%--brain mask with the brain tissuemask_name = ‘C:\Users\Administrator\Desktop\workspace\preprocessed\masks\within_brain_mask.nii‘;M   = load_untouch_nii( mask_name ); % load mask NIFTImask     = double(M.img>0);          % get 3d v%--brain functional 4d datadata_4d = ‘C:\Users\Administrator\Desktop\workspace\preprocessed\4d\func_3d.nii‘;% data_4d = ‘C:\Users\Administrator\Desktop\phycaa_workspace\phycaa_plus_2104_03_27\_PHYCAA_step1+2.nii‘;V   = load_untouch_nii( data_4d ); %--transform 4d array to 2d array, using brain_maskwithin_brain_voxels = nifti_to_mat(V,M); nt_matrix = within_brain_voxels;[V_c S_c temp] = svd( nt_matrix‘ * nt_matrix );% PC-space representationQ_c = V_c * sqrt( S_c );offSet=1;pcs=4;% estimate temporal autocorrelation maximized "sources"Q1 = Q_c( 1:(end-offSet) , 1:pcs ); % un-offsetQ2 = Q_c( (offSet+1):end , 1:pcs ); % offsetted timeseries% canonical correlations on time-lagged data[A,B,R,U,V,stats] = canoncorr(Q1,Q2);% getting stable "average" autocorrelated timeseriesa=[U(1,:)]; b=[U(2:end,:) + V(1:end-1,:)./2]; c=[V(end,:)];tset = [a;b;c];