This is a course assignment of UCSC CMPS263 - Data visuilization. In this assignment, I created a visualization that is similar to the CA Population Density example created by Mike Bostock with some additions:
-
Clickable button shows different color binding and legend.
-
Clickable button toggles state-boundary and census tract boundaries, where county boundaries will always be shown.
-
Download shapefile
From Census Bureau website New York State(FIPS code 36) in ACS 2016 5-year estimate.
curl 'http://www2.census.gov/geo/tiger/GENZ2016/shp/cb_2016_36_tract_500k.zip' -o cb_2016_36_tract_500k.zip unzip -o cb_2015_36_tract_500k.zip -
Convert .shp to GeoJSON
npm install -g shapefile shp2json cb_2015_36_tract_500k.zip -o NY.json -
Apply geographic projection
npm install -g d3-geo-projection geoproject 'd3.geoConicEqualArea().parallels([40.5, 41.5]).rotate([74, 0]).fitSize([960, 960], d)' < NY.json > NY-albers.json -
Enter newline-delimited JSON
npm install -g ndjson-cli ndjson-split 'd.features' < NY-albers.json > NY-albers.ndjson ndjson-map 'd.id = d.properties.GEOID.slice(2), d' < NY-albers.ndjson > NY-albers-id.ndjson -
Download population file and convert to ndjson
curl 'http://api.census.gov/data/2015/acs5?get=B01003_001E&for=tract:*&in=state:36' -o cb_2015_36_tract_B01003.json ndjson-cat cb_2015_36_tract_B01003.json | ndjson-split 'd.slice(1)' | ndjson-map '{id: d[2] + d[3], B01003: +d[0]}' > cb_2015_36_tract_B01003.ndjson -
Join population data to the geometry and compute population density
ndjson-join 'd.id' NY-albers-id.ndjson cb_2015_36_tract_B01003.ndjson > NY-albers-join.ndjson ndjson-map 'd[0].properties = {density: Math.floor(d[1].B01003 / d[0].properties.ALAND * 2589975.2356)}, d[0]' < NY-albers-join.ndjson > NY-albers-density.ndjson -
Convert back to GeoJson
ndjson-reduce < NY-albers-density.ndjson | ndjson-map '{type: "FeatureCollection", features: d}' > NY-albers-density.json -
Convert GeoJson to TopoJson and reduce size
npm install -g topojson geo2topo -n tracts=NY-albers-density.ndjson > NY-tracts-topo.json toposimplify -p 1 -f < NY-tracts-topo.json > NY-simple-topo.json topoquantize 1e5 < NY-simple-topo.json > NY-quantized-topo.json -
Compute county geometry
Resulting NY-merge-topo.json is used to show/hide county boundary with object.counties, whereas object.tracts will display everything
topomerge -k 'd.id.slice(0, 3)' counties=tracts < NY-quantized-topo.json > NY-merge-topo.json -
Compute county internal boarders
Resulting NY-topo.json is used to show/hide state boundary with object.counties, whereas object.tracts will display everything
topomerge --mesh -f 'a !== b' counties=counties < NY-merge-topo.json > NY-topo.json