morgan herlocker

about twitter github posts

Geocolor Improvements

charleston

I have made some signifagant improvements to geocolor over the last week. The api is more flexible now, and I have added support for random coloring as well as custom style options.

var geocolor = require('geocolor')

var cities = {
  // geojson goes here
}

var colors = ['green', 'red']

var style = {
  'stroke': 'white',
  'fill-opacity': .4
}

geo = geocolor.random(cities, colors, style)

I have noticed that geojson performance in the browser (specifically github preview) takes a serious hit after ~500KB. Surprisingly, however, I have been able to share usable geojson all the way past 1MB. With proper simplification, this can represent a good bit of data. Below is a shot of west coast rivers colored by volume.

rivers

Although geocolor is basically "feature complete" now, my next step is to explore more efficient rendering tools. Mike Bostock's bl.ocks.org seems like a pretty good solution, and combined with a cli scaffolding tool, it could be a good way to pass geo visualizations around with larger amounts of data.

sc-green-to-red [South Carolina Counties colored from green to red by population]


3-6-14