Всем привет!
Скоро, 29 ноября в 19:30 MSK, состоится очередной Codeforces Round #216 для участников Div. 2. Как обычно, Div. 1 участники смогут поучаствовать в этом раунде вне конкурса.
Задачи были подготовлены группой авторов в составе: Гриднев Виталий (gridnevvvit), Лось Илья (IlyaLos).
Большое спасибо Гере Агапову (Gerald) и Сергею Сухову (Serega) за помощь в подготовке задач, Марии Беловой (Delinur) за переводы на английский, Михаилу Мирзаянову (MikeMirzayanov) за замечательные системы Codeforces и Polygon.
UPD1: Распределение баллов будет таким: 500, 1000, 1500, 2000, 2500.
UPD2: Это была опечатка с распределением, теперь все верно.
UPD Соревнование закончено, поздравляем победителей!
Удачи!
Friday. And again, and again, and again. No problem, just discoursing :)
Hope This Round to be a bridge to be a Div1 for First Time ,Hope That!
What do you mean by again? there hasn't been a contest on Friday since 25th of October! :D
I think what he means is there is no contest especially for DIV.1. :D
GL && HF!!! I hope this contest will be good for everyone!
Why this blog is not on the Home page ?
An empty contests list !! I hate that ...
Then do something about it! Prepare a contest! No-one gets better from your trash. Sorry.
Всем Удачи!!))) GooD LucK
"Scoring will be next: 500, 1000, 1500, 2500, 2500."
So another Div.2-Only contest where maximum 10 people will solve D and/or E? :)
misprint :)
Scoring will be next: 500, 1000, 1500, 2000, 2500.
HF:)
I thought there are more than 3000 programmers taking part in this contest~Nice weekend~~&&HF
short statement hope .....
Why isn't runtime error considered as a successfull hack?
I tried and it says unsuccesfull hacking attempt
however it should have shown runtime error according to my machine
A program can behave differently on different machines. (On yours, it crashes, but it works fine here.) This happens sometimes, and that's why it's risky to hack on RE or TLE.
There were some sure cases of division by zero. During hack people got caught, no problem. How the passed the final test. In which machine on earth found the result of division by zero successfully!!! please, Read this thread. http://codeforces.net/blog/entry/9771
And very carefully look at the case 19 in the test!!! how can that pass and generate correct output by dividing by zero !!!!
1 1 1 1 1 1 was the case by the way.
Depends on the reason of runtime error. If it's accessing an out-of-bounds element of array, the behaviour is indeed undefined. But if it's division by zero, the hack should be successful. This time, I hacked 4 people on this, and they used different programming languages: GNU C++, MS C++ and Python 2.
Actually it was modulus with 0
something like 0%0.
How solve problem E?
Finally a contest with corner cases that are not covered in the pretests!
I made 10 successful hacks with the case: 5 5 1 1000 100 100
wohoooooooooooooo :D
I've made 6 successful hacks with 3 3 1 3 9 9, but firstly found, that my oun programm can't solve it :D After blocking the solution.. It was offensively..
I made 11 successful hacks with the case : 5 5 1 3 5 5
I made 11th hack in last minute :)
it's unfair :D
Bon appetite!
fucking carelessness !
there isn't any test like your test in system test , is there ? :D
Why so serious. May b u will want leave the earth as soon as u read this http://codeforces.net/blog/entry/9771 Look at the test 19
When I wanted to hack some solution, I found something really interesting...
In GNU C++, this code doesn't get RE
It works well and makes answer equal 0
and also this code
if you get a = 0, the output is 0
but this code will get RE
why??? Is this a bug or...
I did the same thing.
unsuccesful halking attempt.
on users with (sall-sk)%(n-k).
Maybe compiler optimization of the form "0/anything==0"?
any1 knows why i have skipped on all 3 tasks
Я бы хотела помогать переводить. На каких условиях это происходит и что я могу для этого сделать?
Напиши Михаилу Мирзаянову или Геральду Агапову.
What is the approach for solving problem D and E? Thanks in advance
very very awful system test speed
Zazhralis'.
After many rounds.....finally a TRICKY contest!
Two consecutive tricky contests , last was one also full of hacks and failed system test.
what's the algorithm for problem E? i tried to use Fenwick Tree(Binary Indexed Tree) but i couldn't figure out how to use it 8-|. i solved problem just for 1 point... but for couple of points :-?? can anybody give hints? it would be great!
What was special about test 14 in problem C ?
Do you try this case? 4 1 2 2 2 3 1 2 4 2
Answer should be 1 4
My code gets it right
How did you get around Test #14?
