Choropleth Maps: Visualizing Data by Region

A choropleth map shades geographic regions - countries, states, postal codes - by the value of some statistic. Think population density by county or sales by territory. It is one of the clearest ways to show how a metric varies across space.

What you need

Two ingredients: polygon boundaries for your regions, and a value to color them by. The boundaries usually come as a GeoJSON or Shapefile of polygons, and the values come from a spreadsheet - sales figures, counts, percentages, anything numeric.

Join your data to the boundaries

The key step is matching each row of your data to the right polygon, usually by a shared key like a region name or FIPS code. Make sure the join key is spelled and formatted identically in both files - "New York" will not match "New York State," and a zero-padded code like 04001 will not match 4001.

Choose the right value to map

• Use rates or densities, not raw counts - otherwise big regions always look "high" simply because they are big.
• Normalize by population or area so regions are comparable.
• Pick a color scale that matches your data: sequential for low-to-high, diverging for values above and below a midpoint.

Avoid common pitfalls

Large, sparsely populated regions can visually dominate a choropleth even when little is happening there. And too many color buckets make a map hard to read - five to seven classes is usually plenty.