A silhouette plot visualizes the quality of clustering results by showing the silhouette coefficient for each sample, grouped by cluster assignment. Each horizontal bar represents a sample's silhouette score (-1 to 1), where positive values indicate good cluster membership and negative values suggest potential misclassification. This visualization helps evaluate cluster cohesion (how similar samples are to their own cluster) and separation (how distinct they are from neighboring clusters).
# anyplot.ai
# silhouette-basic: Silhouette Plot
# Library: makie 0.21.9 | Julia 1.11.9
# Quality: 87/100 | Created: 2026-09-09
using CairoMakie
using Colors
using Random
using Statistics
Random.seed!(42)
# --- Theme tokens -----------------------------------------------------------
const THEME = get(ENV, "ANYPLOT_THEME", "light")
const PAGE_BG = THEME == "light" ? colorant"#FAF8F1" : colorant"#1A1A17"
const INK = THEME == "light" ? colorant"#1A1A17" : colorant"#F0EFE8"
const INK_SOFT = THEME == "light" ? colorant"#4A4A44" : colorant"#B8B7B0"
const IMPRINT_PALETTE = [
colorant"#009E73", colorant"#C475FD", colorant"#4467A3", colorant"#BD8233",
colorant"#AE3030", colorant"#2ABCCD", colorant"#954477", colorant"#99B314",
]
# --- Data: synthetic clustering result, one pair of clusters overlapping ---
# to produce a realistic mix of strong and borderline silhouette scores.
n_per_cluster = 50
n_clusters = 3
centers = [(0.0, 0.0), (4.2, 0.0), (2.0, 3.2)]
spreads = [1.0, 1.0, 1.3]
feature_x = Float64[]
feature_y = Float64[]
cluster_labels = Int[]
for c in 1:n_clusters
cx, cy = centers[c]
append!(feature_x, cx .+ spreads[c] .* randn(n_per_cluster))
append!(feature_y, cy .+ spreads[c] .* randn(n_per_cluster))
append!(cluster_labels, fill(c - 1, n_per_cluster))
end
n_samples = length(feature_x)
# --- Silhouette coefficient per sample (computed directly: a(i), b(i)) -----
euclidean(i, j) = sqrt((feature_x[i] - feature_x[j])^2 + (feature_y[i] - feature_y[j])^2)
silhouette_values = zeros(n_samples)
for i in 1:n_samples
own_cluster = cluster_labels[i]
same_idx = [j for j in 1:n_samples if cluster_labels[j] == own_cluster && j != i]
a_i = isempty(same_idx) ? 0.0 : mean(euclidean(i, j) for j in same_idx)
b_i = Inf
for c in 0:(n_clusters - 1)
c == own_cluster && continue
other_idx = [j for j in 1:n_samples if cluster_labels[j] == c]
b_i = min(b_i, mean(euclidean(i, j) for j in other_idx))
end
silhouette_values[i] = isempty(same_idx) ? 0.0 : (b_i - a_i) / max(a_i, b_i)
end
avg_silhouette = mean(silhouette_values)
# --- Arrange bars: grouped by cluster, ascending within cluster, gapped ----
cluster_gap = 8
y_positions = Float64[]
bar_values = Float64[]
bar_colors = RGB[]
cluster_center_y = Float64[]
cluster_avg_silhouette = Float64[]
cluster_max_silhouette = Float64[]
y_cursor = cluster_gap
for c in 0:(n_clusters - 1)
global y_cursor
idx = findall(==(c), cluster_labels)
sorted_vals = sort(silhouette_values[idx])
size_c = length(sorted_vals)
append!(y_positions, y_cursor:(y_cursor + size_c - 1))
append!(bar_values, sorted_vals)
append!(bar_colors, fill(IMPRINT_PALETTE[c + 1], size_c))
push!(cluster_center_y, y_cursor + size_c / 2 - 0.5)
push!(cluster_avg_silhouette, mean(sorted_vals))
push!(cluster_max_silhouette, maximum(sorted_vals))
y_cursor += size_c + cluster_gap
end
x_upper = maximum(cluster_max_silhouette) + 0.18
# --- Plot ---------------------------------------------------------------------
fig = Figure(
resolution = (1600, 900),
fontsize = 14,
backgroundcolor = PAGE_BG,
)
ax = Axis(
fig[1, 1];
title = "silhouette-basic · julia · makie · anyplot.ai",
titlesize = 20,
titlecolor = INK,
xlabel = "Silhouette Coefficient",
ylabel = "Cluster",
xlabelsize = 14,
ylabelsize = 14,
xticklabelsize = 12,
yticklabelsize = 12,
xlabelcolor = INK,
ylabelcolor = INK,
xticklabelcolor = INK_SOFT,
yticklabelcolor = INK_SOFT,
xtickcolor = INK_SOFT,
backgroundcolor = PAGE_BG,
topspinevisible = false,
rightspinevisible = false,
leftspinevisible = false,
yticksvisible = false,
bottomspinecolor = INK_SOFT,
xgridcolor = RGBAf(INK.r, INK.g, INK.b, 0.15),
ygridvisible = false,
yticks = (cluster_center_y, ["Cluster $(c)" for c in 0:(n_clusters - 1)]),
)
barplot!(ax, y_positions, bar_values;
direction = :x, color = bar_colors, gap = 0.0, strokewidth = 0)
vlines!(ax, [avg_silhouette]; color = INK_SOFT, linestyle = :dash, linewidth = 2)
for c in 0:(n_clusters - 1)
text!(ax, cluster_max_silhouette[c + 1] + 0.03, cluster_center_y[c + 1];
text = "avg = $(round(cluster_avg_silhouette[c + 1], digits = 2))",
align = (:left, :center),
color = INK,
fontsize = 13,
)
end
xlims!(ax, min(-0.15, minimum(bar_values) - 0.05), x_upper)
ylims!(ax, 0, y_cursor)
# --- Save -------------------------------------------------------------------
save("plot-$(THEME).png", fig; px_per_unit = 2)
Runnable source as JSON, for any HTTP client: https://api.anyplot.ai/specs/silhouette-basic/makie/code. Any spec id and library id listed in llms-full.txt fit the same URL shape; every URL below is complete and callable.
{
"spec_id": "silhouette-basic",
"language": "julia",
"library": "makie",
"page": "https://anyplot.ai/silhouette-basic/julia/makie",
"hub": "https://anyplot.ai/silhouette-basic",
"code_json": "https://api.anyplot.ai/specs/silhouette-basic/makie/code",
"spec_json": "https://api.anyplot.ai/specs/silhouette-basic",
"render_light_png": "https://storage.googleapis.com/anyplot-images/plots/silhouette-basic/julia/makie/plot-light.png",
"render_dark_png": "https://storage.googleapis.com/anyplot-images/plots/silhouette-basic/julia/makie/plot-dark.png",
"quality_score": 87.0,
"license": "MIT",
"guide": "https://anyplot.ai/llms.txt"
}Part of Silhouette Plot on anyplot.ai.