Set up your computer

The course is made to be done on your own computer. The notebooks build on each other: one notebook saves a file or a model, and a later notebook uses it. A local setup keeps all of that in one place. It takes about 15 minutes.

You install two tools:

Git, to download the course and get updates;
uv, which installs the right Python version (3.13) and every package the course uses. You don’t need to install Python yourself.

Everything together needs about 1 GB of disk space.

1. Install Git

Download Git from git-scm.com and install it. On macOS, Git is often already there. Check it in a terminal (on Windows: PowerShell):

git --version

2. Install uv

macOS and Linux:

curl -LsSf https://astral.sh/uv/install.sh | sh

Windows (PowerShell):

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Close and reopen your terminal, then check that it works:

uv --version

Other ways to install uv are in the uv documentation.

3. Download the course

Go to the folder where you keep your projects, then:

git clone https://github.com/BredaUniversityADSAI/introduction-to-ml.git
cd introduction-to-ml

4. Install Python and the packages

From the introduction-to-ml folder:

uv sync

The first time, this downloads Python 3.13 and all packages into a .venv folder inside the course folder. That takes a few minutes. Nothing is installed elsewhere on your computer.

5. Check that it works

uv run python -c "import sklearn, rdkit, xgboost, lightgbm; print('All set')"

If you see All set, you are ready. On macOS, if this fails with a message about libomp, install the OpenMP runtime with Homebrew and try again:

brew install libomp

6. Open the notebooks

Choose one:

Jupyter Lab in your browser. From the course folder:
```
uv run jupyter lab
```
VS Code with the Python and Jupyter extensions. Open the course folder, open a notebook, click Select Kernel in the top right and choose the .venv environment. See Jupyter notebooks in VS Code.

What is where

Folder	Contents
`self-study/`	Self-study notebooks (and the reading pages you see on this website)
`datalab/`	DataLab notebooks, one folder per iteration
`data/`	All datasets of the project (see `data/README.md`)
`models/`	Models you save while working through the notebooks

The order of the activities is the order on this website, not the order of the file names. Follow the menu, or the 8-week plan.

Getting updates

The course is an early release and still changes. To get the latest version:

git pull
uv sync

git pull refuses to overwrite notebooks you have changed. The simplest way to avoid that: work in a copy of each notebook (for example 003_regression_mine.ipynb), and keep the originals untouched.

If something goes wrong

uv or git: command not found. Close and reopen the terminal after installing. On Windows, use PowerShell.
A package is missing or behaves strangely. Delete the .venv folder and run uv sync again.
You want an extra package. Run uv add <package-name> in the course folder and restart the notebook kernel. Don’t use pip install inside a notebook: the package is lost at the next uv sync. Note that uv add changes pyproject.toml and uv.lock, which can make a later git pull complain about these files.
Do not upgrade rdkit. It is pinned to version 2025.9.3 on purpose: other versions compute some molecular descriptors differently from the course data.
Still stuck? Ask in the course Discussions. Say which operating system you use and paste the full error message.

Only a quick look? Google Colab

Every notebook page on this website has an Open in Colab button. Colab runs the notebook in your browser, without installing anything, and the notebooks show how to load the data there. It is fine for a quick look or a single notebook. For the whole course, use the local setup: Colab forgets the files you create when the session ends, and the DataLab notebooks depend on files from earlier notebooks.