Describe › Standalone plots

Violin plot

A violin plot renders a per-group kernel density symmetrically about each category, with an optional median-IQR box overlay.

What is Violin plot?

The violin combines a box plot (median + IQR) with a smoothed distribution shape (KDE). Each group's density is mirrored about the category position, producing a 'violin' silhouette. Width at any y is proportional to the density at that y — wider = more observations there.

Bandwidth comes from Silverman's rule: 0.9 · min(SD, IQR/1.34) · n^(-1/5). Densities are normalised across groups so the widest violin uses 84% of category spacing — group sizes aren't directly comparable from width.

Compared to the box plot: violin shows shape (bimodality, plateau, skew) that the box hides. Use both together when both summary and shape matter — that's why we offer the overlay-box checkbox.

When should I use Violin plot?

When shape matters (bimodal distributions, plateaus).
Effect-of-treatment visualisations where you want to show distribution change, not just mean change.

What data does it need?

One or more numeric columns + optional grouping + Overlay-box toggle.

What does it report?

ECharts custom-series polygons per group, optionally with narrow box + median.

How do I interpret the result?

Bimodal violins reveal a substructure (e.g. responders vs non-responders) that a box plot would summarise as 'wide IQR' and hide.