Charts · Distribution

Shape of a sample

Histograms and box plots answer different questions: the histogram shows the shape, the box plot compares many samples at once. Both take prepared values — Luc UI never bins, and never decides what an outlier is.

Histogram

Pre-binned counts, zero baseline, a reference line at the mean of the sample.

Daily return distribution

34 bins over 900 synthetic sessions.

Box plot

Six prepared five-number summaries with prepared outliers drawn as small hollow points.

Signal dispersion by strategy family

Whiskers are outward, the median is solid, the outliers are not guessed.

Deferred on purpose

Two distribution shapes that did not earn a component this round.

Density / violin

A kernel density needs a bandwidth and a smoothing convention that would have to live either in the chart or in a helper. Neither belongs in a presentation layer, so density renders as a histogram with a prepared outline series today, and a violin waits for a real consumer that needs one.

QQ plot

A quantile-quantile plot is a scatter chart with two prepared series and a 45-degree reference line:

<ScatterChart
  data={qq}
  xKey="theoretical"
  series={[{ key: "sample", tone: "accent" }]}
  annotations={{ referenceLines: [{ value: 0, axis: "y" }] }}
/>