The Nineteenth Byte

Optimizer

7:27 AM

new year eve . Ought to be silent ...

Optimizer

7:40 AM

@Sp3000 started on parser for CJam

Optimizer

9:16 AM

There is no way to close a question as spam ?

ProgramFOX

@Optimizer Just open the flag dialog and select the first option "spam".

After 6 flags, it will be auto-deleted.

Optimizer

ok

ProgramFOX

oh, it already is deleted \o/

Optimizer

community FTW

Sp3000

10:13 AM

CJam parser woo :D

Martin Büttner

10:47 AM

@Doorknob冰 Do you think if we can get some support for this on meta, we could get the devs to activate SoundCloud embeds for PPCG? I think that would greatly increase the feasibility of audio processing challenges.

Optimizer

11:04 AM

Hasn't it been more than one week already ?

Martin Büttner

@Optimizer yes, we extended it by another week

yay, I've finally got a nice array-manipulation idea :D

(although it's stolen...)

Sp3000

Posting it soon? :P

Martin Büttner

yeah, it's in the sandbox right now

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Unflatten an Array code-golfarray-manipulation Inspired by this question on Mathematica.SE. Say you've got a nested list/array of some arbitrary structure (the lists at each level don't necessarily have the same length). For simplicity, we'll assume that the nodes are non-negative integers. As...

Martin Büttner

there you go

the Mathematica solution to this is so ridiculously clever :D

Sp3000

11:12 AM

Ooh that's interesting

You've already come up with an answer?

Martin Büttner

see the linked question

Sp3000

Oh the first link, right

Martin Büttner

I never would have come up with that :D

if I golf that down it's 62 bytes. I hope that should be beatable

I think some recursive Pyth solution will probably beat it

but just in case, I think I'll just post the golfed version of that myself, and declare that I won't accept, so no one can just steal that and kill the challenge

Sp3000

I don't think Pyth has type() inbuilt, so the solution I'm thinking of would take a bit longer

*need more bytes

That might be a good idea

Martin Büttner

44 bytes...

Sp3000

11:18 AM

Can you have an empty list on its own somewhere? eg [[],[],[[1,[]]]

Martin Büttner

oh, that's a good point... hmm...

Sp3000

I was assuming no, but to make sure

Martin Büttner

the mathematica solution can handle it >:D

Sp3000

I'll try in Python when I get back :P

Martin Büttner

@Optimizer you wanna try that one in CJam and JS? ;)

Sp3000

11:23 AM

I think what I have in mind can handle it

Optimizer

@MartinBüttner I might if someone stars this :P

Sp3000

But stack snippet :P

Martin Büttner

@Optimizer that's a bit too desperate ;)

Optimizer

@Sp3000 In progress

Martin Büttner

@Sp3000 but 30 minutes.

Optimizer

11:25 AM

@MartinBüttner Isn't that desperate ? :P

Martin Büttner

lol

even if you did, I doubt it would make 30 minutes, to be honest

Optimizer

I can post 2 answer

Martin Büttner

but I'm not starring messages that ask for stars ;)

Optimizer

@Sp3000 can 2, and you 1 :P

@MartinBüttner I have so many others :P

Martin Büttner

I don't know if self-answers count

Sp3000

11:26 AM

Oh pizza, right

Optimizer

we will have to try

Sp3000

Oh I don't need to use type, there's a golfier way...

Martin Büttner

@Sp3000 added examples for empty lists within the tree

Optimizer

can i have the nested list as a string and the values list as an array in JS function argument ?

Martin Büttner

@Optimizer hm no, consistent representations please, I'll clarify that

@Sp3000 43...

Optimizer

11:32 AM

ugh, you just elongated by 40 byte JS solution

Martin Büttner

can't you toString it manually?

Optimizer

that is what elongation means, I think.

and no.

toString is stupid

46 :(

Martin Büttner

@Optimizer but it wouldn't be 40 bytes

@Optimizer wow, that's actually pretty good

Optimizer

no, a 40 byte solution went to 46

that's not good

Martin Büttner

oh

"my 40 byte solution"... not "by 40 bytes"...

Optimizer

11:35 AM

oh. err.

Martin Büttner

the 46 can handle empty sublists?

Optimizer

anything

Martin Büttner

great

Sp3000

@MartinBüttner Can we regex replace digits with digits?

Martin Büttner

@Sp3000 oh...

Optimizer

11:38 AM

regex was not allowed ?

mine uses regex only :P

Martin Büttner

@Optimizer well what's the point if we do this with regex

it's tagged array-manipulation, not string-manipulation

Optimizer

are you officially banning regex ?

its clever via regex

Martin Büttner

I'm pretty sure I will... this is boring with regex

Sp3000

I just thought about it because I'm using str(L)<":" to check if it's a number or a list

Martin Büttner

@Sp3000 well that's fine by me

hmmmm

@Sp3000 what do you think?

