Well, I’ll probably get a downvote on this, but I’ve searched EVERYWHERE and no tutorial explain this very well. Most of them are very confused and explaining only a small part of the use of these libraries.

So I’m starting a Data Science project and I’m trying to follow the best practices I’ve seem some friends doing, like using Poetry for dependencies and using cookiecutter as a template. However there are a few things which are not clear for me and I couldn’t find an answer anywhere. For instance, should I create the template for Data Science with Cookiecutter, which will create a structure with License, Makefile, notebooks, etc, like we see in https://github.com/drivendataorg/cookiecutter-data-science, and then I create the pyproject.toml inside this folder? If so, where?

Also, in the Poetry website it mentions that we should use poetry in a virtual environment. How does that work? I mean, I should create a virtual environment, install poetry in it and then add the libraries I need there? And I will run my project in this venv? What about the structure created in Cookiecutter?

I know that some people find this simple and dummy, but I’ve talked to many beginners and most are as lost as me.

Thanks in advance