last day (49 days later) » 

3:03 PM
Welcome to this room, where we discuss the OEIS answer chaining challenge: One OEIS after another
 
@NieDzejkob So what's the problem with HyperNeutrino's solution you mentioned?
The blank message history looks so weird...
 
room topic changed to One OEIS after another: For discussion about One OEIS after another (codegolf.stackexchange.com/questions/133754/…) (no tags)
ಠ_ಠ Stupid markdown not working in room descriptions
 
@NieDzejkob "why there's a bigger problem?".
 
@user202729 well, the four repeated bytecounts is something I would consider a bigger problem
anyway, this looks doable in Husk
how do you get a sharelink to a comment?
 
So... that's all? The challenge need some maintenance. @cairdcoinheringaahing What would you decide to do?
@NieDzejkob click at time mark to the right of the comment (not sure what it is called).
 
3:10 PM
@NieDzejkob That would require golfing the Emojicode answer by 36 bytes
 
@cairdcoinheringaahing whoops, misremembered a number
 
@user202729 Yeah, but we're trying to fix the problem without screwing up the chain
The easiest method would be to go back to #134, but that would require deleting half the answers
 
You've seen the graph? For each intersection point you need at least 2 answers changing.
So, probably we need to fix the answers...
 
yeah, that's the plan
how about this in Husk?
 
So... will you give another week for fixing the existing answers?
 
3:14 PM
@NieDzejkob I understand it, so seems good
@user202729 I don't think so, unless it's needed, because later answers shouldn't be invalidated
 
So we will fix existing answers while new answers are written? That's more work to do.
 
@user202729 Well, I;m not answering the challenge anymore (don't understand most sequences), so should be able to get the answers fixed (with a bit of help). Of course, if the next answer is difficult (but not insanely difficult), then that gives us ample time to fix them
Hehe, about 50% of my total edits are on answers on this challenge :P
 
hello
working on the next sequence?
 
not only next sequence. Did you read what happened?
 
in The Nineteenth Byte, 56 mins ago, by NieDzejkob
uh, bytecounts 326, 267, 190 and 85 in the OEIS challenge were used twice, and none of the answers were backup answers...
 
3:22 PM
basically, I decided I would add an option to find backup answers to the snippet, and by accident I discovered that ^
 
Oh hell...
 
you know Husk, right?
 
I do, why?
 
the first answer in the chain that repeated a bytecount was mine, this one
 
Husk is one of the answers that need 'fixing'
 
3:26 PM
the plan is to change its bytecount and then do the new sequence in the same language and bytecount and replace the next answer, which happens to be your Husk answer
 
@cairdcoinheringaahing Oh mine needs fixing? What byte count should I use (lowest possible)?
@NieDzejkob Wait the answer before mine is invalid and I must change it too?
 
@Mr.Xcoder let's focus on mine answer in Emoji and your answer in Husk first. The answers that need fixing are in pairs
@Mr.Xcoder we determined that this is the least disruptive solution to this problem
 
So all I need to do is change the byte count?
 
@Mr.Xcoder Kinda. We're changing the byte count for Nie's Emojicode answer, which is the one before your first Husk answer. So we need to change your Husk answer to calculate the new sequence, while keeping it at 29 bytes
 
@Mr.Xcoder unfortunately, no
@cairdcoinheringaahing s/Emojicode/Emoji/g
 
3:28 PM
@NieDzejkob Oh so I should delete now?
Oh I got it
 
@Mr.Xcoder no, there's no need for that
 
@Mr.Xcoder Basically, can you do this in Husk?
 
Ugh, in 28 bytes?
 
because golfing is hard, I'd kindly ask you to look at the unused sequences starting from A268 and find one that is doable in 29 bytes
 
29 or less (then you can pad it if needed)
 
3:29 PM
@Mr.Xcoder it has a formula, but stirling numbers are a beatch
 
Wait lemme try to do that in 29
 
I'll let Nie talk you through it (if you need it), I seem to be hindering :P
 
No, no way I can do that in less than 30 bytes
Could you pick another sequence? :P
 
A270: (n-3)*a(n) = (n-3)*n*a(n-1) + (n-3)*n*a(n-2) + n*a(n-3)
 
I should be able to do that
crap.
 
