I have two point shapefiles. I want to analyse whether the points in the dataset A are clustering around their respective reference point in the dataset B. Each reference point in the dataset B has a set of related points in the dataset A (linked through a common categorical variable). So, I want to carry out the analysis per sub-sets, based on attribute value, (A) in relation to the specified reference point in B.

Can I do this in ArcGIS “Spatially constrained multivariate clustering"?

  • Can you, please, be a bit more specific regarding your two shapefiles, as well as the variables? Some screenshots from the shapefiles and their attribute tables would be helpful – kowalski Mar 19 '19 at 13:26
  • Hi, thanks. To clarify, both point shapefiles represent events, I am not looking at clustering of values, events only (clustering of a number of events from the dataset A around one point in the dataset B). The only required attribute value is a categorical one which I indicates the subsets. – user138966 Mar 20 '19 at 11:38
  • You should take a look at tools from either pro.arcgis.com/en/pro-app/tool-reference/spatial-statistics/… or pro.arcgis.com/en/pro-app/tool-reference/spatial-statistics/… If I understand correctly what you want to do, I would use Spatial Autocorrelation (Moran's I) per subsets and then calculate central feature or average nearest neighbor (again, per subset) to see how close it is to the corresponding point of the other dataset. However, further clarification is still desired. – kowalski Mar 20 '19 at 19:00
  • Thanks a lot. The tricky thing there is that there are several subsets and thus, this would be quite time-consuming. Therefore, I wonder how could I conduct the analysis repeatedly for several subsets in a time-efficient way. I have the categorical variable in place for that purpose but how to do this in ArcGIS? – user138966 Mar 21 '19 at 11:25
  • 1
    Many thanks for the tips. This is clearer now and I will look at model builder for further information. – user138966 Mar 21 '19 at 13:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy