Old /dpt/ at >>51928020
What are you working on?
1st for parallel implementations of averaging two integers in C
Second for anime
java
C is still shit.
sixth for python for being productive and easy to use
>>51936223
Does anybody know of a better algorithm to find the maximum consecutive sum of a list? This takes nearly a second and goes through nearly 20k iterations for a list 10k items long, and that's a little on the inefficient side for me.def max_subarray(List):
result = None
for j in xrange(length,0,-1):
for i in xrange(length-j+1):
listsum = sum(List[i:i+j])
if listsum > result:
result = listsum
save = i
List = List[save:save+j]
length = len(List)
return result
>>51936590
>Python
>Caring about performance
>>51936590
There's probably some math I'm missing out on due to not enough education, a link to resources or search terms to google would probably be enough.
>>51936590
First see if there is an algorithm that does what you want
Then implement it
Then depending on the language you're using try to optimize it a bit, i.e., see if you're using the most efficient methods to iterate without duplicating memory to minimize allocations and that type of things that slow down the program.
Try to use contiguous memory, I'm not sure how List is implemented in python.
>>51936590
Google lists an O(n) algorithm. Your algorithm looks more like O(n3).
>reading tutorial
>"so here's how you do this"
>10 minutes of reading later
>"remember that code we wrote earlier? well it fucking sucks, do this instead"
I hate this shit
>>51936725
>>51936709
thanks, found it
>>51936739
codecademy?
I think I remember seeing this when running through it to learn ruby syntax
It was either variables or loops, I ended up using them anyway
>>51936590
What do you mean by consecutive? Seems like your algorithm is finding the max sum of any two elements of List.
>>51936739
better to give you the right code from the beginning so you can just paste it in your program instead of understanding how it works amiright
>>51936739
what do you suggest instead?
I think it's a good approach, first the tutorial introduces a naive solution to ilustrate how to solve it and then gradually improve it by introducing more complex features you're trying to learn.
>>51936778
this
>>51936739
It's usually the correct way to introduce things, especially if it's about features that people misuse.
Of course, you need to make an effort to understand why it sucks.
If you were going to an university, they tend to make you learn assembler for literally no other reason than to make you appreciate the luxury of even a simple low level programming language introduced later.
>>51936786
Well I'm going through SFML tutorial now and they have seem to follow this format of
>here's an overview of what it is/does
>here's a common usage
>here's some other things it can do but go read the API to learn about them
>here's a note or two about common mistakes/misuse
>>51936739
I hate reading books that do this too. I go through a chapter writing 5 different iterations of code and I end up coming out just learning useless clutter I could have learned if I skipped it entirely.
>>51936590
probably the same as far as big-O is concerned, but try thisdef max_subarray(l):
return max([sum(l[:i]) for i in range(1, len(l) + 1)])
if it's still too slow try numpy
>>51936858
I don't see nothing wrong with the approach you just described.
And this might be interesting to you: http://it-ebooks.info/search/?q=sfml&type=title
>>51936778
When I'm taking notes from a textbook I don't like writing down snippets from examples only to find it could be simplified or shortened. It's not a problem if they preface it with "this is a naive solution but there is a better way to do this", but when another example just shows up later and it's "oh by the way, this is better", that's annoying. And fuck you I never copy-paste.
>>51936867
This guy gets it
>>51936918
>http://it-ebooks.info/search/?q=sfml&type=title
Thanks. I'm just scratching the surface of it right now but I feel like I'm going to be sticking with it so I'll boomark those.
well dpt?
>>51936881
just testing some shit out but if you're operating on a numpy array then this version is almost instant for a 10000 element array
ex.l = np.array([random.randrange(-100, 100) for x in range(10000)])
def np_maxsub(l):
return np.max([np.sum(l[:i]) for i in range(1, len(l) + 1)])
np_maxsub(l) # almost instant
Are you ready for this /dpt/?
I don't think you're ready.
http://blogs.microsoft dot co dot il/tamir/2008/03/02/how-to-high-performance-graphics-in-wpf/
>>51937020
>homeworklist.Average(); //why use lambda here
>>51937115
It's a meme from one of the last few threads
>>51937106
WHAT THE FUCK
>>51937106
This makes my penis cry
>>51937020List<int> someList = new List<int>() {...};
int average = someList.Sum().Select(x => x/2);
Did I do it right?
>>51937106
>>51937106
>>51936223
So is there any online small easy projects for beginners with C# ?
I tried 2 courses in Udemy one for beginner one advanced, and I am reading some free books right now, but I find my self to learn better if the thing is meaningful, not just some cutoff parts of a "useful" code.
You get the idea, learning while coding something easy. But I still can't think of anything serious myself.
Working on writing a basic compiler. Any books I can sue to help me figure out how to properly write and implement a fsm for my parser?
>>51937152
>x/2
no
>>51937020let avg = fun nums ->
(nums |> List.sum) / float nums.Count
[1.0; 45.6; 3027.2; 22.0]
|> avg
|> printfn "%d"
>>51937106
>>51937186
You're right, I have no idea why I put a 2 in there. It should besomeList.Sum().Select(x => x/someList.length);
>>51937106
eh, comes out fine if you paste into an editor
>>51937020listAvg = \x -> (foldr (+) 0 x) / (fromIntegral (length x))
>>51937115
>LAMBDA
>>51937152
>java
no. and your code is also wrong.
>>51937251
>why use lambda here
>>51937115
>>51937020
or everyone's initial thoughts when they're asked to average a list
float(sum(list))/len(list)
which I suppose could be rephrased to
g = lambda x: float(sum(x))/len(x)
>>51937287
your life is wrong
>>51937132
>>51937144
It's Hack Team Style code. There's a Ruby gem for it.
Why isitertoolsso based?
>>51937287
That's C# mate.
>>51937309
>>51937250
>>51937228
>>51937191
>>51937152
>>51937115
>using built in list and sum functions
Absolutely disgusting. I suppose youre worthy of the cs grad meme.(define (av x)
(/ (foldr + 0 x) (foldr (lambda(y z) (+ 1 z)) 0 x)))
>>51937295
>I cant answer the question
>>51937589
>using built in list [...] functions
>foldr
>>51937020cmd = lambda x: average(x)
average () finds the average of a list, and is defined at the beginning on the program
>>51937589(define (av x)
(/ (foldr + 0 x) (foldr (lambda(y z) (+ 1 z)) 0 x)))
Trying to remind myself how to use pointers by coding a max heap. Man, I'm rusty.
So writing an insertion function for the heap class.
Pseudocode
>if root.data is NULL
>>insert value into root.data and call it a day
>else
>>If value is greater than current node.data, traverse to right child and consider that
>else if
>> value is greater than current node.data, traverse left child and consider that
>when child is null; create new node , attach to parent; fill in data value; prime children as null
Assuming this makes any sense, my questions are; does it matter if I drop left or right when data value is equal to a node's data value? (ie, I'm trying to put 17 into heap, 17 is already there) Do I have to be consistent (always right or always left) or does it not really matter? Should I just add a int dupe to the class to the node's data and keep a tally of how many duplicates are represented by a single node? (wasteful for the ones that aren't duplicated)
And second, more importantly; I want to iterate through the heap, starting at root and descending, to find the bottom where I can attach a new item. How do I write this? I've an idea for what I'm doing, I'm just trying to remember how to write and shuffle through various copies of the struct.
>>51937106
>http://blogs.microsoft dot co dot il/tamir/2008/03/02/how-to-high-performance-graphics-in-wpf/
fucking jews
>>51937589
>instead of using the powerful tools easily available to me I should make my own
>>51937607
>this guy thinks foldr is a built in list function.
kek. its an abstraction of a simple function code. You probably dont even know what it does.
>>51937722
folds are functions, scrublord.
>>51937287
>java
>>51937655
>powerful tools
Funny way of saying crutch
>>51937636
Its by far the most elegant solution to this problem and actually solves the problem in a meaningful way.
>>51937637
Gah. C++. I'm tired.
>>51937742
so you dont know what fold does. to be expected desu senpai
>>51937771
>I'm a strong independent programmer who don't need no library
Enjoy reinventing the wheel.
>>51937799
>I cant make a wheel
Pathetic
>>51937782
>Public:
>Protected:
>capital p
start over
>>51937791
I do know what folds do. They're still functions.
>>51937830
Its an abstraction.
>>51937863
So are functions, what's your point? They are abstractions that aren't functions? Because that's wrong.
>>51937881
Arguing with lisp fags is like arguing with lisp fags is like arguing with lisp fags is like arguing with..
Now we know not to bother.
>>51937830
>doesnt know the difference between regular functions and higher order functions
Are you proud of your cs degree senpai
>>51937901
Oh look. The guy who doesn't know what a tree is.
>>51937881
You dont know the difference between a procedure and a function. Please dont make yourself look this bad.
>>51937942
Uh.. I know what a tree is.
>>51937914
They're still functions. Just admit it, and we can all move on.
>>51937816
Yeah, like I said, pretty rusty.
>>51937722
>You probably dont even know what it does.
>takes CS 101
>acts superior
*tips graduation cap*
I just got an internship job offer and i havent met the guy if be working with. Its with dell for a research side project. I have to sign a bunch of papers and shit, but this would be my first real programming job.
Any first timer job tips?
>>51937954
>admit to something irrelevant
thats pretty sad. Its apparent you dont know what foldr does.
>>51937991
>foldr and lambda in cs 101
Thats some uni you attended. Im a bit jealous.
>>51938010
https://wiki.haskell.org/Fold
>function
>>51938010
It doesn't matter what arguments it takes, it's still a built-in function. You're way too retarded to try to act like a pedantic autist, stop it.
>>51938040
>I can wiki
Is that how the kids are getting their cs degrees now desu sempai.
>>51938042
>it's still a built-in function
This is too good. People who dont know what higher order functions are trying to talk about them.
Your argument is like saying programming languages that arent assembly are nothing more than functions that translate to assembly. You clearly dont understand the concept of abstraction.
>>51938100
higher order functions aren't functions, ok.
>>51938150
>I dont know what a higher order function is
Let's talk about snafucation or averaging two integers in C or something.
This shitposting has gone too far.
>>51938163
>I dont know what a higher order function is
Yeah I noticed.
