Last fall I had the pleasure of visiting Chicago to run my first marathon. While there, I also enjoyed a ride along the waterfront on a Divvy bike-share bike. A few months ago I was excited to learn that a year of trip data had been released as open data for a data visualization contest. Although I did not enter the contest (results are here), I still could not resist the opportunity to transform the .CSV files in to origin-destination maps. With Vancouver poised to launch its own bike-share system in the next year, I hope to learn some transferable lessons that may assist the implementation of our own bike share system.
A friend told me most Divvy trips were quite short, so I explored the data and was surprised to learn the mean trip length is about 20 minutes. Therefore I wanted to look at the travel patterns by trip length, so I decided to produce the map below that includes over 750,000 trips - about 250,000 trips on each frame.
I also added a new graphic below that shows the total minutes of physical activity generated and the average trip length per neighbourhood. The basic story is that people who live on the outskirts of town ride longer trips, but there is a lower cummulative time spent travelling as there are fewer trips. I am very interested in looking at the cummulative health benefit of all these bike trips...I am working with a friend who works with the City of Chicago Public Health Department and she will be writing a guest blog post soon.
Let me know if you would like me to look at any specific questions using this goldmine of big data?