Overview

Questions

• What is a virtual environment?

• How do virtual environments help manage dependencies?

• How do you create and activate a virtual environment?

Objectives

• Learn about virtual environments and their purpose in Python projects.

• Understand how to create and manage virtual environments.

• Get hands-on experience with setting up and using a virtual environment.

Python Environment Management

Managing Python environments is essential for keeping project dependencies isolated and avoiding version conflicts. Python provides several tools for managing environments, including Conda and the built-in venv module. Both help to create isolated environments for Python projects, each with its own set of dependencies.

Conda Environments

Conda is a package, dependency, and environment management tool that simplifies managing different Python versions and libraries. It is especially popular in the scientific community and on Windows platforms.

What is a Conda Environment?

A Conda environment is a self-contained directory that contains a specific collection of Conda packages. These environments help isolate different projects from each other, ensuring that dependencies for one project do not interfere with others.

Avoid installing packages into the base Conda environment. Create a new environment for each project to maintain isolation.

Creating Environments with Conda (Linux/MacOS)

To create a new Conda environment, use the following command:

admin@MacBook~ $ conda create -n <env_name> python=<version#>

For example, to create a Conda environment named ml-climate with Python 3.12:

admin@MacBook~ $ conda create --name ml-climate python=3.12

To activate the environment:

admin@MacBook~ $ conda activate ml-climate

Installing Packages in Conda Environments

Install packages within an environment using conda or pip:

admin@MacBook~ $ conda install <package_name>

Or, if the package is not available via Conda:

admin@MacBook~ $ pip install <package_name>

Exporting and Sharing Environments

Export an environment configuration to a file for sharing with colleagues:

admin@MacBook~ $ conda env export --no-builds --file environment.yaml

This creates a YAML file that lists all packages and dependencies used in the environment. Others can use this file to recreate the same environment.

Deactivating and Removing Environments

To deactivate a Conda environment:

admin@MacBook~ $ conda deactivate

To remove an environment:

admin@MacBook~ $ conda env remove --name <env_name>

Virtual Environments with venv

The venv module, included in Python by default, is another tool for creating isolated environments. While Conda is a more comprehensive solution, venv is lightweight and works well for basic Python projects.

Creating a Virtual Environment with venv

To create a virtual environment, follow these steps:

1. Open a Terminal:
   • On Linux/MacOS, press Ctrl + Alt + T or search for Terminal.

2. Navigate to the Desired Directory: Use cd to move to the project folder:

   admin@MacBook~ $ cd Documents/ml-climate
   

3. Create the Virtual Environment: Use the following command:

   admin@MacBook~ $ python3 -m venv ml-climate
   

4. Activate the Virtual Environment:
   • On Linux/MacOS:

   admin@MacBook~ $ source ml-climate/bin/activate
   

5. Install Packages: Install dependencies with pip:

   admin@MacBook~ $ pip install <package_name>
   

6. Deactivate the Virtual Environment (Optional): When done, deactivate the environment:

   admin@MacBook~ $ deactivate
   

Managing Packages with venv

While in the virtual environment, use pip to install, update, or uninstall packages:

• Install a package:

  admin@MacBook~ $ pip install <package_name>
  

• To list installed packages:

  admin@MacBook~ $ pip list
  

• To uninstall a package:

  admin@MacBook~ $ pip uninstall <package_name>
  

Exercise: Create and Use a Virtual Environment

1. Create a New Virtual Environment:
   • Name it myproject and install requests.
2. Write a Python Script:
   • Create a file fetch_page.py using requests to fetch a webpage.

   Example script:

   import requests
   
   response = requests.get('https://www.example.com')
   print(response.text)
   

3. Activate the Virtual Environment and install requests:

   admin@MacBook~ $ source myproject/bin/activate
   

   admin@MacBook~ $ pip install requests
   

4. Run the Script and deactivate the environment when done:

   admin@MacBook~ $ python fetch_page.py
   

   admin@MacBook~ $ deactivate
   

Creating Conda Environments on Windows (PowerShell)

To set up Python environments using Conda on Windows via PowerShell, follow these steps. Ensure that Anaconda or Miniconda is installed before proceeding.

Installing Conda

1. Download and Install Conda:
   • Download the Anaconda or Miniconda installer for Windows from their official websites.
   • Run the installer and follow the setup instructions, making sure to check the option to add Conda to your system’s PATH.

2. Verify Conda Installation: Open PowerShell and check if Conda is installed:

   PS C:\> conda --version
   

   If Conda is installed correctly, it should return the version number.

Creating and Managing Conda Environments

Once Conda is set up, you can create and manage environments as follows:

1. Create a New Environment: To create a new environment with a specific Python version, use the conda create command. For example, to create an environment called ml-climate with Python 3.12:

   PS C:\> conda create --name ml-climate python=3.12
   

2. Activate the Environment: After creating the environment, activate it using the conda activate command:

   PS C:\> conda activate ml-climate
   

3. Install Packages: Once the environment is activated, you can install packages like this:

   PS C:\> conda install <package_name>
   

4. Deactivating the Environment: To exit the environment and return to the base environment, run:

   PS C:\> conda deactivate
   

5. Remove an Environment: If you no longer need an environment, you can remove it with the following command:

   PS C:\> conda env remove --name <env_name>
   

Exporting and Sharing Environments on Windows

To export the configuration of a Conda environment to a .yaml file (useful for sharing), run:

PS C:\> conda env export --no-builds --file environment.yaml

This creates a YAML file that lists all installed packages, dependencies, and channels used for installation.

Installing Packages from a YAML File

To recreate an environment from a .yaml file, use the following command:

PS C:\> conda env create --file environment.yaml

Here’s the revised exercise with the environment name changed to ml-flow.

Exercise: Conda Environment Setup on Windows

Objective:

• Learn how to create, activate, and manage Conda environments on Windows using PowerShell.
• Practice installing packages and exporting the environment configuration.

Tasks:

1. Install Conda:
   • If you haven’t already, download and install Anaconda or Miniconda from their official websites.
   • Verify the installation by running this command in PowerShell:

     PS C:\> conda --version
     

     You should see the version of Conda installed.

2. Create a New Conda Environment:
   • Create a Conda environment named ml-flow with Python 3.10:

     PS C:\> conda create --name ml-flow python=3.10
     

   • Confirm the environment creation when prompted.
3. Activate the New Environment:
   • Activate the ml-flow environment:

     PS C:\> conda activate ml-flow
     

   • You should see (ml-flow) before the command prompt, indicating that the environment is active.
4. Install Packages:
   • Install two packages: numpy and matplotlib:

     PS C:\> conda install numpy matplotlib
     

   • Wait for the installation to complete.
5. Verify Package Installation:
   • Check if numpy and matplotlib are installed by running:

     PS C:\> conda list
     

   • You should see numpy and matplotlib in the list of installed packages.
6. Deactivate the Environment:
   • Deactivate the environment and return to the base environment:

     PS C:\> conda deactivate
     

7. Export the Environment Configuration:
   • Export the ml-flow environment configuration to a YAML file:

     PS C:\> conda env export --no-builds --file ml-flow.yaml
     

   • Check that the ml-flow.yaml file has been created in your current directory.
8. Remove the Environment:
   • Finally, remove the environment after you’re done:

     PS C:\> conda env remove --name ml-flow
     

   • Confirm that the environment is removed by running conda env list.

Key Points

• Virtual environments isolate dependencies.

• Using venv for creating environments.

• Activating and deactivating virtual environments.