Swarm plot in seaborn

  1. Data simulation
  2. Bee swarm plot in seaborn with swarmplot
  3. Swarm plot by group in seaborn

Data simulation

The data of the following block of code will be used in the examples below. On the right there is a table where you can see the data which will be used in this tutorial.

import numpy as np
import seaborn as sns
import random

# Data simulation
rng = np.random.RandomState(0)
variable = rng.normal(0, 1, size = 400)
random.seed(0)
group = random.choices(["G1", "G2", "G3"], k = 400)
group2 = random.choices(["A", "B"], k = 400)
df = {'variable': variable, 'group': group, 'group2': group2}
Variable
Group
Group 2
1.76
G3
B
0.40
G3
B
0.98
G2
A
2.24
G1
A
1.87
G2
B
-0.98
G2
A
1–6 of 400 rows
...

Bee swarm plot in seaborn with swarmplot

The swarmplot function allows creating a bee swarm plot or swarm plot in Python when using seaborn. Note that you can pass a variable or a variable of a data frame, as shown below.

import seaborn as sns

# Swarm plot
sns.swarmplot(x = variable)

# Equivalent to:
sns.swarmplot(x = "variable", data = df)

Swarm plot or bee swarm plot in Python with seaborn

Vertical swarm plot

If you want to rotate your plot and create a vertical visualization you can input your variable to the y argument of the function, instead of x.

import seaborn as sns

# Swarm plot
sns.swarmplot(y = variable)

Vertical swarm plot in Python

Size of the points

The default size of the points of the plot is 5. However, the size argument allows choosing the size you desire based on your data.

import seaborn as sns

# Swarm plot
sns.swarmplot(y = variable,
              size = 4)

Change the size of the points of the swarm plot in seaborn

Fill color customization

By default, the points of the seaborn swarm plot will be blue, but you can customize its color setting a new color with the color argument of the swarmplot function.

import seaborn as sns

# Swarm plot
sns.swarmplot(y = variable,
              color = "green")

Change the color of the data points of the bee swarm in Python

Border width and color of the points

In addition, it is possible to customize the width (which defaults to 0) and the color of the borders of the points with linewidth and edgecolor, respectively.

import seaborn as sns

# Swarm plot
sns.swarmplot(y = variable,
              edgecolor = "red",
              linewidth = 0.5)

Border color and width of the points of the swarm plot in seaborn

Swarm plot by group in seaborn

Swarm plot by group in Python with seaborn

If you have a categorical variable representing groups you can pass both the numerical and the categorical variable to the function to create a swarm plot by group in Python with seaborn.

import seaborn as sns

# Swarm plot
sns.swarmplot(x = group, y = variable)

# Equivalent to:
sns.swarmplot(x = "group", y = "variable", data = df)

Horizontal swarm plot by group in seaborn

Orientation

Note that if you change the order of the variables you can create an horizontal swarm plot by group.

import seaborn as sns

# Swarm plot
sns.swarmplot(y = group, x = variable)

Order of the groups of the swarm plot by group in Python

Custom order

You might have noticed that the order of the groups on the previous plot was G3, G2, and G1, which is the order of appearance of the groups of the categorical variable (check it seeing the table of the first section). To modify the order you can use the order argument as follows.

import seaborn as sns

# Swarm plot
sns.swarmplot(x = group, y = variable,
              order = ["G1", "G2", "G3"])

Change the color palette of the swarmplot function

Color palette

The palette argument allows customizing the color palette of the plot. You can pass the name of a color palette or a dictionary with a color for each group.

import seaborn as sns

# Swarm plot
sns.swarmplot(x = group, y = variable,
              palette = "pastel")

Color of the swarmplot based on a second categorical variable

Color based on a second categorical variable

If your data set contains a second grouping variable you can pass it to the hue argument to colorize each group based on that subgroups.

import seaborn as sns

# Swarm plot
sns.swarmplot(x = group, y = variable,
              hue = group2)

Dodged swarm plot in seaborn

Dodged swarm plot

In the previous scenario you can also set the argument dodge to True, so the data points will be separated based on the second group.

import seaborn as sns

# Swarm plot
sns.swarmplot(x = group, y = variable,
              hue = group2, dodge = True)
R CODER website
Data visualization

Data visualization

A practical introduction

Buy on Amazon
Better Data Visualizations

Better Data Visualizations

A Guide for Scholars, Researchers, and Wonks

Buy on Amazon

See also

Histogram by group in seaborn

Histogram by group in seaborn
Kernel density plot in seaborn with kdeplot

Kernel density plot in seaborn with kdeplot
Box plot in plotly

Box plot in plotly
Violin plot in matplotlib

Violin plot in matplotlib
Ridgeline plots (joy plots) in matplotlib with joypy

Ridgeline plots (joy plots) in matplotlib with joypy