4 ms·

How much programming requires you to know about Taylor series? I think they're nifty, but not very relevant to most work. I sort of agree with you that experi

How much programming requires you to know about Taylor series? I think they're nifty, but not very relevant to most work. I sort of agree with you that experience matters, but I'm not so sure math is useful unless you're writing programs for scientific computing or you want to be a professor.

moultano 15y ago
That's self-fulfilling. Math lets you solve problems that you wouldn't even have though of as problems before. If you don't know the math, your brain skips over them as a fact of life rather than as something with a solution. Programming doesn't require math because programmers don't necessarily know math, not because the opportunity isn't there.
amcintyre 15y ago
It's not so much knowing a long list of mathematical facts, as acquiring the mindset that lets you take those facts and apply them to new situations. Sure, you can choose a programming career where you don't need to know what a Taylor series is, how large linear systems are solved, or even what sine and cosine are. Maybe that will be a safe choice for the forseeable future, but more and more apps involve 3D, machine learning, computer vision, data analysis, etc. I personally view it as increased job security to learn as much as I can about the basic tools that let me understand how those things work. To not do so is (in my mind, anyway) very much the same as saying that caching, or complexity theory, or how to build a linked list is "not very relevant to most work": it may seriously limit the jobs you're eligible for in the future, because you'll lack the foundation to understand the work at hand at even the most basic level. That said, I've always been interested in math, and probably obsess too much over knowing how things work. I can appreciate that it's just not interesting enough for most people to pursue.
cageface 15y ago
Taylor series are very handy in DSP for finding good approximations to computationally expensive functions.

moultano 15y ago

That's self-fulfilling. Math lets you solve problems that you wouldn't even have though of as problems before. If you don't know the math, your brain skips over them as a fact of life rather than as something with a solution. Programming doesn't require math because programmers don't necessarily know math, not because the opportunity isn't there.

amcintyre 15y ago

It's not so much knowing a long list of mathematical facts, as acquiring the mindset that lets you take those facts and apply them to new situations. Sure, you can choose a programming career where you don't need to know what a Taylor series is, how large linear systems are solved, or even what sine and cosine are. Maybe that will be a safe choice for the forseeable future, but more and more apps involve 3D, machine learning, computer vision, data analysis, etc. I personally view it as increased job security to learn as much as I can about the basic tools that let me understand how those things work. To not do so is (in my mind, anyway) very much the same as saying that caching, or complexity theory, or how to build a linked list is "not very relevant to most work": it may seriously limit the jobs you're eligible for in the future, because you'll lack the foundation to understand the work at hand at even the most basic level. That said, I've always been interested in math, and probably obsess too much over knowing how things work. I can appreciate that it's just not interesting enough for most people to pursue.

cageface 15y ago

Taylor series are very handy in DSP for finding good approximations to computationally expensive functions.