Tango's statistic to perform a general clustering test is expressed as follows:
$$T = (r-p)^{'} A (r-p)$$
where \(r^{'} = [O_1/O_+, \ldots, O_n/O_+]\), \(p^{'}=[E_1/E_+, \ldots, E_n/E_+]\) and \(A\) is a matrix of closeness which measures the cloneness between two zones (the higher the closer).
Tango proposes to take \(A_{ij}=\exp\{-D_{ij}/\phi\}\), where \(D_{ij}\) is the distance between centroids of regions i and j, and \(\phi\) is a constant that measures how strong is the relationship between regions in a general way.
Tango, Toshiro (1995). A Class of Tests for Detecting 'General' and 'Focused' Clustering of Rare Diseases. Statistics in Medicine 14, 2323-2334.
DCluster, tango.stat, tango.boot, tango.pboot