3:36 PM
@Mr.Xcoder What happened?
 
Husk doesn't have e
 
What definition of e?
 
Euler's number
(unrelated)
@cairdcoinheringaahing Question: If Nie changes the sequence, can I change the language too (use something other than Husk, while not affecting the byte count)?
 
@Mr.Xcoder Assuming that the language hasn't been used twice, yeah, sure
 
@NieDzejkob Could you do it in 277 bytes instead?
Please? I'd have to do A000277
 
3:41 PM
I'm back from dinner, yes, 277 is easy too
 
Honestly, at times the sequences on OEIS are incredibly arbitrary
 
Emoji accepts additional whitespace without any problems
 
Now really 3*n - 2*floor(sqrt(4*n+5)) + 5?
Brb adding 3*n - floor(sqrt(4*n+5)) + 5.
Ok, then I'll solve A277
Let me know when you edit
 
I'll pad the answer now, change your answer when you're finished
 
@Mr.Xcoder Remember to pad your answer to 29 bytes (just a final reminder)
 
3:44 PM
Yeah
 
OK, I added padding
 
@cairdcoinheringaahing @NieDzejkob Done too, editing
 
@Mr.Xcoder you sure? I don't see it
 
Done
> editing
 
@Mr.Xcoder Hehe, double header :P
 
3:49 PM
you borked the formatting
 
Fixed
 
And you forgot to update next sequence
 
1 done, 3 to go!
 
I noticed
 
3:50 PM
What are the other 3?
 
@Mr.Xcoder here, have an upvote
 
Thanks
 
@Mr.Xcoder the snippet shows them now
 
Oh got it
I dunno excel though...
 
3:52 PM
Does anyone know Chez Scheme?
 
I don't
 
@Mr.Xcoder We can pad the Excel answer to 104 bytes :P
 
Then do it :P
 
@Mr.Xcoder I'll wait until we get the next answer sorted first
 
Well I think it shouldn't be hard to learn, let's see
It's a dumb prefix language, so
@cairdcoinheringaahing Have any available easy sequence to pad the excel one?
 
3:54 PM
@Mr.Xcoder Looking for the easiest one to go for
 
how about A000120? number of 1's in binary expansion of n
 
Exactly what I was about to suggest :P
 
In Chez... Uhm
Let's try
And we'll just go ahead and edit Giuseppe's answer completely, right?
@cairdcoinheringaahing ^?
 
Maybe just ping him and ask to collaborate?
 
@Mr.Xcoder Yeah. Once we have a solution, I'll drop a comment telling them about the issue
 
3:58 PM
"Last seen: 6 minutes ago"
 
@NieDzejkob Hmm, I'll comment and see
 
Link to the sequence?
The binary one
Oh it's A000120
 
Kinda ninja'd :P
I found this to convert to binary in Scheme
 
4:05 PM
I did too
Not particularly golfy :P
 
Now that Verbosity is on TIO, I've got to find a way to sneak it in somewhere :P "100, 000 bytes" :P
Hey @Giuseppe!
 
hey guys
I'm at work, I'll be able to tweak it in an hour
 
no problem
 
@Giuseppe That's fine, there's no real rush
 
what do I need to do?
 