>>51935771
I wrote an HTTP daemon based on http 1.0
It was a couple of thousands of lines and was a lot of fun. I added a GUI so normies can use it. I stopped maintaining it when I did a stress test with http req floods and it only lasted 3 minutes before total DoS
>>51938195
Admitting you dont know is the easiest path to learning. Some day youll look at your degree with pride. But not today.
>/g/ arguing about something they heard about once and never used
Anyway this is my current little project, map manager for halo 2 (you can only play the first 3 levels on an unactivated copy, but renaming other levels works kek)
>>51938204
nice
Anyone else learn OpenGL with file:///W:/openGL/Tutorial0.3.8/html/index.html ? I like it so far, but I'm curious if anyone has any supplementary resources ?
Implementing stack in Haskell:testStack :: Stack Int ()
testStack = do
push 2
push 3
push 4
push 5
five <- pop
pop
push 6
pop
push 7
push five
toList (testStack)
--equal to [5,7,3,2]
This was fun -- and it was interesting to see how imperative it ended up. That said I much prefer pattern matching on a list for stacklike operations.
>>51938237
>file:///W:/openGL/Tutorial0.3.8/html/index.html
>>51938237
>file://
lmao
>>51938213
Well I definitely learned something new today: that you're not just a faggot you're a higher-order faggot.
>>51938225
>renaming other levels works kek
It's nice when videogames don't pack everything into some massive 69gb game.dat file or something.
>>51938225
>using for in a functional language
This is for comedic purposes right anon...
About to start making an email client with PyQt. Seems ezpz.
>>51937815
>not knowing the difference between can't and shouldn't
>implying anybody in this thread can't make their own sum function
>implying anybody in this thread can't make their own length function
You're just fucking autistic and probably entirely unemployable. If anybody under me made their own sum function when one was already available, I'd start wondering why I hired them.
Someone want to try out this program on their machine and tell me if it works.
https://github.com/ccosm/Security-Cams
Just start up client.py and server.py and enter the info on the server thing into the client thing
You'll need PyQt5, v4l2, and you might need to recompile the shit inside of the huffman folder.
>>51938291
Yeah I love it.
>>51938304
>Ive never used functional language before
>>51938265
Indeed, the PC versions of Halo games were pretty open to modders which is cool.
>>51938274
GUI widgets are objects in Racket, and besides, for/* is clearer in most cases.
http://docs.racket-lang.org/style/Choosing_the_Right_Construct.html#%28part._.Traversals%29
>>51938324
haskell happens to be the only one I've used, but it has both sum and length. I fail to see your point.
unless you're saying I should rewrite the sum and length functions in that too, in which case you're definitely autistic
>>51938323sudo pip install python-qt5 100 ↵
Downloading/unpacking python-qt5
Downloading python-qt5-0.1.10.zip (57.5MB): 57.5MB downloaded
Running setup.py (path:/tmp/pip_build_root/python-qt5/setup.py) egg_info for package python-qt5
The binaries distributed with this version are Windows only,
- If you are on linux, look at
github.com/pyqt/python-qt5/wiki/Compiling-PyQt5-on-Ubuntu-12.04.
Man...
>>51936739
>tutorial/book
>copy paste this code snippet and run it
>look at the result
>now copy another snippet
When will people learn that exercises are about 10000% more useful than just walls of text and code dumps?
>>51938364
>that sum example
Its like they expect the reader not to understand lambda expressions. Youre above that I hope.
>>51936739
>>51938440
I've always been a fan of trial by fire.
Get on one of those coding sites and just attempt to solve the problems. Google when you need to, but never google the actual problem
>>51938438
There should be a package for it in whatever package manager you use. Apt has one anyway.
>>51938378
sum a list in a functional language without the use of functions or higher order functions. I doubt you could without googling the solution.
>>51938474
Only for python 3, actually. Is it for python 3?
If so, you should really ditch /usr/bin/python, as a lot of people have that as python 2.7
>>51938470
Yeah, but when learning a new language that's usually not a great way to go about it since your grasp of even the basic syntax is completely missing. I think a book with good exercises is the best way to learn, because then the exercises can be adapted by the author so that they will always be solvable by the reader while still teaching them something.
>>51938478
(define sum (lambda (x)
(if (nil? x) 0 (+ (car x) (sum (cdr x))))))
Is anybody here good with bit arrays in C? I'm looking at an implementation used for a different purpose (paging) and dont understand why the return value is i*4*8+j instead of i*32+j??
Can anybody explain why??
>>51938456
Yeah, I expect people are smart enough to realize for/fold combines foldr and lambda as one function and it's better to use it lmao
>>51938491
Nope it works on 2 and 3, developed with 2
>>51938510
nil?, car, cdr are all functions, and of course you are also using sum, which is a function (even though you defined it). The post you replied to clearly says don't use functions.
>>51938511
There's literally no difference, it was probably just done to make the 4 bytes/u32int, 8 bits/byte thing more explicit.
>>51938456
>writing more verbose code with less abstraction is better
If you want Java, you know where to find it.
>>51938510
>not tail recursive
>>51938542
Yeah thats what I thought. I saw it on two different sites so I thought there was some odd wizardry behind it.
