Histogram Maker

What a histogram shows

A histogram summarizes one numeric variable by splitting its range into adjacent intervals called bins. The horizontal axis is the value range, and the vertical axis is the number of observations in each range. Unlike a bar chart, the bins are continuous numeric intervals: touching bars show that the underlying values can fall anywhere along the same scale.

For a chosen minimum a, maximum b and k equal-width bins, the width is:

bin width = (b − a) / k

Each input value is counted once. A value equal to the maximum goes into the final bin so no observation disappears at the upper boundary. This tool uses an interquartile-range based automatic suggestion when the data has enough spread, with a Sturges-rule fallback when that width would be unusable. You can always override it with a manual count.

The intervals in a real histogram share borders, so a boundary convention is needed. The visual range labels are rounded for readability; the internal counts use full numeric precision.

Choosing a useful number of bins

Too few bins can hide separate clusters. Too many make random noise look meaningful. Start with the automatic result, then try one modestly smaller and one modestly larger count. A conclusion that changes radically with a small bin adjustment deserves more data or a second diagnostic, not a stronger claim.

Histograms are especially useful for spotting a long tail, a possible outlier, a gap, multiple peaks, or an unexpectedly wide spread. They do not prove a distribution, explain a cause, or replace a statistical test. A single repeated value is displayed in one centered bin so the tool can remain truthful about a dataset with no spread.

Histograms versus bar charts

Use a histogram for measured or counted numeric values such as response time, length, temperature, score or purchase amount. Use a bar chart for named categories such as products, countries or survey answers. Reordering a categorical bar chart is fine. Reordering a histogram is not, because the numeric order and contiguous bins carry the meaning.

All parsing, binning and rendering happen in the current browser tab. The data is not sent to a chart service, placed in the funnel URL or stored in a request to our servers. Funnel settings are kept only in this tab’s session storage and can be cleared with Start over or by closing the tab.

For background, see the NIST/SEMATECH Engineering Statistics Handbook on histograms, the W3C guidance on accessible data tables, and MDN’s SVG accessibility reference.