This has always intrigued me as I see a tonne of posts saying that learning maths helps you become a better programmer.
I bloody love a bit of maths, I frequently go through my old notes and look up more and more complex problems to try widen the arsenal. But, to be honest, as a developer day-to-day I very rarely require any form of higher level mathematics to carry out my job. Every now and then I do a little game dev and that requires a reasonable amount: angles etc.
I’m curious to hear what you do and when you use maths as a developer, let me know in the comments.
Oldest comments (32)
Well, I don't require mathematics at work because I either use librairies that do that for me (Three.js and plugins for animations) or find the solution on the internet because I am not good at math but I guess it depends on the field of work. Someone that has to write a game engine from scratch for optimizations purposes will have to use maths more than I will.
Awesome, what sort of development do you do?
I'm helping my CTO in the development of the website of a luxury brand that is using 3D rendering to allow their customers to configure their jewelry (rings, necklaces, ...). This is both challenging and interesting because I have all sorts of concerns (performance, SEO, design, ...).
Sounds so cool, I’d love to see the final product!
We have a website and here is a page example of one of our rings (note that it is not optimized for low-end devices and is not working properly in Chromium, but is working well under Google Chrome).
That looks amazing, hats off to you!
You cant use Three.js without thinking about maths. In fact Three.js made me actually enjoy learning maths again.
I don't, but I wish I had a better grasp on statistics because it keeps showing up in Machine Learning.
That is true game dev and machine learning are the only times I really ever use it!
I'm using state charts/machines and functional programming more, which are kinda maths.
As I do things for web and some other back end services math is only used sometimes. Most of the time when math is required it is simple add, subtract, multiply, and divide. Occasionally some algebra is needed, but still it is minor.
That being said I think it is important for everyone regardless of profession to be reasonably capable in solving algebraic equations (with the help of a simple calculator) and how to apply them to real life scenarios. In addition one should have and understanding of dimensional analysis as it helps apply algebra to the real world.
I agree completely, I see the world differently the more I learn maths. Just don’t get to apply it nearly enough as a software engineer.
Yes, but I work a lot with algorithms, efficiency, data structures, and manual memory management. I'm also going to be working on a vector graphics rendering project, which involves plenty of linear algebra.
It all depends on what you're doing. The important thing is to have that knowledge in your back pocket, since you never know when it will crop up. Sometimes, it appears in the most unlikely places.
My job is to develop cryptographic primitives. I could not do this job without advanced mathematics.
On a slight tangent, every developer needs to know the basics of cryptography and security. It has nothing to do with whether or not you are writing cryptography software. (You shouldn't; you should use existing libraries.) But no matter what kind of software you are writing, other people will be attacking it. You need to know the principles of defense.
Sounds like a brilliant job! But yeah, I can agree with that, it’s also something you just kind of pick up on the job, but it is taught in computer science degrees a little more thoroughly.
I work on scientific and engineering software. Plenty of math. But still the skills I use more aren't strictly math. They're software design principles.
Yes , I do. My scientific researches always require a level of mathematics.
It's not required per se by my job, but my role in my team is kind of the efficiency/memory management/metrics guy, which leads me to use a lot of the statistics I learned years ago to determine statistical significance and model different approaches to a given problem.
Really depends on your field of work. Most plain old web development does not require mathematics much. You would need it extensively for anything in the field of machine learning, game engine development, operating systems, cryptography, etc.
My background is in mathematical programming and now I am a data scientist, so math is central to my work. Specifically, probability and statistics are very important for what I do. In the past I have also worked on numerical optimization so calculus and linear algebra were needed.
These days the most common algorithms are available in easy-to-use, optimized libraries. I am grateful to the developers of numpy, pandas, scipy, and scikit-learn every day.
I have a web developer friend who knows almost zero mathematics. It takes a variety of people with a diverse mix of skills to develop software.
Studying Mathematics is not just about learning how to use equations and some formulas or additions but rather learning how to approach very complex and highly abstract tasks. Think of how you'd develop a programming language from scratch without any use of toolkits such as LLVM. There, having been challenged with math problems before is off great help.
Try making a small gui based game like tic tac toe or something without using a framework and you will see why you need maths. We generally don't realize this because most of the time it's offloaded to some external library.
A simple easy to follow example would be plotting with d3.js, imagine setting up ticks on an axis without using d3 scale function for a dynamic dataset.
There's a whole new world inside of the cryptographic algorithms, networking, image processing and machine learning that works cause maths does.
I don't use maths for my professional development, where I have to use libraries for extending the system. But one time I developed a radial gradient editor with WYSIWYG controls and for that I had to use geometry to figure out control point positions. So I suppose people in graphics and simulations need math. As for cryptography, it IS mathematics, not development.
Most developers will never use any kind of advanced math and we're doing a huge disservice to our CS students by teaching so much math and theory and so little applicable concepts. I have to tell every new Jr he has just paid hundreds of thousands of dollars to learn things he'll never use and he has a ton to learn before he can contribute.
How about teaching networking fundamentals, git, how to design large projects, architecture, or any other real world skills? As our field grows and diversifies, we really need to be splitting CS programs into specialties. 90% of devs will never need a linked list or to know how to implement a bubble sort but likely 100% are taught these things.