Real artists ship

It’s that time of the semester again when all the classes are reaching their climax, final projects and papers are rearing their ugly heads and you’re wondering where did the last three months go. I still have a good three weeks to go before finals hit, but at least for my independent study projects it’s time to wrap things up and call it a semester. I’ve been doing two independent projects: one pure computer science about parallel computation and another in art about data visualization. As you can probably tell from the title I’m going to be talking about the art project.

I decided to take up this project on something of a whim late last semester. I had been working with a particular art professor a few summers ago doing some interesting things related to computational art. That was an interesting experience but I hadn’t really thought about it for over a year. In that time I became increasingly interested in computer interfaces, design and digital art. The art I was thinking was not really painting or drawings in the generally understood sense of the term, but rather the use of visual elements (possibly interactive ones) to present data and convey a message.

I started the project with the idea that I would work on automated data visualization. There is a lot of data on the internet in the form of blog posts, websites, RSS feeds and of course the Twitter stream. I wanted to look at and come up with ways to build visual forms out of all this data that would allow people to both explore the datasets as well create visually appealing creative works.

As time went on, I realized that I really had no idea what I was doing. The data was out there, it was a bit of pain actually getting to, but with some time invested in strengthening my programming kung-fu I could build what I needed. The more pressing concern was that even with the data in hand, I didn’t really know how to go about turning it into visual form. Jonathan Harris’s wonderful We Feel Fine (and other projects) was really an inspiration for what I wanted to do, but it wasn’t quite enough. I wanted my visualizations to be informative as well as appealing. You would look at them and go “Wow, that’s cool” but you could also use the images created to come to important conclusions about the data that you had just visualized. And I had no clue how to do that.

So I started studying. Luckily for me, at the time the artist Loren Madsen was visiting Lafayette and I got to meet him, sit in on two lectures that he gave and talk to him over dinner. I was interested in his work and really liked the examples that he gave during his talk. When I talked to him over dinner I got to air my ideas and get some feedback from him and professors who were around at the time. It was also around this time that I discovered Edward Tufte. He is a person with really interesting ideas and four really beautiful self-published books. The more I read about him I realized that what I wanted to do was similar to the sort of things he did and talked about in his books. I even went to a class that he was teaching near Philadelphia and spent the better part of a sunny Tuesday learning straight from the man himself (I also got all 4 of his books). I just finished the draft of a 3500 word essay on Tufte and his ideas that will be online in the near future.

While all these ideas, concerns and questions were flying around in my head I had continual input from multiple professors about what to do, what to look at and what other people were doing or had done. A few weeks ago I buckled down, learned some JavaScript and built two simple interactive art pieces I call Typacolor and Readacolor. I’m still thinking about them, planning improvements and spinoffs (especially after talking to a professor yesterday). They don’t quite achieve the goal of allowing visual analysis of data, but they are a step on the way. I feel that my original goal is something that is going to take longer than a semester and significant improvements to my coding chops.

But as I trawled the intertubes looking for inspiration, ideas and things to talk about with my professors, I stumbled on yet another little phenomenon lurking in the wings: blogazines. Blogazines are in many ways a reaction to the template and theme driven look-and-feel of modern blogs and a return to more old school, handcrafted, standalone web pages but with the modern convenience of newer web technologies and better graphics. Heart Directed is a directory of blogazines (or art directed blogs) and looking through them I realized that visual presentation doesn’t have to be linked to actual graphics. The careful use of fonts, colors and images coupled with careful layout can create webpages that are as beautiful and capable of conveying an idea (or presenting data) as any image. I haven’t created any art-directed web pieces yet, but I’m kicking around a few ideas and hope to have at least done before the semester ends.

Coming to the end, I can say that this semester has been quite a journey. I learned a lot about art, graphics and the presentation of data. I learned that we can find beauty when we’re least expecting it. I also learned that I can’t learn new programming languages and technologies at the drop of a hat, but I can do a decent job when I devote sufficient time and energy (the interactive art pieces were built in an afternoon each). All in all, it’s been pretty worthwhile. I haven’t created as much cool stuff as I would have liked to, but I made up for that by learning a lot of things I didn’t think I would and meeting a number of very interesting people (and having lots of interesting conversations). Not too bad for my first art class.

