topic Re: Clustering points funny result in Spatial Statistics Questions

Clustering points funny result

PaulPrevedoros1 — Sat, 21 Aug 2021 06:33:49 GMT

I have performed a Density Based clustering. The tool Produces a a point feature class identifying the points and their identified cluster.

I have attached a map where the points for Cluster ID 40 is identified.

I used Multi-Scale (Optics)

Min no Features 5

Search Distance 100

Would you expect the result, that I got in the attached Map

Re: Clustering points funny result

DanPatterson — Sat, 21 Aug 2021 09:03:51 GMT

Did you see this?

How Density-based Clustering works—ArcGIS Pro | Documentation

perhaps this is a cause

(Note: OPTICS is an ordered algorithm that starts with the feature with the smallest ID and goes from that point to the next to create a plot. The order of the points is fundamental to the results.)

I assume you have no time separator?

Re: Clustering points funny result

PaulPrevedoros1 — Sun, 22 Aug 2021 04:49:27 GMT

Dear Dan

I agree with you however, I was running this algorithm over, a series of search distances, and there are some results are not consistent.

If you look at the entire map there is a huge space from the source and this area.

There is no time separator, as these are ponds in Angkor, and at 100m search separation this sort of behaviour did not occur

Re: Clustering points funny result

DanPatterson — Sun, 22 Aug 2021 08:35:33 GMT

Paul, I moved this question to "Spatial Statistics Questions" to see if others more vested in this type of analysis can weigh in.

Re: Clustering points funny result

EricKrause — Mon, 23 Aug 2021 14:21:22 GMT

Hi Paul, I noticed in the geoprocessing pane that you have some pending edits on the input points. Can you close the edit session and see if this still happens?

Re: Clustering points funny result

PaulPrevedoros1 — Mon, 23 Aug 2021 23:45:34 GMT

Yea I notice the funny behaviour much later when I was doing my next stage.