I'm trying to find the distance between matched points, or to somehow compare the results of one geocoding method with another to see how close the results are. I have tried the method using XY to line here:
using both a geographic coordinate system and a projected coordinate system. The unique ID field that I choose is never populated in the resulting attribute table, and one set of x and y fields is "<NULL>" -- which makes me think that I'm doing something wrong, but I have no idea what. Any suggestions would be appreciated.
Just adding, in case someone else runs into the same problem. I can't pinpoint exactly where I went wrong at this point, but the problem I had seemed to come from not converting between feature classes and/or consolidating my data correctly. I'm pretty new to ArcGIS...
What I ended up doing was creating two shapefiles (each with their own X and Y coordinates), joining them on the unique id, and then exporting the attribute table resulting from the join to a .dbf before using "Run XY to Line", which seemed to produce reasonable results.
I'm planning to look into the Bivariate K function too though. I have only used R slightly + quite a while ago, so it will definitely be a challenge.
R.
Hi Robert,
I would approach this problem with spatial analysis of point patterns. The first thing that comes to mind is the Bivariate K Function. This would test if the two point datasets are different or not. This link should be a starting point: http://wiki.landscapetoolbox.org/doku.php/spatial_analysis_methods:ripley_s_k_and_pair_correlation_f.... I do not believe that ArcGIS has a tool for Bivariate K Function, but you could use R and the package spatstat to do this analysis. You could also potentially program this in python, but using R I think would be simpler. You also might think about using kernel density estimation as well on both datasets.
I hope this helps, Cheers, Luke Kaim
Thank you, Luke Kaim
Thank you Luke Kaim (SIE) Lucas.Kaim@maine.edu (914)263-7866 "Don�??t complain. Just work harder" (Randy Pausch).
