Excercize in exploring bike share data This uses start/stop station ride location data from Chicago divvy bikes to create pseudo best-ride routes using OSM routing software routino.
The final visualization is a heat-map of routes split into regular subscribers / one-off users.
This is a full process file; data cleanup, mapping, routing, routino linux program must to be configured along with OSM maps for city of choice.
GGmap / ggplot plotting of the routes is done by overlaying likely segments from each possible origin-destination.
Mapping builds on Code supplied by james cheshire Feb 2012, blog
Very different use case by regular divvy members vs tourists. Tourists were more likely to take shorter rides, within the loop and along the shore, while regular riders took longer home-work rides that went further into non-loop neighborhoods.
Most gpx files deleted for upload to github, in general not the most efficient way of making a likely route.