oregon-choropleth

Resources

County-level data for Oregon:

• https://www.census.gov/cgi-bin/geo/shapefiles/index.php?year=2010&layergroup=Counties+%28and+equivalent%29

Preview shapefiles:

• http://mapshaper.org/

Geographic projections:

• https://github.com/veltman/d3-stateplane
• http://spatialreference.org/

Procedure

Fetch Oregon county shapes

mkdir formatted raw
mv ~/Downloads/tl_2010_41_county10.zip ./raw/
unzip -o raw/tl_2010_41_county10.zip

Fetch population data by county

curl 'https://api.census.gov/data/2015/acs5?get=B01003_001E&for=county:*&in=state:41' -o raw/or-county-populations.json

Convert county data JSON array to an NDJSON stream

• ndjson-cat: remove newlines
• ndjson-split: split into multiple lines
• ndjson-map: reformat each line as an object

ndjson-cat raw/or-county-populations.json \
  | ndjson-split 'd.slice(1)' \
  | ndjson-map '{id: d[2], population: d[0]}' \
  > formatted/or-county-populations.ndjson

Convert shapefile to GeoJSON and apply a geographic projection

shp2json raw/tl_2010_41_county10.shp \
    | geoproject 'd3.geoConicConformal().parallels([44 + 20 / 60, 46]).rotate([120 + 30 / 60, 0]).fitSize([960, 960], d)' \
    > formatted/or-projected.json

Preview in a browser

geo2svg -w 960 -h 960 < formatted/or-projected.json > formatted/or-projected.svg

Convert the GeoJSON feature collection to a newline-delimited stream of GeoJSON features, then set each feature’s id using ndjson-map:

ndjson-split 'd.features' \
  < formatted/or-projected.json \
  | ndjson-map 'd.id = d.properties.COUNTYFP10, d' \
  > formatted/or-projected-id.ndjson

Join the population data to the geometry using ndjson-join, and remove the props we don't need

ndjson-join 'd.id' formatted/or-county-populations.ndjson formatted/or-projected-id.ndjson \
  | ndjson-map 'd[1].properties = {population: d[0].population}, d[1]' \
  > formatted/or-joined.ndjson

Now, perform some topology-preserving simplification, topoquantize, and delta-encode

geo2topo -n \
  counties=formatted/or-joined.ndjson \
  | topoquantize 1e5 \
  > formatted/or-topo.json

We need a non-linear transform to distribute the colors equitably, so we'll use a threshold scale. Use topo2geo to extract the simplified counties from the topology, pipe to ndjson-map to assign the fill property for each county, pipe to ndjson-split to break the collection into features, and lastly pipe to geo2svg

topo2geo counties=- \
  < formatted/or-topo.json \
  | ndjson-map -r d3 -r d3=d3-scale-chromatic 'z = d3.scaleThreshold().domain([1000, 10000, 25000, 50000, 100000, 300000, 600000]).range(d3.schemeOrRd[9]), d.features.forEach(f => f.properties.fill = z(f.properties.population)), d' \
  | ndjson-split 'd.features' \
  | geo2svg -n --stroke none -p 1 -w 960 -h 960 \
  > formatted/or-counties-threshold.svg

Result