The Nineteenth Byte

Codegolf the Hafnian The challenge is to write codegolf for the Hafnian of a matrix The pHafnian of an n-by-n matrix A is defined as Here S_2n represents the set of all permutations of [1, 2n]. Your code can take input however it wishes and give output in any sensible format but please inclu...

Lembik

A is the input matrix

Mr. Xcoder

Ok let me digest a bit

And what exactly is n?

Lembik

A is an n by n matrix

user202729

3:18 PM

... Sounds easy.

Mr. Xcoder

Oh the half-dimension

@Lembik 2n by 2n

Ok I'll to try to implement it

Lembik

@Mr.Xcoder thanks!

after the code-golf challenge there will be a fastest code challenge :)

user202729

@Lembik "It's guaranteed that for all i,j, A[i][j]==A[j][i] and A[i,j] in [0,1]?"

Lembik

I didn't know the term "half dimension"

@user202729 no.

where does it say that?

user202729

@Lembik But it's an adjacency matrix. Why not?

Mr. Xcoder

3:20 PM

@Lembik I don't know if that's a thing, but it's half of the dimension

user202729

Unlike half-life, this doesn't have any special meaning.

Lembik

@user202729 it doesn't have to be an adjacency matrix. it is defined for all matrices

user202729

But...

Does the definition have mathematical importance for other types of matrices?

Mr. Xcoder

The Wikipedia page seems quite incomplete to me

user202729

@Mr.Xcoder But enough to understand.

Lembik

3:22 PM

@user202729 yes indeed! mathoverflow.net/questions/273324/applications-of-hafnians/…

Mr. Xcoder

yeah

Lembik

someone needs to improve it

Mr. Xcoder

I guess I can assume the matrix is a square matrix of even side-length, right?

Lembik

@Mr.Xcoder yes

I am not sure it is defined otherwise

Mr. Xcoder

> This algebra-related article is a stub. You can help Wikipedia by expanding it.

Lembik

3:27 PM

there should be a challenge for how quickly you can code things too

with a fixed release date for the question :)

user202729

Does Jelly have foldr?

Lembik

or just a "code this in 30 minutes or less"

no idea :)

Mr. Xcoder

That's not a good idea because people could start solving it prior to the official release of the chalenge

moonheart08

^

Lembik

@Mr.Xcoder well they wouldn't be able to if the question has a release time

as in they wouldn't have seen it until then

Mr. Xcoder

3:29 PM

Ah, mhm. That might be fun, but not on PPCG :)

Lembik

no :)

I mean.. yes :)

that's a case where no == yes

My favourite is to implement the best break out you can in 30 minutes

as in en.wikipedia.org/wiki/Breakout_(video_game)

I still think that's a great programming challenge

Mr. Xcoder

Do you have any test case @Lembik?

I want to test my code

moonheart08

no :- \+ yes

Mr. Xcoder

Wait I think that's why you needed the code lol

user202729

Jelly, 29 bytes

Conor O'Brien

3:32 PM

CMC: find largest even integer in a given list of any order

Mr. Xcoder

4 bytes in Jelly

Lembik

@Mr.Xcoder it is !

Thomas Ward

@ConorO'Brien can we assume that the list only contains integers?

user202729

Jelly, something like ḂÐḟṀ

Thomas Ward

or do we need to assume that it might have non-ints.

Lembik

3:33 PM

@user202729 thanks!!

Conor O'Brien

@ThomasWard sure

Lembik

@user202729 I am amazed you did it in jelly

Dennis

@ConorO'Brien Order?

user202729

Jelly, 28 bytes

Pavel

@ConorO'Brien e=>e.Where(i=>i%2<1).Max()

Mr. Xcoder

3:34 PM

@Pavel ==0 => <1

user202729

Who says that golfing languages must be hard to use?

Conor O'Brien

@Dennis as in, you can't assume its sorted

Dennis

Right.

Lembik

@user202729 hmm.. try [[1,0],
[0,1]]

are you sure the answer should be 0?

user202729

Jul 2 '15 at 4:31, by Dennis

I use CJam for everything. It takes much less time to write.

