So this is the power of software engineering

So this is the power of software engineering..

Attached: softwareeng.png (988x400, 209K)

Other urls found in this thread:

npmjs.com/package/is-odd
github.com/jonschlinkert/is-even/blob/master/index.js
npmjs.com/package/is-even
npmjs.com/package/is-integer
twitter.com/SFWRedditImages

>You expect me to remember numbers and shit foreall nigga

well technically ML is just stacking layers in keras. So they should be fine.

You don't have to know all of it, you refresh as much as needed to solve the task at hand, FFS

It's like not knowing basic arithmetic.

That's all shit I learned in elementary school, what the fuck is this industry.

easy money, apparently

Attached: flipped dodge.jpg (3264x1836, 2.06M)

These are the same people who complain about lack of jobs in one of the best fields to be in.

You also have to dance around your workstation while chanting and hitting a tambourine like some shaman hoping that it pleases the machine spirit and it will make you NN work.

Didn't you already make this thread with this same pic? Do you remember how it got locked the first time?

Where did you go? I didn't touch algebra until I was in 7th grade. I didn't even know what a histogram was until recently.

Attached: 1481780262661.jpg (480x454, 33K)

Public Amerilard schooling, but it was a pretty decent school system.
They also let me jump two grades ahead in math (I took pre-algebra in 5th grade) but I recall assisting kids in 6th grade (still elementary school for us) on basic algebra and (linear) graphing. 7th grade isn't far off in any case.

Post warosu archive.

yes

>brainlets ITT
if you didn't do real analysis, complex analysis, fourier analysis, vector spaces, group theory, differential geometry and topology when you were still in kindergarten then you are a brainlet desu senpai baka

Desu, theyre right you know. I cant program for shit but Im doing fine in maths. Theyre completely different subjects, one has words and other one doesnt

Math is easy but I cant learn to program

Attached: 1525454985860.gif (248x225, 203K)

>Math is easy but I cant learn to program
Yeah you can, you have the right mind for it. It's just the consistent application of logic like in mathematics. Don't confuse a bad programming course for your inability to learn it.

Software Engineering and Computer Science courses are more different than people think here.

SE focusses more on engineering software, like software architecture and such things. At my college it involves a lot of drawing diagrams and programming.

CS focusses more on math and data structures.

>is-odd package from npm
Is this for real?

Not him but you need to believe they are seperate. I can't learn to program at all, I've tried many times in many ways. My math professors hate programming and can't do it. Another math student also can't program.

People need to stop this lie that math is so intersected with programming

dude i never did algebra i only did geometry when i was in the 11th grade wtf

Do you know what a set it? In programming they can be types.
Do you know what a function is? One input maps to one output? Congrats, a function in programming.
Do you know how to sum things? Congrats you know recursion.
Have you taken calculus? The differential operator is the higher-order function that, when given a (differentiable) function on the real line, yields its first derivative as a function on the real line. Input a function, return a function. Congrats you know know what a higher-order function is.

Math and programming, are the same thing in a ML derived language like OCaml, ReasomML (dsl built on top of OCaml compiler), Rust, Elm, F#, ect.

They are separate, but if you good at one you should be good at the other. The thing perhaps is that both are actually quite difficult, and there isn't a great deal of overlap in the actual content of programming (outside of computational physics, numerical analysis and so on), so that you actually have to learn an entirely new subject coming from the very beginning. So it can get very frustrating when you're dealing with advanced mathematics and then go to learn programming and you're doing very basic introductory stuff which is still difficult because you've never done it before.

