Simplify Python Imports with Explicit Packaging

When developing in Python, it is always recommended to package your project.

Simply put, if a project is packaged, you can import stuff from it.

Some terminology before proceeding ahead:

• Module: A Python file.

• Package: A collection of Python files in a directory.

• Library: A collection of Packages.

We can package a project by adding an __init__.py file inside a directory.

While Python 3.3+ provides Implicit Namespace Packages — a directory with modules is considered a package by default, it is still advised to create an explicit __init__.py file.

A couple of major benefits of doing this are that it helps in:

• Explicitly specifying which classes/functions can be imported from the package.

• Avoiding redundant imports.

Let’s understand!

Consider this is our directory structure (and we are using Python 3.3+):

• train.py has a Training class.

• test.py has a Testing class.

As we are using Python 3.3+, we can directly import the Training and Testing class in pipeline.py as follows:

While this will work as expected, the problem is that we have to explicitly import the specific class from each of the modules.

This creates redundant imports.

Defining the __init__.py file can simplify this.

Let’s see how.

As depicted above:

• We first explicitly package the directory by creating an __init__.py file.

• Next, we specify the imports directly in this file.

Now, instead of writing redundant imports, you can directly import the intended classes from the “model” package, as shown below:

In other words, specifying the __init__.py file lets you treat your package like a module.

This simplifies your imports.

Also, as discussed earlier, an __init__.py file lets you explicitly specify which classes/functions can be imported from the package, which, otherwise, will not be evident.

This simplifies things for other users of your project.

Isn’t that cool?

👉 Over to you: What are some other cool Python project development tips you are aware of?

Thanks for reading!

Whenever you are ready, here’s one more way I can help you:

Every week, I publish 1-2 in-depth deep dives (typically 20+ mins long). Here are some of the latest ones that you will surely like:

• [FREE] A Beginner-friendly and Comprehensive Deep Dive on Vector Databases.

• Implementing LoRA From Scratch for Fine-tuning LLMs.

• You Are Probably Building Inconsistent Classification Models Without Even Realizing

• Why Sklearn’s Logistic Regression Has no Learning Rate Hyperparameter?

• PyTorch Models Are Not Deployment-Friendly! Supercharge Them With TorchScript.

• DBSCAN++: The Faster and Scalable Alternative to DBSCAN Clustering.

• Federated Learning: A Critical Step Towards Privacy-Preserving Machine Learning.

• You Cannot Build Large Data Projects Until You Learn Data Version Control!

To receive all full articles and support the Daily Dose of Data Science, consider subscribing:

👉 If you love reading this newsletter, feel free to share it with friends!

👉 Tell the world what makes this newsletter special for you by leaving a review here :)