Installing Jupyter for Matlab

This guide assumes you have Matlab already installed. I’ll be assuming you have Matlab R2019_b installed.

To use Jupyter notebooks with Matlab, you need to install Jupyter (which relies on Python) and the Matlab engine for Python.

macOS/Linux

  1. I recommend you install Anaconda to manage your Python environment—it makes installing and managing packages very easy. This comes with the Jupyter notebook.
  2. Create and activate an environment called jmatlab for Jupyter with Python 3.7:
$ conda create -vv -n jmatlab python=3.7 jupyter
$ conda activate jmatlab

Before the second command, you may need to tell your shell (e.g., bash, zsh) about conda, by doing conda init zsh for example. If you get an error with the conda activate command, your terminal should tell you to do this.

You should activate this environment whenever you want to run Jupyter.

  1. Install the Matlab kernel for Jupyter:
$ pip install matlab_kernel
  1. Install the Python engine for Matlab:
$ cd /Applications/MATLAB_R2019b.app/extern/engines/python
$ python setup.py install

If you are using a different version of Matlab, then that path will be different.

  1. Run Jupyter notebook:
$ jupyter notebook

and create a new Matlab notebook with “New” then “Matlab” under “Notebook:”.

Windows

  1. Install Anaconda for Windows
  2. Open the Anaconda Prompt

(to be continued)

Try it out

Now try it out:

[2]:

disp('This is running Matlab!')

This is running Matlab!