Imagine if you're a great programmer and then you try and learn maths but you haven't done it since high school and don't remember anything outside of basic arithmetic, so you have to learn algebra and functions and quadratic equations and simultaneous equations and so on, and there's an awful lot to learn before you even get onto anything interesting or useful and it all seems very difficult even though it's basic stuff, so you'd get frustrated and conclude math wasn't for you.

user, I know the stuff you posted but I cant make a program with it. How on earth could I make a program out of basic bitch math? It doesnt make any sense

How else would a Javashit determine if a number is odd?

npmjs.com/package/is-odd

this is the power of self-entitled low-test code-monkeys who think they are better than they really are because of all the horrible 'youre a rockstar coder" or "youre a member of our team of coding ninjas" that boost their egos until they actually think they deserve to be working for a top tier software company and then get incredibly ass blasted when they fail to understand basic concepts and there is no node library to cover them

Listing off equivalent terminology has nothing to do with actually writing a program that does a thing.

is modulo not available in javascript?

it is, the problem is js is loosely typed which means in theory you could call "gaybois"%2===0
and won't complain until it actually runs

this is why smart people use typescript

he's right though. most of modern 'programming' is gluing shit together. it's basically lego.

crazy
why do people use this language?

Further, for example instead of thinking about a loop that iterates over the elements of an array to generate the sum, you can define a higher-order “reduce” function instead, which in addition to taking the array as an argument, the reduce function takes a binary associative function as another argument. It then sums the array based on that binary function. The reduce function is a tree sum, which is highly parallel, and can be farmed out to separate resources to make summing gargantuan arrays faster than the sequential looping method.

These are all basic mathematical concepts. A binary operator that is associative, reduce applies a function to each element to produce a value much like reducing an equation, basic graph theory that is a tree, differential eq to understand how a function yields a new function, ect. Literally math, and even some languages allow you to write out near math notation such as defining types like you would sets. Then there's Turing Machines, and Lambda Calculus which both explain what computation is. Computation in the lambda calculus consists of applying beta reduction until there is nothing left to reduce, producing a value just like in mathematics. tl;dr you already know how to program

>We never use the shit we use in high school, what's the point in learning this stuff?
>You expect me to use shit I learned in high school in the real world in this specific case? How the fuck am I supposed to remember it?

no alternative
basically languages like coffeescript and typescript are syntactically completely different from JS, they just have compilers that do type checking and ultimately compile to JS
absolutely no reason to write vanilla JS in the current year

And even further, if you know how to program, then you know basic math. If you've ever defined a function, that accepts another function as input, and yields a function, you know what a differential is, and can do Calculus. The scope in Calculus is just your data is confined to the real line/axis. It's the same shit just different data.

Keras's dev is a SJW anti-Trumper libtard just so you know

I should hope so
imagine a techno-primitivistic dystopian society performing rituals to appease the machine spirit before an important calculation
like the priests blessing datacenters tread we had earlier, but even more crazy
you just know that when USA finally folds on itself that kind of thing is going to happen there

Do you remember elementary let statements in math written above equations? Let x = Daniel's allowance for the week of June.
Well in ML derived languages, it works exactly the same way. Let's say somebody asks you to write a program, to calculate factorials of a given number.

let
f(i) = if (i < 2) then i else i × f(i − 1)
in
f(i)
end

You call that function with input 5, entering f(5), it returns the factorial of 5 just like it would in math notation. Now you officially can call yourself a programmer if you've ever written a Let statement in elementary school.

That's just big scary names for absolutely basic stuff. Variables and coefficients are k and x. Linear equations are y=k*x. Graphs of functions is what y=k*x looks like(a fucking line - hence the name linear). Histogram is what you do in Excel when you select a row of numbers and make a plot of it. No fucking way somebody doesn't know that.

did horrible in maths barely passed
have no problem with algorithms and programming in general
explain this

That's where they fuck up. You can't engineer without heavy mathematics. It's the reason most software is complete trash. You need mathematics models like tla+ to write error free code. If you can't do proofs and linear algebra you have no business writing code period.

I cannot believe it.

Attached: Hand.png (1200x1000, 1.05M)

>no problem with algorithms and programming
You are probably not doing it properly.

