/sci/ - Science & Math

>>10107367

>>>/g/68278482
>>>/g/68278616
>>>/g/68278632
>>>/g/68287883
>>>/g/68288244

>>10107393
The absolute state of this fucking board
CStards btfo

>>10107367
A CS minor literally got me my job as an EE.

>>10107367

All I know is that in biomedical engineering, extra programming in MATLAB, Solidworks, etc is required, so I can't see how extra programming can hurt you

>>10107367
also
>C++/C in EE department
can't substitute it
>substituted random process course
can't substitute it
>EE/CpE computer architecture course for a CS computer organization
they all take the same one
>took a watered down CS Discrete Math course
as opposed to taking no discrete math as pure EE
>took a weaker data structure class than CpEs do
they all take the same one, and it's in C/C++
>took a superificial CS Numerical Methods class
they take the math one and it requires DEs

epic

>>10107367
If you minored in Math as an EE, you fucked up.

>>10107411
>can't substitute it
>as opposed to taking no discrete math as pure EE
>they take the math one and it requires DEs
Depends on the college
>they all take the same one
You sure it was computer architecture and not a digital logic class that briefly covered architecture?

>>10107367
>CS ranked higher than EE at my uni
>tfw higher level EE classes are watered down versions of CS classes

>>10107485
larp

>>10107485
>>CS ranked higher than EE at my uni

That's not how it works nigger.

My school's CS and CpE share the exact same classes except for CpE taking 2 cmos classes and a materials class

>>10107711
Then it's a shitty CpE program

>>10107725
all CpE programs are inferior to CS
this is at a top 10

>>10107834
No

>>10107863
haha lmao what is this larp
CpE is literally the biggest brainlet major next to CS. it's the running joke of engineering.

>>10107393
holy fucking shit

>>10107393
>Calculus 1 is basic differentiation and integration in one variable
>Calculus 2 is sequences and series, Taylor approximations, and a few integration tricks
>Calculus 3 is vector calculus
Lel this is all high school shit in EU.

>>10107393
>thread specifically for people to talk about their fuckups
>LOL DEM DAM CSTARDS AMIRITE
Am I missing something? The thread is just drawing in people who can relate to the topic.

>>10107985
Physics majors don't fail calculus ten times
Math majors don't fail calculus ten times
Engineers don't fail calculus ten times
Chemfags don't fail calculus ten times

>>10107992
There probably are some who have, but they just didn't post spontaneously about it on 4chan.

This site's full of fuckups like that.

Why is c/c++ seen as better? Once you know one lang you can know the rest?

>>10107367
>That picture

If you think employers think on that level you're probably never going to get an industry job.

>>10108138
You can learn Java in a weekend, C/C++ is actually moderately difficult.

>>10108138
>It is practically impossible to teach good programming to students that have had a prior exposure to Java: as potential programmers they are mentally mutilated beyond hope of regeneration.
>The use of Java cripples the mind; its teaching should, therefore, be regarded as a criminal offense.
>~Edsger W. Dijkstra

>>10107367
CS guy are the fat guy on /fit/

>>10108138
>no pointers
>no unsigned math because "nobody understands it"
>JVM
>unportable (and much harder to fix than C)
>forcing OOP on everything even when it makes no sense
>needlessly verbose

>>10108138
Java is the lowest common denominator of programming languages.

The issue is not that the language itself is inherently broken (especially for beginners who don't need advanced functionality or extremely precise control), it's just that Java is often taught at a very low level to brainlet future codemonkeys, because Java is the codemonkey language par excellence.

>>10108155
>>no pointers
Wrong. Also, why would that be something bad?
>>forcing OOP on everything even when it makes no sense
Nobody forces you to program object oriented. You can use classes just like structs, with the only difference that instead of allocating them you create them with "new", and don't have to take care of freeing them afterwards
>>needlessly verbose
You need less commands for everything than in C, how is that needlessly verbose?

>>10107367
>CpE requiring more rigorous math than CS
>not the other way around
Where the fuck does this meme come from? CS is literally an offshoot of mathematics, whereas CpE is an engineering discipline and engineers are the ones notorious for disregarding mathematical formality. Not to mention CpE is MORE focused on programming and hardware and thus far more >>>/g/ay than CS.

>>10108178
>CS is literally an offshoot of mathematics

>>10108141
Then what level do they think on?

