Bubble Charts

What is a bubble chart?

A bubble chart is a scatter plot where marker size carries a third variable, so a single point communicates three numbers at once:

• x axis = metric 1
• y axis = metric 2
• bubble size = metric 3 (volume/magnitude)

Add color and you’re encoding a fourth variable too — which is exactly what px.scatter(..., size=...) gives you; a bubble chart isn’t a separate chart type in Plotly, it’s a scatter plot with the size argument filled in.

Example

Bubble chart

import pandas as pd
import plotly.express as px
 
df = pd.DataFrame({
    "city": ["Pune", "Delhi", "Mumbai", "Bengaluru"],
    "sales": [120, 180, 90, 160],
    "profit": [30, 45, 10, 35],
    "orders": [200, 350, 120, 280],
})
 
fig = px.scatter(
    df,
    x="sales",
    y="profit",
    size="orders",
    color="city",
    hover_name="city",
    title="Sales vs Profit (bubble size = orders)",
)
fig.show()

Plotly scales bubble area (not raw radius) to the size column by default, which is the perceptually correct choice — doubling a value should roughly double the visual area, not the diameter.

sketch Bubble size encodes a third metric p5.js Pause Restart

x = sales, y = profit, bubble area = orders — bigger bubble means more orders, independent of position.

Tips

• Keep sizes reasonable; very large bubbles hide others — cap the biggest bubble at a sensible max area.
• Use hover tooltips (hover_name) to reduce label clutter instead of printing every city name on the chart.
• Watch out for overlap: when bubbles are close together, consider adding opacity so overlapping regions stay readable.

Analyst’s tip

People compare area poorly compared to length or position — a bubble twice the area of another doesn’t look “twice as big” to most viewers. That’s why bubble charts work best for rough “bigger vs smaller” impressions, not for reading off precise values. If precision matters, put the number in the hover tooltip.

diagram Bubble chart = scatter + a size channel mermaid

  flowchart LR
  A["px.scatter(x, y)"] --> B["+ size=column
marker area encodes metric 3"]
  B --> C["+ color=column
hue encodes metric 4 (category)"]
  C --> D["Bubble chart:
4 variables, 1 point per row"]

A bubble chart layers a third numeric variable onto marker size, and optionally a fourth onto color.

Next

Continue to: Sunburst Charts for visualizing hierarchical data.

🧪 Try It Yourself

Exercise 1 – Turn a Scatter into a Bubble Chart

pch.exerciseTag pch.exerciseTitle

Exercise 2 – Reduce Label Clutter with hover_name

pch.exerciseTag pch.exerciseTitle

Exercise 3 – Add a Fourth Variable with Color

pch.exerciseTag pch.exerciseTitle