How do I create a density-like display of overlapping polygons?
I have a large data set of polygons (more specifically: ellipses) where the actual event location could statistically be anywhere within their respective ellipse. To perform a density of their center points does not accurately portray the highest concentration of events. I am trying to use these polygons to create a density-like, heat map-ish, clustering display, similar to how spatial analyst creates densities of points. Can anyone please help with this issue?
Mapping Center Answer:
You say that an event is likely to occur anywhere in an ellipse. Given that premise, here is something you could try.
- Convert each ellipse to a raster filled with 1’s
- Sum all the rasters
Note, when converting to raster use the same extent and cell size.
Here is the output when we tried it:
Basically you will be stacking the ellipses, then overlaying the stack with a grid, then counting the number of ellipses that underlie each grid cell. If each ellipse represents a probable area for an event, and if you want to try to find the locations where the event (based on the ellipses) is most probable, then the locations (grid cells) where the largest number of ellipses occur/coincide would be a good guesses.
If you would like to post a comment, please login.