Three options are currently supported by BINGRP,
Both the modes currently implemented perform spatial grouping. The X,Y plane of a dataset is divided into a number of zones according to user input. Each spatial pixel is assigned a zone number. In the 2-D case the output group number is equal to the zone number, but in datasets of higher dimensionality different X,Y planes have the constant NZONE*(I_PLANE-1) added. The result is that pixels are grouped spatially but retain independence in higher dimensions.
Note that when there are more than 2 dimensions the X and Y axes must be the first two axes - use AXORDER to reorder if this not the case.