a loop to get phase synchronization for frequency range from 4-8 hz, but for some reason the script calculates only the value for 8Hz.

What am I doing wrong?

% phase-based connectivity

load 'EC_I.mat'


% names of the channels you want to compute connectivity between
channel1 = 'Fz';
channel2 = 'F8';

% create complex Morlet wavelet
frequencies = [4:8];
time      = -1:1/EEG.srate:1;
half_wavN = (length(time)-1)/2;
all_wavelet = zeros(10, 1001); 
for i_frequencies = frequencies 
    center_freq = i_frequencies ;

wavelet   = exp(2*1i*pi*center_freq.*time) .* exp(-time.^2./(2*(4/(2*pi*center_freq))^2))
all_wavelet(i_frequencies,:) = wavelet ;
end

all_phase_synchronization = zeros(1, 1000); 
for i_frequencies = frequencies 
    center_freq = i_frequencies ;

phase_synchronization = abs(mean(exp(1i*(phase_data(2,:)-phase_data(1,:)))));
all_phase_synchronization(i_frequencies,:) = phase_synchronization ;
end

disp([ 'Synchronization between ' channel1 ' and ' channel2 ' is ' num2str(phase_synchronization) '!' ])

All tips would be very helpful!! Thanks in advance,

Carmen

[Walter Roberson]

for i_frequencies = frequencies 
    center_freq = i_frequencies ;
phase_synchronization = abs(mean(exp(1i*(phase_data(2,:)-phase_data(1,:)))));
all_phase_synchronization(i_frequencies,:) = phase_synchronization ;
end

frequences is 4:8 and you are setting i_frequencies to those. So i_frequencies is content, not index. You will end up writing to all_phase_synchronization(4:8,:) rather than to all_phase_synchronization(1:5, :)

[Carmen Sergiou]

Thank you so much for responding, so now I have changed the following:

frequencies = [4:8];
time      = -1:1/EEG.srate:1;
half_wavN = (length(time)-1)/2;
all_wavelet = zeros(10, 1001); 
for i_frequencies= frequencies
    center_freq = i_frequencies ;

wavelet   = exp(2*1i*pi*center_freq.*time) .* exp(-time.^2./(2*(4/(2*pi*center_freq))^2))
all_wavelet(1:5,:)= wavelet;
%all_wavelet(i_frequencies,:) = wavelet ;
end

all_phase_synchronization = zeros(1, 1000); 
for i_frequencies = frequencies 
    center_freq = i_frequencies ;

phase_synchronization = abs(mean(exp(1i*(phase_data(2,:)-phase_data(1,:)))));
all_phase_synchronization(1:5, :) = phase_synchronization ;
end

disp([ 'Synchronization between ' channel1 ' and ' channel2 ' is ' num2str(phase_synchronization) '!' ])

And get this error :

Unable to perform assignment because the size of the left side is 5-by-0 and the

size of the right side is 1-by-1001.

Sorry, I am very new to matlab coding!!

[Carmen Sergiou]

I am now copying every electrode pair combination in excel, but is there also an option to have it in matlab in table? If I want to do analysis or correlations with it?

And I have to figure out how to differ between my two conditions also, if that is not possible I will make 2 seperate files, if there is documentation for that also very welcome!

Thanks again, really a life saver!!

Best,

Carmen

Answer 1

[Walter Roberson]

frequencies = [4:8];
nfreq = length(frequencies);

time      = -1:1/EEG.srate:1;
half_wavN = (length(time)-1)/2;
all_wavelet = zeros(10, 1001); 
for idx = 1 : nfreq
    center_freq = frequencies(idx);
    wavelet   = exp(2*1i*pi*center_freq.*time) .* exp(-time.^2./(2*(4/(2*pi*center_freq))^2));
    all_wavelet(idx,:)= wavelet;
end

all_phase_synchronization = zeros(nfreq, size(all_wavelet,2)); 
for idx = 1 : nfreq
    center_freq = frequencies(idx);
    phase_synchronization = abs(mean(exp(1i*(phase_data(2,:)-phase_data(1,:)))));
    all_phase_synchronization(idx, :) = phase_synchronization ;
end

but you do not use all_wavelet() after you create it. and your phase_synchronization calculation always calculates over the same data and does not use the center_freq information. You also seem to expect it to be a vector rather than having one for each frequency. I suspect that you should not be using a loop there, just

 all_phase_synchronization = abs(mean(exp(1i*(phase_data(2,:)-phase_data(1,:)))));

as a single statement with no loop and no initialization to zero needed.

[Carmen Sergiou]

Thank you, I have implemented the changes and everything works correctly now.

I see a different phase-synchronization value then before, so I assume it is now measured over 4-8 Hz instead of over 8 Hz. When I look at the all_phase_synchronization table in my workspace, I do see all the same values but I assume that is the averaged over the frequency bands.

I decided to do use the loop, because if I just used the single row the phase value stayed the same as in my 8Hz.

Thank you so much!! I believe I can continue now (see screenshots for the workspace tables)

[Walter Roberson]

all_phase_synchronization is exactly the same for all rows because your code does not use the center_freq or idx information at all. Every iteration of the loop only works with phase_data(2,:)-phase_data(1,:) so it is exactly the same calculation every time.

[Carmen Sergiou]

Ok, well I am still stuck then.

Probably need to do some steps before that.

Thank you for your time!