I'm interested in looking at where hot and cold spots of certain social indicators (such as poverty, housing characteristics, urban sprawl, etc...) are in coastal communities along the East Coast. I have point and polygon features containing the same attribute information. My point data are centroids of the polygons. My first thought was to use the Hot Spot Analysis tool in arcgis but, because I'm looking at coastal communities, my polygons don't have many neighbors (some don't have any) so I don't think my results would be very accurate. So I thought there must be a hot spot analysis that I can do using my centroid points that have the same attribute information but the only analyses I can find have to do with spatial clustering of points- not of their attributes. Does anyone know of an analysis I can do to solve my problem?
-
Right after reading your question the term "spatial aggregation" came to my mind. Before running to the nearest patent bureau I searched for this term and found that it's "in labs" today. :)– Remigijus PankevičiusCommented May 5, 2015 at 20:11
Add a comment
|
1 Answer
If your data density is so low that it is non-contiguous, the obvious alternative, which is frequency hotspot analysis using raster surfaces, won't be so good either. Also raster surfaces are not good near coastlines due to weighting issues (some cells have large chunks of sea, then you have to fiddle with weighted corrections - which requires dense paragraphs of mumbo jumbo in the report and is not so convincing to an audience).
If the data density is low enough that the resulting map won't be crowded, why not go for a bubble map? They are straightforward and easy to interpret, and you can make bubbles of your individual attributes or else indexed distributions (bubbles that represent a combination of factors).
