Customizing Plots

By the end of this lesson, you will be able to customize line colors, styles, and markers in Matplotlib plots using format strings and keyword arguments.

What it is

In data visualization, "customizing plots" refers to altering the visual appearance of graphical elements to improve readability or match specific design requirements. In Python's matplotlib.pyplot, this is primarily achieved through two mechanisms: format strings (a compact shorthand) and keyword arguments (explicit parameters). A format string combines color codes, line styles, and marker symbols into a single string argument. Related terms include line width, alpha transparency, and legend labels.

Why it matters

• Differentiation: Distinct colors and line styles help viewers distinguish between multiple datasets on the same axes.
• Accessibility: Proper contrast and pattern usage ensure plots are readable for individuals with color vision deficiencies.
• Aesthetics: Clean, consistent styling makes reports and presentations look professional.
• Emphasis: Highlighting specific trends (e.g., making a critical line thicker or red) directs viewer attention effectively.

Syntax or steps

The most common method uses the third argument in plt.plot(), known as the format string. The syntax follows the pattern [color][marker][line]. For example, "r--" means red dashed line. Alternatively, you can use explicit keywords like color='red', linestyle='dashed', and linewidth=2. Keywords offer more control but are more verbose.

Example

import matplotlib.pyplot as plt
import numpy as np

# Generate sample data
x = np.linspace(0, 10, 50)
y1 = np.sin(x)
y2 = np.cos(x)

# Create figure
plt.figure(figsize=(8, 6))

# Plot 1: Using format string (Red, Dashed)
plt.plot(x, y1, "r--", label="Sine (Format String)")

# Plot 2: Using keyword arguments (Blue, Solid, Thick, Circle Markers)
plt.plot(x, y2, color="blue", linestyle="-", linewidth=3, 
         marker="o", markersize=4, label="Cosine (Keywords)")

# Add legend and title
plt.legend()
plt.title("Customized Line Styles")
plt.xlabel("X Axis")
plt.ylabel("Y Axis")

# Show plot
plt.show()

Explanation: The first plot uses "r--" to quickly set a red dashed line. The second plot uses explicit keywords: color="blue" sets the hue, linewidth=3 increases thickness, and marker="o" adds circle points at each data interval. The label parameter allows these lines to appear in the legend automatically.

Common mistakes

• Invalid Format Strings: Writing "red--" instead of "r--" causes an error. Use single-letter codes for colors in format strings.
• Overcrowding Markers: Adding markers to every point in a dense dataset creates visual noise. Use markevery=n to show markers only every nth point.
• Low Contrast: Using light colors on white backgrounds or similar hues for different lines makes data hard to read. Always check accessibility.
• Forgetting Labels: Customizing styles without adding label and calling plt.legend() leaves viewers guessing which line represents what.

When to use it

Compare format strings with keyword arguments based on your needs:

Practice

Guided Exercise: Modify the example above to plot a third line representing y3 = x/10. Use a green dotted line with square markers. Ensure it appears in the legend.

Challenge: Change the sine wave (y1) to have a semi-transparent blue fill under the curve using plt.fill_between(), while keeping its line style distinct.

Hint: For the guided exercise, try plt.plot(x, y3, "gs:", label="Linear"). Note that : is dotted and s is square.

Quick check

Question: What does the format string "k^" represent?

Answer: It represents a black (k) line with triangle-up (^) markers. Since no line style character is included, it defaults to solid, but often used just for markers if combined with linestyle='none'.

Summary

Customizing plots in Matplotlib balances brevity and precision. Use format strings for quick iterations and keyword arguments for detailed, publication-quality styling. Consistent use of color, line weight, and markers significantly enhances data interpretation.