The definitive guide to Python virtual environments with conda

This post is about conda, the tool we use to install and manage Python and its libraries in our systems. We use it for both development and production purposes and we strongly believe that conda stands out from other alternatives like virtualenv, poetry, pyenv or pipenv.

We don’t use conda because of habit. We also tried other alternatives and always go back to conda because it is the only, let’s say, full featured solution on the market.

Based on my experience of more than 6 years doing Data Science, conda (and virtual environments in general) is a tool that is often not well understood. You would be surprised how many excellent professionals, even with 5+ years of experience still struggle with chaotic, corrupted, and barely usable Python installations because of that.

The goal of this post is to end this madness once for all. If you get a better understanding of virtual environments we have accomplished our goal.

First, let’s dive into what a virtual environment is and what it isn’t with some examples.

Imagine you want to start a new Data Science project. You want to:

To do that, you want to use Python (or R), with some libraries. In the case of Python, you need:

Related posts:

The words explode into my head like a bullet. I bolt upright in bed, sweat running in rivulets down my chest, pooling into the waistband of my sweatpants. My chest is heaving, I can’t catch my…

If you or a family member has recovered from Covid, congrats on your recovery. Some people are worried about all the hype there is about black fungus, Long Covid etc. while others believe they’re…

Christmas flowers speak to euphoria, happiness, sharing and giving while at the same time symbolizing the expectation in the hearts of everyone that gives or gets them. Flowers and greenery light up…