How to install multiple ipython 3.0 kernels (python 2.7, python 3.4, etc...) with anaconda under linux?

Question

Ipython 3.0 (Jupyter) allows to choose which kernel (python 2.7, python 3.4, etc...) to use when creating a new ipython notebook. How can I install multiple ipython notebook kernels under Continuum Anaconda?

Answer 1

You'll want to make separate conda environments for Python 2 and 3 (see info elsewhere on how to do this), with IPython installed in both of them. Then, in each environment, run:

ipython kernelspec install-self

This registers that kernel so IPython can see it from outside the environment.

If you want more kernelspecs for different environments, look at the files in ~/.ipython/kernels to see how to describe them.

Answer 2

install-self is being deprecated: https://github.com/jupyter/jupyter/issues/23

install-self
    [DEPRECATED] Install the IPython kernel spec directory for this Python.

Try

python -m ipykernel install --user --display-name "Python (current_env)"
source activate myenv
python -m ipykernel install --user --name myenv --display-name "Python (myenv)"

instead.

Reference: http://ipython.readthedocs.org/en/stable/install/kernel_install.html