Sp3000

11:42 AM

The easy way out would be to ban conversion to strings (because you can always just do the regex approach without regex by traversing the string)

Whether it's shorter is another question though

Otherwise banning only regex could work

Martin Büttner

@Sp3000 good idea, done.

Optimizer

if regex was allowed, CJam would have got tough competition from other languages. but not anymore

f=(a,b,i=0)=>a.replace(/\d+/g,x=>eval(b)[i++]) f("[0,[1,[]],[[]],[2,3],[]] ","[1,6,1,8] ") My solution using regex.

Martin Büttner

how much do you need in CJam?

Optimizer

did not try

Martin Büttner

@Optimizer yeah, while it's a nice idea, it really circumvents the actual challenge ;)

Optimizer

11:48 AM

the actual challenge is to get 5 answers in 30 minutes, no ? ;)

Martin Büttner

lol

@Optimizer I don't think a regex-based approach in other languages would be much shorter, except maybe Perl. and Mathematica is still 43, so that 46 was competition for CJam than Mathematica still is ;)

Optimizer

bash

Martin Büttner

hm yeah maybe

Sp3000

115 in Python, I think I'm doing it wrong

Martin Büttner

@Sp3000 recursive solution?

Sp3000

11:52 AM

Yeah, but I defined two functions because the first one returns the list and the next element index

Optimizer

oh i got a nice idea

recursive

Sp3000

@MartinBüttner Surely there's a better way...

Lembik

hi all

happy christmas/new year/everything else

Two binary strings: x and y. Simple matching = count the number of matching digits.

What kind of strings will have the property that the match is better when y is shifted to the left relative to x rather than to the right.

a puzzle to think about :)

Optimizer

44 in JS without regex

Sp3000

Is this cyclically shifted so that the last element goes to the front, say?

Lembik

11:58 AM

@Sp3000 yes

Sp3000

Strictly better or better-or-equal?

Martin Büttner

@Optimizer see ;)

Lembik

In particular (say) the match count is more then 50% for each of the first 10 shifts to left and less than 20% for the first 10 shifts to the right.

@Sp3000 ^^

Martin Büttner

@Sp3000 bloody hell...

Sp3000

:P the only smart thing done there (and I have no idea how to backtick in chat)

Martin Büttner

12:02 PM

sorry about all the pings

Optimizer

pretty sure it can be done much shorter than 44 in python

now if only online pyth compiler was working

Sp3000

Pyth, maybe. I can't figure out how to use D though

Optimizer

what is D ?

define ?

y() ?

Sp3000

@Lembik Hmm interesting, makes me want to try with a bunch of strings but maybe later :P

Define, yeah

Optimizer

its easy

lambda based define and then to use, call y

Sp3000

12:05 PM

Er... I can't lambda with the approach I'm using

Optimizer

oh wait, D does not lambda

its a normal function and u return by ) or R

so say u r creating a function with 2 arguments, you can do D<XYR+XY<3T

will return 13

Sp3000

Er, I get an AssertionError but yeah I'm having trouble specifying the arguments

Optimizer

in my code ?

Sp3000

Yeah

Optimizer

I just wrote that from memory, can't actually compile to check

Sp3000

12:14 PM

Ah, got it. Although I really don't know how you would define functions taking more than 4 arguments...

D:GHbR-+GHb:4 5 6

Optimizer

that's not possible

Lembik

@Sp3000 please do!

Sp3000

45 bytes and can probably golfed a fair bit, possibly with a better approach

@Lembik Ahaha k will do then - do you already have an answer? Out of curiosity

Lembik

@Sp3000 no!

Martin Büttner

so we've got Mathematica, Pyth, Python and JS now? ^^

Sp3000

12:26 PM

Python and Pyth count as 1 :P

Martin Büttner

yeah, I thought so

Sp3000

(I don't like posting separate answers when it's just a port)

Martin Büttner

depends... if it's a port from Python to Pyth, no

but if it's a port from Mathematica to CJam (or vice versa) I'd post two answers

Sp3000

@Lembik Do you have any examples btw? 50%/20% is a pretty hard cutoff to satisfy in my mind

Yeah I'd do that too, but Python -> Pyth :P

Martin Büttner

yeah, that's fine

@Optimizer are you working on CJam or should I just post and hope for the best? ^^

Sp3000

12:29 PM

Give me a bit more time to see if I can work out Haskell

(Hint: I'm terrible at Haskell)

Martin Büttner

you can Haskell?

