.. _jupyter_tutorial:

+------------------------------------+---------------------------------------+
| Author email                       | sebastien.weber@cemes.fr              |
+------------------------------------+---------------------------------------+
| Last update                        | october 2024                          |
+------------------------------------+---------------------------------------+
| Difficulty                         | Easy                                  |
+------------------------------------+---------------------------------------+

Jupyter Lab
===========

JupyterLab is a well known IDE giving access (among other things to
notebooks). Let's install it for practising with the PyMoDAQ's notebooks tutorials__ .

__ https://github.com/PyMoDAQ/notebooks

JupyterLab can easily be installed with conda or pip. Create a new
environment, for instance named *jupyter* and launch the command:

   ``(jupyter) conda install jupyterlab nodejs``

or

   ``(jupyter) pip install jupyterlab nodejs``

Done? You can enter the command: ``jupyter lab`` in your activated
prompt to start it.

.. figure:: /image/jupyter/image61.png
   :alt: python10

   Starting Jupyter lab

In fact by doing so, you’ve started a web server. You can then use a
client, your favorite web browser, to access it. The address of the
server is written in the prompt

.. figure:: /image/jupyter/image62.png
   :alt: python10

   The start page of *jupyter lab* in our browser

Our default web browser is opening into a page starting with *localhost*
that is an internal server. The file browser points to the directory
where we started the server.

To ease the use of this, let's create a shortcut with a *starting* field
pointing to the location of the jupyter notebooks (for instance
``C:\Users\formateur\Documents\pre_formation_pymodaq``)

.. figure:: /image/jupyter/image63.png
   :alt: python10

   Shortcut creation

The jupyter icon can be found in it's environment (we named it
*jupyter*) that could be something like:
``C:\Users\formateur\miniconda3\envs\jupyter\Menu\jupyter.ico``

Rename your shortcut and let's configure jupyter lab to use the python
interpreter from our *form_pymodaq* environment.

To do so, open a miniconda prompt and activate your environment.

Type in

   ``(form_pymodaq) pip install ipykernel``

..

   ``(form_pymodaq) python -m ipykernel install --user --name form_pymodaq --display-name pymodaq``

.. figure:: /image/jupyter/image65.png
   :alt: python10

   Installing and registering our environment as a jupyter kernel

you'll now be able to start a notebook in jupyter lab using the kernel
(the interpreter) of your *form_pymodaq* environment, below is the one
we simply named *pymodaq*.

.. figure:: /image/jupyter/image66.png
   :alt: python10

   Selecting one of the installed kernel (interpreter)

And to open a notebook and configure it using the right kernel:

.. figure:: /image/jupyter/image67.png
   :alt: python10

   Start a notebook with a given kernel

Ok you’re ready to go!