How to Make Stripplot with Jitter in Altair Python?

A stripplot with jitter is an effective way to visualize the distribution of a continuous variable across different categories. In Altair Python, we use mark_circle() to create the plot and transform_calculate() to add jitter, which spreads overlapping points horizontally for better visibility.

Syntax

The basic syntax for creating a stripplot with jitter in Altair involves creating a chart with circular markers and adding calculated jitter ?

import altair as alt

# Basic stripplot with jitter syntax
alt.Chart(data).mark_circle(size=50).encode(
    x=alt.X('jitter:Q', title=None, 
            axis=alt.Axis(ticks=False, grid=False, labels=False)),
    y=alt.Y('continuous_variable:Q'),
    color=alt.Color('category:N')
).transform_calculate(
    jitter='sqrt(-2*log(random()))*cos(2*PI*random())'
)

Creating a Stripplot with Custom Data

Let's create a stripplot using randomly generated data to demonstrate the jitter effect ?

import altair as alt
import pandas as pd
import numpy as np

# Enable Altair to render in notebooks
alt.data_transformers.enable('json')

# Create sample data
np.random.seed(42)
data = pd.DataFrame({
    'values': np.concatenate([
        np.random.normal(10, 2, 50),  # Group A
        np.random.normal(15, 3, 50),  # Group B  
        np.random.normal(12, 1.5, 50) # Group C
    ]),
    'category': ['A'] * 50 + ['B'] * 50 + ['C'] * 50
})

# Create stripplot with jitter
chart = alt.Chart(data).mark_circle(size=50, opacity=0.7).encode(
    x=alt.X('jitter:Q', 
            title=None,
            axis=alt.Axis(ticks=False, grid=False, labels=False),
            scale=alt.Scale(range=[0, 100])),
    y=alt.Y('values:Q', title='Values'),
    color=alt.Color('category:N', 
                   scale=alt.Scale(range=['#1f77b4', '#ff7f0e', '#2ca02c']))
).transform_calculate(
    jitter='sqrt(-2*log(random()))*cos(2*PI*random())'
).properties(
    width=200,
    height=300,
    title='Stripplot with Jitter'
)

print("Stripplot created successfully!")
print(f"Data shape: {data.shape}")
print(f"Categories: {data['category'].unique()}")

Stripplot created successfully!
Data shape: (150, 2)
Categories: ['A' 'B' 'C']

Using Real Dataset Iris Example

Here's how to create a stripplot with the famous Iris dataset ?

import altair as alt
from vega_datasets import data

# Load the Iris dataset
iris = data.iris()

# Create stripplot with jitter for petal width by species
stripplot = alt.Chart(iris).mark_circle(size=60, opacity=0.8).encode(
    x=alt.X('jitter:Q', 
            title=None,
            axis=alt.Axis(ticks=False, grid=False, labels=False),
            scale=alt.Scale(range=[0, 80])),
    y=alt.Y('petalWidth:Q', 
            title='Petal Width (cm)',
            scale=alt.Scale(zero=False)),
    color=alt.Color('species:N', 
                   title='Species',
                   scale=alt.Scale(range=['#e41a1c', '#377eb8', '#4daf4a']))
).transform_calculate(
    jitter='sqrt(-2*log(random()))*cos(2*PI*random())'
).properties(
    width=250,
    height=350,
    title='Iris Petal Width Distribution by Species'
)

print("Iris stripplot created!")
print(f"Dataset contains {len(iris)} samples")
print(f"Species: {iris['species'].unique()}")

Iris stripplot created!
Dataset contains 150 samples
Species: ['setosa' 'versicolor' 'virginica']

Understanding the Jitter Formula

The jitter calculation uses the Box-Muller transformation to generate normally distributed random values ?

import numpy as np
import matplotlib.pyplot as plt

# Demonstrate the jitter formula
np.random.seed(42)
n_points = 1000

# Box-Muller transformation for normal distribution
u1 = np.random.random(n_points)
u2 = np.random.random(n_points)
jitter_values = np.sqrt(-2 * np.log(u1)) * np.cos(2 * np.pi * u2)

print(f"Jitter statistics:")
print(f"Mean: {np.mean(jitter_values):.3f}")
print(f"Standard deviation: {np.std(jitter_values):.3f}")
print(f"Min value: {np.min(jitter_values):.3f}")
print(f"Max value: {np.max(jitter_values):.3f}")

Jitter statistics:
Mean: 0.016
Standard deviation: 1.006
Min value: -3.040
Max value: 3.722

Comparison: With vs Without Jitter

Key Parameters

• size: Controls circle marker size (default: 30)
• opacity: Sets transparency (0-1 range)
• scale range: Controls jitter spread width
• jitter formula: Generates normally distributed random offsets

Conclusion

Stripplots with jitter in Altair effectively reveal data distribution patterns that would be hidden by overlapping points. The transform_calculate() method with Box-Muller transformation provides smooth, normally distributed jitter for professional-looking visualizations.