Lately, I and others have been looking at Foursquare/Gowalla & the like to analyze geolocation data and wrap our heads around what its all about (hence the explosion of Foursquare data in the blog lately). More robust data is being developed, but while this doesn't promise to do any of that, I thought it would be worthwhile to try my hand at another infographic. With this in mind, I thought that the network's potential in generating transport insight might be worthwhile and without providing anything groundbreaking, the takeaways from comparing how people register their travel were pretty interesting.
As mentioned in the graphic, people seem to be registering locations while on the go, as tube stations aren't nearly as popular as train & tube/train stations. I know that some locations are naturally busier, Charing Cross vs. Brixton for example), but the under performance of some tube stations in popular locations (Leicester Square), means that not nearly as many people are checking in after the fact as would be expected.
Aside from this, the variety of places people checked in is quite amazing. From buses (shown in the graphic) to planes & a few boats (not shown), people are dynamically logging information when possible.
Perhaps the most striking fact from the data however, was the lack of dual or superfluous locations. Being that most seek to be mayor of somewhere, you would assume to see more distinctions of tube vs. train stations (for instance, Charing Cross is only listed as one entity). The multitude of people who check in from platforms within a train station however (not shown in final version of graphic), may counteract some of this.
Overall, I think as the service grows, we should see more people fleshing out London's transport network in line with what we would expect. From this sampling of transportation options, the network is already beginning to be represented well.
[Development Notes: I selected a subset of tube stations to represent due to the fact that I was drawing the map underneath it. TFL is rather touchy about the rights to their map, so I wanted to make sure that mine was divergent enough to not be derivative (hence noticeable differences I've made sure are present).
If this goes over well, I'm considering expanding it to a longer version (including everything that I noted was missing), so if you have any thoughts or ideas, just let me know.
.jpg Version (Click to Enlarge):
Very nice. How did you gather the required data?
ReplyDeleteI wrote a .Net aggregation program that queries the API daily and stores it for analysis....
ReplyDeletewhat was the period of data collection?
ReplyDeleteThose were total check-ins as of 13.4.2010.
ReplyDelete