@Lembik Yes of course. The graph with two self-loops doesn't have any perfect matching.

DJMcMayhem

3:36 PM

@Mr.Xcoder Bits, negate, times, Max?

Mr. Xcoder

What's the convension wikipedia uses for indexing?

I mean, 0 or 1 based?

Lembik

@user202729 ah yes!

Pavel

@ConorO'Brien Max@*Select[EvenQ]

Lembik

@Mr.Xcoder a good question.. I imagine there is none

Mr. Xcoder

@DJMcMayhem filterfalse by bit, max

Pavel

3:37 PM

Point-free Mathematica \o/

Mr. Xcoder

@DJMcMayhem That would fail for negatives, no?

Lembik

@user202729 thanks again for this.

user202729

@Mr.Xcoder That is not important, will arrive at the same result anyway.

Lembik

@user202729 do you think I should pose it as fastest-code or code-golf?

or one and then the other

Mr. Xcoder

Why does my code error?

user202729

3:39 PM

I think both.

Mr. Xcoder

@Lembik code-golf.

Lembik

@Mr.Xcoder I am worried it might be too simple for code-golf

already down to 28!

Mr. Xcoder

It's not too simple for code golf...

Lembik

@user202729 I like that idea

user202729

@Mr.Xcoder Because you use 1-indexing for Python.

Mr. Xcoder

3:39 PM

28 in Jelly is huge

user202729

28 bytes is not simple.

Lembik

maybe I could insist it can run for n = 6?

Mr. Xcoder

@user202729 Adding -1 still doesn't work

Lembik

that would cut out most solutions

Mr. Xcoder

@Lembik ಠ_ಠ

user202729

3:40 PM

Then you need to define "can run".

Lembik

@user202729 I was thinking in TIO

before the cut off

user202729

[no onebox] codegolf.meta.stackexchange.com/questions/12707/…

Mr. Xcoder

Should [[0,1],[1,0]] result in 1?

DJMcMayhem

@Mr.Xcoder No? I don't see why it would

Mr. Xcoder

@Lembik Here is my program

user202729

3:41 PM

@Mr.Xcoder Yes.

@Mr.Xcoder Why do you name it telescopic_sum when there is no telescope at all?

Thomas Ward

remind me for CMCs, if I pass a test list in as l in order to be part of a line of code, should that passing in of a test list be included in the bytecount, or can I just put it in the 'header' field of a tio.run, for code bytecount purposes. Concurrently, do I need to provide code as a function, or is it ok for it to just be a one-liner?

Mr. Xcoder

¯\_(ツ)_/¯ I needed a name

3

Thomas Ward

it's been a while since I did codegolf :)

Mr. Xcoder

Hardcoding is not allowed at all.

You should either build a function

Or a full program

DJMcMayhem

@ThomasWard Your code can be a function or a full program, but you are not allowed to assume the input is stored in a certain variable

Thomas Ward

3:43 PM

ok

that's what i thought but it's been a while

@Mr.Xcoder Unfortunate for you that DJ is better at explaining things.

imo

Mr. Xcoder

Nah I didn't really bother :P

DJMcMayhem

Also, an answer that evaluates to the right function works too. (It doesn't have to be named). So if the challenge was find the length of a list then len is a valid solution

This means that if you do a lambda, you can leave out the f=

Thomas Ward

@DJMcMayhem right, I remember that part

DJMcMayhem

@ThomasWard You use mostly Python, right?

Thomas Ward

yep

Mr. Xcoder

3:47 PM

@Lembik Try it online! – A better version, takes input from STDIN

Thomas Ward

@DJMcMayhem it gets tricky when we have to eval(input()) in order to provide input as a list, and what is confusing me is where I need to provide that - i.e. if that can be provided as a 'footer' in a tio.run item (basically: print(f(eval(input())))), and not count against the bytecount or not.

Mr. Xcoder

You should not include print(f(eval(input()))) in the byte count.

