Creating New Environments with Conda

In the previous article, we covered how to install Anaconda and configure its environment variables so that the conda command is readily available in your terminal. In this article, we’ll dive deeper into using conda to create new environments—enabling you to manage different Python projects and their respective dependencies effectively.

What Is a Conda Environment?

In Python development, a conda environment is an isolated workspace containing its own Python interpreter, packages, and dependencies. Key advantages of using conda to manage environments include:

• Isolation: Environments are fully independent, preventing dependency conflicts.
• Flexibility: You can create distinct environments tailored to specific project requirements—including different versions of packages.
• Reproducibility: Team members can replicate identical environments for consistent development and testing.

Creating a New Environment

Creating a new environment with conda is straightforward. Use the following command:

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

• --name: Specifies the name of the new environment.
• python=<python-version>: Optional; specifies the Python version to install (e.g., 3.9, 3.11).

Example

Suppose you want to create an environment named myenv with Python 3.9. Run:

conda create --name myenv python=3.9

Upon executing this command, conda will:

1. Compute and display the list of packages and versions to be installed.
2. Prompt you to confirm installation. Type y and press Enter to proceed—the environment will then be created, required packages downloaded, and configuration completed.

After confirmation, you’ll see output similar to:

## Package Plan ##
...
Proceed ([y]/n)? y

Activating the New Environment

Once the environment is created, you can activate it. After activation, all commands and scripts executed in the terminal will use the Python interpreter and packages from that environment.

To activate an environment, run:

conda activate <environment-name>

For example, to activate the myenv environment just created:

conda activate myenv

After activation, your terminal prompt will typically change to indicate the currently active environment (e.g., (myenv) $).

Optional: Creating an Environment with Specific Packages

You can also specify packages to install during environment creation. For instance, to set up a data analysis environment pre-installed with numpy and pandas using Python 3.8:

conda create --name data_env python=3.8 numpy pandas

This creates an environment named data_env, installing Python 3.8 along with numpy and pandas.

Common Options for Environment Creation

Here are some frequently used options when creating environments:

• --clone <existing-environment-name>: Clones an existing environment—faster than building from scratch, ideal when replicating identical configurations.
• --file <environment-file>: Reads package dependencies from a file (e.g., environment.yml) to create the environment.

For example, to clone the base environment into a new one named my_clone:

conda create --name my_clone --clone base

Summary

In this article, we walked through how to create new conda environments—including core syntax, practical examples, and commonly used options. By leveraging separate environments, you gain precise control over project dependencies, ensuring clean, reproducible, and conflict-free development setups.

In the next article, we’ll explore how to activate and deactivate environments—and seamlessly switch between them—making your development workflow more flexible and efficient. Stay tuned!