Thank you.
>>51938512
if youre just gonna sum a list then (foldr + 0 lst)
is all you need.
> for/fold combines foldr and lambda as one function
My sides. You dont know what a lambda expression is.
>>51937106
>>51938541
What are you trying to prove?
Trying not to use functions in a functional language?
Hey, i'm trying to figure out how to assign globals and locals to a file, i'm copying some of the syntax from a book, and this came up. can anybody help explain whats wrong and why it won't build the code? it's stopping at the (float 11;) part
>>51938646
That the post you replied to gave pointless constraints
>>51938669
you can't name a variable "11". Variable names can't start with digits.
>>51938669
Oh sweet jesus
Holy shit, why is the development environment on Windows such a nightmare? I'm starting to understand why some people just build everything from scratch... it's a miracle any software on Windows works at all. So many conflicting libraries. Building debug versions of older VC runtimes is a horrific endeavor.
>>51937106
>.co.il
>>51938671
>I cant program in functional languages
I bet you cant even implement your own sorting algorithm
>>51938669
>#include <stdio.h.>
No trailing dot. So #include <stdio.h>
>main()
While technically not incorrect in C89, don't do this. Implicit int is a cancer.
>float 11;
A variable name can't start with a number and can't consist only of a number.
Also you forgot the closing semi-colon in main() and there is no prototype for prAgain().
>>51938701
I thought it said l1 except in an atrocious font
>>51938701
the fuck is wrong with this guys book, he keeps making mistakes. god damn it. thanks, anything else i should know?
>>5193866911 = 9.0
>>51938638
Wow! you're actually dumb. All you need is (sum lst).
If you look closely you'll see the lambda is still there, it's the body of the for/ statement
>>51938705
>>51938705
because C was made for unix and everything was built off of C
:^)
>>51938726
>closing semi-colon
Closing curly-bracket I mean.
>>51938728
It obviously says l1 in the book. L for local, g for global.
And you've missed a closing brace after return 0;.
>>51938740
>all you need is sum lst
My sides. I cant find them.
>>51936518
How and why would you want to parallelize that shit?
>>51938749
thank you, that fixed it.. why would he have put prAgain if there is no prototype for it?>>51938726
Is there a non-shit way to do server-side HTML generation/templating? I'm leaning towards using Hamlet.<html>
<body>
<h1>hello, #{name}
renderPage =
let name = "faggot"
in $(hamletFile "website.html")
>>51938740
>If you look closely you'll see the lambda is still there, it's the body of the for/ statement
What the fuck? Do you know what the difference is between an expression an a lambda?
>>51938791
(sum lst) and (apply + lst) are both clearer than your foldr bullshit. enjoy being unemployed :)
>>51938816
add each integer on different cores. Obviously.
>>51938821
Just use C. I got tired of writing my own HTML so I wrote a simple DSL in C to handle code generation for me. It'll work better than any off-the-shelf solution, trust me.
>>51938735
the more mistakes i realise that i made the better this picture becomes
>>51938835
>I dont understand abstraction
kek. Its incredibly clear if you know what fold does. If you dont you dont need to be using a functional language
Python. Paramiko to cisco gear learn autonomously what path to build and build necessary requirements. Keep getting stuck in endless nests I want it to be prettier.
>>51938829
Of course he doesnt. Which is why hes using for in a functional language to begin with.
>>51938829
in lisp there isn't one. I'll even make a simple example so it doesn't hurt your little brain too much. Lets say you have a list of strings, and you want to append something to each.;; /g/ being excessively functional
(define strs (list "a" "b" "c"))
(map (lambda (s) (string-append s "!")) strs)
;; normal clear code
(define strs (list "a" "b" "c"))
(for/list ([s strs]) (string-append s "!"))
see how the body of the lambda is still there?
>>51938887
making stuff more complicated for no reason is actually the opposite of abstraction
What the fuck is a hash, why do I need them, how do I use them.
I want to make something like tender with python but I have no experience managing databases or even making websites with python, I guess django looks fine but I haven't used it yet.
Any help/books would be appreciated.
Where should I start?
>>51939004
A cryptographic hash or a hash for indexing into an array?
>>51939004
Let's say you have a phonebook. A bunch of names, and associated phone numbers. Now I tell you, go find John Smith's number. How do you do it? Do you search through every entry in the phonebook?
No, you know the phonebook is sorted somehow, so you look up where the S's are and start your search there. A hash is just a faster implementation of this kind of thing.
>>51939004
a hash function takes an input and buckets it according to some notion of uniqueness. you might want to a hashing them if you want to have a set of unique values or a key-value set, or another set that needs to be bucketed.
>>51939059
>you look up where the S's are and start your search there
That's not how hashes work. They don't really have a good real-world analogy.
>>51936223
>when you leave little notes in your code of where you're up to so when you go back you can get right into it again
thanks f am
>>51939052
the concepts are the same. the difference is cryptographic hashing functiions need to be difficult to invert.
>>51939087
Well, the computer does some math and figures out magically where the S's start, while a human still needs to flip through the pages.
>>51939004
In cryptography hashed values cannot (without considerable effort and time) be reversed. You store passwords as hashes for example so that if you get hacked the users details are safe, to authenticate them you simply take the users input, hash it, and check it against your stored hash.
