for each feature in the first interval set, find the closest in the second set, then compared the distance distribution to shuffled set 1.