Skip to content

Python and Conda Virtual Environments

1. What is a Python venv

To isolate various Python projects, we create separate Python virtual environments (venv). Each venv acts like a lightweight virtual machine, running independently with its own:

  • Python interpreter
  • Packages and libraries
  • Configuration settings

This ensures that:

  • Projects do not interfere with each other’s dependencies
  • Different projects can use different versions of the same package
  • The development setup is clean, consistent, and easy to reproduce

2. Similar concept to dockers

It is similar to the concept of Docker containers, but limited to the context of Python. While Docker isolates entire applications and their environments at the system level, Python virtual environments isolate Python-specific dependencies and configurations at the project level.

3. Python venv commands

Create a venv

Geheral syntax: python3 -m venv <venv-name>

Example: 

python3 -m venv venv-proj1/

This command creates a new virtual environment in a directory called venv-proj1

  • venv-proj1 is not the actual project folder; it only contains the Python environment (i.e. Python interpreter, installed modules, configuration settings, etc.).

  • The actual project files (i.e. your Python code) are usually placed in the same parent folder as the virtual environment. (such as src, tests, and other files)

For example:

my-project/
├── venv-proj1/
├── src/
├── tests/
└─── <other files/folders for the project>

Activate a venv

General syntax: source <venv-name>/bin/activate

source venv-proj1/bin/activate

Deactivate a venv

deactivate

Deletes a venv

General syntax: rm -rf <venv-name>

rm -rf venv-proj1

4. Conda environment reference

Create, clone and edit a conda environment

Note: --name can be replaced with -n (as short form)

conda create --name myenv python=3.10
or
conda create -n myenv python=3.10
or
conda create -n dataenv python=3.10 numpy pandas matplotlib
or
conda create --name newenv --clone myenv

Using yml files

conda env create -f environment.yml

type following in environment.yml:

name: myenv
dependencies:
  - python=3.11
  - numpy
  - pandas

Activate or deactivate

conda activate myenv
conda activate myenv

View

conda env list

Delete

conda remove -n myenv --all

Python Virtual Environments

Anaconda Environments