Thomas Ward

as you can see it's been a while since I codegolfed :)

that's what I thought

Mr. Xcoder

You only need to include the function body / program in the byte count. If it is a recursive lambda (or if you actually make use of it in another function you define), you should count f=, if it not recursive you should not count f=, if is is a full function (def-style), you should include the name too, because it cannot be anonymous and for a full program you count everything your program needs to execute correctly :)

DJMcMayhem

@ThomasWard Yeah, you would need that for a full program, but not a function.

Also, switching to Python 2 saves 7 bytes: print f(input())

Mr. Xcoder

3:52 PM

@DJMcMayhem It actually saves 7 bytes, I'll let you figure out how

Thomas Ward

then assuming I didn't break any bytecount rules...

DJMcMayhem

@Mr.Xcoder doi thanks

Pavel

input |> eval |> f |> print > print(f(eval(input())))

Mr. Xcoder

Print[F[Eval @ Input$]] >> everything else :P

Pavel

Not Print@F@Eval@Input$?

Mr. Xcoder

3:55 PM

@Pavel Nope, because unfortunately infix stuff is very limited when you transpile to Python, so that would actually be parsed as ((Print[F])[Eval])[Input$]

user202729

(no problem, I viewed the wrong tab)

Idea: Whenever someone ask something in TNB, wait for a while before replying to encourage rubber duck Google searching.

Thomas Ward

@Mr.Xcoder @DJMcMayhem so I would be properly doing bytecount then here? (comes out to 46 bytes, regardless, for Py2 or Py3, only difference is the 'Footer' where you execute the non-recursive lambda)

(this is for Conor O'Brien's CMC there, but I want to make sure I don't break bytecount rules :P)

Mr. Xcoder

@ThomasWard That would be 46 bytes.

Thomas Ward

that's what I thought, just making sure I didn't break a rule :)

DJMcMayhem

@ThomasWard Looks right to me. You have an extra space btw

Mr. Xcoder

3:59 PM

@ConorO'Brien Python 3, 33 bytes (anti-spoiler)

Thomas Ward

you're right.

damn it i was beaten

but that would work :p

and you use my test case list nice

:P

Mr. Xcoder

Of course :)

Thomas Ward

@Mr.Xcoder Also works in Python 2 without changes.

except for the footer :P

Mr. Xcoder

Yay :)

Thomas Ward

i keep forgetting max() is a builtin >.<

DJMcMayhem

4:01 PM

@ThomasWard Xcoder kicked both our butts, but I have a cool hack for you that shaves 3 off: Try it online!

Mr. Xcoder

I think I have 30 bytes

Thomas Ward

@DJMcMayhem nice

Mr. Xcoder

@DJMcMayhem You can drop the -

Poke

@user202729 Personally if I've asked a question in here it's because either I /have/ looked on google already, or I would like more context than I expect google will give (being able to ask followup questions, etc.)

Mr. Xcoder

@ConorO'Brien Can we assume all are positive?

Poke

4:03 PM

delaying a response actively hinders in this case

label a -- goto a

Nobody's on Language Development... I posted a message here, should I repost or can you join yourself? :/

Mr. Xcoder

If all are positive, then 31 bytes

Pavel

@ConorO'Brien Coconut, 21 bytes: max..filter$(i->~i%2)

moonheart08

each in a (i):
a[i] = a[i + 1]

a?

I'm making a thing

Mr. Xcoder

Physica, 26 bytes: ->_:Max[n For n∈_If~n%2]

caird coinheringaahing

4:14 PM

@ConorO'Brien Add++, 24 bytes: L,v2AvbLXBcB%B!B]AvBcB*M

Mr. Xcoder

;'-( Add++ beats Physica :P

caird coinheringaahing

I think I can shorten it as well :P

Thomas Ward

@Mr.Xcoder The question is whether we consider 'largest' to be 'furthest from zero', whereby -10001 is larger than 10000. The way it was worded it wouldn't matter if it's negative or not, I think. (Hence why we need clarification!)

