Bokeh at a Glance


Flexible

Bokeh makes it simple to create common plots, but also can handle custom or specialized use-cases.

Interactive

Tools and widgets let you and your audience probe “what if” scenarios or drill-down into the details of your data.

Shareable

Plots, dashboards, and apps can be published in web pages or Jupyter notebooks.

Productive

Work in Python close to all the PyData tools you are already familiar with.

Powerful

You can always add custom JavaScript to support advanced or specialized cases.

Open Source

Everything, including the Bokeh server, is BSD licensed and available on GitHub.


Common Scenarios


Applications

Build Powerful Data Applications

Python has an incredible ecosystem of powerful analytics tools: NumPy, Scipy, Pandas, Dask, Scikit-Learn, OpenCV, and more.

With a wide array of widgets, plot tools, and UI events that can trigger real Python callbacks, the Bokeh server is the bridge that lets you connect these tools to rich, interactive visualizations in the browser.

Dashboards

Publish Sophisticated Dashboards

Data scientists and developers appreciate Bokeh’s powerful APIs. But when publishing results for a wider audiences, what matters is the ability to generate clean, easy-to-understand presentations.

Bokeh offers its own basic grid and row/column layouts that make getting started a snap. When you need slick, reponsive dashboards, it’s also possible to embed Bokeh plots and widgets into popular templates.

Exploration

Interactively Explore Data in Notebooks

Bokeh works in both JupyterLab as well as classic notebooks.

Sophisticated interactive visualizations to use alongside your notebook explorations are only a call to output_notebook away—and that includes full embedded Bokeh server applications.

Try things out right now with the live tutorial notebooks hosted generously by MyBinder.

Streaming

Visualize Streaming Data

Whether you are streaming data from financial markets, IOT telemetry, or physical sensors, Bokeh has efficient streaming APIs to help you keep on top of things. In a Bokeh server application, it is as simple as passing your new data values to a stream method:

source.stream({'x': new_xs, 'y': new_ys})

But standalone Bokeh output can handle streaming data too, using either the AjaxDataSource or the ServerSentDataSource.

Websites

Add Content to Web Pages

Maybe you’ve created a Flask or Django web app that needs to include reporting charts. Or maybe you’ve written an article for a Jekyll blog that needs some infographics to illustrate your point.

Bokeh offers a variety of methods to embed its content in web pages: server_document for deployed Bokeh server applications, or json_items and components for standalone Bokeh output.



User Showcase

Dask

Dask

Dask is a tool for scaling out PyData projects like NumPy, Pandas, Scikit-Learn, and RAPIDS. It is supported by Nvidia, Quansight, and Anaconda.

The Dask Dashboard is a diagnostic tool that helps you monitor and debug live cluster performance.

Microscopium

Microscopium

Microscopium is a project maintained by researchers at Monash University.

It allows researchers to discover new gene or drug functions by exploring large image datasets with Bokeh’s interactive tools.

Panel

Panel

Panel is a tool for polished data presentation that utilizes the Bokeh server. It is created and supported by Anaconda.

Panel makes it simple to create custom interactive web apps and dashboards by connecting user-defined widgets to plots, images, tables, or text.

Chartify

Chartify

Chartify is an opinionated high-level charting API built on top of Bokeh, created by Spotify.

With smart default styles, consistent tidy data format, and a simple API, it’s easy for you to concentrate on your work.

Mistic

Mistic

Mistic is a software package written in Python and uses the visualization library Bokeh.

Mistic can be used to simultaneously view multiple multiplexed 2D images using pre-defined coordinates (e.g. t-SNE or UMAP), randomly generated coordinates, or as vertical grids to provide an overall visual preview of the entire multiplexed image dataset.

ArviZ

ArviZ

ArviZ is a community-led package for exploratory analysis of Bayesian models in Python.

It Includes functions for posterior analysis, data storage, sample diagnostics, model checking, and comparison. The goal is to provide backend-agnostic tools for diagnostics and visualizations of Bayesian inference.


Where to Connect