This slay the average denizen of Jow Forums

brainlet here, where do I even start learning calculus?

I think there's a link between being able to coast through a calculus series without really understanding the material and being able to pass intro to programming. This leads to a ton of CS undergrads sucking at math because they never put in the effort (probably going all the way back to precalc).

That's because the BS in a science is supposed to set you up for a graduate track. Don't blame your degree; blame yourself for not knowing what you were getting yourself into. I did a double major in math and CS, had a 2 internships, and then did research in mathematical physics and then also in theoretical CS. Found out I like the intersection of physics and CS theory, so I study complexity of physical systems/quantum complexity, quantum information theory, etc. but i like a lot of classical theory and analysis, so I also study topology in the CS grad department.

Get off your ass and go to grad school

Teachers were shit to the point of not knowing any math themselves.
Professors were all turbo autists incapable of talking to a regular human.
Got through all Math classes with decent grades and zero real understanding of anything, promptly forgot anything I learned.
Never used any of that shit outside of university.

I've often thought about restarting low and properly learning math for some real understanding, but I just don't have that time right now and wouldn't even know how to start that journey.

That paper came out in 1976 while the Earley parser was invented in 1968. Nice meme.

Khan Academy, 3b1b's essence of calculus, stewart's calculus, and early transcendentals are all good resources to learn from. I also HEAVILY recommend Paul's Online Math notes for anything to do with calculus 1-3. The /sci/ wiki is also pretty good

I also think CS has, at most schools, a reputation for being a much easier engineering degree. Luckily, I went to a good research institution for undergrad, so we had a good program, but most CS students, and especially most Jow Forums students, want to do the bare minimum to get a well paying job, so it's just a means to an end. They like technology, but have no actual talent/drive to be excellent, so this allows them to do minimal work while still feeling like they're entitled to societal respect.

So I always crack up when I see "learn 2 code" spewed at twitter writers. Like dude, chances are that this guy is shit at what he does, thinks the extent of his degree ought to be the ability to write the least creative, most applied database push-pull workflow code he can, and despite all this, probably has dogshit code smells and cannot think outside the box

Math majors write the most shit code I've ever seen. No regards to programming principles like maintainability, modularity, testability, or even caring about performance at all.

What sort of university do you go to that this is teh case. I can buy this in grade school or high school, but in unversity math departments are filled with some of the coolest people. It's always fun to talk to your professors in office hours

If it takes you the totality of your neurons to see how CS is just mathematics, especially on a theoretical level, then idk what to tell you. I used lots of number theory and modular ring algebra on the software development level when doing cryptography and then lots of physics/calculus/optimization when working on an internship for graphics/systems.

I suggest you download the book of proof, download good resources off libgen, and self study. Use Stanford, MIT, UPenn, ETH Zurich, and Berkeley online resources for math and theoretical CS. I got a lot out of it even as an undergrad in math.

>Get off your ass and go to grad school
Wish I did that earlier. Now I am 30 starting CS masters in the fall. pls let me make it.

I mean, if you want to see how it's progressed:
link.springer.com/chapter/10.1007/978-3-540-31856-9_32

You need notions of distance, space, and meaningful metric space/measure when you want to talk about probabilistic phenomena in automata.

I was a math and CS double major. Math majors write dogshit code in numerical analysis classes, but I've yet to have a problem with their code otherwise. I see them writing good, modular code in architecture classes. Performance is only really a consideration when you're talking about long term scalability for a system running for a while. So, I cared a lot about performance when writing my userspace thread libraries and memory modelling/protection in OS, but I didn't give a shit when I had to do some PDE modelling labs

Either way, this isn't really about that. Coding and CS are two different entities. This is moving the goalposts

Godspeed user. Don't do a meme masters in machine learning. Do hardcore theory or hardcore systems. Graphics is a really cool field. I'm biased towards theory myself, but it's not everyone's cup of tea. PhD in CS is really comfy and has a surprising amount of employability both IN and OUT of academia

why do you care about PDA's so much. CFG's have been studied to death, compilation methods have made their constructions trivial, and they're a dead research field since there are way more interesting languages / machines with more power. Did you see a PDA once in your compilers/language principles class once and then think "wow, this is the pinnacle of CS; calculus is useless!"?

Austria.
Professors are incompetent cunts that are sheltered by virtue of having lifetime chairs and being impossible to remove.
It's not about seeing how this stuff is just math, because it's obvious. It's not about using this shit on the surface, those are just hammer-nail problems, I've been doing that well enough to be successful. It's about truly, intuitively understanding. I can look at languages, music, code or logic problems and on an intuitive level but I just don't get what the fuck is happening during a Fourier Transform or have a true understanding of derivations.
My mathematical base is fucked and I'd need to fix that and work my way up.

