The Nineteenth Byte

Downgoat

12:26 AM

@ASCII-only pls see above code snipet for VSL

ASCII-only

@Downgoat former

Leaky Nun

1:17 AM

CMC: Solve 8x≡-3 (mod p), 0<x<p, where p is a prime greater than 5

f(5)=4
f(7)=4
f(11)=1
f(13)=11
f(17)=6
f(19)=2
f(23)=14
f(29)=25

testcases generated by

for p in [5,7,11,13,17,19,23,29]:
	x = f(p)
	print("f(%d)=%d"%(p,x))
	assert(0<x<p)
	assert(8*x%p == -3%p)

where f is the function I wrote

ASCII-only

@LeakyNun without bruteforcing? :P

Leaky Nun

@ASCII-only I wonder if brute-force would be shorter than my answer

Leaky Nun

1:36 AM

@ASCII-only are you in?

ASCII-only

i have JS: f=(n,i=n)=>8*i%n-n+3?f(n,i-1):i

:/ it does look very ungolfy though

i'm probably missing something completely (or kinda) obvious

seems like it might be possible to use the euclidean algorithm?

Downgoat

@ASCII-only to be fair it is uber-golfy compare to VSL

ASCII-only

2:14 AM

okay so you should be able to use the extended euclidean algorithm but i'm too lazy

user202729

So, 8 = 2³.

And 1/2 = (p+1)/2 (mod p)

So, 8x=-3 <-> x=-3*½³ <-> x=-3*((p+1)/2)³ (mod p)

JS p=>(((p+1)/2)**3*-3)%p probably?

ASCII-only

ah

+p at the end too

user202729

... because it's JS. (truncated modulo)

and x!=0.

ASCII-only

n=>-3*(-~n/2)**3%n+n

n=>-3*(-~n)**3/8%n+n also works

Leaky Nun

2:39 AM

@user202729 that's nice, but there's a small trick that you missed

it will save you... three bytes

ASCII-only

1 byte down, 2 to go: n=>n-3/8*(-~n)**3%n

@LeakyNun is this a math trick

also i'm assuming it's not a modulus hack?

Leaky Nun

@ASCII-only yes, a math trick

what is it with factoring out the /2 lol

@DJMcMayhem hi

DJMcMayhem

o/

ASCII-only

@LeakyNun idk, i think i prefer it because there's less nesting

Leaky Nun

you broke the integrity (integer-ness) of the code

ASCII-only

2:49 AM

oh yeah >_> so just imagine the /8 is back at the end where it should be

i mean JS uses floats so that doesn't change anything, but it's probably more comforting

Leaky Nun

hence integrity

quartata

@Downgoat is x a class or an instance variable?

ASCII-only

can't find the math trick yet. but found n=>n+(-n>>1)**3*3%n instead

Downgoat

@quartata instace

quartata

it's... Kinda weird to have your code split across the class "definition" and the proper constructor

the only difference is that one has access to the arguments

maybe you might want something like C++'s initializer lists if that's what you're going for, just a quick way to init fields

Downgoat

3:01 AM

k

quartata

e.g.

Downgoat

initializer lists are very different from lazy expressions

they are two very different things

quartata

...wait, lazy?

maybe we're not talking about the same thing here

Downgoat

no I don't think so idk what you're talking about :P

quartata

I mean the semantics of class X { let y = ... }

Downgoat

3:04 AM

so you are saying you like C++ initializer lists & duplicating code per constructor versus intial field values?

actually now that I think about it... this answers where they should go

quartata

my concern is that you're splitting the initialization into two separate places. You should be able to look at the constructor and see everything that happens when the class is instantiated

The fact that you had to ask that question is a tipoff that it's not clear, you're not entirely sure what the order of things is just by reading the code

Look at this:

Downgoat

@quartata so you also oppose inheritance where the super-class also performs initialization?

quartata

that's a separate issue

Downgoat

no not really

they have specific inheritance semantics

quartata

class X {
    let x, y : String
    init(y : String) : x=f() {
        self.y = y
    }
}

Now you can have it both ways, but it's all the same place

you don't have to scroll up and keep track of the order

Downgoat

3:10 AM

if you'd like that then sure I suppose you can do that

quartata

this is what C++ does, with slightly different syntax

Downgoat

@quartata this doesn't make sense, with deferred initialization you don't need to keep track of order because it is resolved on-demand without locally captures values

quartata

and as a bonus I think it makes it clear what is a class and what is an instance variable. You init instance variables in the constructor, and nowhere else

