Clean boxes are used in the CLEAN image reconstruction algorithm to define the locations where we expect source(s) to be found. There are two ways to set clean boxes - using clean_box to define rectangular areas, or clean_cw_nop and clean_cw_list to define regions of any shape.
clean_cw_nop, clean_cw_list: This method can handle clean boxes that are any shape. clean_cw_list is a (2,n) array of x,y coordinates that define the vertices of each clean box. clean_cw_nop is a vector giving the number of elements in clean_cw_list that correspond to each clean box. For example, if cw_nop_list is a (2,9) and cw_nop = [4,5], then clean_cw_list[*,0:3] are the vertices for the first clean box (a triangle), and clean_cw_list[*,[4:8] are the vertices for the second clean box (a rectangle).
clean_box: clean_box can only be used to define rectangular clean boxes. clean_box is a (4,n) array, where the 4 elements of the first dimension are xlo, ylo, xhi, yhi, and the second dimension is the box number.
If clean_box, clean_cw_nop, and clean_cw_list are all defined, clean_cw_nop/clean_cw_list take priority.
Note: the graphical method for setting clean boxes (by setting clean_mark_box to 1) uses only clean_cw_nop/clean_cw_list. It does not know about any boxes set via the clean_box parameter.