Values will become the same hash every time, but you should take care and make sure you use a modern hashing technique and add an element of randomness such as a salt.
Hashes are also used elsewhere, but other anons have already touched on them
>i like my notes. The fuck was I doing here.
>i need better variable naming conventions
>sleep.wait it sleeps and waits
>dont enable this. Kittens might die
>>51939120
It's like that except that it's not organized in a clear way (alphabetically) it's organized in a purposely hard-to-predict way.
>>51938712merge :: Ord a => [a] -> [a] -> [a]
merge [] ys = ys
merge xs [] = xs
merge l1@(x:xs) l2@(y:ys)
| x <= y = x : merge xs l2
| otherwise = y : merge l1 ys
memeSort :: Ord a => [a] -> [a]
memeSort [] = []
memeSort [x] = [x]
memeSort xs = go (splitAt' (len `div` 2) xs)
where len = length xs
go (l1, l2) = merge (memeSort l1) (memeSort l2)
splitAt' :: Int -> [a] -> ([a],[a])
splitAt' 0 xs = ([],xs)
splitAt' n (x:xs) =
let ~(meme1, meme2) = splitAt' (n - 1) xs
in (x:meme1, meme2)
Output:λ> memeSort []
[]
λ> memeSort [6,5,4,3,2,1]
[1,2,3,4,5,6]
λ> memeSort [1,5,420,-1,3.14159]
[-1.0,1.0,3.14159,5.0,420.0]
λ>
Sorry for not seeing your post earlier, I was memeing in IRC
>>51939209
hashing functions do not need to be purposefully hard to predict predict. for example h(x) = x mod 10 is a reasonable hashing function that is easy to predict. even in cryptography, hashing functions are frequently easy to compute (but not invert!)
>>51939059
>>51939073
>>51939144
Ugh thank you, everything I've read now completely makes sense. Ive been looking at the wrong hash as the places I've been reading havnt bothered to make the distinction between the two. Which makes sense, but still confused me. Thanks anons.
>>51938988
>see how the body of the lambda is still there?
Yes. The the body is there, the lambda is not. That's my point. Don't go strawman on me anon.
what assembly is worth learning in this day and age?
>>51939250
What does the "~" do?
C#
Why isn't this working /g/?int GetCellIndex(short x, short y) => { return (y-1)*9+(x-1); }
>>51938274
racket devs suggest to use for/* for iteration
>>51939287
This whole argument is a strawman. In >>51938740
I was explaining how for/list eliminates the need for the lambda because the body of for/list IS the anonymous function.
I know you guys wanna be elite and functional and argue about everything, but chill
>>51939482
Not a C# dev, but shouldn't you just get rid of the =>?
>>51939482
>int GetCellIndex(short x, short y) => { return (y-1)*9+(x-1); }
Remove return and the brackets.
>>51938257
BTFO
>>51936590
Literally a wiki article with example code in python for this exact problem:
https://en.wikipedia.org/wiki/Maximum_subarray_problem
Kadane is kind of a traditional example of "dynamic programming" in algorithms classes. It has linear vs your cubic (`sum` takes linear time) complexity.
Not that it has anything to do with the actual problem but naming params with initial caps is considered bad style, initial caps are reserved for class names (note it's class /names/, when passing a class reference you still use snake case, see the numerous examples of "klass" as an argument in the standard library).
>>51939509;; normal clear code
(define strs (list "a" "b" "c"))
(for/list ([s strs]) (string-append s "!"))
there is no lamda/anonymous function here. There is just a block of code that is looped over.
>I know you guys wanna be elite and functional and argue about everything, but chill
This isn't even some obscure concept. Java, C#, JavaScript and Python among others all support lambda's. They are considered functional concepts, but they exist in almost every popular language that's not extremely low level. You need to know what one is to not look like a complete idiot when working on a team in any of the languages.
>>51939482
What are you trying to do? A function?
remove the =>
An anonymous function? Remove "int GetCellIndex"
>>51937020
>use a rope to hammer this nail
Someone from the previous thread suggested that Javascript was more well thought-out than Python. I do not know if this person is still around at the moment, but I would like to remind them that the language was literally created in 10 days.
>>51939521
Thank you based anon.
So I just started learning some Haskell and I want to stop using the command line now.
I've never used anything for programming other than Eclipse, and visual studio. Currently I am on Linux and I see people on /g/ talk about using emacs/vim for text editors and that something is wrong with using IDE's or whatever.
I just installed emacs and it seems like just a weird text editor... what should I know that I obviously don't?
>>51939715
>but I would like to remind them that the language was literally created in 10 days.
Hardly makes a difference does it anon?
Both languages are pretty shit. I guess I prefer python personally. Never really though of one being hugely better than the other, but python has nicer syntax at least.
>>51939625
This is a very neat algorithm. There's a chapter in "Programming Pearls" that talks about it. The author of the book first came up with aO(n^2) algorithm to find the maximal sum (something like >>51936590), and after thinking about it for a while and discussing it with someone else, he refined it to a O(n log n) algorithm. He thought that was the best that could be done and left it for a couple years. Then he mentioned the problem at a conference, and someone in the audience (Kadane) came up with this O(n) solution in like 5 minutes.