@Downgoat I'm talking about how a person reads it. This has nothing to do with implementation, no one cares

Although lazily initializating instance variables sounds like seppuku tbh, I like to use impure functions occasionally in my constructors

Downgoat

@quartata are you seriously saying the semantics of a language dwarfs in importance to scroll distance?

quartata

hell yes

there's no semantic difference though

Downgoat

3:13 AM

.______________.

quartata

I think you have context I'm missing here, frankly

You're seeing something that you've implemented that I don't know

What's the difference between initializations going up there as opposed to down here

Downgoat

what do you mean by 'going up' and 'down there'

quartata

compare the two snippets

Downgoat

If you have something like (hypothetical example):

class BlueView {
    override public let backgroundColor = .blue
}

class View {
    public let backgroundColor = .transparent

    init(in controller: ViewController) { ... }
    init(in parentView: View) { ... }
    init(size: Size) { ... }
}

specifically, even if I were to override an initializer

I wouldn't need to make sure that backgroundColor is being set correctly

very useful for things like Views

quartata

not seeing your point tbh

View's constructor comes first, then BlueView

Assuming BlueView inherits and isn't a mixin

Right?

Mixins are pretty different

...although, this actually provides a great way to handle that

If you make a new constructor in the mixin then it gets that initializer list

Going off track sorry

FWIW here's how I'd structure View in pseudo-C++

user202729

3:27 AM

Can I find corresponding chat account of a main account?

quartata

class View {
    public let backgroundColor: Symbol;

    private init() : backgroundColor = .transparent { ...whatever is common to all three... }

    public init(in controller: ViewController) : init() { ... }
    ...
}

In C++ rather than init() it'd be just View() obviously. That's kinda the equivalent of Java's this(...) constructor but far less ugly. You don't have to repeat the initialization across all three that way

Plus there's inevitably going to be some other code common to all three constructors. Break that off into a private one

@user202729 go to chat.stackexchange.com/accounts/<network profile id>

That will redirect you

Pavel

3:41 AM

@quartata What's ugly about Java constructors? I thought the functioned essentially just like in C++

quartata

"super must be first statement in constructor"

when you type that error message you should reevaluate things

also Java does not have initializer lists

which as previously alluded to are Good

Pavel

Yeah fair enough

New Main Posts

4:24 AM

0

Q: Given a grid, starting at the center and spiraling out, and an ID what is the Position?

Grid: +---------------+---------------+---------------+---------------+---------------+ | id: 20 | id: 19 | id: 18 | id: 17 | id: 16 | | pos: (-2, -2) | pos: (-1, -2) | pos: (0, -2) | pos: (1, -2) | pos: (2, -2) | +---------------+---------------+-----------...

code-golf

ASCII-only

6:12 AM

@Downgoat well a patch that comes with a brick is still a patch :P

New Main Posts

9:28 AM

0

Q: "Hello, name!";

Challenge Your challenge is to make a simple program that takes the input and formats it as "Hello, !" Input Either a function call variable or STDIN, this is the name. Output Hello, ! Example Here would be a program in CJam: "Hello, %s!"[q]e% Probably not the most efficient... lolad ->...

code-golf string

Leaky Nun

10:51 AM

hmm, I thought np.ones(N) would have faster accessing than [1]*N

ASCII-only

11:13 AM

@TonHospel would you mind me asking some perl questions?

Ton Hospel

11:35 AM

@ASCII-only Sure, ask

ASCII-only

11:57 AM

:| sorry internet died for a while

@TonHospel ok so just making sure of something: this doesn't look like it enables any features that may make things golfier, right?

according to here they have shorter quines than PPCG (same with perl 5) - I'm assuming they probably use some method other than q/qw/<>?

