Andrew Seaford

Creating and using a requirements.txt file in Python is a standard practice for managing project dependencies. This file lists all the Python packages your project depends on, along with their versions. Here’s a step-by-step guide on how to create and use it:

Creating a requirements.txt File

Install the Necessary Packages

First, install the packages your project needs. You can do this using pip:

pip install package_name

Freeze the Current Environment

Once you’ve installed all the necessary packages, you can generate a requirements.txt file using the pip freeze command. This command outputs a list of installed packages in the current environment and their versions.

pip freeze > requirements.txt

Manual Editing (Optional)

You can manually edit the requirements.txt file to add or remove packages, or to specify version ranges. The file format is simple: each line contains a package name and an optional version specifier.

package_name==1.0.0
another_package>=2.0,<3.0

Using a requirements.txt File

Create a Virtual Environment

It’s a good practice to use a virtual environment to manage your project’s dependencies. Create a virtual environment using:

python -m venv myenv

Activate the Virtual Environment

On Windows:

myenv\Scripts\activate

On macOS and Linux:

source myenv/bin/activate

Install Dependencies from requirements.txt

With the virtual environment activated, you can install all the packages listed in requirements.txt using:

pip install -r requirements.txt

Best Practices

Pin Versions: Pin the versions of your dependencies to ensure that your project works as expected regardless of updates to those packages. Use == to pin exact versions.

numpy==1.19.5
pandas==1.2.4

Update Regularly: Regularly update your requirements.txt file to keep track of changes in your dependencies.

pip freeze > requirements.txt

Review Dependencies: Periodically review your requirements.txt file to remove unnecessary dependencies that might have been included during development but are no longer needed.

Example Workflow

Install Dependencies

pip install numpy pandas

Freeze Dependencies

pip freeze > requirements.txt

Check the Generated requirements.txt

numpy==1.21.0
pandas==1.3.0

Create and Activate a Virtual Environment

python -m venv myenv
source myenv/bin/activate  # macOS/Linux
myenv\Scripts\activate     # Windows

Install Dependencies in the New Environment

pip install -r requirements.txt

By following these steps, you can ensure that your project’s environment is reproducible, making it easier to collaborate with others and deploy your application.