小提琴图（Julia）

作者

[编辑] Bizard 团队。

修改于

2026-04-04

🤖 AI Skill — Copy this tutorial's skill into your AI assistant

小提琴图将箱线图统计信息与核密度估计相结合，展示数据分布。Julia 的 CairoMakie 可以方便地创建小提琴图，用于比较不同组之间的基因表达、生物标志物水平或临床测量值。

示例

环境配置

系统要求：跨平台（Linux/MacOS/Windows）
编程语言：Julia
依赖包：CairoMakie、DataFrames、Statistics

using CairoMakie
using DataFrames
using Statistics
using Random

数据准备

Random.seed!(42)
n_per = 80
groups = vcat(fill("Tumor", n_per), fill("Normal", n_per), fill("Adjacent", n_per))
expression = vcat(
    randn(n_per) .* 1.5 .+ 8,
    randn(n_per) .* 1.2 .+ 5,
    randn(n_per) .* 1.8 .+ 6.5
)
group_idx = vcat(fill(1, n_per), fill(2, n_per), fill(3, n_per))
df = DataFrame(Group=groups, Expression=expression, GroupIdx=group_idx)

240×3 DataFrame

215 rows omitted

Row	Group	Expression	GroupIdx
	String	Float64	Int64
1	Tumor	7.45496	1
2	Tumor	8.37761	1
3	Tumor	7.52752	1
4	Tumor	7.53312	1
5	Tumor	9.22446	1
6	Tumor	8.71511	1
7	Tumor	6.71067	1
8	Tumor	5.79607	1
9	Tumor	4.8285	1
10	Tumor	8.06567	1
11	Tumor	6.762	1
12	Tumor	9.26043	1
13	Tumor	8.65083	1
⋮	⋮	⋮	⋮
229	Adjacent	8.67822	3
230	Adjacent	5.35136	3
231	Adjacent	6.22416	3
232	Adjacent	3.53355	3
233	Adjacent	8.33217	3
234	Adjacent	6.73736	3
235	Adjacent	11.0713	3
236	Adjacent	6.77344	3
237	Adjacent	7.34419	3
238	Adjacent	5.31451	3
239	Adjacent	5.72756	3
240	Adjacent	3.46341	3

可视化

基础小提琴图

fig = Figure(size=(700, 500))
ax = Axis(fig[1,1], xlabel="Group", ylabel="Expression Level",
          title="Gene Expression Distribution",
          xticks=(1:3, ["Tumor", "Normal", "Adjacent"]))
violin!(ax, df.GroupIdx, df.Expression, color=(:steelblue, 0.7))
fig

带箱线图叠加的分组小提琴图

fig2 = Figure(size=(700, 500))
ax2 = Axis(fig2[1,1], xlabel="Group", ylabel="Expression Level",
           title="Violin + Box Plot",
           xticks=(1:3, ["Tumor", "Normal", "Adjacent"]))
colors = [:red, :steelblue, :green]
for (i, g) in enumerate(["Tumor", "Normal", "Adjacent"])
    mask = df.Group .== g
    vals = df.Expression[mask]
    violin!(ax2, fill(i, sum(mask)), vals, color=(colors[i], 0.4), side=:left)
    boxplot!(ax2, fill(i, sum(mask)), vals, color=(colors[i], 0.7), width=0.3)
end
fig2

半小提琴（雨云）图

fig3 = Figure(size=(800, 500))
ax3 = Axis(fig3[1,1], xlabel="Group", ylabel="Expression Level",
           title="Raincloud Plot",
           xticks=(1:3, ["Tumor", "Normal", "Adjacent"]))
for (i, g) in enumerate(["Tumor", "Normal", "Adjacent"])
    mask = df.Group .== g
    vals = df.Expression[mask]
    violin!(ax3, fill(i, sum(mask)), vals, color=(colors[i], 0.5), side=:right)
    scatter!(ax3, fill(i, sum(mask)) .- 0.15 .+ randn(sum(mask)) .* 0.03,
             vals, color=(colors[i], 0.3), markersize=5)
end
fig3

参考文献

Danisch, S., & Krumbiegel, J. (2021). Makie.jl: Flexible high-performance data visualization for Julia. JOSS, 6(65), 3349.

📋 AI Skill: 小提琴图（Julia）

示例