Data & Big Data

Practical AI Data Science Tools: The Code and Commands That Really Matter

AI Data Science Tools: The Essentials in One Article — Real Code, Diagrams and Concrete Steps, Extracts from a 43-Lesson Course.

REHOUMA Haythem

• 10 min read
Practical AI Data Science Tools: The Code and Commands That Really Matter

No endless theory here: open the terminal and practice. Here is the essentials of Data Science AI Tools, extracted directly from a complete 43-lesson course — with real code you can copy-paste right now.

tl;dr
  • Introduction and Setup
  • Jupyter Notebook and JupyterLab
  • Managing Python Environments
  • VS Code for Data Science
  • Git and GitHub for Data Scientists
~$ cat ./parcours.md # AI Data Science Tools — 10 chapters
01
Introduction and Setup
→ Course presentation and data toolkit→ Install Anaconda, VS Code and Git+ 1 more lessons
02
Jupyter Notebook and JupyterLab
→ Jupyter Notebook vs JupyterLab, choose your interface→ Cells, kernels and useful magic commands+ 2 more lessons
03
Python Environment Management
→ venv vs conda vs poetry, when to use what→ Create, activate and export conda environments+ 2 more lessons
04
VS Code for Data Science
→ Essential extensions, Python, Jupyter, Pylance→ Step-by-step Python debugging in VS Code+ 2 more lessons
05
Git and GitHub for Data Scientists
→ Basic Git, clone, add, commit, push, pull→ Branches, merges and pull requests on GitHub+ 2 more lessons
06
Cloud Platforms Colab and Kaggle
→ Google Colab, free GPU and limits→ Kaggle Notebooks and dataset access+ 2 more lessons
07
Docker for Reproducibility
→ Docker in 10 minutes for data scientists→ Dockerfile for a Python + Jupyter project+ 2 more lessons
08
Data and Model Versioning
→ Why Git alone is not enough for data→ DVC, Git for data+ 2 more lessons
🏁
Final project (+ 2 chapters along the way)
→ You leave with a concrete and demonstrable project

Create, activate and export conda environments

NOTEObjective — Master the full lifecycle of a conda environment: creation, activation, installation, export and recreation identical on another machine.

Learning objectives

TIPAt the end of this module
  • Create a conda environment with a specific Python version
  • Activate and deactivate an environment
  • Install packages from conda-forge
  • Export an environment to environment.yml
  • Recreate an environment from the file

Basic intuition: a recipe you can replay anywhere

A conda environment is like a recipe written in black and white. You list the ingredients (packages) and their quantities (versions) in an environment.yml file. Anyone, on any machine, can then recreate exactly the same dish by following the recipe. This is the heart of reproducibility.

The cycle is always the same: create, activate, install, export. The export produces the recipe; the recipe enables recreation elsewhere.

Create and activate

Export from-history

Keep only what you explicitly requested. More portable between Windows, macOS and Linux.

TIPTip: To share between different OSes, prefer --from-history. To guarantee reproduction down to the exact package on the same OS (CI, production), use the full export.

Recreate an environment elsewhere

On another machine, or after cloning a project, recreate the environment in one command from the file.

Install Anaconda, VS Code and Git

NOTEObjective — Properly install the three pillars of your data workshop (Anaconda, VS Code, Git) on your system, then verify that everything works before going further.

Learning objectives

TIPAt the end of this module
  • Install Anaconda (or Miniconda) on your OS
  • Install Visual Studio Code and its basic data extensions
  • Install and configure Git with your identity
  • Verify each installation with a control command
  • Choose between full Anaconda and lightweight Miniconda

Basic intuition: lay the foundations before building

Before writing a single line of data code, you must lay three foundations. Anaconda provides Python and an environment manager (conda). VS Code will be your main editor. Git will version your work. Once these three building blocks are in place, you have everything you need to start any project in the course.

Install once and verify immediately. Verification is essential: 90% of beginner problems come from an incomplete installation or a misconfigured PATH discovered too late.

Anaconda or Miniconda?

Anaconda and Miniconda both provide conda, the environment manager. The difference is the number of pre-installed packages.

CriterionAnacondaMiniconda
Download size~ 1 GB~ 70 MB
Included packages250+ (pandas, numpy, scikit, etc.)The bare minimum (conda + Python)
Graphical interfaceYes (Anaconda Navigator)No, command line
Ideal audienceComplete beginnerUser who wants control
TIPTip: To start this course, full Anaconda is simplest. If you are comfortable with the terminal and want lightweight environments, Miniconda is preferable. The conda commands are identical in both cases.

Install Anaconda according to your OS

Windows

Download the .exe installer from anaconda.com, run it, and check “Add to PATH” even if discouraged, then open “Anaconda Prompt”.

macOS

Download the .pkg (Intel or Apple Silicon depending on your chip) and follow the wizard. The terminal will recognize conda after restart.

Linux

Download the .sh script, then run it and accept initialization of your shell.

First project, folder structure and environment

NOTEObjective — Create a professional data project skeleton, associate it with a dedicated conda environment, and make your very first Git commit, all in under ten minutes.

Learning objectives

TIPAt the end of this module
  • Adopt a standard folder structure for your data projects
  • Create a conda environment dedicated to a project
  • Understand the role of each folder (data, notebooks, src)
  • Write a .gitignore suited to data science
  • Initialize a Git repository and make a first commit

Basic intuition: a place for everything

A data project without structure quickly becomes chaos: notebooks named final_v2_really_final.ipynb, CSVs everywhere, scripts mixed with data. The solution is a standard structure where each file type has its place. You no longer waste time searching, and anyone can understand your project by looking at the folder tree.

The golden rule: one project = one folder = one environment. Each project lives in its own folder with its own conda environment. No mixing, no conflicts.

The standard structure of a data project

go-further

This article covers the most useful excerpts — the complete Data Science AI Tools course (11 chapters, 43 lessons, corrected exercises and final project) takes you all the way.

./access-the-full-course free course: Mastering Claude Code

FAQ

How long does it take to learn Data Science AI Tools?
With structured progression (11 chapters, 43 short practical lessons), you reach an operational level in a few weeks at 30 to 60 minutes per day. The key is to practice each concept immediately.
Are there any prerequisites?
Basic computer knowledge is enough. If you can use a terminal and read simple code, you are ready.
Where to start concretely?
Reproduce the commands in this article, then follow the full Data Science AI Tools course: it chains the 43 lessons in order, with exercises and a final project.
./read-also
→ AWS Data Engineering Bootcamp explained simply (with diagrams and real code)→ Get started with AWS Real-Time Data: your first concrete step today→ Python Data Science: the 9 key steps to go from zero to operational

📬 Want to receive this type of guide every week? Subscribe for free — real code, zero fluff.

Stay up to date

Enter your email
Subscribe