Python virtual environments

This is the edited Python Virtual Environments tutorial from the sharepoint.

Why should you use virtual environments for your projects

Every Python Project should be started with its own virtual environment. This way, it is easy to restore older projects, share them or work on them with others. The benefit comes from a clear definition of the packages and versions (!) used in the project. So older projects can run with older versions of the packages, i.e., deprecated commands will be no issue.

How to setup your PC to use virtual environments

There are several ways to do it.

conda

I would recommend using the Miniforge installer.

Installation

• Linux:

  curl -L -O "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
  bash Miniforge3-$(uname)-$(uname -m).sh
  

• Windows: Download the installer from the Miniforge website and run it. This will include the “Miniforge Prompt” in your start menu. You can run any conda commands from there and more.

Working with conda

Here you can find the convenient Conda cheat sheet. You can create a new environment via

conda create --name <env_name> python=<version>

You can then activate this environment via

conda activate <env_name>

Once activated you can install and configure your python packages.

Anaconda Navigator

You can get the Anaconda Navigator from the Anaconda website. This is probably the option where you have to use the command line the least, because it comes with a GUI and easy integrations into your IDEs.

virtualenv

Use the Python package virtualenv together with virtualenvwrapper (on Linux). On Windows you need virtualenvwrapper-win instead. These packages can be installed via pip.

Installing the package

• Linux (virtualenv Linux):

  pip install virtualenv virtualenvwrapper
  

• Windows (virtualenv Windows):

  pip install virtualenv virtualenvwrapper-win
  

Configuring virtualenvwrapper

For the configuration of virtualenvwrapper you need to set up your $PATH variable to include the folder where virtualenv is located.

You can configure a path where your environments will be stored. For this, use the $WORKON_HOME variable. Make sure to use an existing folder.

Working with virtualenvwrapper

Create a new virtual environment via mkvirtualenv <env_name>. You can then activate this environment via workon <env_name>. Inside the environment you can install and configure all Python packages you need for your project. It is highly recommended to save the required packages into a text file via

pip freeze > requirements.txt

inside your project folder. You will be able to restore all required packages later (or on another computer) via

pip install -r requirements.txt

To leave the environment after your work is done, just use deactivate.

venv

venv