Thursday, November 17, 2016

Data Visualization in R

Thought I'd update some R visualization I've recently completed. Although I prefer Python, I have to admit, R is an incredibly powerful language for translating data into graphs and visualization and I might just be warming to it, especially for visualization.

The data is the journey to work census conducted by the NSW Bureau of Transportation. The data is a snapshot of daily commuter behavior and which suburbs people travel to and from in a single weekday in Sydney. The first image is a heatmap of the origin and destination suburbs that workers travel to. It does a nice job of clustering together commercial hubs in the Sydney area and showing the suburbs that people travel from to get there.



The next is a network connection map that is showing the number of people travelling between each destination, with the transparency of the line indicating more people. While not really quantifying much information, it shows hundreds of connections in a single image. The human eye can process a whole lot of information, and this drops all that on you with a glance, which I think is pretty amazing.