4:07 PM
@Giuseppe Solve a new sequence, while preserving your byte count.
 
we can easily pad the offending answer to get A000120
 
ewwww what's the new sequence?
 
then, if you can solve it in the same amount of bytes in Chez Scheme, we're at home
 
85 I guess?
 
number of 1's in binary representation
 
4:08 PM
oh huh
ok
 
@Giuseppe This
Maybe tough in 68
 
now I have to go and re-learn chez scheme syntax haha
 
Keep in mind that we can change the sequence if needed
 
yeah no worries this is a good challenge
 
You can change the language too, if it hasn't been used twice
 
4:10 PM
in python, it would be something like
    def a(n):
        k=0
        while n:
            k+=n%2
            n%=2
        return k
(untested)
fuck, formatting got screwed up
 
chez scheme has bitwise operators
 
or lambda a:bin(a).count('1') :P
 
Oh cool
 
@Mr.Xcoder that's less portable
 
4:11 PM
hence the :p
 
you can also replace n%2 with n&1 and n%=2 with n>>=1
 
(for binary only)
How to sum?
 
we've got 68 bytes
 
I know
I am golfing as much as I can
 
@Mr.Xcoder It needs to be reversed
 
4:16 PM
@cairdcoinheringaahing Huh?
 
Hahaha
got it
 
Yeah, forgot that order doesn't matter :P
 
But sum is invariant over reversal
I think define can be a lambda
 
is it possible to do a loop?
 
4:18 PM
nvm
 
Holy crap, have you just seen ais532's edit to a challenge?
 
No, link?
 
LOL ಠ_ಠ
 
Holy crap, but this is not TNB
 
4:21 PM
Ok, back to golfing Scheme :P
 
There has to be a replacement for (remainder n 2)
Shoot
 
In case you can't golf Scheme below 68 bytes, here is A000120 in SML (which has only be used once, as far as I see): Try it online!
 
@Laikoni Nice!
 
Replaced quotient with fxmpr or smth and then just fx/
 
4:25 PM
Shame that Scheme doesn't auto-close brackets
 
:p
Also we have to sum, and I have no idea how to do this
 
64 bytes!!!! (define(d n)(cond((< n 1)'())(else(cons(even? n)(d(fx/ n 2)))))) and returns truthy/falsy... If we can simply sum that up
 
@Mr.Xcoder You also have to not each
 
Wait we are recursing, so no chance to sum it this way!
Also, I can still use Swift :-)
On a slightly less related note, the challenge will end if nobody solves the next sequence in... 2 days
 
4:31 PM
61 bytes: Try it online!
 
@Mr.Xcoder Yep. I think user202729 said they'd be willing to do it if needed
@Laikoni Very nice!
 
@Laikoni That is amazing
 
Time to add padding :P
 
Actually 59 bytes, one pair of the parents was superfluous.
 
4:33 PM
@Giuseppe When you finish work, you can just update with what Laikoni did in 69 bytes.
 
@Mr.Xcoder Come on, the only real way to pad Scheme is to add even more parenthesis. :D
 
@cairdcoinheringaahing Any of my other answers involved in this? Cuz I'll have to go soon
 
@Mr.Xcoder Not as far as I can tell
Should I edit the two answers, or wait for Giuseppe?
 
I think you should edit
 
go ahead and edit
 
4:39 PM
ಠ_ಠ I'm trying to add more parenthesis to the code and it keeps failing :P
 
well you can wrap it with (+...) twice to get to 67 bytes Try it online!
 
Edited
2 gone, 2 to go!
I'll unpin that after we're done
 
So what's next?
 
good luck, y'all, and thanks @Laikoni for fixing the code :)
 
4:52 PM
BRILLIANT!
I can handle the next one all on my lonesome :P
Sh*t I made a mistake
@NieDzejkob For this would it matter if you changed 4.188790204786390 to 4.18879020478639?
 
@cairdcoinheringaahing Let me check
@cairdcoinheringaahing it doesn't matter
 
5:12 PM
@NieDzejkob Perfect
3 done, 1 to go!
It seems fairly quiet right now, I think fixing the final one can be done tomorrow
And A330 seems to be a good target
 
what's the last one
 
Oh no... But the next answer is Hexagony
 
I'll look into it
 
5:39 PM
I got it, let me update the answer together with the explanation
 
6:04 PM
@user202729 could you pad your answer?
 

  last day (49 days later) »