So I have a few thousands trees. Some of them have disease (just an example).
Assuming something in the area cause it I would like to know what is the chance to have the disease in a new tree location.
I do a kernel density on the trees. Then on just the trees with the disease and divide them to get the results.
This works nice for areas that have many trees.
If I have a big gaps between group of trees I see a problem in these gaps.
The values in the kernel of the sick trees go from 2 to 1 (again, just an example) but the kernel for trees go from 10 to 3 so the value of the result change from 0.2 to 0.33 - it does not look right.
I am thinking about just clip all the value below 1 from the sick tree kernel out and have to results there.
Anybody have any idea about this problem?
Thanks
