Reveal dense patterns with a hexbin chart.

Build a hexbin plot for dense two-dimensional data

This free hexbin chart maker groups nearby X and Y observations into a regular hexagonal grid. Color represents the number of observations in each occupied bin, revealing clusters, gradients, gaps, and unusual regions that may disappear in an overplotted scatter chart.

Import two numeric variables

Upload CSV, TSV, TXT, JSON, XLSX, or XLS data, or paste a table. Select one numeric X column and one numeric Y column, review the preview, and exclude rows that should not enter the analysis. Your data stays in the browser.

Choose a useful hexagon radius

Smaller hexagons preserve local detail but may produce many bins with only one observation. Larger hexagons create a smoother overview but can merge distinct clusters. Adjust the radius until the important structure remains visible without becoming noisy.

Select the color scaling for the density range

Linear scaling preserves proportional differences in counts. Square-root scaling gives medium-density bins more contrast, while logarithmic scaling is useful when a small number of bins contain far more observations than the rest.

Interpret the pattern carefully

Hexbin position follows the original X and Y scales, while fill intensity represents local frequency. The hexagons do not change individual values; they summarize spatial concentration. Enable raw points when you need to see how observations sit inside the bins.

Bin size changes the story

Every binned visualization depends on resolution. Test several radii before concluding that a cluster, gap, or boundary is real. The Pearson coefficient summarizes linear association but does not capture every nonlinear pattern visible in the heatmap.

Export at a readable size

Customize titles, axes, colors, opacity, grid lines, counts, and legend before downloading PNG or SVG. Larger exports help preserve axis labels and individual hexagons when the bin radius is small.

Practical example: reveal crowded trip patterns

A large set of trips can place distance on the X axis and duration on the Y axis. A raw scatter plot may become a solid cloud where observations overlap. Hexagonal bins count nearby points and use color to reveal the most common combinations, secondary clusters, sparse tails, and regions with no observations.

Compare counts only under consistent exposure

Darker bins mean more records, not necessarily a higher rate or risk. If one group covers a longer period or has more opportunities to generate observations, normalize or separate the datasets before comparing density. Keep both axes in meaningful units and note any filters that changed the source population.

Check patterns at more than one radius

A small radius preserves detail but creates many bins with low counts. A large radius is stable but may merge distinct groups. Inspect several nearby settings and compare with the scatter plot. Patterns that remain visible across reasonable resolutions deserve more confidence than a single isolated hot bin.

Protect privacy in exported density maps

Binning reduces direct point visibility but does not automatically anonymize sensitive records. Rare bins, precise coordinates, labels, or contextual axes may still identify individuals or locations. Aggregate appropriately and follow the rules that apply to the source data before publishing.