Installing Jupyter

Sometimes Jupyter may be required to run a Notebook kernel. In such circumstances VS Code will attempt to locate a Python environment that has Jupyter installed in it and can start a notebook.

The first step is to verify you are running the Python environment that you have Jupyter installed into.

If you would like to change from using the saved Python interpreter to a new interpreter for launching Jupyter just use the "Python: Select interpreter to start Jupyter server" VS Code command to change it.

The second step is to check that jupyter isn't giving any errors on startup.

Run the following command from an environment that matches the Python you selected: python -m jupyter notebook --version

If this command shows any warnings, you need to upgrade or fix the warnings to continue with this version of Python. If this command says 'no module named jupyter', you need to install Jupyter.

How to install Jupyter

You can do this in a number of different ways:

Anaconda

Anaconda is a popular Python distribution. It makes it super easy to get Jupyter up and running.

If you're already using Anaconda, follow these steps to get Jupyter

1. Start anaconda environment
2. Run 'conda install jupyter'
3. Restart VS Code
4. Pick the conda version of Python in the python selector

Otherwise you can install Anaconda and pick the default options https://www.anaconda.com/download

Pip

You can also install Jupyter using pip.

1. python -m pip install --upgrade pip
2. python -m pip install jupyter -U --force-reinstall
3. Restart VS Code
4. Pick the Python environment you did the pip install in

For more information see http://jupyter.org/install