>Don't do a meme masters in machine learning
Time to change my specialization. :(

This. Change the way math is taught and more people will get it.

Ah, I see user. I suggest looking at the first part of for good resources. Also use the resources I talked about when replying to you the first time. Libgen is your best friend from going from zero to hero

ML theory is cool (and it's all measure/probability), but base ML is oversaturated and is losing employability quickly. I know a lot of people who did ML only to never use it on their jobs, or to become dead weight on their development teams. I suggest doing systems with a focus on learning techniques to solve specific problems. You could try something like applying learning on routing ad hoc networks under specific, hard-to-account-for constraints/conditions

I work with ODEs and PDEs all the time and I still don't know where exactly calculus fits into CS, other than implementing finite difference schemes and solvers, and analyzing time/space complexity of algorithms. However, my background with programming is entirely self-taught, so I know there is a lot I don't know. Programming seems to be much more about logic and algebra, rather than analysis.

I'll look at it.

It doesn't

Just a bunch of jealous people from /sci/ who thought they were superior for studying math and are now stuck without jobs at graduation.

Fourier/Harmonic analysis is becoming popular in complexity on anything with communication complexity and operations on the boolean cube. Whenever you want some tricky bounds, analysis using saves the day, and a lot of the best general TCS papers (imo) touch analysis along with enumerative structures/counting arguments/generating functions. This is why I'm a big fan of analytic combinatorics, which was spawned to get precise bounds on average-case complexity. Algebra is indispensible to any TCS researcher, but analysis is slowly but surely making its way into more areas of the canonical study of CS

There are several examples of posts where people admit it has. These are exceedingly disturbing numbers
I studied math and CS, had a job, but then went to grad school since it was way more interesting. Everything on grad CS shits on industry as far as a balance between fun and interesting problems and quality of life.

Programming and CS are two different entities. i do agree that programming is heavier on the logic, which is why it's more important to talk about type theories/category theories/classical mathematical logic when studying things like type verification and compiler construction. Algebra is also useful in these areas, since you want to impose structure on various semantics

I feel like if anything OP should be bitching at people who complain about learning Linear Algebra, which is insanely important for fucking everything. Anyway, if OP is a grad student, I feel bad for him being suckered into a dead end.

Ah, I didn't think about relating harmonic analysis to communications. Never heard of the boolean cube nor analytic combinatorics, they sound like interesting topics to research.

Right, I suppose I've never quite "gotten" CS as a discipline. I've always just looked at programming as a (mostly) fun way to solve problems intelligently). I'm actually trying to learn about compiler construction techniques right now since they are supposedly applicable to many other domains, but I keep getting bored out of my skull reading these textbooks.

Lol, OP here. Theory research, especially in CS, isn't a dead end at all. Why is grad school associated with this? I've got like 5 papers done in my first two years, a lot of conferences, a lot of connections, and my standard of living is fine. Being "suckered" into grad school is mostly a meme among people who don't see how mathematics/science is an interesting career path.

Compilers are very interesting. I suggest starting with MIT's SICP to do interpretation first, going through Scott's programming language pragmatics for basic language theory, and then tackling the dragon compilers book, skipping anything you already have familiarity with.

Oh, also on the topic of linear algebra: i generally see CS students shut up and accept that linear algebra is important for everything. I just wish they were forced to learn about vector spaces, transformations, and operations on different fields. LA is relevant in CS once you have those down (most of pagerank uses a stochastic matrix, and a lot of CS applications use [math]\mathbb{Z}_2[/math] anyway)

Does Jow Forums not have LaTeX support? That's annoying

Eh, I have a biased perspective from engineering. I've seen a lot of grad students burn out after spending years pursuing a PhD. My general impression of most grad students is that they do a lot of bitch work for their advisor for dubious benefits.

I've actually already written a very shitty LISP interpreter a few years ago. It started out as a calculator with LISP notation (S expressions are too easy to parse) and mutated into something more powerful but crippled compared to stuff like racket or guile. I'm pretty sure I should learn the theory to make something that isn't so shitty.

Interestingly enough, one of my big research projects prior to grad school was with a physics professor and engineering professor about modelling wave scattering without using PML since it breaks easily. That was a lot of fun, and I got along with them really well. I've had nothing but positive feedback while working with other engineering researchers. That being said, here in theory, your adviser generally gives you interesting directions to explore and helps you with derivations. I can count on my hand the number of times my adviser has asked me to do anything related to labor rather than research, but it hasn't been anything that I didn't finish in like 2 days.

