fmri_corrmat

Description

FMRI_CORRMAT Column-wise pattern similarity matrix between two fMRI datasets. Computes the Pearson correlation between every pair of columns across two multivariate datasets, yielding an output matrix of size Ncol1 × Ncol2. Columns can represent time points, trials, conditions, or group-level patterns. If a single dataset is provided, the result is the autocorrelation matrix of its columns (e.g., T × T for time points).

Usage

c = fmri_corrmat(d1)       % autocorrelation of columns (Ncol × Ncol)
c = fmri_corrmat(d1, d2)   % cross-correlation between two datasets (Ncol1 × Ncol2)

Inputs

D1: Multivariate data (e.g., voxels × columns); correlations are across rows
D2: Optional second dataset (same number of rows). Default: D2 = D1

Outputs

C: Similarity matrix of Pearson correlations between columns of D1 and D2

Examples

% Temporal similarity between time points
c = fmri_corrmat(data);           % T × T matrix
imagesc(c, [-1 1]); colorbar
% Cross-condition / trial / group-level similarity
c = fmri_corrmat(cond_data1, cond_data2);  % C1 × C2 matrix
imagesc(c, [-1 1]); colorbar