Chat:World/2021-06-22
TinyKitten: sup
Mr_BlueBird: hi
DavidDeLord: is it just me or are the test cases for most of the clash of code problems awful?
BlaiseEbuth: You're not so awful...
Alihussain: CAN ANYONE BEAT ME
Westicles: DavidDeLord, recent approvals get more play, and they are indeed awful
DarkPhoenixBoi: yo i have a question
DarkPhoenixBoi: what is "process has timed out " means? my code worked for all test cases except for the one with so many inputs
Bbaaii: that means that your code is inefficient
Phurinat: Your code is too slow, you should look to optimize it. Maybe in your loop
DarkPhoenixBoi: oh OK thanks guys now i get it <3
MiyamuraIzumi: Hi is there anychange in how server run validate
MiyamuraIzumi: most of my randomize algorithm return different result with the same seed :(
MiyamuraIzumi: my seed in optimize game actually
eulerscheZahl: do you have some kind of (while time < xx) tryRandom()?
MiyamuraIzumi: ofcourse yes
eulerscheZahl: to be deterministic, you have to use a fixed number of simulations
eulerscheZahl: you won't get the very same number of iterations. you won't get the same number offline on your PC either
MiyamuraIzumi: you are right, I ain't think about that
MiyamuraIzumi: it is just that I used to get the same result with same seed
MiyamuraIzumi: but now it is gone :(
PatrickMcGinnisII: MiyamuraIzumi execution time varies a bit, depends on server machine i guess
MiyamuraIzumi: I just try my old code
MiyamuraIzumi: I will keep my old rank then :D
PatrickMcGinnisII: eh, never give up
Westicles: Anyone done Cross the Lines puzzle?
PatrickMcGinnisII: not I
Westicles: Not sure how to approach it.
Westicles: https://imgur.com/a/8kUNhS6
MiyamuraIzumi: then don't approach it
Westicles: If 69 people can do it then I can do it
Westicles: I guess maybe you look for polygons then even/odd number of sides
PatrickMcGinnisII: u could easily beat me at connect4
PatrickMcGinnisII: at least until i finish my bitboarding
PatrickMcGinnisII: ;)
GarethEddies: hi
1natoz: sup
latteddy: hi
Calimero90: hi all ! I've done test in Javascript and now I aim in Java... How to return in Javascript now ?? Don't found it !! Tk
nickeyminasj: u cannot return it now u have to learn java to continue
Calimero90: how ? cant find menu or buton for this !
Uljahn: there should be a drop-down menu atop of your code
BlaiseEbuth: There's a dropdown list at the top of the IDE.
Calimero90: ha ok. So I have to go in puzle to chose.. Ok and Tank !
1natoz: IM GAY
Simone-Riva: hello guys im from england america and my president is michael jordan
1natoz: IM GAY
AlessandroSpaziani: NICE
Siro_Sisso: I like shrek
1natoz: WHO WANT MY LIL COC
Siro_Sisso: no bro
Siro_Sisso: O_O
Siro_Sisso: everybody dance now
KiwiTae: 1natoz happy pride month ~
BlaiseEbuth: Sry kiwi, already banned him :3
KiwiTae: ><
KiwiTae: well done BlaiseEbuth
BlaiseEbuth: :sunglasses:
AlessandroSpaziani: :slight_frown:
AlessandroSpaziani: Are you homophobist?
BlaiseEbuth: Worst. I'm a moderator.
Simone-Riva: ¢f7uzjkvxélokicvuzzuplto
genthios: can you report a moderator for being a moderator? lol
derjack: oO
ableflyer: how do you write int(i) for i in input().split()] using stub generator input
ableflyer: anyone?
ableflyer: how do you say input().split() but in stub generator
nothansk: what is stub generator
BlaiseEbuth: You don't.
BlaiseEbuth: The stub generator is 'generic' what your asking for does not exist in all languages.
BlaiseEbuth: Read the doc.
BlaiseEbuth: https://www.codingame.com/playgrounds/40701/contribute---help/stub-generator-syntax
thanveershah: input :8372?9514
output : 6
thanveershah: does anyeone know ?
BlaiseEbuth: And the guidelines too https://www.codingame.com/playgrounds/40701/contribute---help/contribution-guidelines
BlaiseEbuth: thanveershah, no, I realy don't see the link between all the numbers between 1 and 9 except 6 and 6... :thinking:
Nekko551XD: missing digit
LazyMammal: puzzle of the week is nice. can be done in ~ 1 hour.
LazyMammal: Does anyone have the link for lookup up CoC puzzles? I want to avoid duplicate submission.
nothansk: where do i find puzzle of the week
KiwiTae: took me 10min :p i liked it though
nothansk: Kiwi where is puzzle of the week
KiwiTae: https://www.codingame.com/training/medium/connect-the-hyper-dots
KiwiTae: ^ nothansk
nothansk: how do you even do this :(
LazyMammal: Orthant is like Quadrant in more dimensions. No biggie.
