Bokeh

3.9.0 · active · verified Thu Apr 09

Bokeh is an interactive visualization library for modern web browsers. It helps create versatile, interactive plots, dashboards, and data applications from Python, targeting web browsers for rendering. It is currently at version 3.9.0 and typically releases minor versions every few months with major versions less frequently.

Warnings

breaking Bokeh 3.x introduced significant breaking changes, most notably the removal of `output_file()`, `output_notebook()`, and `output_server()`. The `show()` function now handles output implicitly based on the environment (e.g., Jupyter, script).
Fix: Remove explicit `output_file`/`output_notebook` calls. Rely on `show()` to display the plot. For Bokeh server applications, run `bokeh serve your_app.py`.
gotcha For interactive features like callbacks, linked panning, or dynamic updates, data should be stored in a `ColumnDataSource`. Direct Python lists or NumPy arrays are static and will not update interactively without manual re-rendering.
Fix: Always wrap your data in `ColumnDataSource(data=...)` and reference columns by string names (e.g., `p.line('x', 'y', source=source)`).
gotcha Building Bokeh server applications (`bokeh serve`) requires specific structuring using `curdoc()` and callbacks, and cannot be run like a standard Python script that calls `show()` at the end. The server continuously updates the document.
Fix: Structure your application logic to add components to `curdoc().add_root()` and define callbacks using `curdoc().on_change()` or `source.on_change()`. Run with `bokeh serve your_app.py`.
deprecated The `gridplot` function is largely superseded by the more flexible `layout`, `row`, and `column` functions for arranging multiple plots or widgets.
Fix: Use `from bokeh.layouts import row, column, layout` and compose your plots/widgets using these functions instead of `gridplot`.

Install

pip install bokeh Install Bokeh

Imports

figure
```
from bokeh.plotting import figure
```
show
```
from bokeh.plotting import show
```

ColumnDataSource

from bokeh.models import ColumnDataSource

curdoc
```
from bokeh.io import curdoc
```
`curdoc` was moved to `bokeh.io` in Bokeh 3.x.

Quickstart

This quickstart creates a simple interactive sine wave plot and displays it in your default web browser. For more complex interactions or live updates, consider using Bokeh server.

from bokeh.plotting import figure, show
from bokeh.models import ColumnDataSource
import numpy as np

x = np.linspace(0, 4*np.pi, 100)
y = np.sin(x)

source = ColumnDataSource(data=dict(x=x, y=y))

p = figure(width=600, height=400, title="Simple Sine Wave",
           x_axis_label='x', y_axis_label='y')
p.line('x', 'y', source=source, line_width=2)

show(p)

view raw JSON →