Mr. Xcoder

25 bytes if positive-only: ->_:Max[~n%2*n For n∈_] or this: ->_:Max[~n%2*n For n In_]

@ThomasWard largest usually means greatest

caird coinheringaahing

22 bytes: L!,2abLXBcB%B!B]aBcB*M. Takes n arguments, representing a list of size n

Thomas Ward

4:19 PM

@Mr.Xcoder i know, that's why i haven't complained about your solution to that thing :P

but even if there's negative values i assume yours would work

Mr. Xcoder

My 33 byter would work nonetheless, the 31 byter fails if there are negative values.

Thomas Ward

@Mr.Xcoder the 31 byter works so long as the negatives are at the end of the list, but you're right, because I was able to get it to tell me 6 instead of 7 in the 31byter :P

Lembik

@Mr.Xcoder can I add your code to my question?

Mr. Xcoder

Sure, go ahead!

Lembik

thanks!

Mr. Xcoder

4:23 PM

You're welcome!

caird coinheringaahing

As a side note, I love challenges where I can use other entries to test my entry's validity. — Magic Octopus Urn 2 hours ago

My thought's exactly :P

Lembik

ok so thanks to this chat room, the question is awesome now :)

Mr. Xcoder

I am working on outgolfing user202729... :P

Suggested test case: [[0, 1, 1, 0], [0, 1, 0, 1], [1, 0, 1, 0], [1, 1, 1, 1]] -> 1.25

caird coinheringaahing

TFW you tie Dennis in a restricted-source challenge where he follows the restrictions and you don't

Mr. Xcoder

Lol... :P

Lembik

4:28 PM

@user202729 hmm.. try [[1,2,3,4,5,6,7],[1,2,3,4,5,6,7],[1,2,3,4,5,6,7],[1,2,3,4,5,6,7],[1,2,3,4,5,6,7]‌, [1,2,3,4,5,6,7], [1,2,3,4,5,6,7]]

user202729

@Lembik "Challenge"? (not "question")

@Lembik That (the side length) is odd.

Lembik

@user202729 The Hafnian one

oh oops

I see your code doesn't like that

my mistake

user202729

Great, we had some on-topic (about a code-golf challenge) discussion recently.

Lembik

@user202729 hmm.. try [[1,2,3,4,5,6,7,8],[1,2,3,4,5,6,7,8],[1,2,3,4,5,6,7,8],[1,2,3,4,5,6,7,8],[1,2,3,‌4,5,6,7,8], [1,2,3,4,5,6,7,8], [1,2,3,4,5,6,7,8]]

I get [2144.713699438102, 4289.427398876204, 6434.141098314307, 8578.854797752409, 10723.568497190512, 12868.282196628614, 15012.995896066717, 17157.709595504817] from your code which makes so sense

Mr. Xcoder

For me it doesn't even halt

Lembik

4:32 PM

hmm

your code finishes very quickly for me with 551.25 which I don't think is right

@Mr.Xcoder what do you get for 1,2,3,4,5,6 repeated 6 times?

Mr. Xcoder

And idk but I get syntax errors when I run it with your test cases

> 551.25

Lembik

really?

as in the input has syntax errors?

or your code does?

Mr. Xcoder

The input for some reason

Lembik

oh!

Mr. Xcoder

Maybe it's SE weirdness

caird coinheringaahing

4:34 PM

SE inserting ZWNJ spaces

Mr. Xcoder

Yeah :/

caird coinheringaahing

Cause that's always helpful /sarcasm

Lembik

33673.5

fixed it

some weirdness in the input

caird coinheringaahing

CMP: Has anyone here ever played Little Alchemy?

Pavel

yes

Mr. Xcoder

4:36 PM

s/every/ever/

And yes

H.PWiz

I played a similar game that came before it afew years ago

Lembik

ok thinking of posting the challenge now as code-golf

now that it has examples :)

Mr. Xcoder

Hopefully they're correct

Lembik

well the jelly and python versions agree :)

so how could they possible be wrong :)

Pavel

@cairdcoinheringaahing The best variation on this is Elemental, which is unfortunatly currently offline. It's basically crowd-sourced little alchemy, when you find a combination that doesn't exist yet you're asked to say what you think it should make. Then the community votes on your proposal.