KiwiTae: BlaiseEbuth ^ you got another one for lunch
BlaiseEbuth: That's gluttony...
ableflyer: how can i make an array using stub generator
BlaiseEbuth: Have you even read the doc?
ableflyer: i did
ableflyer: but nothing mentioned a list
ableflyer: i read it like 3 times
BlaiseEbuth: So you suppose there's things that are not mentioned in the doc but still possibles?
ableflyer: maybe?
BlaiseEbuth: No
ableflyer: oh
ableflyer: i tried loopline z z:int but it doesn't work
BlaiseEbuth: The purpose of the stub generator is to provide the basic structure of code to read the inputs. Not to store them.
BlaiseEbuth: You can check every puzzle or multi, when there's loops reading inputs, the base code does not include arrays or lists. It just have a simple variable that is overwritten each turn
ableflyer: ok thanks
IamFish: ello
AndersLearmonth: I started the mars lander 2 challenge in 2018, now just finished it 😎 lol
AndersLearmonth: #RapidDevelopment
KiwiTae: gg
IamFish: proud of you
KiwiTae: u can start episode 3 now
Uljahn: or ML2 optimization
LazyMammal: True, there is achievement for having 400L fuel in ML2 validation test case 5.
Wontonimo: hey LazyMammal , how's it going?
LazyMammal: hey hey. I made my first contribution. https://www.codingame.com/contribute/view/7001fa1becaece3de7ec0846cef2af2d1dbf
LazyMammal: welcome
kostis14: really nice contribution @LazyMammal
LazyMammal: thx!
snoyes: LazyMammal: In your contribution's sample code, checking a span up to 10 may be overkill. The span seems to be either 1 or 2 for all values of C well into the tens of millions
snoyes: (I was trying to find a value for which your brute force approach would time out, 'cause that's the way I roll)
LazyMammal: Thanks for checking. Flaaamort provided a directly solution that seems to imply the same. My brute force solution was due to ignorance :D
Flaaamort: Yep, I can prove that the span is 1 for odd c and 2 for even c, for all c.
kyoya34: hello
kyoya34: anyo?
kyoya34: anyone?
Wontonimo: bueller bueller?
kyoya34: what?
Wontonimo: watch "ferris bueller's day off"
LazyMammal: hello? <Adele singing ...> "is this thing on?"
Wontonimo: you'll thank me
Wontonimo: yeah, nice contribution LazyMammal. I upvoted
LazyMammal: Thanks. It still needs work. The feedback has been very useful to sort out the corner cases.
KiwiTae: achievements#PZ_CLASHOFCODE_500 wuhu i can finally stop coc
KiwiTae: :grin:
KP56: who has finished this challenge? https://www.codingame.com/ide/puzzle/connect-the-hyper-dots
LazyMammal: I liked this puzzle of the week. I had to look up "Orthant" but the rest was straight forward.
KP56: why is there a space after a colon in test case #10
KP56: since there is still an a on this position: a 14 -2 -4 9
KP56: which i believe is the same orthant
LazyMammal: what are the vector coords of previous letter and letter you are asking about?
KiwiTae: criss-cross, ['-', 8, -5, -1, 6] ['-', 1, -1, -1, 1] criss-cross, a ['-', 9, 1, 3, 8] ['-', 1, 1, 1, 1]
KiwiTae: thats why
KiwiTae: 2coordinates cahnging sign
LazyMammal: +/- signs don't change (or it is zero) thus same orthant
LazyMammal: yep, if they do change (not zero) then orthant changed
LazyMammal: yes!
KP56: isn't this a meant to be in a "sauce" word?
KiwiTae: KP56 u got the wrong a
KP56: but there should be no 'a' in the same orthant as ','
KP56: so how does that make sense
KiwiTae: u go to the closest point
KiwiTae: thats it no need to check for pts in same orthant
KP56: yes but there is a space only when going to a different orthant and the 'a' and the ',' are seperated
KP56: i think im just too dumb for this
KiwiTae: find the nearest point, and then the nearest unused point from that point, and so on until all points have been connected.
KiwiTae: u got a list of points
KiwiTae: u start from origin
KiwiTae: é
KiwiTae: just get the closest point till the list is empty
KP56: dont worry i understand these basics, everything works except for the case #10
KiwiTae: criss-cross, a ['-', 9, 1, 3, 8] should be where u land
KP56: thank you for trying to help, ill have a look at it again tomorrow maybe i will understand at that time
LazyMammal: KP56 you are so close. Good luck tomorrow.
KP56: i think i figured it out Kiwi sent me his entire debug
KP56: what i did wrong is i would iterate through every orthant seperately
LazyMammal: Oh sure. When I did it, I separately compared each pair of numbers at each position in the two vectors.
LazyMammal: If *any* pair doesn't match sign, then it's a new Orthant.
KP56: fortunetly the way i implemented it makes it very simple to reformat
codeing: https://www.codingame.com/clashofcode/clash/182761462c4eb4128c208e179ee69d23a63864b
jacek: oO
jacek: Oo
jacek: reCurse do you use pitting in your totozero?
jacek: i thought the difference between alphago zero and alphazero is that there is not pit and nn is updated regardless
codeing: blaiseEbuth why you ban me from france channel ?
