Most commutes cluster between 20 and 40 minutes , a long tail of extreme commuters persists Figure 40.1 , Most commutes cluster between 20 and 40 minutes

A Histogram visualises the frequency distribution of a single continuous variable by dividing its range into equal-width intervals (bins) and drawing a bar for each bin whose height encodes how many observations fall within that interval. Unlike a Bar Chart, the x-axis is a continuous numeric scale , the bars are adjacent, with no gaps, because the underlying data has no categorical gaps either. The perceptual mechanism is area: each bar's area (width times height) is proportional to the count or relative frequency it represents.

The message requires knowing where values concentrate across a continuous range , not how categories compare, not how something changed over time. Commute time is continuous: 23.4 minutes and 23.6 minutes are neighbouring values on a scale, not separate categories. The histogram is the correct chart when the variable is continuous and the question is distributional. The right skew in this data , mean pulled above median, long tail to the right , is the actual finding. A summary statistic (just the mean) would hide this completely.

Bin width is the single most consequential design choice in a histogram. Too narrow: every bar contains one or two observations; the chart looks like noise and the shape is unreadable. Too wide: genuine structure (bimodality, gaps, secondary peaks) gets smoothed away. The right bin width reveals the distribution shape without overfitting to sample noise. This implementation exposes the bin-width slider deliberately , it is a teaching tool. Slide from 2 minutes to 20 minutes and watch the same data tell different stories.

A Bar Chart requires discrete, named categories. Applying it to continuous data requires first binning the values and assigning labels , at which point you have rebuilt a histogram with gaps inserted for no reason. The gap signals a categorical break that doesn't exist in the data. A Box Plot summarises the same distribution in five numbers (min, Q1, median, Q3, max). It is more compact and better for comparing distributions across groups.

Paste this into Claude Code to generate a working version of this chart, plus its data file. The result will not be a perfect replica , the goal is that the reader can run the prompt, get a chart of this type, and read its source. Generate a complete, self-contained histogram in D3 v7. Two files:

The ideas in this chapter didn't appear from nowhere. Herman Hollerith developed the punch-card tabulating machine for the 1890 US Census , automating the production of frequency counts that became, when printed, the first machine-aided histograms. His company eventually became IBM. Herman Hollerith, circa 1890. AI-generated portrait based on a public domain photograph (Wikimedia Commons).

The ideas in this chapter didn't appear from nowhere. Herman Hollerith developed the punch-card tabulating machine for the 1890 US Census , automating the production of frequency counts that became, when printed, the first machine-aided histograms. His company eventually became IBM. Herman Hollerith, circa 1890. AI-generated portrait based on a public domain photograph (Wikimedia Commons).