The semester is coming to an end in less than a month, but my art expedition isn’t stopping. I’m heading to Italy for three weeks to be immersed in a completely different kind of art but at the same time. I’m going to be doing research on campus over the summer but I should have enough time to take up some interesting projects. In particular, I’m very impressed with the power of JavaScript to create art on the web (especially with Processing.js) so that is definitely something I’ll continue to do. And at some point I’ll finally get around to writing some sort of data visualizer for the web like I had originally planned to. My last independent study sparked an interest in programming languages that is probably going to take me to graduate school and I think this one is going to have a similar impact on my life in the next few years.

Lessons learned from a course with Edward Tufte

Ever since I became interested in data representation and visualization a few months, I’ve been actively trying to seek out interesting people in the fields so that I can learn directly from them. A few weeks ago I had a chance to meet and talk to Loren Madsen

who’s done some really interesting data-driven art pieces. Yesterday I went down to Philadelphia to attend a one-day course taught by Edward Tufte: an Emeritus Professor at Stanford and author of four great books on data analysis, presentation and visualization. He’s also a great observer of user interfaces and presentations and a sworn enemy of PowerPoint, Excel and the “lowest common denominator” school of design. The course is a bit pricey ($200 for students and $380 for everyone else) but you get all four of his books and you get to attend one of the best presentations you’ve attended. Here are some of the things that I learned (to learn everything, you really need to take the course).

If you’re going to a course taught by a famous person, get there early

I was under the impression that the course would be a fairly small affair, say 50 to 100 people. Blame it on going to a liberal arts school with very small class sizes.  There turned out to be more like 400 people there and I happened to arrive just as the thing was getting started. I ended up getting a seat right at the back, though I quickly made it up to the middle. Edward Tufte is a great presenter and the way he conducts the course means that it doesn’t matter very much where you sit, but it’s nice to actually see the man as he talks. Also take something to write with.

Keep your ears, eyes and mind open

The course is presented in a rather informal way. It’s not disorganized, but there’s no simple outline either. He says a lot of simple but important things, and if you try to write down everything, you’ll be writing a lot and not really taking in what’s being talked about. Write down the important things but also pay attention. Open the books and look at the examples he points at. There’s a lot you can learn by just following along. Since you’ll have all the books you can pop them open later and refresh your memory later.

Leave your preconceptions at the door

The course teaches you a lot of things about how to think about and visualize data that go contrary to what the popular opinions. You’re not going to get much out of the course unless you go in open to give other ideas a try. Also don’t worry immediately about how you are going to apply what you learned to your problem. This will probably prevent you from getting the most out of the general principles taught. Think about and absorb the principles first and then think about the specifics. If you’re someone who keeps an eye on the internet, especially in the web and interface design worlds, you’ll also find some of his advice conflicting with what you read online (such as emphasizing content over design). Do be your own judge, but make sure you’ll judging on the basis of actual merits as opposed to hearsay and group-think.

You’ll need to deal with your problems yourself

This course isn’t about giving out prepackaged solutions. Like many high-level thinkers, Tufte is more concerned about identifying the overarching principles and then applying them, rather than focusing myopically on niche issues. He will give you some solutions (especially on regarding preparing and giving presentations) with some specifics (like how to use paper and avoid Powerpoint) but they’re templates that certainly need to be filled in with information specialized to the task. I also think that it’s important that you think about your own problem and bring your sense of creativity to the issue (without which you’ll be cloning someone else’s stuff).

Read, read, read and think. A lot.

Edward Tufte is a very well read and very intelligent man. He draws on examples from people all in all sort of fields through history (from Euclid to Feynman people you’ve probably never heard of). It’s not expected that you know everything about all the things he shows you (if you did you wouldn’t be going to the course). But if you want to understand things the way he does and come up with new ideas of your own, you’re going to want to keep reading about the things he refers to. It’s also important to keep exploring other things and actively playing around with and implementing the things you learn. And that means going out there and actually giving presentations and creating graphics based on what you’ve learned.

I’m at the stage where I can understand most of the principles that he’s talked about, but I’m not sure about how to apply them to the problems I have at hand. Some of the issues he talked about are similar to ones that I’ve had myself (and some I haven’t encountered at all). I love all the great historical examples he used and I intend to read up on them more. What I need to do now is to look harder at my own problems with Tufte’s examples as a guideline. As he said, it’s generally a good idea to take a strong model and copy it for your own purposes.