Panel documentation#
Panel is an open-source Python library for building data apps, from two widgets beside a plot in a notebook to a multi-page application served behind your own authentication. The same objects do both. A Panel component renders in Jupyter, JupyterLab, VS Code, Colab and marimo, and panel serve turns the file it lives in into an application without a rewrite.
Updates are declarative rather than script-level. You bind a function, or a component’s state, to the widgets that feed it, and when one of those changes only what depends on it runs again, so an app that loads four million rows does not load them again because a slider moved. Panel renders what your plotting library already produces, whether that is Matplotlib, Plotly, Bokeh, HoloViews, Altair, ECharts, Deck.gl or an ipywidget, and the same app grows from a bound function into components made of components as it gets larger, which is refactoring rather than rewriting.
For a tour with runnable examples, see the Panel home page. This site is the documentation: guides, reference pages, and the API.
Learn#
Install Panel, build your first app, and get an overview of how its pieces fit together.
Guided, end-to-end walkthroughs that build a real application step by step.
Why Panel works the way it does: the reactive model, the two APIs, components, and design.
Look something up#
Every component that ships with Panel, one page each, with the code that produced the example.
Step-by-step recipes for the specific problems that come up: layout, auth, deployment, testing.
The classes, parameters and methods that make up Panel’s public interface.
Also here#
Complete applications, each editable and runnable in the browser.
What changed between major versions, and what to do about it.
The questions that come up often enough to have a written answer.
Where to ask, how to report a bug, and how to contribute.
For usage questions or technical assistance, head over to Discourse or our Discord server. If you have any issues, feature requests, or wish to contribute, you can visit our GitHub site.
Sponsors#
The Panel project is grateful for the sponsorship by the organizations and companies below:
