Use Multiple Columns in a Matplotlib Legend

Legends helps to understand what each element in the plot represents. They help to understand the meaning behind different elements like colors, markers or line styles. If a plot contains many labels a single-column legend may:

• Take up too much vertical space
• Overlap with the plot and
• Look messy or cluttered.

By arranging legend items into multiple columns we can make our plot more compact and visually appealing. matplotlib makes it easy to create multi-column legends and helps us improve the readability and layout of our plots and helps in organizing them more effectively.

Step-by-Step Guide to Create Multi-Column Legends

Step 1: Import Required Libraries

To get started ensure we have Matplotlib installed. If not install it using pip:

pip install matplotlib

Next import the necessary libraries like numpy and pandas :

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

Step 2: Create a Plot with Multiple Lines

Let’s begin by creating a simple plot with multiple lines. We create 4 different lines: sine, cosine, tangent and exponential decay.

• label= "argument" gives each line a name for the legend.
• plt.legend(ncol=2) splits the legend into 2 columns.
• loc='upper right' places the legend in the top-right corner.

x = np.linspace(0, 10, 100)
y1 = np.sin(x)
y2 = np.cos(x)
y3 = np.tan(x)
y4 = np.exp(-x)

plt.plot(x, y1, label='Sine')
plt.plot(x, y2, label='Cosine')
plt.plot(x, y3, label='Tangent')
plt.plot(x, y4, label='Exponential')

Output :

The above plot effectively displays four distinct mathematical functions - Sine, Cosine, Tangent and Exponential on the same graph. The use of a 2-column legend minimizes vertical space and keeps the layout tidy.

Step 3: Customize the Legend Appearance

We can further customize the legend to suit our needs:

• Font Size : Adjust the font size using the fontsize parameter.
• Border and Background : Modify the border and background using frameon, edgecolor and facecolor.
• Spacing : Control spacing between columns using columnspacing.

Example with additional customization:

plt.figure(figsize=(8, 5))
plt.plot(x, y1, label='Sine')
plt.plot(x, y2, label='Cosine')
plt.plot(x, y3, label='Tangent')
plt.plot(x, y4, label='Exponential')

plt.legend(
    ncol=3, 
    loc='upper center', 
    fontsize=10, 
    frameon=True, 
    edgecolor='black', 
    facecolor='lightgray',
    columnspacing=1.5
)

plt.title('Customized Multi-Column Legend')
plt.xlabel('X-axis')
plt.ylabel('Y-axis')
plt.grid(True)
plt.tight_layout()
plt.show()

Output:

This plot visualizes four mathematical functions (sine, tangent, cosine and exponential decay) with a customized multi-column legend placed at the top-center. The legend is styled with distinct colors and labels for each function enhancing readability while the gridlines provide context for the data values.

Step 4: Handle Long Labels and Move Legend Outside Plot

If our legend labels are long, multi-column legends can help prevent them from overlapping. We simulate long labels and shift the sine wave for variety.

• bbox_to_anchor=(0.5, -0.25) positions the legend below the chart and keeps the plot clear and uncluttered.

x = np.linspace(0, 10, 100)

labels = [
    'This is a very long label for sine',
    'Another long label for cosine',
    'Yet another one for tangent',
    'And the last one for exponential'
]

plt.figure(figsize=(10, 6))

for i, label in enumerate(labels):
    plt.plot(x, np.sin(x + i), label=label)

plt.legend(
    ncol=2, 
    loc='lower center', 
    bbox_to_anchor=(0.5, -0.25),  
    fontsize=9, 
    frameon=True,
    columnspacing=1.2
)

plt.title('Handling Long Labels with Multi-Column Legend')
plt.xlabel('X-axis')
plt.ylabel('Y-axis')
plt.grid(True)
plt.tight_layout()
plt.show()

Output:

This plot demonstrates four sine waves with slightly different offsets, each labeled with a long descriptive label. The legend is placed at the bottom of the plot in a multi-column layout to accommodate the lengthy labels ensuring clarity and readability while maintaining an organized appearance.

Using multi-column legends in Matplotlib is a simple yet powerful way to enhance the readability. Whether you're working with line plots, bar charts or scatter plots the ncol parameter allows us to organize our legend efficiently.