I watched a few videos about it and nobody really explained it to my satisfaction. The way I understand it, the algorithm is based on the fact that you can "prune" a subarray on the left end of the main array if you know that the subarray will never contribute to the maximal subarray, i.e. because the sum of that subarray on the left end is <= 0.
>>51937020average = lambda { |list| list.reduce(:+) / list.size }
>>51939762
To suggest that any language that one has spent only 10 days of thinking on is in any way "well thought-out" is to suggest one has no standards at all.
>>51939785
>someone in the audience (Kadane) came up with this O(n) solution in like 5 minutes.
Some people are not meant for this world.
>>51939743
Vim or bust
Or sublime text 3 if youre a sane normal man
>>51939858
that doesn't really help me.
like other than making text files what does this program do?
>>51939882
>like other than making text files what does this program do?
towers of hanoi.
>>51939638
Yup, and if you don't want to get laughed off the team you'd follow the documentation recommended style and not use lambda all over the place when for/* forms are available.
>>51939882
Those are all text editors.
A text editor alone isn't a replacement for an IDE, since an IDE usually comes with your compiler/build system/debugger etc. built in, and locked behind its GUI somewhat.
For Haskell, all you need is GHC and a text editor. Your "project" is specified with a cabal file.
>>51939901
Seems like an IDE would be less of a pain in the ass then.
Why don't people just use IDE's?
Being serious I saw a guy do this
https://youtu.be/D1sXuHnf_lo
It seems that when he types something it happens immediately.
>>51939840
>To suggest that any language that one has spent only 10 days of thinking on is in any way "well thought-out" is to suggest one has no standards at all.
I recall he based it a lot of off other languages he has worked on in the past. It's certainly for from the worst of languages out there. Miles miles better than PHP. It's far less verbose than Java. And The Good Parts(TM) of it are actually pretty nice to work with in my experience, if you can deal with the dynamic typing and weak typing shite.
The current version has certainly come on a long way since v1.0 too.
It's generally retarded reasoning to claim a language must be shit simply because not much time was spent on it. Might as well claim a movie is shit because it had a low budget.
Also the anon didn't claim it was well thought-out. Simply better thought out than python.
>>51939954
Personally, I don't use an IDE. I find that IDEs always make it more difficult to do "advanced" things like configuring the compiler settings or doing non-trivial library linking.
For Haskell, I don't think an IDE can really get you anything. Cabal automates most of the build process, and if you really want it, most programming-oriented text editors support linting and autocomplete.
>>51939954
>Why don't people just use IDE's?
Almost everyone in industry does. The only people who don't are generally just control freaks. They also generally tend to lack understanding of the benefits IDE's provide to productivity. Or they use text editors that have so many plugins there are basically IDE's anyways.
>>51940096
One of the best programmers I worked with used vi (not vim). As far as I know she didn't use any macros or anything either, just straight vanilla /bin/vi.
I use emacs myself. The only reason I'd use an IDE is for the GUI designer stuff. Most things in an IDE are just a distraction, at least for the purpose of writing code.
>>51940096
desu the only reasons I use vim are:
1) I can have multiple files tiled on my screen
2) it literally matches my firefox and 4chan css
3) its in the terminal
tooltips etc like when you mouse over functions and shit seems awesome though
>>51940195
(cont)
I also don't like having my projects bound to IDEs, having to integrate git/hub with the IDE, etc. It's all just a frontend to what I can do myself in the term, which I'm more comfy with
>>51939954
>Why don't people just use IDE's?
For me, it's two reasons:
1. Interface feels clunky. I do a lot of my editing in Sublime, which has a number of useful features purely as an editor, without having to take up much more than around 10% of screen real-estate for the menu bar and tab bar.
2. Many IDEs have a build system that is over-complicated/difficult to understand. I tend to use a plain old Makefile (albeit with GNU Make extensions) for most things, because there's not a lot of learning curve to understand what everything does. If I have to deal with a convoluted project system that isn't used on any other IDE, then that IDE better damn well be the best damn thing since sex. So far, nothing has met my standards.
Is sml '97 actually viable nowadays? What's the best implementation? MLTon? Any benchmarks available? Are spaces significant in sml like they are in haskell?
>>51940361
To be honest, family, I really like JetBrains IDEs. IntelliJ is considerably better than Eclipse (at least, imo) and CLion is great.
>>51939341
I thought I would need it to make the pattern match lazy (so splitAt' would be lazy), but I tried it in ghci without the ~ and it looks like it works lazily anyway.
Reading https://wiki.haskell.org/Lazy_pattern_match , it looks like let-expressions match lazily by default. So if instead I had been matching on (,) as a function argument, or in a case-expression, I would have needed the tilde, but not in this case.
>>51939954
Non-IDEs are more general and can be made specific on a per-purpose level, such as adding addons to make vim able to perform live, "perfect" code completion with clang. IDEs, on the other hand, are typically tailored to a specific language or environment, making them less versatile. There is great utility in learning one tool completely instead of 40 poorly.
