A packed bubble chart displays data as circles where size represents value, and circles are packed together without overlap using physics simulation. Unlike scatter or traditional bubble charts, position has no meaning - only size and optional grouping matter. This visualization efficiently uses space for comparing values across many categories.
""" anyplot.ai
bubble-packed: Basic Packed Bubble Chart
Library: matplotlib 3.10.9 | Python 3.13.13
Quality: 85/100 | Updated: 2026-05-29
"""
import os
import matplotlib.collections as mcoll
import matplotlib.patches as mpatches
import matplotlib.patheffects as pe
import matplotlib.pyplot as plt
import numpy as np
# Theme tokens — Imprint palette, theme-adaptive chrome
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"
INK_MUTED = "#6B6A63" if THEME == "light" else "#A8A79F"
# Imprint palette — 8 hues, canonical order
IMPRINT_PALETTE = ["#009E73", "#C475FD", "#4467A3", "#BD8233", "#AE3030", "#2ABCCD", "#954477", "#99B314"]
# Data — department budget allocation (in thousands USD)
labels = [
"Engineering",
"Marketing",
"Sales",
"Operations",
"HR",
"Finance",
"R&D",
"Customer Support",
"Legal",
"IT",
"Design",
"Product",
"Data Science",
"Security",
"QA",
]
values = [950, 420, 680, 310, 160, 280, 820, 200, 130, 370, 230, 580, 470, 145, 175]
# Group assignments — organizational structure
group_map = {
"Engineering": "Engineering",
"IT": "Engineering",
"Data Science": "Engineering",
"R&D": "Engineering",
"Marketing": "Business",
"Sales": "Business",
"Product": "Business",
"Design": "Business",
"Operations": "Operations",
"HR": "Operations",
"Finance": "Operations",
"Customer Support": "Operations",
"Legal": "Compliance",
"Security": "Compliance",
"QA": "Compliance",
}
# Groups mapped to first 4 Imprint palette positions (canonical order)
group_order = ["Engineering", "Business", "Operations", "Compliance"]
group_colors = {g: IMPRINT_PALETTE[i] for i, g in enumerate(group_order)}
colors = [group_colors[group_map[label]] for label in labels]
# Scale values to radius (sqrt for area-proportional sizing)
min_radius = 0.30
max_radius = 2.0
values_array = np.array(values, dtype=float)
radii = min_radius + (max_radius - min_radius) * np.sqrt(
(values_array - values_array.min()) / (values_array.max() - values_array.min())
)
# Sort by size (largest first) for better packing
n = len(labels)
order = np.argsort(-radii)
radii_sorted = radii[order]
labels_sorted = [labels[i] for i in order]
values_sorted = [values[i] for i in order]
colors_sorted = [colors[i] for i in order]
groups_sorted = [group_map[labels[i]] for i in order]
unique_groups = group_order
group_ids = np.array([unique_groups.index(g) for g in groups_sorted])
# Initial positions in spiral pattern for tighter convergence
angles = np.linspace(0, 4 * np.pi, n)
spiral_r = np.linspace(0, 3, n)
positions = np.column_stack([spiral_r * np.cos(angles), spiral_r * np.sin(angles)])
# Physics simulation with group-aware clustering
for iteration in range(500):
progress = iteration / 500
pull_strength = 0.06 * (1 - progress * 0.8)
group_pull = 0.04 * (1 - progress * 0.5)
group_centers = {}
for gid in range(len(unique_groups)):
mask = group_ids == gid
if np.any(mask):
group_centers[gid] = positions[mask].mean(axis=0)
for i in range(n):
dist = np.linalg.norm(positions[i])
if dist > 0.01:
positions[i] -= pull_strength * positions[i] / dist
gc = group_centers[group_ids[i]]
to_group = gc - positions[i]
gd = np.linalg.norm(to_group)
if gd > 0.01:
positions[i] += group_pull * to_group / gd
for i in range(n):
for j in range(i + 1, n):
delta = positions[j] - positions[i]
dist = np.linalg.norm(delta)
same_group = group_ids[i] == group_ids[j]
gap = 0.06 if same_group else 0.20
min_dist = radii_sorted[i] + radii_sorted[j] + gap
if dist < min_dist and dist > 0.001:
overlap = (min_dist - dist) / 2
direction = delta / dist
positions[i] -= overlap * direction
positions[j] += overlap * direction
# Center the layout
bbox_min = positions.min(axis=0) - radii_sorted.max()
bbox_max = positions.max(axis=0) + radii_sorted.max()
positions -= (bbox_min + bbox_max) / 2
# Plot — square canvas for symmetric bubble chart (2400×2400 px at 400 dpi)
fig, ax = plt.subplots(figsize=(6, 6), dpi=400, facecolor=PAGE_BG)
ax.set_facecolor(PAGE_BG)
