How To Make Bubble plot with Altair in Python?

Prerequisite: Introduction to Altair in Python

Altair is a simple and easy to use statistical visualization library for python. It contains many types of built-in plots and various options to modify the properties and generate other plots. Bubble Plot is a very useful visualization for bivariate analysis of data with respect to a third variable. It is not readily available in the Altair library but can be made by doing some simple modifications to the scatter plot.

What is a Bubble Plot?

Bubble Plot is basically a scatter plot between two variables/data columns where in place of the data points, there are bubbles/circles of varying sizes indicating the third variable. The third variable can be of a quantitative, ordinal, or nominal type, but the best type to be used in bubble plot is the ordinal type, i.e. data having a specific ordering. The legend shows which circle size corresponds to which data value.

A bubble plot can help us see the relationship between two variables with respect to a third variable. The bigger the bubble, the bigger value of data it corresponds to.

Creating a Bubble Plot

To make a bubble plot, the user simply has to map a suitable variable from the dataset to the size encoding in a simple scatter plot.

The datasets used in these articles are from the Vega_datasets library. 

# Python3 program to illustrate 
# How to make a bubble plot
# using the altair library
  
# Importing altair and vega_datasets 
import altair as alt 
from vega_datasets import data 
  
# Selecting the cars dataset 
cars = data.cars() 
  
# Making the base scatter plot 
alt.Chart(cars).mark_point().encode( 
  
  # Map the sepalLength to x-axis 
    x = 'Acceleration', 
  
  # Map the petalLength to y-axis 
    y = 'Displacement',
  
  # Map the Cylinders variable to size
  # and specify it as a nominal variable
      size = 'Cylinders:N'
)

Output:

Customizing the Bubble Plot

You can do the following customizations to the bubble plot:

• Color: You can change the default color of the bubbles by setting the color parameter of the mark_point() method.
• Opacity: You can change the default opacity of the bubbles by setting the opacity parameter of the mark_point() method. It ranges from 0 to 1.
• Filled: This is false by default, but you can change the filled parameter to true, thereby filling the bubble with the specified color.

Example:

# Python3 program to illustrate
# how to customize a bubble plot
# Importing altair and vega_datasets
import altair as alt
from vega_datasets import data
# Selecting the cars dataset
cars = data.cars()
# Making the base scatter plot
# and adding the customizations
alt.Chart(cars).mark_point(color='green',
                           filled=True,
                           opacity=0.4).encode(
  
    # Map the sepalLength to x-axis
    x='Acceleration',
  
    # Map the petalLength to y-axis
    y='Displacement',
  
    # Map the Cylinders variable to size
    # and specify it as a nominal variable
    size='Cylinders:N'
)

Output: