Cluster width for split clusters
I'm looking into DUT data with large rotations currently. In this case, it might be important to allow split clusters, i.e. clusters with one or multiple missing pixel hits. For instance by using:
[Clustering4D]
neighbour_radius_col = 2
neighbour_radius_row = 2
Now I checked the cluster object and saw that the cluster width in col/row does not take this into account appropriately because it's only counting how many columns/rows are part of a given cluster but not if they are touching.
Do you agree that it would make sense to change this? I would argue that in the following example the cluster should have a width of 4, not 3:
p = (1,10)
p = (1,11)
p = (1,13)
What do you think?