One post tagged with "conda"

uv has quickly become the go-to Python package manager. It's fast, handles lockfiles, and manages virtual environments out of the box. For pure Python projects, it works great.

But data science and AI projects rarely stay pure Python. Many key packages depend on compiled C/C++ libraries that must be installed at the system level.

uv can install the Python bindings, but not the system libraries underneath them.

In this article, we'll explore three tools, each adding a layer on top of the previous:

• conda creates isolated environments with both Python and non-Python dependencies, including R, C++, and system libraries
• pixi adds lockfiles, multi-platform support, system requirements like CUDA versions, and a built-in task runner on top of conda-forge
• nebi layers version control and team collaboration on top of pixi workspaces