Jupyter Interactive Widgets (ipywidgets)

8.1.8 · active · verified Sat Mar 28

ipywidgets is a Python library that provides interactive HTML widgets for Jupyter notebooks and the IPython kernel. It enables users to create interactive controls like sliders, text boxes, and buttons, bringing notebooks to life and allowing interactive exploration of data and models. The current version is 8.1.8, and it maintains an active development cycle with regular patch and minor releases, and major versions typically every few years.

Warnings

breaking The `FileUpload` widget API changed significantly in version 8.x. The `.data` and `.metadata` traits were removed, and the `.value` trait was revamped to be a list of dictionaries, each containing file information (e.g., `content`, `name`, `type`, `size`, `last_modified`).
Fix: Rewrite usage of `FileUpload` to access file content and metadata through the new `.value` structure. For example, to get content bytes: `[f.content.tobytes() for f in uploader.value]`.
breaking The `description_tooltip` attribute, used for tooltips on some widgets, was deprecated in favor of a universal `tooltip` attribute available on all widgets inheriting `DOMWidget` in version 8.x.
Fix: Replace `description_tooltip` with `tooltip` in your widget instantiations and property assignments.
breaking The `description` field of most widgets now sanitizes HTML content by default. If you rely on custom HTML in descriptions, it will be stripped or rendered as plain text.
Fix: Set `description_allow_html=True` explicitly for the widget if you are in full control of the HTML content and require it to be rendered.
breaking The method for setting titles on container widgets like `Accordion` or `Tab` changed. Direct calls to `container.set_title(index, title)` can cause `IndexError` if not handled correctly. It is now recommended to initialize with a `titles` tuple or assign to the `_titles` trait.
Fix: When initializing `Accordion` or `Tab`, provide titles using the `titles` argument (e.g., `Accordion(children=[...], titles=('Title 1', 'Title 2'))`) or assign to the `_titles` trait after creation.
gotcha Direct `print()` statements within functions linked to widgets (especially with `@interact` or `widget.observe()`) may not always display reliably or may appear in the console instead of the notebook output area in newer Jupyter environments.
Fix: Use an `ipywidgets.Output` widget as a context manager to capture and display output reliably. Wrap `print()` calls within a `with output_widget:` block and use `clear_output()` for controlled updates.

Install

pip install ipywidgets Install ipywidgets

Imports

ipywidgets
```
import ipywidgets as widgets
```
interact
```
from ipywidgets import interact
```
display
```
from IPython.display import display
```
Button
```
from ipywidgets import Button
```
VBox
```
from ipywidgets import VBox
```
Output
```
from ipywidgets import Output
```
Output is a class, 'out' is often used as an instance variable name.

Quickstart

This quickstart demonstrates creating an integer slider, observing its value changes, and displaying updates in a dedicated output area within a Jupyter notebook. It highlights the use of `ipywidgets.IntSlider` and `IPython.display.display` along with `ipywidgets.Output` for controlled output handling.

import ipywidgets as widgets
from IPython.display import display, HTML, clear_output

# Create a simple IntSlider widget
slider = widgets.IntSlider(
    min=0,
    max=100,
    step=1,
    description='Value:',
    value=50
)

# Create an Output widget to capture print statements
output = widgets.Output()

# Define a function to be called when the slider's value changes
def on_value_change(change):
    with output:
        clear_output()
        print(f"Slider value changed to: {change['new']}")

# Observe changes in the slider's value
slider.observe(on_value_change, names='value')

# Display the slider and the output area
display(HTML("<h3>Interactive Slider Example</h3>"))
display(slider, output)

view raw JSON →