Coding organization part 2: Saving, sharing, and importing environments

Read Time ~ 4 Minutes

I wanted to revisit a topic of a previous edition of AI Insights, Coding organization: A quick look at virtual environments. If you haven’t already, I would suggest reading that one first as this article is a continuation of that one.

Software development is often times a collaborative effort and it’s helpful to keep everyone on the same page when it comes to the packages you should be using. This is where saving, sharing, and importing environments comes in handy!

The way it works with virtual environments is you don’t actually share the packages and libraires themselves, but instead share a list of the ones you have used in your environment. So instead of someone using your exact packages, they instead simply download those package versions that match yours. This makes it easy to pass around what packages and versions you used without having huge project file sizes.

In the last article, I showed you two ways of setting up a virtual environment. One with the Python included virtual environment called venv and another with Anaconda. So let’s continue exploring those here today.

Python venv

Let’s start by activating our virtual environment, I am going to use the one I created in the previous article:

If you want to see what packages you currently have in your virtual environment type “pip list” and it should look similar to this. NOTE: your output might be different depending on what packages you have installed:

For this example I only have numpy and pip currently installed but most of the time this will be filled with dozen of different packages. In order to share this list just type of the following:

What this does is use pip to create a text file of all the packages installed inside of your virtual environment. Technically, you can use this on the “base” environment of your computer too. The file will be saved where ever you initiated this command. For me that would be inside my env_example folder. If I open it up with a text editor you should see something like this:

This file basically says that for this environment I am using this specific version of numpy. Also, keep in mind you can name this file whatever you like. It doesn’t have be called requirements.txt but there is a loose naming convention within the community that encourages you to use “requirements” as the file name.

Now you can take this file and share it with anyone you like. The most common way this is done is to include it in your project folder and post it on a remote repository like Github. That way, when people download your project they have a list of all the correct packages to download as well. Speaking of that, let's take a look at that next.

So, let’s say you want to check out someone else’s project and just downloaded it to your local machine. In order to set everything up we first have to create a new virtual environment. If you have forgotten how to do that refer back to my previous edition of AI Insights.

Once you have your new virtual environment set up, activate it and navigate to where the requirements.txt file is. Then simply type the following:

This will install everything listed inside of that requirements.txt file to your activated virtual environment and that’s basically it, nice and simple!

Anaconda

Using Anaconda to save and load environments is very easy and straight forward. To export your environment, first load up Anaconda and activate which ever environment you want to save. Then in the “Environments” tab towards the bottom you will see the following row of buttons:

Simply hit the “Backup” button and click on the button in the pop up window labeled “Backup” as well. From here simply save the file where ever you want and name it whatever you want as well. The file that you save out is actually in the form of a YAML file and not a txt file. A YAML file is a human-readable data serialization language and is what Anaconda uses to save its environments.

Again, you can share this file the same way as the requirements.txt from venv.

Installing the packages from a YAML file couldn’t be easier with Anaconda as well.

Simply hit the “Import” button and navigate to the YAML file you want to use. From there it will ask you what the name of the environment should be and it will create it as well as install any packages listed in the YAML file.

There is a reason I have been using Anaconda more and more as it really simplifies a lot of things for you.

Wrapping it up

While this edition is a bit shorter I hope you enjoyed the continuation of virtual environments and all the headaches they can solve.

Until next time.

Andrew-

Have something you want me to write about?

Head over to the contact page and drop me a message. I will be more than happy to read it over and see if I can provide any insights!