Question: what are you studying to do with Boris bikes?

Martin Zaltz Austwick answered on 12 Mar 2012:

The first thing I did was to animate them. We had a big data set that shows the start and end times, and start and end locations for every bike journey in a six month period. One of the people I work with used a computer program to make a decent guess at their routes. Then I animated all of these bikes travelling over one day, just to show how interesting and cool this data is!

The analysis is a bit more complicated and I’ve been using something called Network Theory. Network theory does things like tell me where the most important locations, which the important routes are, and whether there any clusters – in this context, parts of the city that tend to keep themselves to themselves.

The new research comes from the fact that this network of journeys varies by time of day, and different patterns appear over weekdays and weekends. Also you can’t ignore that short distances are more likely because no-one wants to cycle further than they have to! If you’re not careful your analysis will just tell you people cycle to things nearby, which is very obvious. So I’m trying to take account of that to spot when people go further than you would expect, and try to work out why. Does that help? 🙂