Conda Environments

We recommend that you use conda environment to package Python dependencies inside notebook sessions.

Each conda environments that you create in notebook session can correspond to a different notebook kernel in JupyterLab. Use conda environments to run notebooks in different kernels. Each kernel has a set of Python libraries associated with it. The base install has a minimal set of libraries installed. The service is designed to use conda environments.

The notebook session environment includes the odsc conda CLI tool and the conda Environment Explorer.

The odsc conda CLI tool allows you to install, browse, search, and publish conda environments. You can access the odsc conda CLI documentation by executing odsc conda -h in a terminal window tab of a notebook session.

The Environment Explorer in JupyterLab helps you browse and search conda environments.

Although the conda CLI is available in a notebook session, we recommend that you use odsc conda to browse, install, clone, publish, and delete conda environments. preinstalled in notebook sessions and is available in a terminal window tab. The odsc conda CLI installs the necessary dependencies in a conda to make it available as a kernel in JupyterLab and creates the required manifest file that's necessary for each conda environment.

The Python3 conda environment is preinstalled in the notebook session. This conda environment is a Python 3 based conda environment and has a minimal set of libraries installed. We recommend that you install at least one Data Science conda environment or create your own.