How To: Be Compatible with JupyterLab 4 and Notebook 7

Overview

Seeq Data Lab now leverages JupyterLab 4.x for its Advanced Mode and Notebook 7.x for its Notebook Mode interfaces. These updated versions bring improved performance, with the Notebook Mode featuring a redesigned user interface. Although these enhancements elevate the Data Lab experience, you may notice changes in how widgets are displayed, especially in Add-on Mode. Also, some Jupyter extensions may need to be upgraded or are not compatible.

Solution

JupyterLab is the base of Data Lab. All JupyterLab extensions and widgets need to be compatible with this version.

To get the precise version please run

pip show jupyterlab

Jupyter Extensions

To ensure optimal performance, all extensions, whether installed locally or globally, may not work as intended unless updated to JupyterLab 4-compatible versions. Refer to the extension's documentation to verify if an upgrade is necessary, and proceed with reinstalling it if needed. For guidance on installing Python packages locally and globally, visit Installing Python Modules. If a compatible version for JupyterLab 4 is not yet available, consider alternative extensions.

Widgets and Python Visualization Packages

The Python package ipywidgets has been updated to version 8.x in Data Lab's pre-installed packages for both Python 3.8 and 3.11 kernels, ensuring full compatibility with Add-on Mode. It's important to note that widget and Python interactive packages like ipywidgets and ipyvuetify might display differently when executed inline in a notebook or through Add-on Mode. To ensure seamless compatibility, it is recommended to test your code in both environments - notebook and Add-on Mode - and make any required adjustments accordingly.

Note: Please be aware that ipyvuetify is no longer pre-installed in the Python 3.11 kernel.

Unsupported Packages

Below is a list of Python visualization packages that are no longer supported in Data Lab due to JupyterLab 4 incompatibility. However, there exists a suitable alternative package.