BlaiseEbuth: I didn't
codeing: you ok ?
Marchete: do you use pitting jacek?
jacek: no
Marchete: then you just create selfplays from your last training?
jacek: i have chart current nn vs baseline as sanity check after each iteration but thats it
jacek: yes
jacek: dunno if its better or worse than pitting
Marchete: I sometimes went downhill
Marchete: like one iteration having 7% winrate
jacek: the chart shows some regression, but it is improvement over time
jacek: 7%? i think i never had that drastic regression
Marchete: well, that's an extreme case
Marchete: but I see this too:
codeing: https://ibb.co/kJr49CN
Marchete: let me check :D
codeing: stop bro
codeing: you ugly
Marchete: Pit_gen0035_gen0021_27.777779
codeing: hell
BlaiseEbuth: Hell! \o/ :imp:
codeing: go hell ugly :rage:
jacek: :upside_down:
codeing: its for Ebuth
codeing: sorry
Marchete: talk like that to a moderator, great idea
codeing: moderator that ban anybody ?
codeing: don't nee to stay moderator
darkhorse64: people that do not behave have no place here.
nuggetbucket54: L
ChargedParticle: whats going on in here?
nuggetbucket54: idk some people are getting mad and some people are trolling
ChargedParticle: then i'll stay away
nuggetbucket54: lol good idea
jacek: yeah, trolling is very bad
fvla: imagine if codingame could actually optimize tail call recursions lol
fvla: no but seriously, i swear that they have C++ on options for zero optimization
jacek: for c++ use #pragma things
fvla: i wish we could set them up ourselves
jacek: #pragma GCC optimize "Ofast,omit-frame-pointer,inline,unroll-loops"
- pragma GCC option("arch=native", "tune=native", "no-zero-upper")
- pragma GCC target("rdrnd", "popcnt", "avx", "bmi2")
jacek: add that to first lines
fvla: oh, thanks!
fvla: this makes me pretty angry, just discovering this
fvla: i actually could have used this in spring challenge :P
jacek: :shrug_tone1:
DavidDeLord: are macros allowed?
DavidDeLord: expecially for clash of code
jacek: anything is allowed within language syntax
DavidDeLord: i meant like input macros
DavidDeLord: press a keyboard shortcut to include bits/stdc++.h
jacek: oh. i dont know
fvla: you could just copy/paste a bunch of preprocessor macros? unless that's unallowed
Marchete: extreme lazyness
DavidDeLord: for some contests people have these HUGE macro sheets
DavidDeLord: like these: https://gist.github.com/kodekracker/e09f9d23573f117a5db0
jacek: O:
PatrickMcGinnisII: o:q
Marchete: :O
Marchete: include <bits/stdc++.h>
DavidDeLord: BuT iTs NoN sTaNdArD
fvla: please never use it in an actual project
fvla: not because of the portability, but for build time
DavidDeLord: fvla: https://en.wikipedia.org/wiki/Precompiled_header
fvla: i was referring to bits/stdc++.h
Marchete: I use it on every and all projects
Marchete: in CG
struct: main reason I dont use it its because its a gcc thing
struct: Cant use it on vs :(
AllYourTrees: whats vs?
struct: its easy to set it up to get it to work
DavidDeLord: https://i.redd.it/pgq7pxrqykb51.jpg
struct: but I never bothered
Marchete: I did, it's like adding 1 file on a folder
GreenPineapple: Has anyone tried this puzzle yet? "https://www.codingame.com/ide/puzzle/divisibility-of-fibonacci-numbers-sum" I am failing to optimize around the 5 digit mark :/
LazyMammal: Not yet, but looks interesting.
BobLob: What are you trying to optimize? Generating the number from the sequence?
Westicles: that one is tricky
KiwiTae: GreenPineapple theres a trick to get it
KiwiTae: can find in the fibonacci properties
fvla: GreenPineapple hint: exponentiation is achievable in linear time in terms of the exponent
Husoski: @GreenPineapple I'm doing this one now, pure Python or bust. If you read the rest of the Wikipedia article (link in problem statement) there's some useful info.
Wontonimo: Hey Marchete, I saw a link to your write up on OwareZero. The link was to page 2 and for some reason I thought that was the whole write up and I though it was great
Wontonimo: then i saw you had several pages of awesome detail. It's a fabulous write up!
Wontonimo: Thanks so much for sharing
Wontonimo: Westicles, did you do the divisible fibonacci puzzle?
Westicles: yeah
Westicles: You need to find two large values in the fib sequence to solve. There is a log2(N) alg for finding them
fvla: by log2(n) you mean by the size of n or the number n itself?
Westicles: the nth term
fvla: ok
TheBigBadMaster_164: hello, did anyone solve the lucky numbers of 6s and 8s with Java?
Husoski: Yay! Got "Divisibility of Fibonacci number sum" ...100% on first clean compile!
Husoski: ...and no numpy!
Plantchant: I'm now realizing how inefficient Lua is
Boraz10: poop
TeachMeSenpai: can i see previous submissions of my clash of code?