VS Code has a Python Interactive window now, which keeps a lot of the functionality of Jupyter Notebooks without the horrific issue of output being saved in your source file. I also use Jupyter Lab when I want a more haphazard quick-and-dirty analysis environment, or I want too build out a report. The support is growing as well, so there’s always someone who’s come before me with a solution to nearly every problem I’ve faced. It’s very extensible, free, and looks nice. In terms of working environment, I’m a happy Visual Studio Code user. Poetry helps with that in to-date, the nicest fashion. Maintaining a reproducible environment for others to build out is essential to good Data Science and Engineering. As it it, I will discuss only the parts of it I find useful for my workflow. That’s how I use it, but it includes some very nice package publishing capabilities that I may explore one of these days. Poetry is a dependency and package manager, and a virtual environment manager. Jump to the tldr for the steps and not the explanation.

I’ve been skeptical of migrating since I haven’t had issues with pipenv yet, but the momentum of Poetry seems to be gaining and I’ve recently updated my tooling to use Poetry. I’ve been a very happy pipenv user for a while now (technologically speaking), and have had very few issues with it, however, it’s pretty much a defunct package and Poetry is the latest solution on the scene. There’s Conda, pip, pipenv, virtualenv, etc. If you’ve been in the Data Science and Python world for more than a day, you’re probably very familiar with the plethora of ways to manage Python environments. These are my current set of tools that I use routinely for Machine Learning, Data Analysis, and Software Engineering. Tools change often, and if I decide there is a better way, I’ll update this post. I’m writing this blog post to myself, mostly to remind myself what I did last time to configure my environment. remembering how the hell you set it up last time.static checking, run-time error catching.

multiple Python environments, maintaining independency between different projects and versions.reproducibility, keeping track of installed packages.

How to build python in visual studio