Estimating coverage with (possibly overlapping) rectangles

[sponsoredlinks]

hi, just a quick question on recommended usage of the new RectClip functionality - apologies if this has been asked elsewhere (I couldn't see it anywhere)

We have multiple rectangles (which themselves possibly overlap) that we are using to clip multiple polygons in order to estimate coverage of the subject pattern. Currently I just use Difference() with lists of (clip rectangles, subject polys) and calculate the area of the resulting clip - this appears to deal with overlaps automatically

If we were to use RectClip it seems that there would be either unions (clipper overlaps) that need to be taken into account manually when calculating total coverage as a sum of the individual rectclips - is there a good way that you know of to do this? Or a method to generate the equivalent of Difference that uses any optimizations contained within the new functionality?

edit: corrected description

[AngusJohnson]

Or a method to generate the equivalent of Difference that uses any optimizations contained within the new functionality?

Not that I'm aware of, no.
The ClipRect function could be modified to return difference rather than union.
However that'd be quite a bit of work and it's not something I'd likely be persuaded to do myself.

[sponsoredlinks]

Thanks yes this was what I suspected. Out of interest I came up with the following to try to do difference for overlapping clips and non-overlapping subjects - not sure if this makes sense

Iterate over all clips (rectangles) and for each ascertain whether they overlap any previous clip
If so, add each to a list of clip overlaps (keep to only the overlap between the two clips)

for each clip look at the overlap with each subject and sum each of these
Then for each overlap look at the overlap with each subject and subtract each of these from the sum

I think this works to do difference (it's possibly similar to how you do it already?)