>>10108138
It's not, /sci/ just doesn't know any real programming languages besides c/c++/java(R, Matlab, Python doesn't count).
None of this matters anyways because any decent CS-fag should have mastered several languages by the time they graduate along with being able to be decent at a new one in a week or two.

>>10108138
>Why is c/c++ seen as better?
Because it's superficially harder than the java version, and the whole point of threads like this is to show off how badass you are by how difficult your courses are. To the sort of people who visit these threads, if people invent an easier way to explain their favorite topic, that's a very bad thing because they lose badassery points that way.

>>10108437
>C++ is superficially harder than Java

>>10107992
>math majors are too autistic to pick up people larping as retards for fun
Most of this site is satire btw

>>10108512
I guess they failed the classes because they were LARPing as retarded

>>10108550
Some people are committed to role play ya know

>>10108429
> R, Matlab, Python doesn't count
Spotted the triggered brainlet

>>10108429
>>10108577
The only real programming languages are assembly, C, and FORTRAN. If you disagree, I hate to break it to you, but you're a brainlet in the programming domain

>>10108586
lol

>>10108586
*blocks your path*

>>10108178

>>10108150
Dijkstra also doesn't like to use picture to explain proof sketches and uses retarded notation for quantifiers

>>10108601
D^-1

>>10107992
>Math majors don't fail calculus ten times

I'm a math major, and I know a math major who failed calc 2 four times. It happens more than you think.

It's all a larp. School isn't hard until you do grad school, period, and even then, that's more the material than the actual class.

>>10108622
I know this is a meme and all, but nobody talks about basic sorting in grad school. All of the papers I've read in sorting actually come from pretty insane and interesting mathematical motivation.

That being said Complexity Theory is a pool of incredibly difficult math classification problems

>>10108138
Most people will tell you that it has something to do with difficulty or incompetence. The answer is a little more complicated.

It's the issue of driving with an automatic vs a stick shift. Java reinforces OOP design in everything it does, and because it doesn't delve much into low abstraction computer organization (where all the deep magic and black magic lies), it often doesn't prepare introductory students for learning about solving hard problems. They are usually told to use a library to solve a hard problem or reduce anything they can't immediately use a library for into a few data structures or basic algorithms. It also doesn't mesh well when trying to tell students to think in a different way.

The problem is that Java does this by design in order to ensure reliability on the market. The abstraction and simple (yet inefficient) way you turn all data into objects lends itself well for symbolic computing on objects makes banging out a rapid prototype really easy. When the work you're doing is interesting yet sensitive, you might want to work it out in a language where there aren't a million things to think about in order to see if it works in principle.

C# and C++ have these features as well, with the former being basically Microsoft's Java and the latter being bogged down with a million libraries.

Pure C has a simple ruleset and philosophy, and it allows you to go as far as your imagination if you're willing to put in the work. That being said, as much as I love C, I have to recognize that it isn't the right tool for all jobs.

>>10108183
Undergrad CS and actual CS are two different things. Most of my favorite papers from professors at my school have been CS researchers collaborating with math researchers or vice versa

>>10109084
>not getting that all the name dropped subjects are trivial as fuck

>>10108577
I will say that MATLAB is very bad for /sci/. I think it paints a very reductive view of programming in most engineering students, as they use it as a big calculator by matching packages to problems. I would say that MATLAB is a good collection of FORTRAN libraries for matrix manipulation, but it's not really programming out novel solutions. There's a difference between using rref on a matrix and implementing a scheduler that has a provably efficient throughput, provably solves deadlock, gives a close approximation to NP problems like register allocation, etc.

>>10109103
Oh shit, all of algebraic topology and HEP is trivial as fuuuuuuck

>>10107992
What kind of uni let you fail same course 10 fucking times? If I fail same course twice I'm instantly out.

>>10109106
>as they use it as a big calculator by matching packages to problems
They literally just want to solve problems. What the hell

>>10109201
They don't want to solve problems as much as offload it to pre-existing methods and libraries. It's fine using MATLAB, but it's pretty stupid to think that you've got it all figured out just because you write commands and basic Package.SolveProblem() all day in scientific calculator++;

>>10109106
>I would say that MATLAB is a good collection of FORTRAN libraries for matrix manipulation

found the retard talking out of his ass

>>10107393
cs majors confirmed most based

I get that CS hate threads are mostly self-hating CS grads pretending to be math PhDs, but why is it the only subject? With all the selfhating subjects there are, CS seems to be the only that makes everyone mad. Is it popularity?

