Violin plots

seaborn.violinplot() visualizes the distribution of a numerical variable for one or more categories — similar to a boxplot, but with a smoothed kernel density estimate (KDE) on each side.

You can think of it as a boxplot + density curve combined.
It shows:

• The median (white dot by default)
• The interquartile range (IQR) (thick black bar)
• The full data range (thin line)

Syntax:

Parameters:

• data = DataFrame containing the data
• x, y = Variables for categories and numeric values
• hue = Adds subcategories (split violins)
• order, hue_order = Category order
• palette = Colour scheme
• bw = Bandwidth for KDE smoothing
• cut = Extent to which the violin extends beyond data range
• scale = 'area', 'count', or 'width' — how violins are sized
• inner = What to show inside the violin: 'box', 'quartile', 'point', 'stick', or None
• split = If True, split violins for hue categories
• width = Violin width
• orient = "v" (vertical) or "h" (horizontal)

Basic example

Shows the distribution of total_bill for each day.

Add hue (subgroups)

• Each day has two violins, one for each sex.
• By default, they’re placed side by side.

Split violins (compare distributions directly)

Splits the violins in half (left vs. right) for each subgroup. Great for direct visual comparison.

Horizontal violin plot

Flips the violins horizontally — good for long category labels.

Customise inner display

Options for inner:

• 'box' = Small boxplot inside each violin (default)
• 'quartile' = Horizontal lines for quartiles
• 'point' = Dots for data points
• 'stick' = Small vertical lines for each observation
• None = Removes internal marks

Adjust smoothness & width

• bw: bandwidth for KDE (controls smoothness).
• cut: how far violins extend beyond actual data.
• scale: how violins are scaled ('area', 'width', 'count').

Overlay raw data for context

Shows the smooth distribution (violin) + individual data points (stripplot).