Multi-figure Layouts with Matplotlib

Welcome to our lesson on multi-figure layouts! Visualizing data effectively is crucial, and the ability to present multiple plots within a single figure is a valuable skill. This approach not only simplifies the representation of complex data but also helps in showcasing multiple relationships at once.

By grouping related plots, you can weave a compelling narrative, conveying intricate stories clearly and concisely. In this lesson, we'll explore how to leverage Matplotlib to craft these layouts, equipping you with the skills to enhance your data storytelling.

To create multi-figure layouts in Matplotlib, you'll primarily use the plt.subplots() function. This function helps you design organized layouts by letting you pick the number of rows and columns in your figure. When you use plt.subplots(), you get a figure and some axes. Each of these axes is a different subplot, giving you the freedom to customize each plot.

The plt.subplots() function helps you create organized layouts by defining the number of rows and columns for your plots.

Here's how to use it:

import matplotlib.pyplot as plt

# Create subplots with specified rows and columns
fig, ax = plt.subplots(n_rows, n_columns)

Replace n_rows and n_columns with numbers to set up your grid size, like plt.subplots(2, 3) for two rows and three columns.

The function returns two objects:

• fig: Represents the entire figure containing all plots.
• ax: An array that holds each individual subplot.

To customize each subplot, use indexes, similar to a list: ax[0, 0] for the first row, first column plot. You can then use methods like .hist(), .scatter(), .set_title(), .set_xlabel(), and .set_ylabel() to modify each plot as required. This approach allows you to arrange multiple plots in one figure for easier data comparison and analysis.

Here's a simple example using the Penguins dataset where we'll place two plots side by side for easy comparison:

import matplotlib.pyplot as plt
import seaborn as sns

# Load the Penguins dataset
penguins = sns.load_dataset('penguins')

# Create a horizontal layout with two subplots
fig, ax = plt.subplots(1, 2, figsize=(8, 4))

# First subplot: Histogram of bill lengths
ax[0].hist(penguins['bill_length_mm'], color='blue')
ax[0].set_title('Bill Length Distribution')
ax[0].set_xlabel('Bill Length (mm)')
ax[0].set_ylabel('Frequency')

# Second subplot: Scatter plot of flipper length vs. body mass
ax[1].scatter(penguins['flipper_length_mm'], penguins['body_mass_g'], color='red')
ax[1].set_title('Flipper Length vs Body Mass')
ax[1].set_xlabel('Flipper Length (mm)')
ax[1].set_ylabel('Body Mass (g)')

# Adjust and display layout
plt.tight_layout()
plt.show()

In this example, ax[0] represents the first subplot on the left (the histogram), and ax[1] represents the second subplot on the right (the scatter plot). The use of plt.tight_layout() is crucial here as it automatically adjusts the spacing between the plots, ensuring they fit nicely within the figure without overlapping.

The horizontal arrangement allows simultaneous examination of the histogram and scatter plot for effective analysis.

If you want the plots stacked one above the other, try this vertical layout:

# Create a vertical layout with two subplots
fig, ax = plt.subplots(2, 1, figsize=(4, 6))

# First subplot: Histogram of bill lengths
ax[0].hist(penguins['bill_length_mm'], color='blue')
ax[0].set_title('Bill Length Distribution')
ax[0].set_xlabel('Bill Length (mm)')
ax[0].set_ylabel('Frequency')

# Second subplot: Scatter plot of flipper length vs. body mass
ax[1].scatter(penguins['flipper_length_mm'], penguins['body_mass_g'], color='red')
ax[1].set_title('Flipper Length vs Body Mass')
ax[1].set_xlabel('Flipper Length (mm)')
ax[1].set_ylabel('Body Mass (g)')

# Adjust and display layout
plt.tight_layout()
plt.show()

With this configuration, we've changed the subplot arrangement to be vertical by using plt.subplots(2, 1), which stacks the plots one on top of the other. Adjusting the figsize to (4, 6) ensures that the vertical layout is well-suited for this orientation, providing ample space for each plot.

Here, ax[0] represents the histogram on the top, while ax[1] shows the scatter plot below, offering a sequential view that facilitates a clear and comprehensive analysis.

For a more comprehensive visual storytelling, consider arranging your plots in a 2x2 grid. This approach utilizes two dimensions to effectively structure your visualizations, even allowing for unused spaces to maintain balance.

# Create a 2x2 grid layout to fit three plots
fig, ax = plt.subplots(2, 2, figsize=(8, 6))

# First subplot: Histogram of bill lengths
ax[0, 0].hist(penguins['bill_length_mm'], color='blue')
ax[0, 0].set_title('Bill Length Distribution')
ax[0, 0].set_xlabel('Bill Length (mm)')
ax[0, 0].set_ylabel('Frequency')

# Second subplot: Scatter plot of flipper length vs. body mass
ax[0, 1].scatter(penguins['flipper_length_mm'], penguins['body_mass_g'], color='red')
ax[0, 1].set_title('Flipper Length vs Body Mass')
ax[0, 1].set_xlabel('Flipper Length (mm)')
ax[0, 1].set_ylabel('Body Mass (g)')

# Third subplot: Line plot of flipper lengths
ax[1, 0].plot(penguins.index, penguins['flipper_length_mm'], color='green')
ax[1, 0].set_title('Flipper Lengths Over Index')
ax[1, 0].set_xlabel('Index')
ax[1, 0].set_ylabel('Flipper Length (mm)')

# Set empty plot (ax[1, 1]) to keep layout balanced
ax[1, 1].axis('off')

# Adjust and display layout
plt.tight_layout()
plt.show()

In this setup, we use plt.subplots(2, 2) to create a grid that positions subplots across two dimensions, allowing for varied visualization arrangements. The figsize is set to (8, 6), optimizing space to accommodate the plots clearly.

To maintain a balanced appearance while only placing three plots, ax[1, 1] is intentionally left empty, and the axis is turned off using ax[1, 1].axis('off'). This ensures that the layout remains clear and visually appealing, avoiding clutter and making it easier to focus on the data presented.

In this lesson, you have explored how to create and customize multi-figure layouts in Matplotlib using the Penguins dataset. You’ve learned about structuring layouts with plt.subplots() and customizing subplots to deliver clear and informative visualizations. As you prepare to tackle the practice exercises, I encourage you to experiment with different datasets and plot types to strengthen your understanding.

Congratulations on reaching the end of this course! You’ve developed a comprehensive skill set in data visualization, from color customization to advanced multi-figure layouts. As you apply these skills in your projects, remember the power of visualization lies in its ability to convey complex data insights clearly and effectively. Keep experimenting and refining your visual storytelling techniques.