Algorithms are discrete math.

you worked harder at programming than you did at math
maybe because you enjoy it more

>learned x, but never needed x in real life
>now you need x in real life
>complain

Faceberg switched to ReasonML, which essentially is syntax sugar that interprets to OCaml and then compiles to js using bucklescript. OCaml type system prevents any kinds of these bugs which are notoriously hard to pin down, for example that modulo test you could write tests/guards in js in order prevent anything except an integer from being tested but the bytecode compiler will happily toss it out for optimization reasons, then you spend forever trying to figure out what the fuck the problem is after thousands of hours of your js running without a problem.

Any half-decent CS program is supposed to drill all that shit into your head before handing you a diploma. What the hell is happening to education?

It's not taught in CS. It's taught in elementary.

>coffeescript
I wouldn't talk about coffeescript anymore, its a pile of shit.

What if i need to check if a number is even tho?

How does an absolute mathlet catch up with the times? Haven't done shit since high school and calculating pocket change is a struggle.

Attached: 1511880773527.jpg (1218x1015, 212K)

!is_odd(number)

>Keras's dev is a SJW anti-Trumper libtard just so you know
Who cares?

Don't be an idiot. You're reinventing the wheel for no reason, when there npm modules that solve this without any effort.

github.com/jonschlinkert/is-even/blob/master/index.js

>and won't complain until it actually runs

False. It won't complain at all because that's perfectly valid Javascript.

Attached: Capture.png (206x55, 1K)

>Faceberg switched to ReasonML,

I'm really not convinced that's meaningful. The last ReasonML update basically changed the syntax so much every existing piece of code is a rewrite. Facebook would only pull that off if they were *barely* using the language.

Every single fad has someone saying Facebook is moving there. There's a guy constantly posting on Reddit asserting the whole Facebook backend is being rewritten in Haskell, and the problem is there's on Facebook blog where they allude to doing one thing in Haskell that gives it credibility.

It's like how everyone says Wordpress is "moving to NodeJS", just because they added some frontend Javascript.

I have no doubt there's a few people looking to push ReasonML who sent some ReasonML code into the existing code base, but that's a stretch to say it "switched".

It's only a pile of shit because it's obsolete by Typescript. It was a valid solution in its times.

they extracted React from Standard ML too, the faceberg devs post on HN. Bucklescript was made by Bloomberg. Rust also targets bucklescript, and a dozen other langs. Compilers producing bytecode for another language that allow for type checked programs are pretty common at fberg and wall street.

it was a 50/50 valid or nan

This, essentially.

Kotlin -> JS

Programming and Algebra are practically the same thing at its core. You are attributing values to objects and calculating objects with values.

Which fucking adult that has fucking finished high school doesn't atleast know coefficients and linear motherfuckin equations?

>Which fucking adult that has fucking finished high school doesn't atleast know coefficients and linear motherfuckin equations?
adults in the US

I don’t get why they wouldn’t just revise that stuff if ML was something they were interested in. It would take less than a week to get the jist of it. Why just say “nah, not doing it.”?

And that's why you need maths for ML

Attached: 1508861591906.jpg (749x1200, 694K)

That's cool and all, but completely useless for writing an ERP or even a payroll system.

The beauty of American education is that it only gets worse over time.

I imagine they are like the kids constantly asking Jow Forums how to hack and then retreat back into the woodworks when told that they need to actually learn and program shit for it.

ML =/= deep learning/NN duh

There are brainlets on Jow Forums that can do advanced maths but cant programming.

As far as math goes, the math used in deep learning ML is so fucking boring I don't think I could bare to do it for a living. You basically only need to know linear algebra, vector calc, and basic stats. You can skip the vector calc if you are a brainlet who doesn't want to know how gradient descent and backprop work in detail.

