search

scc-pi / hexsheff

Resources to help create hexagonal cartograms for Sheffield
0 stars 0 forks source link

readme

hexsheff

Resources to help create hexagonal cartograms for Sheffield.

Hex maps vs choropleths

Choropleth maps are a great way to illustrate differences in a city. However, Sheffield is part national park and population density varies more than most cities. Often, city sub-areas are of approximately equal population sizes, which means Sheffield sub-areas are of significantly different geographical sizes. For Sheffield choropleths, the differences in geographical sizes can detract attention from the main statistic of interest.

Hexagonal cartograms, or “hex maps”, are one alternative to choropleths. They still provide an indication of location, but mute differences in geographical size.

ODI Leeds have done some interesting work on hex maps.

Building a hex map template

Once a template for Sheffield sub-area hexagons is defined, e.g. by a GeoJson file, it can be reused to illustrate different statistics. The process for creating the LAC (Local Area Committee) hexagons was:

  1. Drop the data/lac_asc_xref.csv file into the ODI Leeds hex map builder, hexify, and save hexes as HexJson.
  2. Drop the data/lac_asc.hexjson file into Oli Hawkins HexJSON Editor, edit, and download as GeoJson to file data/lac-asc-hex.geojson.

The HexJSON format defined by ODI Leeds is a great idea. However, use of the format hasn’t been huge, so we’ve stuck with GeoJson, which plays well with the ggplot2 R package.

geojson.io is useful for viewing and editing GeoJson files.

Different Sheffield hex maps

So far, this repository includes a hex map definition for LACs: data/lac-asc-hex.geojson

ASC (Adult Social Care) localities have the same boundaries as LACs. The LAC hex map definition includes ASC locality cross-references, so it can be used for that purpose too.

If hex maps prove useful, other Sheffield hex maps that could be created include:

  1. Wards (28)
  2. Neighbourhoods (100)

The more the audience of Sheffield hex maps use them, the easier they will be to read. Ideally, data analysts producing the hex maps will agree on a single hex map for each type of sub-area. New or improved Sheffield hex maps, or suggestions, are encouraged via pull requests or issues.