ghostnetsaustralia

spatial auto-correlation & hotspot analysis

Discussion created by ghostnetsaustralia on Oct 14, 2012
Latest reply on Nov 7, 2012 by tmackey-esriaustralia-com-au-esridist
Hi,
I've been working with a point data set of roughly 11000 points. I want to demonstrate on a heat map where the clusters are.
I watched the video's and took the following steps
1) I integrated the points with an xy tolerance of 50m - some of my points overlap, some are widely dispersed I ran the nearest neighbour analysis which showed them to be clustered
2) then I used 'collect events' to get a group number for clusters
3) then I ran an incremental spatial auto correlation to get the threshold distance (distance band).
4) Using that I ran the hot spot analysis ....but the results are not what I was expecting which leads me to believe that my steps though correct are missing something.

Could someone please shed some light for me on where I may be going wrong, and why I cannot get the analysis to reflect the hot spot areas i know exist?

I would like to cross reference this by using the fishnet instead of the collect events tool. I created several fishnets but cannot work out how to run the spatial join to give me the grouping column I need to use this data for the hot spot analysis.

Any and all assistance much appreciated.

Outcomes