0

I have a huge number of polygons that contain attribute field that represent number of offices inside each polygon. Now I want to merge all adjacent polygons which count (sum) of offices values is between 1000 and 2000. I want to do that using ArcGIS. I have idea to use Dissolve tool, Integrate tool and Merge tool, but I am not sure about steps in order to solve this.

6
  • 1
    Possible duplicate of Graph/Network building and analysis of linked polygons in ArcGIS for Desktop?
    – FelixIP
    Commented Dec 10, 2017 at 17:42
  • @FelixIP very similar, but not same.
    – user104851
    Commented Dec 10, 2017 at 22:53
  • 100 percent identical. If you dissolve result shown by rcvnode . Store you Shape s somewhere in Dropbox, I solve it in 10 minutes.
    – FelixIP
    Commented Dec 11, 2017 at 0:08
  • @FelixIP This is the link: dropbox.com/s/0c62n0cy5nebwl6/TEXAS.zip?dl=0 If you can describe step by step what have you done in my example that would be great. JOINT_COUNT field is a field depending on which count I want to merge the polygons. Thx
    – user104851
    Commented Dec 11, 2017 at 15:29
  • @FelixIP This is great. But could you please send me a script that you have used in my example (with my parameters). Sorry, I am new in coding.
    – user104851
    Commented Dec 12, 2017 at 0:00

1 Answer 1

3

I took your polygons and converted them to single parts. For each part I recomputed number of offices pro rata of parts count in parent polygon. I applied procedure and script from the post mentioned in my comments with tolerance 0.05 and 316 groups, i.e. I tried to create 316 groups from 3059 original polygons with each group containing 1500 offices (total of offices=475169 ):

enter image description here

This is statistics of dissolved:

enter image description here

And these are results. Note I removed islands because they have no neighbours in your definition.

For more information on algorithm, see this post.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.