Hierarchical clustering is used based on the correlation of all data at hand.

<p>A correlation matrix together with clustering (i.e., Pearson uncentered) of the feature points is presented from R = −1 negative correlation (white) to R = 1 positive correlation (blue).