# Draw circles with PatchCollection for efficient batch rendering
circles = [mpatches.Circle((positions[i, 0], positions[i, 1]), radii_sorted[i]) for i in range(n)]
collection = mcoll.PatchCollection(
circles, facecolors=colors_sorted, edgecolors=PAGE_BG, linewidths=2.0, alpha=0.90, zorder=2
)
ax.add_collection(collection)
# Labels inside circles (if large enough)
small_circles = []
for i in range(n):
label_chars = len(labels_sorted[i])
min_r_for_label = 0.48 + label_chars * 0.018
if radii_sorted[i] > min_r_for_label:
font_scale = min(1.0, radii_sorted[i] / 1.6)
label_fontsize = max(9, int(11 * font_scale))
value_fontsize = max(8, int(9 * font_scale))
# Contrast-appropriate text color (WCAG relative luminance)
bg_hex = colors_sorted[i]
rgb = [int(bg_hex[j : j + 2], 16) / 255 for j in (1, 3, 5)]
luminance = 0.2126 * rgb[0] + 0.7152 * rgb[1] + 0.0722 * rgb[2]
text_color = "#1A1A17" if luminance > 0.35 else "#F0EFE8"
# RGBA tuples — portable across matplotlib versions
stroke_fg = (0, 0, 0, 0.15) if luminance > 0.35 else (1, 1, 1, 0.15)
# linewidth=1.0 at 400 dpi ≈ 6 px — crisp without garish stroke
stroke = pe.withStroke(linewidth=1.0, foreground=stroke_fg)
# Wrap multi-word labels to reduce horizontal text extent within the circle
words = labels_sorted[i].split(" ")
if len(words) > 1:
display_label = "\n".join(words)
label_y_offset = 0.05
value_y_offset = -0.35
else:
display_label = labels_sorted[i]
label_y_offset = 0.12
value_y_offset = -0.22
ax.text(
positions[i, 0],
positions[i, 1] + radii_sorted[i] * label_y_offset,
display_label,
ha="center",
va="center",
fontsize=label_fontsize,
fontweight="bold",
color=text_color,
path_effects=[stroke],
zorder=3,
)
ax.text(
positions[i, 0],
positions[i, 1] + radii_sorted[i] * value_y_offset,
f"${values_sorted[i]}K",
ha="center",
va="center",
fontsize=value_fontsize,
color=text_color,
alpha=0.85,
path_effects=[stroke],
zorder=3,
)
else:
small_circles.append(i)
# External labels with leader lines for small circles
# Scan 16 candidate angles and pick the direction with maximum clearance from
# all other circle edges — avoids the angle-from-origin pitfall where a small
# circle near the cluster centre gets a label that points into a large neighbor.
for i in small_circles:
cx, cy = positions[i, 0], positions[i, 1]
r = radii_sorted[i]
offset_dist = r + 0.65
best_angle = np.arctan2(cy, cx)
best_clearance = -np.inf
for test_angle in np.linspace(0, 2 * np.pi, 16, endpoint=False):
lx = cx + offset_dist * np.cos(test_angle)
ly = cy + offset_dist * np.sin(test_angle)
clearance = min(
np.sqrt((lx - positions[j, 0]) ** 2 + (ly - positions[j, 1]) ** 2) - radii_sorted[j]
for j in range(n)
if j != i
)
if clearance > best_clearance:
best_clearance = clearance
best_angle = test_angle
angle = best_angle
lx = cx + offset_dist * np.cos(angle)
ly = cy + offset_dist * np.sin(angle)
ax.annotate(
f"{labels_sorted[i]}\n${values_sorted[i]}K",
xy=(cx + r * np.cos(angle), cy + r * np.sin(angle)),
xytext=(lx, ly),
fontsize=9,
fontweight="bold",
color=INK_SOFT,
ha="center",
va="center",
arrowprops={"arrowstyle": "-", "color": INK_MUTED, "lw": 1.2, "shrinkA": 4, "shrinkB": 0},
zorder=4,
)
# Symmetric axis limits centered at origin
all_x, all_y = positions[:, 0], positions[:, 1]
max_r = radii_sorted.max()
half_extent = max(all_x.max() - all_x.min(), all_y.max() - all_y.min()) / 2 + max_r + 0.75
ax.set_xlim(-half_extent, half_extent)
ax.set_ylim(-half_extent, half_extent)
ax.set_aspect("equal")
ax.axis("off")
# Title
title = "bubble-packed · python · matplotlib · anyplot.ai"
title_fontsize = max(8, round(12 * 67 / len(title))) if len(title) > 67 else 12
ax.set_title(title, fontsize=title_fontsize, fontweight="medium", color=INK, pad=12)
# Legend for group colors with theme-adaptive frame
legend_handles = [
mpatches.Patch(facecolor=color, edgecolor=PAGE_BG, linewidth=1.5, label=group)
for group, color in group_colors.items()
]
leg = ax.legend(
handles=legend_handles,
loc="lower right",
fontsize=9,
fancybox=False,
borderpad=0.8,
handlelength=1.5,
handleheight=1.2,
)
leg.get_frame().set_facecolor(ELEVATED_BG)
leg.get_frame().set_edgecolor(INK_SOFT)
plt.setp(leg.get_texts(), color=INK_SOFT)
fig.subplots_adjust(left=0.02, right=0.98, top=0.91, bottom=0.02)
plt.savefig(f"plot-{THEME}.png", dpi=400, facecolor=PAGE_BG)
Part of Basic Packed Bubble Chart on anyplot.ai.