Also, in that case, I suggest just skipping SICP's interpreter and going to programming language pragmatics. It's a very easy, top level overview of language basics. One of my favorite professors in undergrad was a guy who wrote multiple C compilers in the 90s and corresponding symbolic debuggers

Thanks for the recommendation. Would you still recommend the dragon book? I've heard people say it's somewhat out of date and recommend Engineering a Compiler instead.

I liked it as a reference text. Engineering a Compiler is probably a strictly better text. I suggest downloading both on libgen, starting with Engineering a Compiler, and then switching between books whenever you hit something you don't understand / are bored with

>Se^x
l-lewd

Thank you, user. This thread has turned out much better than I expected.

Calculus is actually fucking fascinating and amazing. It's one of the most interesting puzzles I've come across in my entire life. I fucking hated Calc101. They make it as dry and irrelevant as possible.

Think of it like this; most people like birthdays. Get a bunch of people together, good food, good drinks, good music, good night! But what if instead I put you in a room with no windows or tvs or anything and made you sit there all night then came in and said happy birthday? That sucks right? That's basically what academia has done to math.

So, to answer your question, people have trouble with calc because how they teach it.

>If you can't take a derivative, you sure as hell won't be able to prove (or even utilize sometimes) the pumping lemma for CFG's, talk about PDA/CFG equivalence, etc. You're making a case against yourself here
My CS course did all of those things without taking a single derivative.
That said, I of course know how to take a derivative, in fact I'm currently taking vector calculus.
Not the person you're replying to btw.

>What the fuck does Calculus have to do with pushdown automata?
yo imma jump in here, it's irrelevent. what does english have to do with it? nothing, but knowing english is going to make it easier to work with a team on it. what does calc have to do? nothing directly, but being able to clearly and accurately express yourself with a shared knowledge and vocabulary is critical to efficient teams.

if you're working solo then nothing matters at all. education doesn't make you smart, it makes you better able to understand and communicate with other educated people. that's it.

Funnily enough, the way they initially teach it is pretty valuable when it comes to implementing differentiation and integration numerically. I didn't really hate my course though, so my experience may have been different than yours.

I coincidentally am doing Calc II right now, then III and DE (which is covered in II but also has its own course) over the summer.

>all of CS is pure math in a different context
Strictly speaking this statement is correct. Computing Science is the study of algorithms and procedural knowledge, and is basically all math.

... however
When most people say CS they mean IT & Software Dev. And before you say "well it doesn't mean that and those people are wrong" I'll tell you it doesn't matter, if everybody else is using the word differently than you then you won't understand them and they won't understand you. You're right, CS is math, but colloquially CS isn't computing science.

tl;dr you're strictly correct but colloquially wrong

>Performance is only really a consideration when
Probably not related to the discussion here, but I would like to add performance on mobile means longer battery and a more responsive app, so it's more important than many think.

this times infinity

>I feel like if anything OP should be bitching at people who complain about learning Linear Algebra, which is insanely important for fucking everything
I do hobby gamedev (just released on Steam, yey!) and can say linear algebra is required if you're making anything besides a shitty clone. I also do contract app dev, like xamarin and such, where I almost never use anything related to Linear Algebra.

The problem is the culture of complaining among CS and engineering students, desu. Make no mistake; many professors are pretty mediocre, but also with rising admissions to these programs, you have people who demand:
1) calculus be interesting, not cut and dry
2) calculus be passable
So you have students who complain all the time about their program requirements and the difficulty of classes, ask for curves, etc, but also complain about the material. If you want interesting material, it's gonna be tough. If you want passable, easier material for everyone to take, you have to make it rote.

My argument was in the class, you don't need a direct derivative, but if you cannot pass calculus I, which a surprising amount of people here cannot, then you don't have the maturity to tackle basic proofs in discrete topics.

It gets useful down the road. Everything from analysis in theory CS to applications in brain inspired computing

Arguably, calculus across summations and series are some of the most important things you need for mainstream theory CS. Series are an indispensable, invaluable tool in all CS, since it shows up everywhere in algorithms and correctness

>so my experience may have been different than yours.
I suspect the

I'll agree with you there. I suppose I think that even so, the subtext of my statements is that a CS education should remain rigorous. If we care only about software skills in industry, then there should be only an SE degree with CS separate. The truth, however, is that SE degrees are usually seen as less useful than CS because the big industry positions are about solving problems using the basics. So I'd argue that CS, as the premier degree for both grad school study and as a path to solving problems in industry jobs, should stay rigorous in order to push out quality students rather than people who can open a terminal. My frustration comes from the fact that I think the program could cater to both sides

Thanks. I got tired of being a skilless NEET that felt better than other skilless idjits because I had muh superior blackpill ideology and browsed Jow Forums while being a techlet. I'm going to school for mathematics for a pretty low price and hoping to either become self-sustaining via learning CS or fucking off and getting into the trades since I've got connections in my trade of choice.

