Quantcast
Channel: Forums - Geoprocessing
Viewing all articles
Browse latest Browse all 1952

Semi-random point distribution with customized point count per polygon

$
0
0
I'm trying to disperse the clustering that is caused by zip code-based geocoding. Our statewide geocoding process results in an average of 5 percent of the geocodes ending up at the zip code level. I want these geocodes to be scattered throughout the zip code area as opposed to clustered to the zip code centroid. I do not want the points to be distributed entirely random however, but instead distributed semi-randomly within each zip code using the area's population density as a distribution weight.
The website spatialecology.com provides two tools that are very close to what I am looking for. Unfortunately, both tools appear to be short one feature that the other tool provides. The genrandompnts (Generate Random Points) tool allows the distribution to be weighted through the use of its 'probsurf' option, which allows a raster layer (in this instance, a raster containing the population distribution of the state) to be used as a probability density surface. However, applying this option does not allow the use of a polygon (in this instance, the zip code tabulation areas feature class for the state) to be used as strata for stratified random point distribution, which is something that the tool genstratrandompnts (Generate Stratified Random Points) does allow. However, the latter tool does not allow the use of a raster layer as a probability density surface.
Does anyone know of any tools that might provide a combination of the features of both aforementioned tools? Or an effective method for applying both tools to get the desired end result? The best that I can think of is creating population distribution surfaces for each zip code and applying the genrandompnts tool to every zip code individually on a piece meal basis. Any suggestions would be much appreciated.
Thanks!

Viewing all articles
Browse latest Browse all 1952

Trending Articles