Lembik

hi.. can I advertise this awesome answer codegolf.stackexchange.com/a/157980/9206

it deserves lots of upvotes and has only one

Ton Hospel

@ASCII-only only "say" looks useful and we allow that here too :-)

ASCII-only

so the golfers there must be geniuses

Ton Hospel

@lembik yeah, i decided not to bother after seeing that c++ answer. The only way ot further improve it is to vectorize it

Lembik

@TonHospel Oooh that would be cool!

but I am really impressed more than one person could actually understand the paper!

it's not an easy read

Ton Hospel

12:06 PM

@lembik: Yeah, I didn't get it either without seeing the code. They explain very badly

Lembik

I definitely couldn't understand it without the codegolf answers

Ton Hospel

@ASCII-only Or they found a good cheat. These automatic online judges (does that site use one?) are sometimes easy to fool

Lembik

@TonHospel I am still hoping to pose a challenge with complex arithmetic to see how well people can vectorize it

as compilers do a terrible job currently

ASCII-only

@TonHospel easy to fool?

Ton Hospel

@ASCII-only calling external stuff (shell commands), reading your own scource code, that kind of thing

or only working for the cases tested while you know there are subtle things where your program is wrong etc.

ASCII-only

12:16 PM

oh yeah i think that site allows errors (only checks STDOUT), not sure if that changes anything though

Ton Hospel

I never played on them for that reason and that their penalty for commandline options is way too high (lthough here the commandline option penalty is too low)

Ton Hospel

1:14 PM

@ASCII-only Ok, after a bit of searching I found the quine (see me on the scoreboard). It's in fact a fair one

It is indeed say that makes it possible

Leaky Nun

2:03 PM

is (((a+b)%n+c)%n+d)%n faster or ((a+b)%n+(c+d)%n)%n?

Ton Hospel

2:16 PM

I would expect (a+b+c+d)%n to be fastest :-)

since add is fast and modulus is slow

Neil

@ASCII-only did I mention that MapAssignRight doesn't like to be followed by a separator? Try it online!

New Sandboxed Posts

2:45 PM

0

A: Sandbox for Proposed Challenges

Minimal viable file corruptor Your totally legal ROM collection, as vast as it may be, has started feeling a bit mundane recently. Why not spice it up a notch? Write a program that takes in the following arguments: Path of the file to be corrupted (or the contents of the file itself, if your l...

betseg

how do computers even calculate modulus

quartata

Division algorithm

A division algorithm is an algorithm which, given two integers N and D, computes their quotient and/or remainder, the result of division. Some are applied by hand, while others are employed by digital circuit designs and software. Division algorithms fall into two main categories: slow division and fast division. Slow division algorithms produce one digit of the final quotient per iteration. Examples of slow division include restoring, non-performing restoring, non-restoring, and SRT division. Fast division methods start with a close approximation to the final quotient and produce twice as many...

some decent reading there

Downgoat

3:03 PM

Poll: how do you pronounce malloc?

user202729

3:14 PM

(especially Jelly atoms)

I don't pronounce what I don't need to pronounce.

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Roguelike pathfinding Your task will be, given a two-dimensional array of the elements described below, which represents a dungeon, to output or return a single number representing the amount of gold pieces the rogue can collect without waking up any monsters. The elements of the array are as f...

Cows quack

3:34 PM

@user202729 how does this work?

unless you meant 1/2 ≡ p+1/2 (mod p)

user202729

So now I'm able to boot into WinRE, what next?

(1) chkdsk /f /r --> https://social.technet.microsoft.com/Forums/office/en-US/2527e82d-e4d7-44d9-a5a6-a1fad3b49743/chkdsk-unspecified-error-75736e6a726e6c2e-4fa?forum=win10itprohardware
(2) sfc /scannow --> found bug but can't fix (I don't remember the exact details)
(3) a partition is undetected
(4) I read somewhere https://www.kapilarya.com/fix-page-fault-in-nonpaged-area-bsod-in-windows-10 that I need to disable Windows automatically managing pagefile.sys but I can't boot so how can that help?

@Cowsquack Yes, but I can't find congruent symbol on my keyboard...

Cows quack

so 1/2 is the modular multiplicative inverse of 2 and not the fraction?

or is ½ the inverse?

user202729

... Yes, modular multiplicative inverse.

(as F_p is a field, dividing 1 by 2 automatically produce the inverse of 2 modulo p, anyway...)

Cows quack

~~what is F?~~ where does F come in?

user202729

3:49 PM

Looks like my mathematical notation is a bit rusty. The set of integers modulo p.

