I've been using optimized hotspot analysis and the hotspot maps I am getting back frequently have rings of perfectly circular hotspots.

Is this because of the way Getis-Ord Gi* tests calculates neighborhood sums?

Natural events don't often happen with perfect symmetry and when I look at the underlying data, it doesn't appear to be set out in circles. I feel like it must be due to the statistical processes underlying hot spot analysis.

Circular Hotspots Example



Browse other questions tagged or ask your own question.