>>10107888
>HS level math in EU is equivalent to college intro courses in USA
>EU students still flock to America to go to college

hmm, makes you think

>>10109477
Because it actually is a joke

>>10109483
No they don't. They just go there for research gibs because certain things are monopolised.

>>10107367

Hey /sci/, CS major here. We all know where your animosity comes from. It's the fact that CS majors are going to replace you. Yeah, everyone one of you. You see your brain power cannot compare to the computational power of distributed computing. Mathematicians and physicists will be without a job once the AI (designed and programmed by CS majors btw) solves all your fancy "problems". HAH!

/sci/ fears computer science, because your collective brains just aren't as powerful as distributed computing. HAH!

>>10109672
>computing

have u ever heard of halting problem?

>>10109100
>Undergrad [X] and actual [X] are two different things
They say this about every fucking subject. It's pretentious.

>>10109696
Grad student here, it's pretty true that only my upper div courses even slightly resembled grad school

>>10109672
That meme was already posted here: >>10108622

>>10109696
I think that's because it is in fact true for every fucking subject.

>>10109826
>PhD students are the ONLY ones who can say they have any knowledge of [X]
>t. PhD student

>>10109696
That's cause it's true, bud. You really don't even glimpse the current discussions in any given field until you're at graduate level. You only really begin to learn meaningful knowledge in your subject at 400 level courses at the earliest (although there's exceptions to the rule).

Your first 3-4 years of study are simply laying the groundwork.

>>10109828
No, it's that once you hit a good grad school, you realize just how incredibly foundational yet incredibly basic undergrad education is. If you aren't ready to accept that grad school and research/graduate work are the de facto standards to the actual quality of a field, then you're the pretentious and immature one. I'm not even in CS. I get that most CS hate threads are about venting about stupid undergrads, but people on /sci/ take the meme too seriously and forget that the field has merit in spades.

Well. except ML. Most theorists on all sides (CS included) I know want ML craze to die anyway

>>10110050
I know machine learning is largely over-hyped in the media, but you can't deny that it has shown great potential compared to classical algorithmic approaches in areas such as image recognition, game AI (AlphaGo/Zero), machine translation and so forth. It certainly is not going to die; if anything, ML will only get more powerful as computation power increases and the existing techniques are slowly improved.

I don't know who these "most theorists" are who you are referring to, but they clearly don't know what they're talking about.

t. CS PhD student specializing in ML

>>10107393
Why did you link your own posts?

>>10109526
Yeah, but why is this subject more funny than, say, your average Teh Bingo bongo theory joke about biology? It has to be popularity

>>10110164
I have a great deal of respect for ML/AI theory. The ML that I (and many others) abhor are the ones where researchers tweak parameters without much understanding of why things are going on and call it a paper. I don’t mean to say that ML is inherently a bad field. However, it’s annoying when conferences push the current trends and reject papers that don’t have some spin to the trend. It’s been incredibly annoying dealing with “aaaaand this is how it connects back to ML”

The problem isn’t ML; it’s that the ML craze has allowed the good ML papers to be buried under garbage. I know that a good amount of ML is done on experimentally alone, but I feel like there needs to be a stronger long term plan than just performing ML techniques on every problem we can think of.

It’s an important and interesting field, but the (notable) over saturation mixed in with publish or perish has facilitated a significant dip in quality. At least with theory (including ML theory papers), there’s some trace of methodological rigor.

tl;dr ML is a fine field but the good is buried under the heaps of bad, and I might just be salty that I have to write more grants

>>10110314
My department feels the same way

>>10109281
"Let's make life deliberately difficult for myself"
Tell me about how a client would rather you bill 20 hours because you didn't want to use a prebuilt design
"Pfft, battery drill? I prefer using screwdrivers, also your house is going to take another 18 months to put together"

>>10110880
That’s not even what I was referring to.
It’s fine to use libraries. What I’m getting at is that the use of libraries doesn’t make you proficient at writing code or solving problems. A lot of people tend to confuse the former for the latter. MATLAB has its uses, but you have to realize that it’s stupid to think you’re on top of the world and a proficient coder if all you can do is run premade libraries

>>10110880
To complete the analogy, I’m complaining about people who use the battery drill and then claim they understand all the inner workings and subtleties of design of the drill

>>10107397
Doing what? Software?

>>10107367
>>>/g/68320097