Introducing Julia with a comparison to Python and R

Understand why Julia is popular and how it compares to Python and R.

Motivation

If you have been following the Data industry, you must have probably heard of Python and R, two of the most popular programming languages in the Data field.

A few years ago, a new yet efficient programming language, Julia was introduced for scientific computing and seems to have a bright future in the Data Science, and Machine Learning community.

These are some of the questions that might cross your mind:

How is Julia different from Python or R? Is learning Julia worth the learning?

In this conceptual blog, we will start by looking at what Julia is, and some of the features that make it popular. Then we will explore how it has evolved over the last couple of years. Finally, we conclude by giving our viewpoint of the future of Julia.

What is Julia?

Julia is a dynamic open-source programming language introduced in 2012 by Alan Edelman, Jeff Bezanson, Stefan Karpinski, and Viral B. Shah. It is created for high-performance scientific computing with high-level code.

The goal behind Julia is to have a language that can combine the following benefits:

• Speed of C.
• Dynamism of Ruby.
• Flexibility of Python.
• Perform statistics like R.
• Perform linear algebra like Matlab.

What makes Julia so Popular?

These are some of the main benefits that make Julia popular over Python, and R.

1. Interoperability with other languages

Julia is interoperable with other languages, meaning that you can include any other programming language such as Python, R, C, or C++ in your Julia program. As Data Scientists, this interoperability can provide a lot of benefits, because this makes it possible for Julia to integrate with existing code bases and extend its capabilities through the use of libraries from other languages.

Additionally, because Julia is open-source and has a growing community of users and developers, it is supported by a wide range of tools and libraries that enable compatibility with other languages.

2. Speed and performance

Data Scientists have increasingly adopted Python and R because of their simplicity and flexibility, compared to more complex and performant languages like C and C++, which lack suppleness.

Julia provides Data Scientists with a much better tradeoff by combining the benefits of both worlds. This is because Julia is a compiled language similar to C and has been programmed for speed since the beginning. This gives Julia a C-like level strength, hence providing much better speed when it comes to dealing with larger programs.

On the other hand, Python and R and interpreted scripting languages, and therefore have only a quick execution time for small programs.

Julia has even been added to the list of programming languages that have achieved peak performance exceeding one petaflop per second.

The following quadrant illustrates how Julia compares to other scientific computing languages, based on speed and execution time.

3. A one-language approach

Developers often adopt the two-language approach, where they prototype their applications with a slow dynamic language such as Python or R. Then, when it comes to scaling those applications into production, they rewrite them using a much faster language.

Using Julia, developers don’t have to go through this programming cycle anymore, because it solves the issue with its dynamic aspect and ease of use.

4. Parallel processing

When it comes to speeding up big data preprocessing, Julia can distribute processes into multiple memory spaces or servers, and even runs natively on GPUs.

5. Dynamic and easy to understand

Julia is a dynamically-typed language that does not require variables to be declared with a specific type before they are used. Instead, the type of a variable is determined at runtime based on the value it is assigned. This allows the type of a variable to change over the course of a program, making Julia flexible and easy to use.

Additionally, Julia’s syntax is easier to understand, because, according to the co-founders Alan Edelman:

We have created Julia to allow not only humans to talk to computers, but to allow humans to talk to humans.

This tells a lot more about how easy it is to understand Julia’s syntax.

6. The use of multiple dispatch

Multiple dispatching is using the process of naming multiple functions the same way and their behavior changes based on the arguments they receive. This is by far one of the best features of Julia. It is important because it enables the language to support polymorphism, which is the ability of a function to work with multiple types of data. This is particularly useful in scientific and numerical computing, where data is often represented in different forms, such as matrices, vectors, and scalars.

Additionally, multiple dispatch allows users to define new methods for existing functions, which enables them to extend and customize the behavior of the language to suit their specific needs.

How is Julia growing over time?

Julia’s adoption has grown over the last couple of years at different levels, such as the number of downloads, cumulative Github stars, the number of YouTube views, and the cumulative number of visits on the official website as of January 2022.

Industries and professions using Julia today

Julia is used by more than 10.000 companies in the world, from big technology and aerospace companies to healthcare industries. Below are some of the examples:

• NASA achieved a 1000 times speedup to catalog 188 million astronomical objects in 15 minutes by using Julia in a supercomputer in their process of analyzing the “largest batch of earth-sized planets ever found”.
• Pfizer used Julia to improve the calculation time of a cardiac model from nearly 24 hours to only 40 minutes, which is more than 95%.
• Microsoft included JuliaPro to Windows Data Science Virtual Machine (DSVM) to make it available on Microsoft Azure.
• Many other companies such as Amazon, Google, IBM, KPMG, Capital One, AstraZeneca, etc. use Julia as well.

In addition to the previous companies, below are some of the academic institutions using Julia. This list is by no means complete or even comprehensive.

Comparative table of Julia, Python and R

So, the ultimate question is — Is Julia worth it?

You will barely find Julia listed in most Data Science and Machine Learning job postings, plus this programming language is young compared to Python and R, which have been around for decades with a very large community.

Even though Julia is still in its early stages compared to Python and R, I believe that it has a lot of benefits to offer the Data Science and Machine Learning community and would in the next 5 to 10 years become one of the most demanding skill sets, and would say, YES! It is worth learning Julia.

Does that mean you have to give up on other languages? Absolutely not, because each programming language has its pros and cons. Whether you are very experienced with other languages or figuring your way to learning your first language, Julia can be a good bet, because it combines the benefits of all the other languages.