Replicated Territories
![Replicated Territories](/content/images/size/w2000/2020/01/image-18-1.png)
The other day I saw this great bit of plotter work on Twitter, and loved the look, so I wanted to see if I could replicate something similar digitally.
Territories #axidraw #processing #plottertwitter #generativedesign #creativecoding #generativeart pic.twitter.com/3CyCeuY7bA
— Julien Gachadoat (@v3ga) January 6, 2020
First I made a quick streamgraph in Rawgraphs. This is the data from Beautiful News' "Foreign Aid Has Exploded" graphic. It originally comes from OECD. The colours are different categories of aid.
![](https://blog.duncangeere.com/content/images/2020/01/image-11.png)
I exported it as an SVG and imported it into Figma, my favourite vector graphics editor. Then I made a big canvas with a bunch of lines. I spaced them by hand.
![](https://blog.duncangeere.com/content/images/2020/01/image-12.png)
Then I just used the shapes as a mask for different sets of lines, rotating the lines between different examples. Here's the first output. It's okay, but the lines are too sparse.
![](https://blog.duncangeere.com/content/images/2020/01/image-13.png)
To fix this, we need a bit more line density. Let's make a a new set.
![](https://blog.duncangeere.com/content/images/2020/01/image-14.png)
![](https://blog.duncangeere.com/content/images/2020/01/image-18.png)
That's looking nicer. It's pretty unreadable as dataviz, so I took off the axis. But it's getting towards a nice piece of data art.
Improving it further would mean increasing the density of lines again, I suspect. There's definitely a code way I could do that which wouldn't mean aligning them by hand. Getting closer to the original would also mean finding a dataset (or faking one) which has more categories. Perhaps I'll look into those things another day.