How to Make Cyberpunk “Dark Mode” Data Visualizations in Python

Neon Lines and Dark Designs, An Introduction

Apr 8, 2024

3 min read

I’ve always loved a dark background on a chart with neon lines for their aesthetic, and also their improved accessibility for certain types of vision impairments – in this article we’ll be discussing how you can make some very cool and aesthetically pleasing cyberpunk-style charts in Python.

(ps. shoutout to one of my early Analytics managers who said my charts were ugly and unprofessional – you’d hate this article! 👋 )

At Chime, my friend Maia Bittner and I used to create charts that looked like this in Google Sheets and Excel, by manually selecting all of the neon colors to make our data eye catching + engaging. With this method, you can easily generate them in Python too!

Now, let’s take a look at how we can create some cool cyberpunk style neon data visualizations using the Palmer Penguins dataset which you can access directly in your python notebook by installing palmerpenguins

💫 Enter mplcyberpunk, a "Python package on top of matplotlib to create ‘cyberpunk’ style plots with 3 additional lines of code."

First, make sure to install the packages we’ll need for this tutorial before proceeding, and load the penguins dataset like so:

pip install palmerpenguins
pip install mplcyberpunk

from palmerpenguins import load_penguins
import matplotlib.pyplot as plt
import mplcyberpunk

# Load the penguin dataset
penguins = load_penguins()

Now, we’re ready to incorporate cyberpunk into our charts, this package works best with line charts, so let’s take a look at the distribution of body mass in different species of penguin.

Image created by the author. Code to generate this data visualization below:

# Create the plot with the cyberpunk style
plt.style.use("cyberpunk")
fig, ax = plt.subplots(figsize=(10, 6))

# Define the colors for each species
species_colors = {
    'Adelie': 'cyan',
    'Chinstrap': 'magenta',
    'Gentoo': 'yellow'
}

# Plot KDE for body mass by species
for species in penguins['species'].unique():
    subset = penguins[penguins['species'] == species]
    sns.kdeplot(subset['body_mass_g'], ax=ax, lw=3, color=species_colors[species])

# Enhance with cyberpunk style
mplcyberpunk.add_glow_effects()

# Manually create legend handles and labels
handles = [plt.Line2D([], [], color=species_colors[species], label=species, linewidth=3) for species in penguins['species'].unique()]

# Create the legend
ax.legend(handles=handles, title='Species')
legend = ax.get_legend()
plt.setp(legend.get_texts(), color='white')

# Set the title and labels
ax.set_title('Body Mass Distribution by Species')
ax.set_xlabel('Body Mass (g)')
ax.set_ylabel('Density')
plt.show()

Another example of how to create unique + eye-catching cyberpunk styled charts in a situation where you want to visualize time series data (chart created using dummy data) – the glow effect available in this package lends itself very well to time series and percent change charts in particular:

The Cyberpunk package works best with line charts, but here’s an example of how we can use it to visualize the correlation of Penguin stats in a scatterplot, which still looks pretty cool!

# Create the plot with the cyberpunk style
plt.style.use('cyberpunk')
fig, ax = plt.subplots(figsize=(8, 6))
scatter = ax.scatter(penguins['flipper_length_mm'], penguins['body_mass_g'], s=30,
                     c=penguins['island'].astype('category').cat.codes, cmap='cool')

# Enhance the plot with cyberpunk style glow effect
mplcyberpunk.make_scatter_glow(ax)

# Set the title and labels
ax.set_title("Penguin Stats- Flipper Length vs Body Mass By Island")
ax.set_xlabel("Flipper Length (mm)")
ax.set_ylabel("Body Mass (g)")
plt.show()

In summary, this package is a great addition to your Python data viz tool kit to up-level boring + basic matplotlib charts. A large part of our job as data scientists and analysts is to engage our audience, and one way to do that is to create eye-catching charts. I’d love to hear if you try this out! Let me know if you give it a go ✨

Big shoutout to the creator of this package, Dominik Haitz. You can find his work on Github! 👇