One of our favorite projects these past few months has been partnering with the Markus Covert Lab at Stanford. The Lab has synthesized decades of research to create a computational “whole-cell” model that allows researchers to run simulations of a single cell. Having worked with them a bit last fall, we were thrilled to have a chance to work with their team again and build a tool to support their research.
Throughout our exploration of large document sets, we keep returning to Wikipedia. It’s intriguing for a lot of reasons—the sheer size (5.5 million English articles), the diversity of authors (anyone), and the thematic range (from completely mundane to the wildly esoteric).
Curious to see the breakdown of articles by subject, we ran various topic modeling algorithms on the archive. After trying a few approaches, we built an interactive tool that enables users to explore the high level topics and progressively dig deeper and discover subcategories of interest.
While the tool enables a thematic top-down exploration, we also wondered about what it would look like if we focused on the landscape. What does a “cartographic” view of Wikipedia look like?
We started with the topics and sub-topics we’d generated, and then selected individual articles to illustrate. (Read more on our process here). The result is a poster that depicts a visual inventory of Wikipedia—chock-full of interesting and obscure people, human creations, discoveries, and systems (a favorite: the bendy banana law).
What originally started as a few team members experimenting with shaders on Friday afternoons has now become a weekly tradition of sharing and teaching skills within the studio. While these Friday workshops range from React JS to pinhole cameras, it’s still called “Shader Friday.” Last week, it returned to its roots, with Mark breaking down the basics of rendering patterns and 3D shapes with GLSL.
We’re always looking to bring clarity to complex ideas, and this was no exception. For the last few days Mark has been tweaking and simplifying the tutorial, and it’s now posted it here. You can also check out the examples, or download the code on GitHub.
Last month, we hosted a great group from Catherine D’Ignazio’s data visualization course at Emerson College. We always love getting to share our projects in person and having guests come by to show us their latest work. Reach out if you’d like to stop in for a visit!
Happy Halloween! We’ll see you in November.
*As of 2022, we no longer sell our posters online, but office visitors may get a special goodie bag...