Processing Data Visualizations

I’ve seen a lot of interesting data visualizations lately but have struggled to figure out how to visualize my own data.  It seems like there is a vast chasm between creating pie charts in Excel and Hans Rosling’s TED Talks.  The I stumbled upon Processing.

Processing was used to create the genetics simulation I described in an earlier post.  After looking into it some more, I learned that Processing was developed out of a project at MIT’s Media Lab.  It is an object-oriented programming language conceived as a way to sketch out images, animations and interactions with the user.

Examples of of Processing projects include everything from a New York Times data visualization of how articles move through the internet and visually representing data in an annual report to more esoteric and artistic works.

To get started, download the application at http://processing.org and go through some of the tutorials on the site.  There are lots of examples included with the download so you can also open them up and start tweaking and hacking them, if that’s your preferred method of learning.  Once your code is complete, or after you’ve made a minor tweak, click on the play button to open a new window and see it looks.  Once you’ve completed your project, you can export it as an applet, which can be uploaded to a web server, or as an executable file for a Mac, Windows, or Linux computer.

I’ve been through the first half-dozen tutorials and am to the point of making lines and circles dance around.  I can even make the colors and sizes vary based on mouse position.  I have also opened up some of the more advanced examples and started picking away at them to see what I can understand and what I still need to learn more about.  Once I can import data from an external source, it will be really exciting to see the different ways to represent it.

I haven’t had a foreign language learning experience in a while.  I am learning (and re-learning) many valuable lessons as I try to express myself in this new language.  Not surprisingly, I’m finding that I need a balance between instruction (going through the tutorials) and practice / play (experimenting with the code I’m writing or hacking together).  I’m also a bit frustrated by my progress because I can see what can be done by fluent speakers (see examples, above) but am stuck making short, choppy utterances (see my circles and lines, which really aren’t worth sharing.)  I plan to both work my way through the basics (L+1) as well as dabble with some more advanced projects (L+10) to see if I can pull them off.  If not, I’ll know what to learn next.

Fortunately, I have one or two friends who are also learning Processing at the same time.  They are more advanced than me (in programming languages, but I hold the advantage in human languages), but it has been helpful and fun to bounce examples and ideas off of one another.  We plan to begin a wiki to document our progress and questions as they arise — a little like a students notebook where vocabulary and idioms are jotted down so they can be reviewed later.

Watch for more updates as projects get pulled together as well as notes on other ways to visualized data in the near future.

Visualizing Presidential Speeches with Tag Clouds

I’ve written about word clouds before, but many more options for visualizing information have become available since my first post about Wordle.  Some newer applications include Tagxedo, Tagul, WordSift, TagCrowd, each of which has slightly different features and ways to customize the look of your tag cloud.

Why are these applications so popular?  A lot can be gleaned from looking at a text in this format.  There are many more complex ways to analyze a text (one of my favorites is Xiaofei Lu’s Synlex, which can analyze a wide range of features from the frequency of structures to the complexity of the text) but word clouds are simple and straightforward.

A great example is the US Presidential Speeches Tag Cloud page by Chirag Mehta.  It’s based on Tagline, the Timeline-based Tag Cloud Generator that he developed.  This is actually a series of tag clouds with a slider bar that allows the viewer to scroll though 200 years of presidential speeches.  It’s interesting to see how the most frequent words and themes change over time, something that is very easy to see as you scroll through the tag clouds.

How can students use these tools?  The more complex tools, which allow students to target specific features, might be the best option for analyzing one’s own writing.  On the other hand, tag clouds would be a better option if a student just needs a snapshot of a text.  For example, by feeding a reading assignment into a tag cloud generator, it would be very easy to pick out the most frequent terms and themes prior to reading it — a little like having the text skimmed for you.

Do you use tag clouds and other text analyzers with your students?  Leave a comment to share your tips and ideas.