Thank you Thom, Thank you Jeffrey.
Thom,
You�??re right the hardest part is to redistribute the points from K-Mean cluster to the nearby cluster. The idea is we have 10k hydrants scattered in the town for inspection. We need to slice the inspection area into sub-areas that each have 50 hydrants close to each other, so that we can allocated the inspection works to our scopers. I know this might sound like a ArcGIS Logistic or similar study but we don�??t have access to these extension/tools. I guess there�??re some other scenarios that have similar required as mine, each create census areas based on each of 200 nearby households?
Jeffrey,
There�??re no overlapping in the dataset. As mentioned above the issue in my scenario is there�??s no such �??desired distance�?� exists.