I'm currently creating a workflow from former case studies that employs the use of K-Nearest Neighbors algorithm to determine the distance band or epsilon value for use as input in DBSCAN. The nature of my work requires that I adhere to this methodology, so I have to use it.
Unfortunately, I cannot find a specific tool in ArcGIS Pro's toolbox that uses the K-NN algorithm specifically as proximity search tool based upon a minimum number of feature. The closest I've come to this is using a combination of tools in Model Builder:
- Create Near Table (N=3) -> Table to Table w/Select by Attribute (Near Rank = 1) -> Export Table -> Add Field (To identify duplicates)-> Field Calculator (Duplicates =1) -> Delete all rows which are duplicates
I've also scripted this process outside of ArcGIS Pro in python using SciKit Learn package.
Is there an easier way to do this with an off the shelf tool? Or will I have to create a custom tool using Python and use the script to build the tool in Model Builder?
I'm using ArcGIS Pro 2.8
