I'm trying to combine point data into clusters and I think I've come up with a method that will work for what I want to accomplish. However, I don't understand why it works the way it does and I'm hoping someone can explain it to me.
Basically I have a set of wells that are operated by different companies and are scattered throughout PA. I want to determine how clustered the wells are by operator in terms of how many clusters there are for each operator and how many wells are in each cluster (clustering at a distance of 5 miles). It appears that if I run integrate on my point data for the wells it will do this. Then I can collect event values to determine how many wells are in each cluster. What I don't understand is why some of the data is broken into more than one cluster.
[Image Description] The three highlighted points are the integrated points for the company represented by the light yellow well sites. When you open the attribute table for the integrated data it is clear that the group on the right is one cluster and that integrate has determined that the group on the left is two clusters split on the blue line.
What I don't understand is why the points immediately above and below the line have been assigned the clusters that they have been. They are all within 5 miles (the integrate thresh hold) of the lower integrate point. And the upper ones are more than 5 miles from the upper integrate point. So why split it there?
Any thoughts would be helpful.
To help clarify I've added a screenshot of the process that I used to integrate all the points into the clusters that I would like to use.