In June 2010 I attended Flashbelt in Minneapolis, Minnesota. This annual conference focuses “around Adobe Flash and related technologies." Two of the speakers that I found most informative, Jeremy Thorpe and Wes Grubbs, seemed to focus not so much on Flash, but on their use of a “related technology” called Processing.
Just a few weeks later I am finding Processing useful to help me understand the large amounts of data to which I have access: membership data, underwriting data, major giving data, pledge data, web analytics data, etc. Sure, I could create graphs with a spreadsheet or other simple interfaces. However, I would like to see data to help gain a different perspective. We have so much data that sometimes it would be nice to pick out certain pieces and look more closely.
I have examined a few other data visualization tools such as Gephi, Protovis, and Parallel Sets. While these tools look quite promising, Processing seems to be backed by a much larger community, and the literature about Processing seems to be much more vast (which suits me).
One of my first forays into Processing started with a question about comparing data from two different sources, our website and our membership database. This certainly did not mark the first time that I wanted to compare data from disparate sources. I wanted to know how a particular web initiative affected the number of pledges at a certain level -- to help my organization make informed choices. These days I’m reading more and more blogs -- such as datavisualization.ch, eagereyes.org, infosthetics.com -- and books about visualizing data, and I am well aware that there may be holes in my methodology.
That said, in this case I used Processing to display two types of data over a ten-year period (in this example I am using fake data). Although I’ve done a fair amount of programming, it still took me quite a while -- even following along with Ben Fry's excellent book, Visualizing Data: Exploring and Explaining Data with the Processing Environment -- to get this first project working how I wanted it. I am hoping that time invested now will prove useful for my next question to be answered with Processing. In other words, I didn't do this project just to answer the question, I also did this project to help learn Processing.
Part of what I like about using Processing is that I can control every element: fonts, colors, width, height, labels, etc. Processing can create Windows apps, Mac apps, Linux apps, and interactive web apps. The app I shared with my coworkers is interactive and they can use their mouse to hover over any data point. The image above shows my (scrambled) results.
I realize this may not be mind blowing – I made a graph with two lines. Yippie. In the coming days, I plan to show other data that I have “visualized” with Processing. It probably won't be nearly as cool as what Jer Thorpe did with Processing for his recent visualization for Wired UK or any of the other exhibitions on Processing.org. Instead, I hope to show a few uses for Processing that might be relevant for people working in public radio or fundraising.