New Main Posts

4:12 PM

-1

Q: In a C ++ programming language, when you use a word (void) in the function

In a C ++ programming language, when you use a word (void) in the function

array-manipulation c++

New Sandboxed Posts

4:49 PM

0

A: Sandbox for Proposed Challenges

Title needed proof-golf Well I think it is about time we have another proof-golf question. This question is about Łukasiewicz's third Axiom Schema, an incredibly elegant set of three axioms that are complete over propositional logic. Axioms Łukasiewicz three axioms are: φ → (ψ → φ) (φ → (ψ ...

3

user202729

Looks like WW starred their own sandbox post.

Leaky Nun

@Cowsquack "1/2" is the unique solution to 2x=1 (mod p)

and note that (p+1)/2 is an integer, and it satisfies that modular equation

btw my python solution is lambda p:(~-p>>1)**3*3%p

@TonHospel hmm, interesting

NieDzejkob

is there a way to make make stop when a command in $(shell ...) fails?

Cows quack

@LeakyNun nice, do you have any more similar CMCs?

Leaky Nun

5:08 PM

@Cowsquack given prime p>3, find 1/3 mod p

Cows quack

so 1/3 mod p is the solution to 3x=1 (mod p), right?

Leaky Nun

yes

New Main Posts

5:24 PM

0

Q: Is it a valid 7-seg solution?

Background Download and save as .rar A save is 140 characters, each two be a group, the ith(0-based) group mean the cell a[(i div 7) + 1][i mod 7]. Let x mean (ASCII(first byte in a group)-48)*64 + (ASCII(second byte in a group)-48), then the k+4th(0-based) lowest bit of x means it takes a[i di...

code-golf

Blue

6:13 PM

@Downgoat neither of those ways. ma-l-uk

betseg

6:28 PM

/mɑɫɫot͡ʃ/ for me because the letter c makes the t͡ʃ sound in my language

Erik the Outgolfer

well, I think you forgot where to stress it...

caird coinheringaahing

6:44 PM

ಠ_ಠ I just calculated that I need to ask 46 more questions that aren't deleted, closed or have a negative score to get the Curious badge

betseg

@EriktheOutgolfer /ˈmɑɫɫot͡ʃ/ ?

Zacharý

@betseg ... I thought your velarized lateral approximants were lateral fricatives at first.

betseg

wow

Zacharý

My reaction was: what language is THAT?

HyperNeutrino

7:23 PM

CMP: I'm currently doing a thing where you have different content that has various tags assigned to it. I'm currently working on the part that allows you to look up content by tags (match things with all listed tags). Should all user-entered tags do fuzzy find or partial (start with) match or should only the last one (the one the user is currently typing) do semi-match?

I'm probably not going to do fuzzy find because that might confuse my users lol. I'm probably just going to stick with partial matching (either start-with or contains). Should it be last-tag-only partial or all-tags partial?

NVM figured it out myself, so much more natural to do All Partial Find.

DJMcMayhem

7:51 PM

@cairdcoinheringaahing Ouch

Pro-tip: Don't delete your challenges

Bálint

Hey there

I haven't been here in a while

Zacharý

9:03 PM

@DJMcMayhem ... ?

DJMcMayhem

9:23 PM

@Zacharý It's really buggy

Zacharý

Ah, so it isn't an actual hate, I presume

DJMcMayhem

It's close lol

Leaky Nun

10:41 PM

CMC: Given prime p>3, find 1/3 (mod p)

Destructible Lemon

print(1/3)?

Anonymous

@LeakyNun Actually, 5 bytes: ;¬3ⁿ%

Anonymous

In a finite field GF(p^n), 1/a = a^(p^n - 2)

Leaky Nun

@Mego oh right, there's this trick that I forgot, lol

New Main Posts

10:58 PM

1

Q: A running gene crossover algorithm

Your task is to accept as input two gene sequences, and a sequence of "cross over points", and return the gene sequence that results from the indicated cross overs. What I mean by this is, say you have the sequences [1, 2, 3, 4, 5, 6, 7] and [9, 8, 7, 6, 5, 4, 3], and cross over points of 2 and ...

code-golf array-manipulation

H.PWiz

@LeakyNun Does it have to be in the range 0..p-1?

Leaky Nun

@H.PWiz yes

H.PWiz

@LeakyNun Python 2, 20 bytes: lambda p:p-p*(p%3)/3. I chose Python because in Haskell, Mego's formula wins

Leaky Nun

@H.PWiz 19 bytes: lambda p:-p%3*p/3+1

H.PWiz

11:13 PM

Nice

Transcript for