Lembik

4:38 PM

I am still tempted to add a speed element...

hmm

Mr. Xcoder

@Lembik Please don't combine swiftness and golfy-ness for this challenge. I love it as plain code golf :|

user202729

@Lembik 8 columns and 7 rows?

Lembik

@user202729 that's not in the sandbox example is it?

@Mr.Xcoder ok

Manish Kundu

somebody could give feedback on this? : codegolf.meta.stackexchange.com/questions/2140/…

Lembik

@Mr.Xcoder I see you edited your own code :)

Manish Kundu

4:46 PM

nice challenge. feels good to see math challenges on codegolf

Lembik

thanks.. you wait until the fastest-code version :)

user202729

@ManishKundu You know our rules about built-in methods?

> If a challenge is easy, banning built-in functions won't change that.

(can't remember the source)

Manish Kundu

like, do x without y is not allowed?

New Main Posts

3

Q: Codegolf the Hafnian

The challenge is to write codegolf for the Hafnian of a matrix. The Hafnian of an 2n-by-2n matrix A is defined as: Here S2n represents the set of all permutations of the integers from 1 to 2n, that is [1, 2n]. The wikipedia link talks about adjacency matrices but your code should work for any...

AdmBorkBork

It's not disallowed, just discouraged.

user202729

4:50 PM

2 hours ago, by New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Codegolf the Hafnian The challenge is to write codegolf for the Hafnian of a matrix The pHafnian of an n-by-n matrix A is defined as Here S_2n represents the set of all permutations of [1, 2n]. Your code can take input however it wishes and give output in any sensible format but please inclu...

2 hours ago. That's...

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Introduction As we all know, there are many number systems - the most common ones being the binary, decimal, octal and hexadecimal number systems. A number system with a base 'n' would consist of exactly 'n' unique characters to represent numbers starting from 0 till n-1 (in decimal). For inst...

Manish Kundu

i hate that rule lol. but okay, i removed it..

Mr. Xcoder

@user202729 It's ok for a quick sandbox.

They just wanted to receive some quick feedback

NieDzejkob

Does anybody have problems connecting to SE (connection timed out)

Mr. Xcoder

Nope

Manish Kundu

4:51 PM

nope

user202729

@ManishKundu I understand, but then it's not clear to define/check that.

Also I think that it may be a dupe. It's so simple...

Manish Kundu

but i dont think any challenge has asked for a random base conversion

most challenges were like decimal to binary, decimal to octal, etc.

user202729

There is even a tag for it, base-conversion.

Manish Kundu

5:11 PM

lol

Mr. Xcoder

Is GitHub down for anyone else?

NieDzejkob

@Mr.Xcoder responds to pings but not to ssh push or web https for me

just like Reddit, whole SE and wikia for me

Manish Kundu

yeah its down for me too

NieDzejkob

it's probably a datacenter problem

Esolanging Fruit

@Mr.Xcoder Works for me

NieDzejkob

5:21 PM

it's up again for me

but many sites still get stuck on "Performing a TLS handshake" for me...

Manish Kundu

yeah works now

J. Sallé

Akamai probably dun goof'd for a second there.

NieDzejkob

akamai... or cloudflare... or ovh... we'll never know

J. Sallé

Yeah I just guessed akamai because it's slightly more likely it's them

Neil

5:35 PM

CMC: given a list of floats, return the one with the largest absolute value

ETHproductions

@Neil Japt, 4 bytes: ña o

Giuseppe

@Neil R, 43 bytes: function(l)l[order(abs(l),decreasing=T)][1] tio.run/##K/qfpmCj@z@tNC@5JDM/…

ETHproductions

Or JS, 28 bytes: a=>a.sort((x,y)=>y*y-x*x)[0]

Pavel

@Neil C# e=>e.Select(Math.Abs).Max()

tfbninja

@Neil can the output be the absval of the biggest?

