Skip to content

Custom Conda Environment in JupyterLab

This tutorial walks you through creating a new conda environment and making it available as a selectable kernel in JupyterLab.

Step 1: Start a JupyterLab Session and Open a Terminal

Start your JupyterLab session.

image

Open a new Terminal.

image

Step 2: Create a New Conda Environment

In the terminal, create a new conda environment. Replace myenv with your preferred environment name and specify your desired Python version. If prompted to proceed with installing new packages, type y.

conda create -n myenv python=3.11

💡 Tip: You can name your environment anything you like. Using a descriptive name helps when managing multiple environments.

Step 3: Activate the Environment

conda activate myenv

Your terminal prompt should update to reflect the active environment, e.g., (myenv).

image

Step 4: Install ipykernel and Other Packages

With the environment active, install the ipykernel package (required) along with any other packages needed in your conda environment, replacing the example packages below with whatever your project requires:

pip install ipykernel numpy pandas matplotlib

💡 Tip: Whenever you need to install additional packages into this environment in the future, open a Terminal in JupyterLab, activate the environment with conda activate myenv, and then run your pip install or conda install command.

Step 5: Register the Environment as a Jupyter Kernel

Run the following command to make your conda environment visible as a selectable kernel in JupyterLab:

python -m ipykernel install --user --name myenv --display-name "Python(myenv)"
Flag Description
--user Installs the kernel for the current user only
--name Internal name used to identify the kernel
--display-name The name shown in the JupyterLab UI

Step 6: Select Your Kernel and Start Working

After a moment, open a new tab in JupyterLab. In the Notebook section of the Launcher, you should now see your new environment listed as a kernel option alongside the default:

  • Python 3 (ipykernel) (default)
  • Python(myenv) ✅ (your new environment)

To start a new notebook using your environment, click the Python(myenv) icon in the Launcher.

To switch the kernel in an existing notebook, click the kernel name in the top right corner (next to the small bug icon) and select Python(myenv) from the dropdown.

image

Summary

conda create -n myenv python=3.11
conda activate myenv
pip install ipykernel numpy pandas matplotlib
python -m ipykernel install --user --name myenv --display-name "Python(myenv)"

Once registered, your environment will persist as a kernel option in JupyterLab until you explicitly remove it.

To install additional packages in the future:

conda activate myenv
pip install package_name