Hi, I'm Louis!
The Inspiration
A few months ago, I chanced upon the official Star Wars Galaxy Map (Thanks to my buddy Kevin Wee) that was created to complement the 2009 book Star Wars: The Essential Atlas. Being a huge fan of the franchise, this was very exciting to see how much depth and lore the story was built upon. I wanted to share my excitement, but over 6,000 planets is a bit much to explain to someone new isn't it? I think I say this for every dataviz nerd, but we get so much satisfaction from being able to simplify data for non-data folks.
The Idea
I landed on an idea to visualize the planetary travel for each character, but obviously that would be a crazy endeavor. So I wanted to test this out with AI. I started off asking ChatGPT to do deep research on the topic, feeding it context like the official Star Wars databank and Wookieepedia. The initial D3 visualizations it came out with was pretty lackluster, but it had the right idea of showing travel through animated lines. I didn't get too far with ChatGPT because it kept breaking the animations as I tried to add more design to it, so I migrated to Claude.
A coding partner
To be completely honest, I'm still a novice in D3 but I understand CSS and code structures. Having Claude as my dedicated coding sidekick as I prompted through requests from like adding particles, fog system, cinematic camera control to page snap scrolling... was liberating. However, I relied heavily on my design and UXUI knowledge, often questioning and correcting course when Claude tries to do something weird. It's definitely not 100% accurate nor my best work, but it's a close replication of what I had in mind. And the best part? I didn't have to scrap any data either. But it did take burning through maybe ~$50 worth of Claude tokens due to the verbose nature of the franchise 😅