I think I have posted these maps here before, but since someone asked...
Here are a couple of maps I made by plotting about 6 months of raw locations as reported by the GPS in my Nextel cellphone. At a major highway scale this starts to look pretty good: http://gadgeteer.org/maps/AllenRegion.gif But with no grouping of the data points the map breaks down at a street level: http://gadgeteer.org/maps/Allen03Detail.gif The two biggest clusters are around my house and the neighborhood pub. You can also start to see smaller clusters around intersections that are controlled by stop lights. The raw data is in the format Lat, Long, speed(km/hr), heading, date, time N 38.809216,W 77.022565,98,5,5/09/2003,17:43:00 N 38.838075,W 77.009200,105,30,5/09/2003,17:45:10 N 38.864971,W 76.992640,102,46,5/09/2003,17:47:21 N 38.885733,W 76.962661,100,17,5/09/2003,17:49:33 Anyone is free to grab the whole file from here: http://gadgeteer.org/maps/Allen1.txt http://gadgeteer.org/maps/Allen2.txt http://gadgeteer.org/maps/Allen3.txt http://gadgeteer.org/maps/Allen4.txt On 11/30/06, joshua <[EMAIL PROTECTED]> wrote:
This feels like a clustering problem, to me. Is there any data that you have to share? Perhaps a netflix-style competition might be in order. Couldn't one analyze this on a per-region segmented datasets? (grid partitions or whatever) You can also probably identify time-of-day and day-of-week traffic flow rates, and so on. This project seems like an enormous amount of fun.
Allen Smith [EMAIL PROTECTED] Track my cell phone at http://www.gadgeteer.org
_______________________________________________ Geowanking mailing list [email protected] http://lists.burri.to/mailman/listinfo/geowanking
