Hi, I hope someone can share some ideas for an approach.
I have a bunch of data for driving from point a to b and b to a (home & work). It is GPS data, so it is lat/lon & speed. I am trying to see what the data says about me, i.e. optimized time, least time waiting, total milage, etc. What I don't have is any indication of what paths I took, so I have no idea how to average several trips / path. From my own thoughts, there has to be many distinct paths; but finding them seems to be a very hard problem (no silver bullet), so I know I'll be doing a lot of hard work. But, here is current thoughts for an approach: -Find an area for each type of path that the other paths don't go near (perhaps an intersection, town, road, etc). This would allow me to mark/name that type of path, then move on to the rest. As I am mixing path segments, this doesn't seem to work well. -Given the first path x, find each point in y, to the line in x. Find the total distance away per points. Determine some cutoff point to determine if x is really y. If not y becomes its own path (which will then be compared to the other path). Has anyone done something like this that they can share their approach, or can think of an easier/better way? Thanks all. -- View this message in context: http://www.nabble.com/Finding-Categorizing-Paths-from-A-to-B-tp17296462p17296462.html Sent from the PostGIS - User mailing list archive at Nabble.com. _______________________________________________ postgis-users mailing list [email protected] http://postgis.refractions.net/mailman/listinfo/postgis-users
