Divvy Data Analysis and Visualizations

Here is some analysis and some visualizations I've been working on for the Divvy Data Challenge - mapping data from Chicago's bike sharing program.

  1. Routes Analysis
  2. Directions & Times
  3. Nearest Station

1. Routes Analysis

I used the Google Bicycling Directions API to estimate the detailed route for the 17,000+ most common Divvy routes (consisting of nearly 700,000 trips).

This information can be used to estimate the density of Divvy riders at all points in the city:

(In-progress: Overlaying these on top of an interactive map.)

From this, we can also view how other information varies across the city. For example, for every point on the map, is the Divvy rider more likely a Member or a 24-hour-pass user? The first image below shows this (red=member, blue=pass), (The next image shows them as probabilities).

We can see that the lakefront is dominated by guests and inland is primarily Divvy members. However, there are some notable exceptions, such as Clark and Lincoln. Clark street is blue up until Waveland (Wrigley Field) when it turns red going north from there. Lincoln stays blue until Southport and Wellington.

Color-coded by average age of rider,

I plan on overlaying this data (and more) on a map, but in the meantime, I have made a map allowing you to visualize the top 30 most popular routes. Click thumbnail below,

2. Directions & Times

This is an attempt to display the average direction that bicycles are traveling as they leave stations and how this changes throughout the day.

The color of each region displays the average direction that riders are traveling from that station (at the current time). The intensity of the color encodes the amount of activity (such that white represents no Divvy bikes leaving). Hover with the mouse and hit "Play" to how things change for different hours of the day.

Instead of stations as single points, stations are represented by regions containing all locations for which that station is the closest station (a so-called Voronoi Tesselation).

3. Nearest Station

Move your mouse anywhere in the city to quickly visualize the nearest Divvy Station.

As in the previous map, this uses a Voronoi tesselation to visualize the regions represented by each station.




Acknowledgments

These make use of the following,





Created by Ken Schutte, 2014.