About How to Construct the Feature Vector (fv) for Custom Analyses

[batiyilmaz]

Hi David,

Upon your recommendation, in order to implement the time- and frequency-resolved MVPA, TGM, and MVCC, I first decomposed the EEG time series data using complex Morlet wavelet convolution and now I have the downsampled data for each participant in separate conditions with these dimensions: 63 (channels) x 70 (frequencies) x 36 (timepoints) x 123 (trials).

I saved the data to use it as the input for the mvpalab_mvpa() function to perform MVPA analysis so
input.data = 4-D data;
input.times = -0.1:0.02:0.6; % in seconds
input.fs = 50; % Hz
input.format = 'mvpalab';

I am trying to loop over frequencies as you also suggested before in this way:

for freq_idx = 1:70

fv{1, 1}.X.a = squeeze(data{1, 1}.condition_a(:, freq_idx, :, :));

% Compute MVPA analysis:
[result,cfg] = mvpalab_mvpa(cfg,fv);

% Save cfg file:
mvpalab_savecfg(cfg);

end

However, I receive an error, my question is that how should I construct the feature vector (fv) parameter to avoid these errors and successfuly implement MVPA and run statistical analyses on them via mvpalab_permaps() and mvpalab_permtest(), and finally plot the results using the functions of MVPAlab without any problems?

If you could guide me through this issue, I would greatly appreciate it.

Thank you so much for your time in advance.
Also, Happy New Year!

Best,
Bati