I'm joining 20 data-minded journos at a techie workshop next week, led by the good people at UC Berkeley's Knight Digital Media Center. It's about learning tools and comparing ideas for mining next year's Census data dump at our news sites--we're talking GIS mapping, data cleanup, and schmancy visualization tools.
For instance, it would be an interesting exercise to approximate one of my favorite visualizations, NYT's Flash-based Immigration Explorer, using the Protovis choropleth map method. Anyone have experience and tips for handling large data sets in Protovis?
Inspired by next week's training and this interesting visualization of employment data from Minnesota (it's really fun to see the seasonal variance in industries like "Amusement Parks and Arcades" and "School Bus Transportation"), my first Protovis attempt is below. Not good times in Santa Clara in the early 2000s . . .
Was easy enough to get started, though I think some smarter .csv organization would have made the data input smoother--the Protovis tutorials I've come across don't really explain how to best store your data set, and I'm very interested in hearing other people's tips!
Update: figured out how to load a csv into an external JSON file using very handy Mr. Data Converter app, and then pull the data objects using