>if you cannot pass calculus I, which a surprising amount of people here cannot, then you don't have the maturity to tackle basic proofs in discrete topics.
Anyone with an IQ above room temperature can learn basic calculus. Most of the initial difficulty is because teachers are lazy cunts and don't put in the tiny extra effort to help visualize it.
I probably learned more about calc from trying to teach myself programming than I did in 4 years of high school math.

Calculus isn't tough. This is why it's appalling people can't pass. This is why everyone makes fun of CS majors

the integral of e^x is e^x
fuck you

Attached: 1546702335732.gif (250x241, 770K)

and what is the integral of e^x^2, big boy?

a waste of fucking time

This is literally the easiest calc shit someone could ask, pathetic

math major here. i lol'd

>hurr durr I’m a big boy, I know integrals
>but fuck you, anything I can’t immediately figure out is useless!

The absolute state of vir/g/ins

knowing calculus is useless tho

I agree that for codemonkeying it is useless, but when you're in charge of running something and making it efficient, calculus is useful.

Calculus is pretty cool but acting like the dumbest motherfucker you know couldn't easily learn calculus is pretty stupid. The basics of calculus are easy as fuck. Nobody is impressed that you can remember a few integral limit and derivative rules.

>knowing calculus is useless tho
lmao
Knowing basic bitch calculus is important if you want to bound harder complexity problems. You can offload something that's nasty in the discrete case to something that's really easy to solve as an integral. It also helps a lot to remember shit like Stirling's approximation.

This is why nobody respects entry level software development. You have the most self important kids who are obsessed with this idea of practicum in the short term that they fail to see just how applicable the math they're supposed to learn actually is. Is calculus always useful in software dev? Depends, since some fields rely on it heavily and others (entry level, or codemonkey shit usually) don't. But saying "calculus is useless" is beyond stupid.

I thought that calculus was something everyone could do, especially in stem, but I was wrong after going on Jow Forums, and meeting Jow Forums-tier techbros and biology majors

Calculus is easy but differential equations is pure cancer

II can be a pain from memorization of trig rules

>first week calculus
>calculus is easier than trig or algebra.

As it should be. Also
>trig
>algebra
??
differential equations are pretty cool. I'd say they're the least important part of classical CS, but they have their uses

>Austria
Oh hey, didn't expect to see another here. You were at TU? I've just begun studying, but 'till now the profs are all pretty decent.

>barely passed calc II in univ (CS fag), and have forgotten everything related to it since
>family is full of electrical engineers who can do hard as fuck engineering math in a breeze, i am the only black sheep
>always feel like a hack even though make software that people actually use
>see this thread
Never has my impostor syndrome been so strong

Attached: 0053 - 1qfwZyT.jpg (709x765, 72K)

>Most of you wouldn't be able to even understand these theory/systems results, much less apply them.
Nigger I literally majored in applied maths....

you sound like an absolute faggot, holy shit dude...

And I double majored in pure math and CS.
This is cope

No, main university.
I made a terrible mistake. My saving grace was doing my Master's somewhere else.
I only heard some stories about TU, supposedly about some classes that have like a 94% failure rate, but I'm sure you'll manage. At least you actually get some assistance classes instead of pointless ramblings and being told to just figure it out yourself within two weeks.

Ah, yeah, I heard that technical classes at the main uni were pretty bad and/or easy. Guess it's more of a liberal arts university. But yeah, besides some prof writing like he's about to go super saiyan and not recording his lecture, all have been pretty nice and helpful when asked. As for the difficulty, I'm definitely feeling it on the maths side. Three weeks in and we're already doing proof and shit. And not to go full Jow Forums here, but hearing some foreigner attempt to structure a german sentence and the prof attemping to decipher it every 15min is getting kind of annoying. But since I've started in the summer I managed to skip most of the winter masses, so even this isn't much of a problem.

its just a simple guess and check like you would do with sin(2x)

Wanting people that study at an austrian university to speak german or, let's be real, english at a decent capacity has nothing to do with being Jow Forums. You shouldn't go study outside of your fucking home country unless you are doing well in english and are willing to decently learn the local language. That's why I did when I went to Japan.