Non-IDEs do not come with 5 billion features you'll only use 1% of the time. One can add features as needed, incrementally. This greatly reduces the program's bloat level, increasing dramatically the starttime, responsiveness and resource footprint. You're also almost certain to be working with standard project management tools that can be used from the command line, or in general from any toolset other people are using, instead of IDE-specific project files. This is a huge bonus for portability, reproducibility and sharing.
Since non-IDEs aren't IDEs, they need only focus on the core task they're meant for (typically editing text), making them significantly more efficient to use for any language even before any specific tailoring.
TL;DR: use vim, pleb.
>>51940417
You know, I've used IntelliJ in Android Studio, and to be quite honest, I was not a fan of the fact that it was constantly swapping memory to the fucking disk. By comparison, Sublime Text is only using 4 MiB of memory at this current moment.
It took me 3 hours to implement A* from scratch in C, from the pseudo-code on Wikipedia. Am I retarded? I feel like I'm not cut out to be a programmer if it takes me this long to reimplement someone else's algorithm.
>>51940550
You're already better than 99% of programmers.
Most people employed as a programmer today couldn't even implement A* if their life depended on it.
>>51940498
Compared to Sublime Text it's fuckhuge, but compared to other IDEs, it's pretty nice.
>>51937020
Professional Enterprise Programmer reporting in.Func<IEnumerable<double>, double> average =
list => list.Sum() / list.Count();
// Example
Console.WriteLine(average(new[] { 1.0, 2.0, 5.5, 27.3 }));
Why aren't you using GNU Nano for all your editing needs?
>>51940664
GNU Vim is better.
>>51940664
It's shit.
Emacs (or Spacemacs if you prefer Vim keybindings) is pretty much better in every way, and easy to configure for literally *any* language.
>>51940687
What kind of configuring do you need aside from syntax highlighting?
>>51940706
If it's an interpreted language, don't you want a REPL?
>>51940708
No, because I don't write in scripting languages.
>>51940723
>he spends 10x as long writing code that isn't even performance-oriented
>>51940723
Opendylan is a compiled language with a repl. Checkmate, masterb8.
>>51940738
>he fell for the interpreted language meme
>>51940751
isn't this literally like running a bash 1 liner that runs make and executes your toy program over and over?
>>51940596
Nah mate, even Visual Studio lags less.
>>51940708
Does one really need a REPL in their editor? I mean, I've got an extension in Sublime to evaluate Ruby in the editor, as a way of generating code programmatically, but that sees very little use. When I just want a REPL, I alt-tab into my terminal.
>>51940849
It's a matter of efficiency. Alt + tab, paste, enter vs. C-x-e
Eval-buffer and Eval-region are also invaluable when I'm hacking scripts together.
>>51940849
>Nah mate, even Visual Studio lags less.
Works for me, m90.
Android Studio is pretty shitty. Don't know what the dev environment is like, I only used it to save snapchats, lel, but it super bloated.
>>51940849
>to evalueate Ruby
Is ruby your goto language for random stuff? What are you working on?
I've always had the impression that ruby is only good for web shit
>>51940965
He's working on his doctoral thesis: machine learning which adequately selects best waifu.
Anyone here C#?
I have 5 decimal objects and need to make sure that their sum isn't greater than exactly 1, but this seems a bit tricky to implement. Any advice?
>>51940778
A REPL lets you give commands to define new values, evaluate expressions, etc., and it can load definitions from a file, so you can mess with them, like if you aren't finished but want to see if the output of the stuff you have so far looks right.
But if you're in bash, you might as well use the standalone REPL program that you probably have for whatever language.
Here's an example of what repl output might look like:codesnips ghci sort.hs
GHCi, version 7.10.2: http://www.haskell.org/ghc/ :? for help
[1 of 1] Compiling Main ( sort.hs, interpreted )
Ok, modules loaded: Main.
λ> 5
5
λ> 3 + 6
9
λ> let meme s = s ++ " is a meme"
λ> meme "haskell"
"haskell is a meme"
λ> memeSort "Memes are for the masses" -- memeSort defined in sort.hs
" Maaeeeeefhmmorrsssst"
λ>
>>51941058
Couldn't this be done in literally one if statement or am I retarded? (I know nothing about C#)
>>51941058double sum = 0;
double array[5] = {0.2, 0.3, 0.4, 0.5, 0.6};
for(int i = 0; sum < 1.00 && i < 5; i++)
{
sum+=array[i];
}
>>51941137
>>51941094
It can be, for some reason I was having a moment of mental retardation and thought I had to use .Compare
Thanks senpai, I was being a retard
>>51941216
>no line numbers in emacs
>that green term background
>not doing -t when you know your term is too small for screenfetch
-1/10
also get banned
>>51941216
>white background
>tool-bar
>menu-bar
>that word wrap
0/10. Try again.
>>51941216
4chan is 18禁
>>51940965
Ruby is my goto language for random stuff (when not using C or C++), and I've even used it in a few homework assignments. Not a few weeks ago, I used it for an assignment to implement el gamal encryption using the discrete logarithm problem, and also using elliptic curves. It is useful for a lot more than just web shit, although I am recently finding myself using it for web shit. I've got a friend who wants to port over a game played on /b/ using simple gets to a web interface, with a chat, some accounts for keeping track of wins/losses/etc... and he's willing to pay a little cash for it, so I'm seeing what I can hack up for him this winter break. Web development's not really my forte, but it is at least decent that I can toss something together in Ruby real quickly for that purpose if need be.
