In this book, when we mention high-level, we are referring not to some assessment of value or improvement over something, but rather to layers of abstraction, or more precisely, layers of interaction. When engaged in high-level plotting, we expect that users and developers will be creating visualizations of complex data with fewer commands or steps required than by using matplotlib's basic functionality directly. This is a result of complex tasks wrapping a greater number of smaller, simpler tasks.

Plotting itself is a high-level activity: raw data and often calculations on that data are combined, processed, some more in anticipation of user consumption, arranged or grouped in ways suitable for conveying the desired information, and then applied to some medium in ways that one hopes will render greater insight. By our definition, each activity upon that original raw data is, in some way, high-level.

Before we look at examples of modern high-level plotting, let us gain some...