Python’s Dominance, Explained

by Tom Helvick

Over the past 20 years, Python’s popularity as a programming language has steadily grown. While it’s still popular for its original uses as a general purpose scripting language. It has also grown to support popular web frameworks, libraries for scientific and statistical computing, and many of the most popular libraries for AI and machine learning development.

In this article, we’ll examine why Python has become so popular and whether you should consider using it within your organization on future projects.

Python’s Beginnings & Early Uses

If you’re not familiar with Python, you’d be forgiven for thinking it’s a new language. After all, it does seem to have modern syntax and a growth trajectory that has spiked rapidly in recent years. However, the Python language is actually over 30 years old, with its origins in the late 1980s.

Python’s early uses focused around general purpose scripting. As an interpreted language, Python’s uncompiled nature became popular for quick programming tasks where development speed was a priority.

While other languages like Bash and Perl were also popular for scripting, Python’s clean, easy-to-read syntax made it easy to learn. Non-programmers in other fields could learn and use Python easily, leading to a wide user base across many industries.

This wide user base began building libraries in Python for many different use cases. The open source community of easy-to-use libraries is one of the defining features of Python’s growth.

Popular Teaching Language

Python’s clean syntax also made it a great choice for teaching programming in secondary schools and universities. Features like runtime interpretation, inferred variable types, and the read-evaluate-print loop allow students to write expressive code quickly in Python compared to boilerplate code and type specification in other languages.

This interpreted, on-the-fly coding style is not without its drawbacks, since it abstracts away many of the fundamental concepts of computer science. Nevertheless, Python has introduced thousands of people to writing code. Many of those people are not students of computer science, but they can use Python in their work in scientific research, urban planning, actuary science, analyzing patient outcomes in healthcare, and more.

The ease of learning and use is one of Python’s greatest accomplishments, democratizing coding and making programming skills available to more people across more industries.

Scientific Computing & The C API

One popular criticism of Python is that it’s relatively slow at CPU-heavy computational tasks. That’s true, but Python libraries have overcome many of these criticisms in order to get the benefits of Python’s development speed while still being computationally quick.

The secret to this success? Python is built atop the C programming language, one of the fastest possible languages you can write code in. For many Python libraries, especially libraries that deal with big data or complex calculations, this means that they can drop down into lower level C code in order to complete certain tasks and then return to Python for the higher-level needs of the end user.

When you use the NumPy, SciPy, or pandas libraries in Python, for instance, you never have to write C code yourself. Instead, write your code in expressive Python and allow those libraries to drop into C code behind the scenes in order to perform your calculations at blazing fast speed.

The C interface in Python is one of its greatest features. It’s what has allowed Python to gain a foothold and now dominance in the fields of big data and machine learning. Python’s rapid growth in recent years is largely due to the C API and what it means for Python’s performance on CPU-heavy tasks.

Growth of Big Data & Machine Learning

Thanks to Python’s solid foundation of libraries and interfaces for data analysis, scientific computing, and statistical modelling, it was the obvious choice for machine learning applications when they began to become popular.

Today, the major libraries for machine learning — PyTorch, Tensor Flow, and more — are all Python implementations of the best algorithms for ML models. While other languages have machine learning resources, the majority of models and the most recently released models appear in Python implementations first.

As big data and machine learning continue to grow, expect Python’s dominance to continue to grow as well.

About Intertech

Founded in 1991, Intertech delivers software development consulting and IT training to Fortune 500, Government and Leading Technology institutions. Learn more about us. Whether you are a developer interested in working for a company that invests in its employees or a company looking to partner with a team of technology leaders who provide solutions, mentor staff and add true business value, we’d like to meet you.

Originally published at https://www.intertech.com on February 24, 2020.