>>51941026
1. Master's degree, not PhD
2. I'm mostly studying operating systems and security.
3. If I wanted to focus on machine learning, I'd be much more interested in making a program that can fake being human and act like a waifu, rather than simply choosing a subjectively good waifu.
>>51941334
I know, I'm just having a laff.
>>51941345
Aye.... although I will be studying some computational linguistics next quarter as an elective. Alas, I will still not end up with the knowledge to make a waifu bot...
(not that any of that matters because even if I make a waifu bot, I will never be able to knock her up, thus making it all pointless).
How do I implement a tiny stack for storing integers?
Do I preallocate the space for the stack and then push and pop things, or do I grow and shrink it as it's used?
>>51941392
>women are only good for getting knocked up
Didn't know you were such a misogynist, Ruby.
>>51941392
What would a half-anime half-human half breed look like?
>>51941421
https://en.wikipedia.org/wiki/Dynamic_array
I've implemented a ton of them before. However, THIS IS WHY I USE C++: SO I CAN JUST TYPE "std::vector<int>" AND LITERALLY NEVER HAVE TO DEAL WITH THIS SHIT.typedef struct {
int* data;
size_t size, cap;
} ist;
void ist_push (ist* st, int a) {
if (st->size >= st->cap) {
st->cap *= 2;
st->data = realloc(st->data, st->cap * sizeof(int));
}
st->data[st->size++] = a;
}
int ist_pop (ist* st) {
// TODO: stack underflow?
return st->data[--st->size];
}
void ist_init (ist* st) {
st->size = 0;
st->cap = 8; // you choose initial capacity
st->data = malloc(st->cap * sizeof(int));
}
int main () {
ist stack;
ist_init(&stack);
ist_push(&stack, 2);
ist_push(&stack, 3);
ist_push(&stack, 4);
while (stack.size > 0)
printf("=> %d\n", ist_pop(&stack));
free(stack.data);
return 0;
}
>>51941474
might have fucked up my size check, should be "if (st->size + 1 >= st-.cap)"
I'm trying to use this FOSS(?) java program:
- runs as a web service, uses at least 500MB of RAM
- impossible to predict how much it takes to process inputs, outputs sometimes after X milliseconds, sometimes after 3X ms.
- started a feature request in their issue tracker 2 days ago, proposed a patch, still haven't got a response
this is shit
>>51941523
hentai@home?
>>51941421
>How do I implement a tiny stack for storing integers?
Using an array or vector and an index for the "top" of the stack.
>Do I preallocate the space for the stack and then push and pop things, or do I grow and shrink it as it's used?
Depends on your use case.
>>51941444
Not quite my views. It's more the point that I want kids in the future, and as such, I've decided it's not something I'm going to compromise on when choosing someone for a life partner. But that doesn't mean that I'm going to look only for this trait. If a woman's only redeeming quality is that she can and is willing to birth me a few kids, I'm not going to take her. I'd like to think I'm worth more than that. Good companionship and loyalty are also highly important. It's just that the kids thing is something I'm going to absolutely put my foot down about.
>>51941558
>Depends on your use case.
I'm just trying to write a stack for this codeeval challenge.
https://www.codeeval.com/open_challenges/9/
Working through a MEAN stack tutorial and enjoying winter break.
Been learning Java for about 5 months, and want to try a small project to really grasp it. Basically, I'd like to help my bosses make schedules at work. It's retail and we have about 30 employees with different availability for each, and shift times/demand vary a lot.
I was thinking of making a scheduling program for practice. GUI, display a spreadsheet for the week, and create createable "profiles" for each worker, showing their availability. The spreadsheet would read the profiles, and color-code the spreadsheet accordingly (pic related). They could just click a cell, and type in a shift.
Thing is, I have no idea where to begin studying to do this. Never worked with it, but would JTable be a good spreadsheet look for this? And for the profiles, how might I go about doing that, arrays of an "Employee" object and have the table read it somehow? Any advice is welcome.
>>51941568
I would say that you should make it resizable then. You have no idea whether you will have 100 or 1,000,000,000 integers to read.
>>51941646
>is called ruby
>gives sensible advice
my head hurts
Learning PyQt to make a simple email client
this is such a meme
>>51941627
If it was me, I'd look at other systems that have already solved this problem. Now you don't necessarily have to use these programs but you can certainly study them to find out how they work.
http://easyappointments.org/
https://github.com/caldav4j/caldav4j
>>51941692
Alright, thanks. Noticed there were lots, and I can grasp the idea, just not all the specifics. Those source codes are helpful.
>>51936223
>when your incredibly complex and long code works perfectly first run
just open sourced my first lib ever /g/, feels good
>>51941748
What's the library do?
>>51941748
>works perfectly the first run
>leaks on the second run
>entire system crashes on the thousandth run
>NASA fires you for ruining their billion dollar space mission
I have learned python and ruby, yet I am having trouble thinking of projects to apply them to. Any suggestions?
