Violin Plot with Embedded Box Plot — Altair

A violin plot with an embedded box plot inside, combining the distribution shape visualization (KDE) with traditional quartile statistics. Shows both the probability density and summary statistics in one plot.

Violin Plot with Embedded Box Plot rendered with Altair

Python source (Altair)

""" anyplot.ai
violin-box: Violin Plot with Embedded Box Plot
Library: altair 6.1.0 | Python 3.13.13
Quality: 90/100 | Updated: 2026-05-12
"""

import os

import altair as alt
import numpy as np
import pandas as pd


THEME = os.getenv("ANYPLOT_THEME", "light")
PAGE_BG = "#FAF8F1" if THEME == "light" else "#1A1A17"
ELEVATED_BG = "#FFFDF6" if THEME == "light" else "#242420"
INK = "#1A1A17" if THEME == "light" else "#F0EFE8"
INK_SOFT = "#4A4A44" if THEME == "light" else "#B8B7B0"

IMPRINT = ["#009E73", "#C475FD", "#4467A3", "#BD8233"]

# Data - Generate realistic response time data for different server tiers
np.random.seed(42)

groups = ["Basic", "Standard", "Premium", "Enterprise"]
n_per_group = 80

data = []
# Basic tier - higher latency, more variance
data.extend([(np.random.exponential(120) + 80, "Basic") for _ in range(n_per_group)])
# Standard tier - moderate latency
data.extend([(np.random.normal(100, 25), "Standard") for _ in range(n_per_group)])
# Premium tier - lower latency, tighter distribution
data.extend([(np.random.normal(60, 15), "Premium") for _ in range(n_per_group)])
# Enterprise tier - lowest latency, bimodal (some cached, some not)
enterprise_cached = np.random.normal(25, 8, n_per_group // 2)
enterprise_uncached = np.random.normal(55, 12, n_per_group // 2)
data.extend([(v, "Enterprise") for v in np.concatenate([enterprise_cached, enterprise_uncached])])

df = pd.DataFrame(data, columns=["Response Time (ms)", "Server Tier"])
df["Response Time (ms)"] = df["Response Time (ms)"].clip(lower=5)

# Create violin plot layer using transform_density
violin = (
    alt.Chart(df)
    .transform_density("Response Time (ms)", as_=["Response Time (ms)", "density"], groupby=["Server Tier"])
    .mark_area(orient="horizontal", opacity=0.6)
    .encode(
        y=alt.Y("Response Time (ms):Q"),
        x=alt.X(
            "density:Q",
            stack="center",
            impute=None,
            title=None,
            axis=alt.Axis(labels=False, values=[0], grid=False, ticks=False),
        ),
        color=alt.Color("Server Tier:N", scale=alt.Scale(domain=groups, range=IMPRINT)),
    )
)

# Create box plot layer
boxplot = (
    alt.Chart(df)
    .mark_boxplot(
        extent="min-max",
        size=25,
        median={"stroke": INK_SOFT, "strokeWidth": 2},
        box={"fill": INK_SOFT, "fillOpacity": 0.3},
        outliers={"size": 60, "strokeWidth": 2, "stroke": INK_SOFT},
    )
    .encode(y=alt.Y("Response Time (ms):Q", title="Response Time (ms)"), x=alt.value(0), color=alt.value(INK_SOFT))
)

# Layer violin and box plots first, then facet
layered = alt.layer(violin, boxplot).properties(width=280, height=600)

# Apply faceting after layering
chart = (
    layered.facet(
        column=alt.Column(
            "Server Tier:N",
            header=alt.Header(titleFontSize=20, labelFontSize=18, labelOrient="bottom"),
            title=None,
            sort=groups,
        )
    )
    .properties(
        title=alt.Title("violin-box · altair · anyplot.ai", fontSize=28, anchor="middle", offset=20), background=PAGE_BG
    )
    .configure_axis(
        labelFontSize=16,
        titleFontSize=20,
        gridOpacity=0.0,
        domainColor=INK_SOFT,
        tickColor=INK_SOFT,
        labelColor=INK_SOFT,
        titleColor=INK,
    )
    .configure_view(stroke=None, fill=PAGE_BG)
    .configure_legend(
        titleFontSize=18, labelFontSize=16, symbolSize=200, orient="right", titleColor=INK, labelColor=INK_SOFT
    )
    .configure_title(color=INK)
    .resolve_scale(x="independent")
)

# Save outputs
chart.save(f"plot-{THEME}.png", scale_factor=3.0)
chart.save(f"plot-{THEME}.html")

Part of Violin Plot with Embedded Box Plot on anyplot.ai.

Other implementations