Neil

5:41 PM

@tfbninja of course not

Pavel

Oh, dur

tfbninja

@Neil Pyt, 2 bytes Å↑ this is what I mean

basically does it have to keep the negative sign

Pavel

e=>e.MaxBy(Math.Abs)

betseg

@Neil Jelly, 3: tio.run/##y0rNyan8/9/x4c5FD3eu@v//f7SZnqmOgq4liDSOBQA (my first Jelly answer to anything! :D)

tfbninja

@betseg nice!

label a -- goto a

5:44 PM

My question on superuser has no upvotes, downvotes, comments nor answers after 2 hours! That's gotta be my new best...

betseg

There is a badge for this

label a -- goto a

I know

tfbninja

I think its funny that tumbleweed is a bronze badge, it's extremely hard to get, although I suppose its so people don't try to get it with boring questions

betseg

You don't have SU on your network profile? Wut @labela--gotoa

Erik the Outgolfer

@betseg correction

betseg

5:46 PM

@EriktheOutgolfer ah right

Erik the Outgolfer

@labela--gotoa do you have a superuser account?

label a -- goto a

@betseg Perhaps I've got it hidden, since I don't use it, I got account because I made a single-time action or something.

yes

Lembik

so the next challenge will be to use this algorithm for the Hafnian arxiv.org/pdf/1107.4466.pdf

which is somewhat trickier I think

Erik the Outgolfer

@labela--gotoa yeah if you didn't join you posted under an unregistered account

label a -- goto a

@EriktheOutgolfer I joined to make the action, but I hidden the account since I don't really use it

You can hide communities in the settings

Erik the Outgolfer

5:52 PM

well, I respect your reasons, although I'd say you don't need to hide (and most probably change the profile details too) if there isn't something you actually don't want to connect to your SE profile

label a -- goto a

I don't really enjoy having my profile spammed with 101 rep communities

recursive

@Pavel MaxBy() is not standard C#. I know there's one in Jon Skeet's MoreLinq package.

Pavel

@recursive Yes it is. Observable.MaxBy.

recursive

@Pavel Oh, I see. I did not know about that. Of course, that means the argument must be an observable list.

Pavel

Well, yes.

DJMcMayhem

6:14 PM

Thoughts?

J. Sallé

@Neil APL (Dyalog Unicode), 3 bytes ⌈/|

Erik the Outgolfer

nope, you have to return the number with its original sign

J. Sallé

@DJMcMayhem I think the only flaw I see there is that "noteworthy" is a single word

@EriktheOutgolfer I failed to notice that >.>

recursive

@DJMcMayhem This looks really good. Maybe a part about creating an esolangs wiki entry?

Erik the Outgolfer

noteworthy is a word, that's no fault

recursive

6:23 PM

I haven't done that for stax yet

J. Sallé

@EriktheOutgolfer he wrote note worthy

Erik the Outgolfer

ah, and you don't have the rep to edit that yet, I see

DJMcMayhem

@recursive Ah, good idea!

thanks

Christopher

6:42 PM

Hello

Laikoni

@DJMcMayhem Regarding the disclaimer for language of the month nominations: An important point is the stability of the language. There shouldn't be any breaking changes while people start learning the language.

Christopher

@DJMcMayhem quine in painflak

Erik the Outgolfer

that looks like you're demanding something from somebody you have control over or are very tight acquaintances with...

what are you referring to?

Lembik

what does VTC mean?

DJMcMayhem

Vote To Close

J. Sallé

6:54 PM

@Lembik Vote to Close

Lembik

oh wow!

DJMcMayhem

26

A: What are the PPCG specific abbreviations and terms?

See also Stack Exchange Glossary - Dictionary of Commonly-Used Terms. Abbreviations marked with a star (*) are chat specific. Catalog: A type of simple on-topic challenge where the challenge's aim not so much to find a winner as it is to create a catalog of solutions in many languages. CG: Cod...

4

J. Sallé

@DJMcMayhem one of the most useful meta posts in the site for sure.

Lembik

oh someone kindly came to my rescue :)

(re: my open question)

Transcript for