Violin Plot

The Violin Plot widget estimates the distribution of numeric objectives or parameters as Gaussian kernel density curves, drawn as filled violins. With no category selected, every numeric column of the chosen source (objectives or parameters) is drawn side by side, one violin per column. With a categorical parameter selected, a single numeric column is split by that parameter's levels and one violin is drawn per level, so the distributions can be compared across categories.

Every violin is scaled to the same maximum width, which makes the shapes comparable regardless of how many trials fell into each column or level. A Normalize [0,1] toggle min-max normalizes each column for display, and a median marker is drawn inside each violin. Columns or levels with fewer than two finite, non-identical values cannot produce a density estimate; they are skipped and reported below the plot.

• Theory and details: Violin Plot