Sp3000

Just what I know from the uni assignment last semester (aka not much)

Martin Büttner

oh, there's someone who could add golfscript to the mix :D

Sp3000

Oh I know how I should be doing this in Python, silly me :P

Optimizer

I am in a meeting

so no cjam

Martin Büttner

12:39 PM

@Optimizer then why are you even in chat? :D

Optimizer

tab is open

Sp3000

@MartinBüttner Can I specify which list is inputted first?

Martin Büttner

sure

Sp3000

That makes things shorter :)

Martin Büttner

yeah changing the order of arguments often helps in Mathematica too

Sp3000

12:46 PM

I just means I can print and call the function at the same time

I think I'm ditching Pyth, I like my new Python solution too much to post a Pyth one

Martin Büttner

haha

how long is it?

Sp3000

73 bytes

Martin Büttner

and if you ported it? :P

Sp3000

It uses iter, which I'm not sure Pyth has (I've never figured out how to use Python in Pyth)

It'd probably be shorter anyway with some workaround

Nevermind, got another approach. I'm porting.

Optimizer

i got the hat, but no idea which message was starred

its not even showing on the right

Martin Büttner

12:52 PM

it is for me ;)

23 hours ago, by Optimizer

Optimizer

ah thanks!

Martin Büttner

it's the bottom-most starred message for me

Optimizer

so post ur question in an hour ?

Martin Büttner

but the length of the list might depend on the window resolution

Optimizer

i dont have it. bottom most is SE one by sp3000

Martin Büttner

12:53 PM

@Optimizer yeah, roughly

hm I've got two more

Sp3000

I think Pyth can be about 30 or so bytes but I cbf debugging right now

Optimizer

cbf ?

Martin Büttner

can't be bothered

except with a different verb

Optimizer

oh

what is f then ?

Martin Büttner

well... the different verb :D

Optimizer

12:59 PM

wtf f ?

Martin Büttner

yes ^^

Optimizer

we are not kids here, I think

you could have said that :P

Sp3000

I should really use cbb, not sure why I use cbf instead

Martin Büttner

I'm not sure if chat has autoflags :D

Sp3000

@Lembik Random string generation's turning up nothing :/

Martin Büttner

1:00 PM

also there are kids in here :D

Optimizer

there are ?

who ?

Martin Büttner

if you consider 13 year-olds kids...

Optimizer

still who ?

Martin Büttner

oh, they're actually 14 and 15 now

Doorknob and cjfaure at least

Sp3000

@MartinBüttner Awkwardly, my code's shorter if I take one list as STDIN input but the other as function argument... I'm guessing that's not allowed? :P

Martin Büttner

1:01 PM

@Sp3000 nope, I mentioned that explicitly ^^

Sp3000

Thought so :P

I don't get how to check types in Haskell easily, so I don't think I'll be doing that :/

Lembik

@Sp3000 :(

Sp3000

:(

Lembik

@Sp3000 is it easier if one string is longer than the other so you don't need to worry about cyclic shifts?

Sp3000

Hmm I think it's more the 50%/20% and ten shifts

Probability-wise I'm not sure what the chances of hitting a case that fits is

Lembik

1:15 PM

@Sp3000 well it would be interesting is one exists!

if one

Sp3000

I'll try less rotations

Lembik

fewer :)

Sp3000

@MartinBüttner Porting my nested solution would be easy if Pyth had a pop which actually worked like a Python pop

@Lembik Well corrected :)

Martin Büttner

@Sp3000 heh...

Sp3000

I like my solution anyway :) it's on 68 now

(Python that is)

@Lembik Even 2 rotations is given my program a hard time. 1 is pretty trivial though

Lembik

1:24 PM

interesting

Sp3000

Do you have any examples?

001111000011100
100001110001111

^ It found one for 2 rotations

Lembik

@Sp3000 hmm.. I will try to make some

I don't have any here sorry

that's a nice example"!

Optimizer

doorknob is a kid :O

7

Lembik

and an interesting potential pattern

Sp3000

001110000011110
000111111000011

also

000110000111000
110011100011111

Weirdly they have the same length and I can't tell if that's coincidence

Oh, happy new year - it's half past 12

Martin Büttner

1:33 PM

lol how did you miss that?

Sp3000

Too busy golfing

@MartinBüttner If we write a function, can the function destroy the values list?

Martin Büttner

sure

time to post it I think

Sp3000

:) k

Not gunning for pizza?

Martin Büttner

we'll see

Optimizer

I can defs do the JS one right now.

output as a returned array is fine ?

@MartinBüttner ^

Martin Büttner

1:43 PM

yes

Optimizer

kewl

post it

I'll start on cjam in the mean while

meh

Martin Büttner

0

Q: Unflatten an Array

This challenge was inspired by a question on Mathematica.SE. Say you've got a nested list/array of some arbitrary structure (the lists at each level don't necessarily have the same length). For simplicity, we'll assume that the nodes are non-negative integers or empty arrays. As an example [[[1...