Every neural net program I have written consists of a few repetitive steps. Get the data from a CSV or a DB, clean the data using a whole bunch of functions in numpy/pandas/scipy and basically doing whatever the fuck it takes to turn the data into nice numbers that can get shoved into the NN. Then you have a whole fucking stack of constants like learning rate, number of nodes, what activation functions etc... Then you shove the data through the NN and use whatever basic stats shit you want to calculate your loss function.

The choice of the shit that actually matters, namely the constants, shape of the net, loss function is to a large extent total fucking shit as far as being a rigorous science. A lot of it is legitimately just picking shit and seeing what works.

The big question right now is why a lot of deep learning techniques work at all, so evidently actually writing programs to do it is fucking tedious and confusing as all hell.

it's more that i simply don't remember either

that, and the french names for some of those things are totally different.

>anti-Trump
all jokes aside is anyone actually pro-trump, I get that he’s funny and he triggers libruls and all that, but anyone even moderately intelligent can see he is completely out of his depth, and probably doesn’t even want to be President. However amusing his twitter is I don’t see how that could make up for the fact he consistently embarrasses the US on the international stage.

I dislike him, but only at the same level of dislike I have for every president since at least JFK. All Zionist stooges, all of them.

don't worry a genius wrote a npm package called is-even with the dependancy is-odd, both have upwards of 400k monthly downloads

If you're writing solid code then you're probably using a ton of complex high level mathematical concepts without even realizing it.

thanks, it has been some time since i laughed that hard, I'm literally crying

Unironically yes.
Not because he's good, but specifically because he's a post turtle.
He cannot get fucking anything done, and that's perfect. Obama did an okay job, and now we get 4 years of whatever he left standing.
Way, WAY better than the alternative of Hillary being in office, because Hillary is a wheeler and dealer, she breaks people and gets her way through lying and cheating and manipulation.
Trumps is just an idiot. Which is just what we needed.

>thinking the is odd package is a joke
> npmjs.com/package/is-odd
> 250k weekly downloads

Please get me off this ride. Can we go back to computer programming being an eccentric job that recluses did?

At worst, it'll take you a week to re-familiarize yourself with this 6th grade algebra.
At.
Fucking.
Worst.

also, if you program AT ALL, you have a working grasp of variables and coefficients

npmjs.com/package/is-even

has is-odd as dependency

i didn't touch anything more complex than x2 + x - 2 = 0 until college

Which is on par with what's on OP's pic.

i'm gonna set up a patreon page for is-integer that has both is-odd and is-even as dependencies. Please support this valiant effort!

Hell, it's more complex than what's on the pic.

Is NULL odd or even?

Attached: 1526604209568.jpg (526x423, 20K)

I'm not a programmer, is there any usefulness for this?

the hell, how'd you survive standardized testing?

Luckily a kind soul already made npmjs.com/package/is-integer
This only depends on is-finite

I guess if you want to check if a string is an odd/even number real quick without turning it into a number yourself

well standardized testing didn't require anything more complex than 2degree equations.

(i live in northern brazil and that was about 10 years ago, things have changed)

Have I got the setting for you...

Attached: adeptus-mechanicus.jpg (672x372, 108K)

AHHHH PLUG NUMBER IN TOOOO HARD GRUGG NOT WILL TO PUT AWAY 15 MINUTE TO LEARN

Attached: file.png (900x900, 386K)

Shut the fuck up, it's not scary names it's what the damn things actually are.

>Obama did an okay job

fucking americans

you only measure how well a president did his job by the number of dead sandniggers divided by the number of dead americans.

As an employed dev I'd like to say two things:
1) Learning programming, programming, improving your programming skills generally do not require a math background or any excessive math knowledge. The notion of "you MUST learn math before even thinking about learning programming!" is autistic.
2) Dropping the whole idea of learning a math-reliant programming branch due to not having the required math knowledge is even more autistic - just study nigga, what are you doing?

I don't even know how to solve this. At first I thought x might be 1.

This stuff is literally told to 12-year-olds in France.
Amerimutts can't do math.