I spent a good portion of my weekend creating version one of my visualization.

The concept is to take the current weeks’ Top 40 Radio chart and mash it up with YouTube to see if the music that the recording industry is pushing on the airwaves is in sync with what people consume when they can self-select their music in video format. Thumbnails from the videos are displayed in a collage and their image sizes correspond to the total video views they have received. This means that a video with 100,000,000 views will have 50x the surface area of a video that has only 2,000,000 views, where a video with only a few thousand views will appear to be only a few pixels large.

I’m doing all the scraping (BeautifulSoup), data crunching and processing (PIL) in python and sending the result to HTML via a templating (Jinja2) engine. Right now everything works locally. I’m going to work on the interaction and aesthetic and hopefully have a version that works on web rather than just my machine, but that may not be feasible (each request requires ~40 API calls and a good deal of image processing).