nbagg

nbagg, short for Notebook Agg, is a backend for Matplotlib designed to display interactive plots inside Jupyter notebooks. It renders figures in the notebook frontend by communicating between the Python kernel and a JavaScript-based browser interface, allowing user interactions to be reflected in the displayed figure.

The backend uses the Agg rendering engine for drawing while providing interactive capabilities such as pan, zoom, and event handling. When enabled, user actions in the notebook (for example, dragging to pan or scrolling to zoom) are sent to Python code, which can update the figure in real time. The overall workflow relies on the notebook’s communication protocol to synchronize drawing commands and figure state between the kernel and the browser.

To use nbagg, users typically activate it with a magic command or backend switch, for example by running %matplotlib nbagg or matplotlib.use('nbagg') in a notebook cell. It is compatible with the classic Jupyter Notebook interface, but support in newer environments such as JupyterLab has been limited or requires additional extensions.

In practice, nbagg provides an early, browser-based interactive plotting experience within notebooks, but it has been largely superseded by newer solutions. Modern workflows often prefer ipympl, the Matplotlib Jupyter widget, or the WebAgg backend for more robust integration and broader environment support. nbagg remains a reference point in the evolution of interactive plotting in Jupyter notebooks.