JavaScript, which originated in the 1990s to enable webpage editing, serves as the foundation of the web. Over the years, it has skyrocketed in popularity and scope, and is now considered as one of the most widely utilised programming languages in the field.
As the demand for JavaScript continues to rise, the question arises as to whether it is ideal for data analysis and data scientists. Unfortunately, the answer isn’t straightforward. To gain more clarity, let’s explore the pros and cons of JavaScript as a viable option for data science related work in this area.
Why JavaScript Has Received Criticism
Python, R, Scala, and Julia are generally regarded as the most suitable programming languages for Data Science. It is not recommended for beginners to start with JavaScript as their language of choice.
Do keep in mind that JavaScript is known for its diverse application in web development and is a language worth considering for front-end web development.
Acknowledging the enormous popularity of JavaScript, the author of the book, David Beazley, shares in the preface that he had initially thought of naming it ‘JavaScript Vs. Data Science’.
JavaScript is known to have a number of peculiarities, particularly when it comes to handling numerical values. The term NaN (Not-A-Number) is frequently used to indicate values which cannot be expressed using the numeric data type, and was first introduced as part of the IEEE 754 floating-point standard.
In JavaScript, when one tries to divide by zero, the output received is ‘NaN’. However, this result is not particularly informative or useful. The issue is further complicated by the fact that within the language, ‘NaN’ is classified as a number, which makes it difficult to track and locate the error’s origin.
Although a minor inconvenience, this issue highlights a larger problem with JavaScript’s dynamically typed nature, where it has a vague method of determining whether a variable holds a textual or numerical value. While it may not be a significant concern, as a precautionary measure, protective programming strategies must be put in place to address this.
Handling large quantities of data can be challenging, with JavaScript not being an ideal language for such tasks, as it lacks precision with larger numbers and does not support multithreading or parallel processing. Also, when dealing with computation-intensive, CPU-bound tasks, both JavaScript and Node.js can have performance issues.
While there may be ways to work around some of the issues, the associated opportunity cost could make it impractical for a data scientist to spend time learning JavaScript. Other languages may be able to deliver comparable, if not better, results with greater efficiency, offering less complexity and more convenience.
Choosing to learn JavaScript may mean taking away time from mastering other programming languages. However, this may not be a wholly negative consequence after all.
Why Consider JavaScript?
In his publication, Beazley argues that the challenges associated with JavaScript have been adequately addressed, and the data science community’s interest in JavaScript has expanded markedly in the past few years. As a result, there are now resources available to support its suitability as a programming language for data science.
JavaScript’s accessibility and readability make it an ideal choice for programming. For instance, if you are currently on a personal computer, opening the JavaScript console is as simple as pressing the F12 key.
Learning JavaScript is straightforward, given its enormous popularity, and there are a wealth of resources available to support the learning process. A cursory look at the figures on StackOverflow gives an indication of the vast amount of information readily available on JavaScript.
An increasing number of companies are using web technologies based on a Node-based stack for their products, which is advantageous. When data scientists and product development teams share a common programming language, it improves the former’s ability to communicate effectively.
When the same technology is widely used, integrating new products and services is faster and more straightforward. It’s similar to conversing with someone who speaks the same language as you.
Microsoft’s TypeScript is a superset of JavaScript that addresses one of the most significant criticisms of the language: weak typing. TypeScript’s stricter typing system is even more rigorous than that of Python, which is also a popular language among data scientists. Using statically typed languages like TypeScript leads to better programming practices and a reduction in code bugs.
For those in need of multithreading support, Microsoft’s Napa.js may be an appealing choice, even though it is still in its early stages and may not be the most optimal option. Nonetheless, this is a testament to the increasing recognition of JavaScript as a versatile language that is capable of fulfilling roles beyond data analysis. However, this is not the only benefit.
Enhanced Data Science Resources
It’s commonly understood that JavaScript doesn’t provide the same breadth of data science libraries as more robust languages like R and Python, and we concur with this assessment. It’s evident that to become a proficient data scientist, one must possess a variety of other skills, regardless of their admiration for JavaScript.
The adoption of JavaScript in data science has grown remarkably in the last five years, with the noteworthy introduction of the TensorFlow JavaScript library.
It’s worth noting that a burgeoning data science landscape using JavaScript is emerging. D3.js is a well-known data visualization library that provides a wide range of browser-based tools for constructing dashboards, generating reports, and presenting data in a compelling manner.
TensorFlowJS effectively showcases the capabilities of Machine Learning. TensorFlow, which is a widely-used Machine Learning library, has been integrated into JavaScript, allowing the deployment of ML algorithms on the browser and/or Node.js server.
In response, I would question why one would opt for this approach. Although working within a web browser may not be optimal, it is a practical option for prototypes, small projects, and non-memory-intensive programs. The time and effort required to establish a virtual environment may not be proportional to the benefits when a web browser would suffice.
We now have a range of tools available to work with the language that powers the internet and online applications. Browser-based data science enables us to explore novel approaches for processing and presenting data in a user-friendly way.
Users with smart devices and internet access can access the outcomes of a data narrative presented through a web application created with JavaScript, HTML, and CSS, all within seconds.
The increasing prevalence of JavaScript in Data Science showcases the industry’s evolution. Data scientists are no longer just individuals who analyze data in isolation; they are storytellers who must devise creative methods for communicating their findings and promoting data-driven organizations.
Just Another Approach to Utilize
It’s apparent that mastering JavaScript is currently not an essential requirement for data scientists; nevertheless, it could be a valuable supplement to their skillset. Although this knowledge may not drastically change the field, it could unlock numerous new possibilities. Therefore, it should be viewed as a favorable advancement.
