Jeff at Vector One asks, “Why is there so little geospatial analysis?” I can think of any number of possible reasons:
- People aren’t aware of it, or what it can do; for some people, geographic analysis ends when you put a point on a map
- It can involve advanced analytical and statistical techniques that are challenging to learn
- Commercial software can be expensive, and difficult to learn; ArcGIS’s Spatial Analyst and Geostatistical Analyst extensions list at $2500 apiece.
While there are many freeware programs available for geospatial data analysis (I’ll post on some of them on the future), there’s one program I’d select as the best choice to address the above issues: GeoDA, spearheaded by Luc Anselin at Arizona State University. And that’s as much for its teaching materials and documentation as it is for its capabilities. There’s a 100+-page user’s guide, a 200+-page workbook that’s a mini-course on analyzing spatial data, multiple publications, dozens of sample datasets, even a QuickTime movie. It’s by no means complete; for example, there’s no geostatistical interpolation capabilities in this program (e.g. variograms, Kriging). But just by itself, it’s a great introduction to many techniques for analyzing spatial data correlation. It’s only for Windows now, but they’re working on a cross-platform open source port for Windows, Linux and Mac; no word on when that might become available.