1

enter image description hereI want to aggregate points by count, for example, groups of 10 or 20. I use ArcMap 10.5 and ArcGIS pro, and so far i have only found tools that will aggregate by proximity and sum number of points. I need the aggregated points to be equal in weight/count.

Gunnar B
  • 11
  • 3
  • 1
    Can you provide an example of how the points should be aggregated, please? You have a point dataset with features distributed over a geographic area. What exact principle would you like to use to assign them into individual groups? And what happens with their geometries after they have been assigned to a certain group? – Alex Tereshenkov Feb 20 '18 at 09:42
  • I've added a picture. I would like to make groups of perhaps 10 points and represent them by a midpoint. The grouping should be based on witch 10 points are closest to each other, but eatch group has to have the same number of points. – Gunnar B Feb 20 '18 at 09:59
  • I have added an answer, but on review it may not be the best for your solution. How do you determine the starting point for your groups? Reason I ask is each point is near another point. So as you try and generate a result for each point it will result in duplication as points will effectively double counted. Do you have any criteria in your datasets which are effectively the central points for your groupings? – Keagan Allan Feb 20 '18 at 12:59
  • 1
    It is still unclear what you are trying to achieve. People will have a hard time trying to help you. – Alex Tereshenkov Feb 20 '18 at 13:10

1 Answers1

1

You could use the GenerateNearTable in the Toolbox.

Open the Tool via "Analysis Tools" -> "Proximity" -> "Generate Near Table"

enter image description here

From here select your points as the "Input Features" and the "Near Features", the one layer is both inputs.

Choose that you want the Location and untick all other options. Select that you want the "Maximum number of closest matched" to be 10. Run the tool and you will generate a Table.

enter image description here

The table has an "In_FID" and a "NEAR_RANK". You can use the "NEAR_X" and "NEAR_Y" fields to plot these points, using the "IN_FID" as the identifier for the groups.

Granted there will be duplication, but this is a starting point for grouping you points by counts.

Keagan Allan
  • 3,706
  • 1
  • 15
  • 25