I am using ArcGIS pro and this is the first time I've done something like this. I don't have a "proper" GIS background and have just been plotting point maps and choropleth maps. I've been asked to plot some points on a map, draw a buffer of a fixed distance around them, then determine the number of overlaps for each buffer, and then determine which points have the most overlaps.

I think I do something like this:

1) draw buffer of 400m around points 2) self join this buffer layer to itself - I think this is a one to many join (as I want to see all the buffers a single
buffer overlaps and I think I set this to have a match option of Boundary touches)

3) dissolve this joined layer so I can see which point buffer intersects with which point buffers and how many. each point has a unique ID. So if I dissolve by the field Target_FID (which is generated by the join) and the unique ID and sum the Join_count field, I should generate a layer where each point has a sum join count.

I'm having problems visualising it.

If I've not explained anything clearly, I'm really sorry, I've only really got a vague idea of what words to use to explain what I want. What I want to do is Draw some buffers on some points - can do this

Determine their overlaps with the other buffers on the layer - can do this, but do I use intersect, boundary touches, what set up? and I know I need one to many so that I can count ALL the overlaps

Calculate the number of overlapping polygons - this is probably the field Sum Join_count (but do I subtract one from this value as the polygon probably overlaps with itself?) can do this

Map this to determine the "catchment area" (is that the right term?) for each point, so that I can show for each point the number of overlaps and determine if there is any clustering or if there are any hotspots. Can sort of do this, I can change the symbology to graduated colours but at the moment they're complete circles, what I think I want is more intensity at the greatest point of overlap and more transparency where the overlaps are least. (if that makes sense?)

This document https://www.newham.gov.uk/Documents/Environment%20and%20planning/EB01.%20Evidence%20Base%20-%20Cumulative%20Impact%20V2.pdf

Explains exactly what I want to do: assessing the degree of spatial proximity of each point to other equivalent points by looking at the number of others within 400m (5 minute walk).

Is there a specific ArcGIS Pro tool to do this, is my workaround of self join, sum, display the right way? I'm spinning my wheels generating "stuff" but not the "thing" I need.

  • in python this would be really easy. do you use it? – Leo Jul 12 '19 at 10:03
  • Barely....I'm at the beginning of more complicated work in GIS. And I'm at the point of R vs python. I'd be happy to do it in python. – damo Jul 12 '19 at 10:09
  • you will need the shapely library for it shapely.readthedocs.io/en/stable/manual.html – Leo Jul 12 '19 at 10:10
  • I suppose this could be done using sf in R? – damo Jul 12 '19 at 10:13
  • probably, I have never used it though – Leo Jul 12 '19 at 10:20

If you want to get into Python you can use the shapely library for it. https://shapely.readthedocs.io/en/stable/manual.html

they have a buffer function https://shapely.readthedocs.io/en/stable/manual.html#object.buffer

and spatial analyses of which "intersects" will be useful for you https://shapely.readthedocs.io/en/stable/manual.html#object.intersects

Probably you want to write attributes to your spatial objects and for that the geopandas library is the way to go http://geopandas.org/

| improve this answer | |

Your Answer

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

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