To me, when I write it I’m thinking: okay, starting with df1, and I’m augmenting with df2.

Starting with the data and layering operations is more intuitive to me.

Thanks! Yeah you can find a lot of code in this repo: github.com/Pecners/rays...

My inspiration for this map was a combination of ice and stars. Ice because, Alaska is way north, but also the reflective quality highlights the sparse population.

Stars because the sparse population centers look a bit like stars in the sky, and as a reference to Alaska's flag.

How does that 14% compare to the rest of the US? Alaska has more National Park land than the rest of the US, combined.

Among that surfeit of unpopulated land is a large amount of National Park land. About 14% of land in Alaska is managed by the National Park Service.

While there isn't much population, there's a lot of land. In fact, Alaska's land area is about equal to that of these 12 states.

There are just a handful of identifiable population centers. Top 3 most populous cities are visible: Anchorage, Fairbanks, and Juneau.