What Is a Hexbin Map?

A hexbin map divides space into a regular grid of hexagons and summarises observations inside each cell. The hexagons can show counts, densities, averages, totals or other statistics.

It is useful when a point dataset is too dense to draw directly but you still want the aggregation units to remain explicit.

Why hexagons?

Hexagons have six neighbours at equal edge distance and are visually less aligned to horizontal and vertical axes than squares, which can produce a balanced-looking grid for neighbourhood-style comparison.

But hexagons are not inherently more truthful than squares. The important analytical choice is the cell size and grid placement, not the aesthetic superiority of one shape.

Choosing cell size

Large cells produce a smoother, more general pattern, whereas small cells preserve local variation but can become sparse or noisy.

The appropriate size depends on point density, positional uncertainty, map scale and the geographic process being studied. A cell size that works for citywide events may be meaningless for national data.

Changing cell size can change the visible pattern, so it should be treated as part of the analysis rather than a purely visual setting.

Counts, densities and summaries

A cell can encode the number of points it contains, but it can also aggregate an attribute. For example, a hexbin map might show:

number of incidents per cell;
average measured temperature from sensors;
total sales associated with customer points;
median travel time for observations inside each cell.

The legend should make clear which statistic is being mapped.

Hexbin vs heat map

Both reduce point clutter, but they do so differently.

A heat map creates a smooth intensity field in which point influence extends beyond exact aggregation cells, whereas a hexbin map keeps the summarisation units visible, making counts or statistics easier to reproduce and inspect.

If the reader needs to know that “this cell contains 84 events”, hexbins are usually clearer. If the purpose is exploratory perception of broad concentration, a heat map may feel more natural.

Where hexbin maps can mislead

Grid origin and cell size affect which points fall together, so a small shift in the grid can change local counts. Cells near the edge of a study area may also contain less available land or population, making raw counts difficult to compare, while positional uncertainty can move observations between neighbouring cells.

For exposure or population questions, normalising by an appropriate denominator may be necessary rather than mapping raw point counts.

When should you use one?

Use a hexbin map when you have many points, want a regular spatial summary, and need the aggregation units to remain explicit.

Keep the original point map when individual locations matter. Use a heat map when continuous-looking concentration is the intended visual message. Use administrative aggregation when the question itself is about those administrative units rather than a regular grid.

Hexagons create a new geography

A regular grid replaces administrative areas with equal-sized analytical units. That can reduce the visual influence of unequal polygon sizes and make point density easier to compare, but the grid is still a chosen partition of space.

Shift the grid origin or change the cell size and local counts can change. This is a grid-based version of the broader problem that spatial aggregation depends on the units used, which makes the cells analytically convenient rather than natural geographic objects.

Exposure can require normalisation

If more people, roads or sensors exist in some cells, higher counts may reflect greater opportunity for observations rather than higher underlying risk. A hexbin of crashes, for example, might need road length or traffic volume as a denominator if the question concerns crash risk.

The regular geometry makes cells comparable in area, but equal area is not the same as equal exposure.