I am trying to make a semi-automatic pipeline for data generalization to be used on tactile maps. I have a point layer representing natural resources.
I want to make points from within the same group (resource type) being close to each other (within specified threshold - say 50 km), aggregated into a single point representation. The final location could be a centroid of the polygon created around the set of points (clusters).
I know the Aggregate Points tool but it ignores the symbol categories and requires me to do the further processing of setting centroids and so on.
Is there any other way to group these symbols? Most of the original point symbols do not overlap but on tactile maps large distances between signs have to be maintained.
In the image there is more or less the generalization result i want to achieve. Thanks in advance!
EDIT: Oh, and I am aware of the Disperse Markers tool but as far as I know it doesn't have an option to reduce the number of resulting points...?
Thanks Dan, it looks like there is no tool implemented for such operations.
But regarding Aggregate Points - is there a way to force the tool to consider point groups separately based on attribute value (different resources types) or do I have to split the original point layer into a number of layers (or maybe subtypes) and run the tool separately?
Including a screenshot of the data in question. Thanks in advance!
It would be a useful addition, but Split by Attributes may need to be a precursor to attain what you want. If you have many attribute grouping that would be problematic, but modelbuilder could facilitate the workflow (split, aggregate the split groupings, merge the results)
Split by Attributes should work in my case. But there's another problem... Aggregate Points requires minimum of 3 points to create a polygon (this makes sense) but there are cases, where only 2 objects need to be aggregated. Is there any way to include such scenarios?
Perhaps there is a way to determine clustering? Sort of the way heat maps work but within ArcGIS?