Optimizer

you could have waited for a bit before posting your own solution

Sp3000

I thought the point of posting that one was to rule out copying from Mathematica.SE?

Optimizer

if you are giving credits to someone else, make it a wiki answer ? (but maybe after 30 minutes)

should have waited for 30 minutes. what if wiki answers dont count ?

Martin Büttner

1:52 PM

@Optimizer already did

@Optimizer hm, maybe... well

Optimizer

I know

Martin Büttner

@Sp3000 yes?

Sp3000

(that was a reply to Optimizer)

Optimizer

so , only 2 (or 3) answers to go

Martin Büttner

damn that Ruby one is good

@Sp3000 oh

Lembik

1:54 PM

@so

@Sp3000 ok I have a harder version :)

Martin Büttner

@Sp3000 :D

boom HNQ :D

Sp3000

That was fast

ProgramFOX

ugh, C# doesn't allow arrays/lists that contain both other lists and integers

Martin Büttner

@ProgramFOX not even as objects?

ProgramFOX

oh, forgot about that. Thanks!

That's going to become looong code :P

Martin Büttner

2:00 PM

@ProgramFOX can you use the try/catch trick from Ruby?

Sp3000

"As long as it's within 16 minutes," thinks Martin, as he watches over the HNQ

ProgramFOX

@MartinBüttner I don't think C# has a built-in Shift method that works like that.

Martin Büttner

@ProgramFOX I don't think that methods depends on the unshift

Sp3000

I wish try/except and return were shorter in Python or I'd have done that :/

def g(N,L):
 def f(N):
  try:return map(f,N)
  except:return L.pop(0)
 return f(N)

Martin Büttner

oh yeah, that return...

grc

2:05 PM

happy 2015 everyone :D

ProgramFOX

Happy 2015!

Optimizer

:D

Sp3000

Happy 2015 :) one timezone at a time

Optimizer

@MartinBüttner got hat ?

Martin Büttner

there's only one real time zone

Optimizer

2:05 PM

more than 3 timezones have it right now

Martin Büttner

@Optimizer not yet, they take a bit

@Optimizer I don't think my self-answer counts tbh

Sp3000

@grc nice one!

grc

@Sp3000 actually I'm just over an hour late

Martin Büttner

@Optimizer how do you split a CJam array on multiple elements?

Sp3000

@grc Also same :P 1am here

grc

2:06 PM

did you watch the fireworks?

Sp3000

Yup

Man, I keep forgetting that numbers are smaller than lists

That would have been better

Also that list comp instead of map :/ I feel silly now

Optimizer

@MartinBüttner its hard

recursive answer in a stack based language :D

Martin Büttner

I've got 24, but it's also string manipulation (but without using regex)

Optimizer

these empty arrays represntation like "" bit me

Martin Büttner

@Optimizer why is that a problem?

Optimizer

2:16 PM

if it was represented like [] only, then the code would be '[#~ but now it is "[\""&

Martin Büttner

aaaaand top of the HNQ :D

and 30 minutes are over

Optimizer

got hat ?

Martin Büttner

not yet

patience

maybe the answers need to be from 5 different users (not counting myself) though

Sp3000

I have a question idea myself but I'm not sure if it's been done yet: Given a permutation, find the next one lexicographically, wrapping back to the first permutation if the input is the last one

Martin Büttner

@Sp3000 if anyone can tell you it's @Peter ;)

@Optimizer got it

thanks guys :)

Optimizer

2:20 PM

nice!

i hope you remember this for my question this week :P

ProgramFOX

Fun, my first C# attempt gave a StackOverflowException :/

Martin Büttner

@Optimizer sure ;)

@Optimizer now go and beat 24 ;)

Optimizer

well, you should have banned string manipulation

as it is array-manipulation question

double standards :P

Martin Büttner

then how would you have determined the type? :P

Optimizer

my JS one does not do string manipulation at all

and so does the CJam one

Martin Büttner

2:23 PM

what's the backticks then? :P

Optimizer

its not manipulating :P

just stringifying then set operations

Martin Büttner

lol

talk of double standards...

Optimizer

no manipulation happening :P

ProgramFOX

Okay, at least my C# code doesn't give errors anymore! Now checking the correctness... :P

New Sandboxed Posts

2:39 PM

0

A: Sandbox for Proposed Challenges

Next Lexicographical Permutation code-golf array-manipulation Not sure if this has been done before so I thought I'd post here first Given an array representing a permutation of the first n >= 1 nonnegative integers, find the next lexicographical permutation. If the input was the last possible ...

ProgramFOX

Well, running another array gives an error again :/

ah, used a wrong variable name

The structure of the result already seems okay, but the values are incorrect.

Transcript for