For example, if you use estEllipse, could you grab the centroid of the polygons and estimate the direction of the difference (+/-).
From Peter Breslin:
If we wanted to estimate not only the degree of difference between saguaro microsites and random microsite, but also the direction (+ being preference, - being avoidance) I wonder how rKIN might get to that? I have been playing with the percent overlap of ellipses, etc, and those definitely get to degree of separation between saguaro microsite features and random ones, but I'm wondering about how to quantify that positive or negative direction. Might need a different approach.
For example, if you use estEllipse, could you grab the centroid of the polygons and estimate the direction of the difference (+/-).
From Peter Breslin: If we wanted to estimate not only the degree of difference between saguaro microsites and random microsite, but also the direction (+ being preference, - being avoidance) I wonder how rKIN might get to that? I have been playing with the percent overlap of ellipses, etc, and those definitely get to degree of separation between saguaro microsite features and random ones, but I'm wondering about how to quantify that positive or negative direction. Might need a different approach.