how-to-create-environments.md

title	How to set up an environment

In this guide we summarize some key commands to set up an environment with different tools that you might encounter in the scientific python ecosystem. An environment is a workspace into which you can install Python libraries, separate from what is being used by your operating system.

The environment managers that are covered in this how-to guide include:

• venv
• conda
• mamba
• uv
• pixi

In each of these examples we'll create a new virtual environment related to our project called science (you can use whichever name you prefer!). We'll activate the environment, install some dependencies, and see an example of installing dependencies from an existing file. You may encounter files like requirements.txt, environment.yml or pyproject.toml that specify needed dependencies for a project.

Set up a virtual environment with venv

With venv to create environment associated with a project folder called science.

python -m venv science

Start using it by activating it as follows:

source science/bin/activate

You are now ready to install Scientific Python packages using pip! For example:

pip install ipython numpy scipy

Often you'll interact with projects that have a specific list of dependencies (for development environments, testing environments, or the project itself). You can install the list of dependencies with pip in your venv using:

pip install -r <path/to/requirements.txt>

Remember to re-activate your environment every time you open a new terminal, using:

source science/bin/activate

You can find more information on using venv for packaging here.

Set up an environment using conda

With conda, we can create a new environment named science (-n is the same as passing --name):

conda create -n science

Start using your environment by activating it:

conda activate science

You are now ready to install Scientific Python packages using conda! For example:

conda install ipython numpy scipy

Some projects distribute environment files with listed dependencies with an environment.yml file. The first line of this file sets the environment's name. To create an environment and install the dependencies with this file, use:

conda env create -f <path/to/environment.yml>

Remember to re-activate your environment every time you open a new terminal:

conda activate science

You can find more information on using conda for environments here.

Set up an environment using mamba

With mamba, like conda, we can create a new environment named science (-n is the same as passing --name):

mamba create -n science

Start using your environment by activating it:

mamba activate science

You are now ready to install Scientific Python packages using mamba! For example:

mamba install ipython numpy scipy

To install a specific environment from a .yml file, use:

mamba create -f </path/to/environment.yml>

Remember to re-activate your environment every time you open a new terminal:

mamba activate science

You can find more information on using mamba in the mamba user guide.

Set up a virtual environment using uv

To create a new environment using uv in a project folder called science, navigate to that folder and execute:

uv venv

Start using your environment by activating it:

source .venv/bin/activate

You are now ready to install Scientific Python packages using uv! For example:

uv pip install ipython numpy scipy

To install dependencies from a requirements file, use:

uv pip install -f </path/to/requirements.txt>

Remember to re-activate your environment time you open a new terminal:

source <path/to/science/>.venv/bin/activate

You can find more information on using uv for environments here.

Set up a virtual environment using pixi

To initialize a new project with pixi in our project called science, execute:

pixi init

You are now ready to install Scientific Python packages as dependencies in this project! From the science directory, execute:

pixi add ipython numpy scipy

To install dependencies from a file like environment.yml, use:

pixi init --import <path/to/environment.yml>

Remember to re-activate your environment when you re-open a terminal. Navigate to the science folder, and execute:

pixi shell

This will drop you into the default environment for the pixi project, with all dependencies in that environment accessible to you in that shell.

A pixi project may have multiple environments defined in the pixi.toml file. To load a specific environment:

pixi shell --environment=<envname>

You can find more information on using pixi here.