Skate_FFT.m

%CONDITIONS...
...preferred, clockwise - non-preffered, CCW
%%    
ccc
%
exp = 'Skateboard';
%subs = {'100' '101' '102'};
subs = {'100'}; %to test on just one sub 

nsubs = length(subs); 
conds =  {'P_CW';'P_CCW'; 'NP_CW'; 'NP_CCW'};
conds_lab = {'Preferred Clockwise'; 'Preferred Counterclockwise'; 'Non-preferred Clockwise'; 'Non-preferred Counterclockwise'};
%{'P_CW';'P_CCW'; 'NP_CW'; 'NP_CCW'};
nconds = length(conds);
Pathname = 'M:\Data\Skateboard\Winter2019/';
[ALLEEG EEG CURRENTSET ALLCOM] = eeglab;


for i_sub = 1:nsubs
    for i_cond = 1:nconds
        
        Filename = [subs{i_sub} '_' exp '_' conds{i_cond}];
        EEG = pop_loadset('filename',[Filename '_Corrected_Standard.set'],'filepath','M:\Data\Skateboard\Winter2019\segmentsFFT\');
        [ALLEEG, EEG, CURRENTSET] = eeg_store( ALLEEG, EEG, 0 );


    end
end
eeglab redraw

%% loop through the sets and do a wavelet analysis
% Version 4 - NEXT - loop this through electrodes
% Version 4 - compare left vs right hemisphere


wavenumber = 6; %wavelet cycles
freqs = [1:1:30]; %wavelet frequencies
power_out = [];
i_count = 0;
n_electrode = EEG.nbchan;
for i_sub = 1:nsubs 
    fprintf(['Subject - ' num2str(i_sub) '. \n']);
    for i_cond = 1:nconds
        fprintf(['Condition - ' conds_lab{i_cond} '. \n']);
        power = [];
        i_count = i_count+1; %which data set in ALLEEG to use
        n_trials = ALLEEG(i_count).trials;
        for i_trial = 1:n_trials
            for i_electrode = [2 3 4 5 6 10 11 12 13 14] %pairs of contralateral electrodes
                tempdata = ALLEEG(i_count).data(i_electrode, :,i_trial) ;
                [temp_power temp_times temp_phase] = BOSC_tf(tempdata,freqs,EEG.srate,wavenumber);
                power(:,:,i_electrode,i_trial) = temp_power; %take log 
            end
            
        end
        power_out(:,:,:,i_sub,i_cond) = nanmean(power,4); %save the power data
 
        
    end
 
end



%% plot spectrograms for each subject (rows) and condition (columns)
 figure;
 i_count = 0;
 CLim = [-1e6 1e6];
for i_sub = 1:nsubs
    maxval = max(max(max(max(abs(power_out(:,:,:,i_sub,:))))))/10; %computer a colorbar plot range for each subject based 1/10th the max value
    CLim = [-1*maxval maxval];

    for i_cond = 1:nconds
        i_count = i_count+1;
        
        subplot(nconds,nsubs,i_count)
        imagesc(EEG.times,freqs,squeeze(power_out(:,:,1,i_sub,i_cond))-squeeze(power_out(:,:,2,i_sub,i_cond)),CLim); %
        colormap(redblue)
        set(gca,'Ydir','normal');
        title(['Sub ' num2str(i_sub) '-' conds_lab{i_cond} '- Left-Right'])
        colorbar
    end
end
xlabel('Time (s)');
ylabel('Frequency (Hz)');



%% Collapse over time to get power spectra, average over subjects

power_spectra = squeeze(mean(power_out(:,:,:,:,:),2)); %collapse over time can pick time range here in 2nd dim.
power_spectra_std = squeeze(std(power_out(:,:,:,:,:),[],2)); 

if nsubs > 1 % if you have more than one subject take the mean over subjects
    mean_power_out = squeeze(mean(power_spectra,3));
    stderr_power_out = squeeze(std(power_out,[],3))./sqrt(nsubs);
    
else %if only one subject just plot mean and std over trials
    
    mean_power_out = power_spectra;
    stderr_power_out = power_spectra_std;
end

%% Plot spectra accross conditions
figure;
boundedline(freqs,mean_power_out(:,6,1),stderr_power_out(:,6,1), 'b', ...
    freqs,mean_power_out(:,6,2),stderr_power_out(:,6,2), 'g', ...
    freqs,mean_power_out(:,6,3),stderr_power_out(:,6,3), 'r', ...
    freqs,mean_power_out(:,6,4),stderr_power_out(:,6,4), 'k', ...
    freqs,mean_power_out(:,12,1),stderr_power_out(:,12,1), 'b--', ...
    freqs,mean_power_out(:,12,2),stderr_power_out(:,12,2), 'g--', ...
    freqs,mean_power_out(:,12,3),stderr_power_out(:,12,3), 'r--', ...
    freqs,mean_power_out(:,12,4),stderr_power_out(:,12,4), 'k--');
axis tight
xlim([0 30])

xlabel('Frequency (Hz)');
ylabel('Power (uV^2)');
title('Baseline power Standard Trials Electrode Left');
legend(conds_lab,'Location','NorthEast');