Unveiling the Hierarchies of Plots, Axes, and Figures in Matplotlib

The Conundrum of Matplotlib Plotting

Many programmers struggle to decipher the complexities surrounding the different techniques for creating plots in Matplotlib. The concepts of figure, axes, and plot can be confusing, leaving some uncertain about the underlying mechanisms. This article aims to clarify these distinctions, providing a comprehensive understanding of their roles and applications.

Deciphering the Objects

At the core of Matplotlib lies the figure, representing the canvas on which graphs are drawn. Similar to a painting canvas, the figure establishes dimensions, background colors, and other attributes. The axes is akin to a Swiss Army knife, offering tools for plotting, scattering, and histograms. Multiple axes can reside within a single figure.

The plt Interface: A User-Friendly Facade

The plt interface provides a simplified method for creating figures and axes, mirroring the MATLAB™ interface. It acts as a bridge between the user and the underlying objects. Every plt command internally translates into a call on the respective methods of these fundamental objects.

Illustrating the Differences

Let's delve into the three plot creation methods you provided:

1st Method (plt.plot):

Using only the plt interface, this method creates a single axes within a figure. While efficient for quick data explorations, its flexibility is limited.

2nd Method (plt.subplot):

Employing a convenience method from the plt namespace, this method assigns a name to the axes object. While it offers greater control over plot attributes, it still creates a single axes per figure.

3rd Method (figure.add_subplot):

This approach bypasses plt convenience methods and directly instantiates a figure using the object-oriented interface. It provides complete customization and control, but requires manual tweaking for interactive features.

Recommendations for Usage

For interactive data exploration, the bare plt.plot method proves efficient. For complex, customized subplots or embedding Matplotlib in a program interface, the object-oriented approach is preferred.

In conclusion, understanding the relationships between plots, axes, and figures in Matplotlib is crucial for effective graph creation. The choice of method depends on the specific use case, with plt offering simplicity and object-oriented programming providing customization and flexibility.