--------------------------
MATLAB kernel for Jupyter
--------------------------

To run MATLAB code within Jupyter notebooks, set up the MATLAB kernel in your
environment.

Prerequisites
=============

- Ensure you have **MATLAB R2025b** installed (required for Python 3.12+
  compatibility).
- Your conda environment should be using **Python 3.12**.

Installation steps
==================

1. Activate your environment:

   .. code-block:: bash

      conda activate neurowaves-lab-documentation

2. Install Jupyter and the MATLAB kernel:

   .. code-block:: bash

      conda install notebook jupyter_client
      pip install matlab_kernel

3. Install the MATLAB Engine for Python. Navigate to your MATLAB installation
   folder (for example ``C:\Program Files\MATLAB\R2025b\extern\engines\python``)
   and install the engine:

   .. code-block:: bash

      cd "C:\Program Files\MATLAB\R2025b\extern\engines\python"
      python -m pip install .

Using the kernels
=================

Two kernels are available in Jupyter:

- **Matlab**: starts a new, independent MATLAB session for the notebook.
- **Matlab (Connection)**: connects to an already running MATLAB instance,
  allowing you to share the workspace.

To use the **Connection** kernel:

1. In your standalone MATLAB command window, run:

   .. code-block:: matlab

      matlab.engine.shareEngine

2. In Jupyter, select the **Matlab (Connection)** kernel. You can now access
   variables from your standalone MATLAB session using commands like ``who`` or
   simply by typing the variable name.