Install packages at runtime
When your environment lacks required packages, you can install them dynamically using either mamba or pip commands directly in a notebook cell or terminal.
Installation methods
Mamba installation
!mamba install pandas
Pip installation
!pip install bqplot
Supported features
The following features are supported:
- Conda package installation from channels:
conda-forgeemscripten-forge(use!mambacommand)
- Pure Python packages from
PyPI(use!pipcommand) - Persistent installations: Installed packages remain available between sessions
tip
Notebook.link supports dynamic loading of JupyterLab extensions! For example:
!mamba install ipyleaflet
from ipyleaflet import Map
Map(center=(52, 10), zoom=5)
Environment reset
To return to the original environment state:
- Open the
Project Panel(cloud icon in the left sidebar) - Select the reset option
For detailed instructions, refer to the Storage section in Project panel documentation.