Notes for April 7

Today we'll work with graph data once again. However, this time we will be placing emphasis on the edges between nodes rather than the structural relationship between nodes. For example, while it might be useful to spot character clusters in Les Misérables, you need a different visual metaphor if you want to show more clearly the intensity of interactions *between* characters. Today we'll try out a chord diagram to expose the edges more saliently. While we won't get to in in class, there are also a wide variety of hierarchical visualizations available when you want to show nesting and structural details in a network.

Today's dataset comes from a larger dataset of character co-occurrences in the Game of Thrones book series. I aggregated data into a final, sparse dataset of character relationships that show high co-occurrence (>50 interactions across 5 books). If you're a fan of the books and it makes you feel better, I promise I'll update the dataset for Winds of Winter when it comes out.

Here is an Observable example of a chord diagram. This one attempts to add some interactivity.

The d3 chord library has lots of helpful functions. We'll also be making use of the arc generator from d3-shape.

To make a color palette for 20 different categories, I used the Colorgorical color optimization tool. Viz Palette is a decent way to test out different color scheme ideas.

HTML for today:
Code for today: