This small side project was developed after taking a crash course in data visualisation with Marius Watz at the Copenhagen Institute of Interaction Design. In this page I collected two of my favourite visualisations which came out of those intense but rewarding 7 days.
What fascinated me about Marius Watz's art is the possibility of painting with code. His ability to generate stunning visuals based on seemingly random rules and numbers is just mind blowing.
After being introduced to his work I started thinking about my own experiments. I decided I was not interested in telling a "story" with my data, trying to uncover some magical unseen detail of my digital self.
Instead I wanted to see how the data I gathered about me could be used as my own digital brush. The process I followed was rather simple.
I decided to use Facebook as a data source for my work. Facebook API gave me access to large sets of data that could be as flexible as I needed them to be.
The second step was to take the data apart. More than looking for emergent patterns I started looking how this data sets might interact together in unexpected ways.
Finally I started creating rules based on those interaction I thought were interesting. This is where the magic and fun happened :)
For these experiments I used Processing. I was already familiar with the tool, and although you could pick your weapon of choice for this, Marius Wats is a Processing master I wanted to learn from. Plus, Processing has an amazing community of users creating libraries for almost anything, which meant I could focus on experimenting and not bang my head trying to solve technical challenges.
The image on the right is based on two identical sets of data from my girlfriend's and my Facebook accounts.
The thicker green and yellow lines in the middle are the data points representing the number of posts made over the previous 6 months. Each of the post is then connected to the red arches representing the group of friends that interacted most with us in that same period.
This initial visualisation still shows a bit of a story. The first thing that jumps at you is the significant difference of activity between the two sets. Granted, it was not designed to be readable, which is why no key is presented to the reader. But as far as abstraction, I started to take a less literal 1:1 interpretation of my data, and instead started introducing more abstract procedural rules.
The second rounds of experiments was taking the abstraction one step further. The new set of rules were introducing random ranges of values based on the interaction between friends and posts.
So, for example, the more activity a post would get, the thicker the lines would become. Or if the post happened at certain times of the day, then the slant of the line would increase or decrease accordingly, and so forth.
All this seemingly random rules where used as a procedural way to create unexpected visual patterns, regardless of the relationship between data points (much like with Sol LeWiit art, minus the talent).
The end of the summer school course was marked by an final show case of our collective work.
For my visuals I decided to make laser cut wood prints, so I turned the graphic into grayscale images and fed those to the laser cutter.
The way the cutter works is it converts the grayscale into a depth map, so each shade of gray gets a specific engraving depth.
As for material I experimented a little with wood finishes to find the wood that would work best for each of the prints. For the first visual I chose an inexpensive wood made of wood pulp pressed together. This wood has the advantage of not turning brown once engraved and it gives out a silky finish to the touch.
For the second print I chose a sandwich structured wood. The light color of top sandwiches gives way to a rich deep brown once it is engraved, making the shapes really pop.