feronis

[feronis]

Cargo shorts are one of the ugliest clothes ever invented. Next to overalls.

[feronis]

Lab views :]

[feronis]

I'm going crazy writing an abstract for a conference and studying for a midterm!

[feronis]

I wish you

Clear skin

4.0 GPA

Focus

Clear goals

Love

Affection

Peace

Self love

Good things from the universe

[feronis]

Santa uses quantum physics to deliver to everyone in just one night. Provided no one observes him, he can be in multiple chimneys at the same time.

[feronis]

[feronis]

#Straight People™

[feronis]

[feronis]

Why beehives are hexagons.

[feronis]

[feronis]

This probably isn’t that true.

If I wanted to write a physics engine or a program to find find resonance structures of chemicals, I would probably need to know physics or chemistry.  However, a lot of coding isn’t that, and you could easily be a programmer without ever needing that.

It’s similar to math.  Yes, what you do can be described mathematically (as most things can be).  But I know a lot of programmers who hate math and are great programmers.  And while creating a machine learning library or a package is math intensive, USING that library once created isn’t.  That’s the whole point, and making a product that uses such libraries instead of reinventing them yourself doesn’t make you less of a programmer.  

I have a bs in physics and bs in math and am getting my master’s in CS.  So while there are many things I love to do that are in the intersection of all those, you don’t need math to code. Math makes a lot of things easier to code but in some aspects it really doesn’t.  There is a ton of programming that isn’t mathematical at all (SQL queries, UI programming, and even pattern design).

Anything can be encapsulated as well.  

 You can’t derive how to work with a technology mathematically (unfortunately, it’s the most unnerving thing I’ve ever faced).  And being good at that is probably more useful than knowing abstract algebra or vector calculus (unless of course you are doing graphics and linear algebra is probably the most useful thing).  Stats is probably a lot better to know.  And no math class has ever covered data structures.  And knowing how to prove something won’t necessarily make you better at creating an algorithm. Critical thinking does that (and yes math does help critical thinking but it’s sufficient not necessary). 

The amount of “math” capabilities and payoff is probably a fairly sharp sigmoid, where hidden in that is people who are good at math are probably also good at logic and puzzles which makes for better programming.

Yes, Virginia, you do need math to be a programmer

The sentiment is probably related to the statement “Everybody should learn to code“: When newbies ask in forums if they need to be good at math to be a coder, they get the answer “No“. This either means “Don’t worry, you can pick up the required math as you go along“ or “I never needed to solve for X when I built Drupal themes“ or even “I got a C in Category Theory II in uni. As long as you remember the basics from high school, like proofs and equations, you’ll be fine���. It does not mean “No”.

You will definitely benefit from knowing math, because the skills used in math in school are very similar to those you need when learning to program. You need to manipulate formal symbol systems in your head. You need to build an abstract mental model of a deterministic but ridiculously complex machine. Skill in math is strong evidence of talent for programming.

Even “word problems“ are important. You need to formalize requirements from the real world in a way that a computer can understand them. This is what “word problems” are all about.

All programmers use certain mathematical concepts all the time, even when they don’t call it math. Set theory, classical logic, proofs (mostly by induction), polynomials and limits hide in every program.

Beyond that, you need probability, statistics, linear algebra and calculus if you want to do any kind of machine learning.

You need vector algebra and trigonometry in computer graphics or robotics.

You need to understand linear and non-linear dynamical systems if you want to design economy-heavy video games.

You need to understand statistics if you want to understand the performance under stress of operating systems, databases, load balancers and routers.

You can learn what you need while you go along. As I said, you don’t need to have an A in Category Theory to start learning. But even Category Theory really helps. Maybe you don’t see now how this abstract nonsense will benefit you, but I use the math I learned all the time.

[feronis]

[feronis]

[feronis]

Florence + The Arts 3/?

Florence Welch in the cover of “Lungs” vs Lucien Levy Dhurmer’s “Eve”

[feronis]

good selfies are hard to come by nowadays

[feronis]

Time to go do allllll the machine learning