Installation

First clone this repository on your machine.
In the terminal in the folder you desire do the following:

Please ask if you don't know how to use the terminal.

git clone https://github.com/ai4er-cdt/induction-jupyter.git

pip vs conda

pip and conda are package manager for Python. They allow to install packages.

pip

If you want to use pip, make sure you have the latest pip

pip3 install --upgrade pip

(Using pip instead of pip3 should work if you don't have Python 2.)

Conda / Mamba

Anaconda is another package and environment manager.

The other option is mamba

mamba: a Python-based CLI conceived as a drop-in replacement for conda, offering higher speed and more reliable environment solutions

I prefer mamba as it is lighter.

how to use them

When doing a project it is better to keep to one package manager to install everything. Of course, it will come a time when your project in conda needs a package that's only available through pip. In this case it is fine to install it with pip, but try to do it for every other packages.
What I like to do is install all the pip dependencies at the end.

virtual envs

I would highly recommend using virtual environment, as it packages all the dependencies for a project and avoid having a mess of different version.

Note: You do not need to this on Google Colab as it creates a Linux instance every time you launch it, but more on that later.

venv

I personally have never used the venv to create a virtual environment, but you can read up on it here.

Conda

I have only used the conda/mamba virtual environment manager as I prefer it. For this workshop you can simply do the following:

conda env create -f environment.yaml

conda activate ai4er-ws

Now this is the explanation if you wanted to do it from scratch.

conda create -n $name python=3.11

Where $name is the name of the environment (do NOT use spaces).

Then you will need to activate the environment:

conda activate $name

Installing Jupyter

pip

pip3 install jupyter

Conda

conda install jupyter

Opening the notebook

Start a server

To use jupyter notebook you need to first launch the jupyter server.
In a terminal (with the correct environment activated):

jupyter notebook

(To stop simply type Ctrl + c in the terminal).

Now open your browser and go to http://localhost:8888

If you are using Colab will do all that for you.

Local vs colab

You can either run jupyter locally on your computer, or you can use Google Colab. However, Colab comes with some caveat, you are limited in runtime, it now runs Python 3.10 (no easy way to change it), and the terminal access is granted through the Pro version.