Any update on this?
I'm stuck to this for hours now :/
Условие 4-ой задачи улыбнуло:)
Бывали и позабавнее. На одном контесте была задача про раскрашивание яиц. Все фразы были специально так подобраны, чтобы у всех фантазия разыгралась :)
Была и задача, в которой надо было вывести "I'm too stupid to solve this problem", если решения нет (а оно было всегда)
Это неловкое чувство, когда после контеста не помнишь ни одно из условий.
Это нормально? 2 задача валилась на 4 тесте, сейчас посмотрел ошибку: wrong answer sum of scores of the team doesn't equal to 892330, 892248 попробовал свой код для входных данных на 4 тесте, посчитал сумму. Сумму выводит правильно. Спрашивается, каким образом тут вылезает эта ошибка.
А, нет, нормально все, не тот код проверял :)
Тогда может кто сказать, в чем принципиальная разница между этими двумя кодами:
5306082
5305733
Вроде как они одинаковы, а ответы разные дают.
тест: 3 2 3 4 10 7 первое решение вернет ответ 3 3 3(7/2 +7/2!=7), что неверно, нужно 4 3 3
Я думал что sum_k должно быть кратно k
Ну да, ясно, я неправильно условие понял...
Почему?
Условие просто неправильно понял)
Обьясните кто-нибудь идею решения задачи С. Спасибо.
Подкину идейку.
Я это читал. Я хотел бы услышать суть решения, а не сделай так, раскрась так и выйдет так.
А суть в том, что нужно найти все "самые нижние" проблемные дороги, такие, от которых если идти вниз по дереву, не встретишь больше проблемных. И в ответ записать нижнюю вершину каждого такого ребра. Это так, потому что если ниже какой-то проблемной дороги есть ещё одна проблемная дорога, то верхняя будет покрашена по пути из нижней. Посчитать же все такие "нижние" можно, например, как в разборе.
Test 32 on D is the killer of reds (me too :D), what could it be about...
My solution of E is as follows: sort the intervals by li; sort the points of all queries together and for each of them, remember all queries that include it. Now, process the points in that order; in a BIT, remember how many intervals whhich start before that point and end at position i after that point (remembering their endpoints in a set also helps remove those that end before it). Check all queries asking for the current point, and add to their answers the intervals which won't be counted for any later point of that query — that is, end before the next point of it (or infinity, if that point is the query's last), which is just a sum from the BIT. The query's next point can be found just by remembering a pointer to it and increasing it when necessary.
Here's the simplified 32 test on D.
3 2
100 0 50
Solutions that fail on 32 test do not find the way to kill all fools and output 4 while the correct answer is 5. The main idea behind this test is that if one alive fool remains in front of a suffix of alive fools with a gap between them and he kills the first fool from suffix and dies himself at the same time, it may produce one more answer.
Here it looks like that:
round 1: The second fool is shot. The other two are alive.
round 2: Remaining fools shoot each other simultaneously.
Solutions that do not consider this case get WA 32.
I see, thanks. It really took a lot of casework, this one...
I liked the problems. Thanks for the contest. :) Even though I think D might be a bit hard, I have no clue on how to solve it. Hopefully, the tutorial will be out soon!
Problems statistics (official round participants only):
can u share with us how u got this data? it might help during future contests. thanks!
I've parsed contest standings pages content and aggregated data from cells using javascript. It's draft, and I plan to extend the script in order to provide more interesting information in a pleasant view :)
Good idea.
Standings :
1- Dshovn : Registered 11 Hours ago
2- WhitedarkWalker : Registered 15 Hours ago
5- Ronnoc_2 : Registered 3 Days ago
6- LinXinya : Registered 8 Hours ago
7- marcorezieho : Registered 10 Hours ago
All those newcomers are pretty good i think.
мне одному задача С напомнила цитату коммента с хабра?
"Alexey2005: ...Кстати, асфальта там тоже почти нет. Поэтому местные жители на выборах мэра всегда выбирают того, который, во-первых, живёт совсем не там, где предыдущий, и во-вторых, такого, чтобы жил как можно дальше от местной мэрии. Потому как первое (и обычно единственное), что делает новый мэр — прокладывает асфальт от своего дома до мэрии. Так-то вот постепенно и прирастает дорожная сеть."
Задача была придумана как раз по мотивам этого коммента)
Shouldn't it be allowed to lock unsolved problems in order to hack other guys' wrong solutions?
i dont know, that seems unfair to me
You may not know how to solve the problem in an acceptable time and memory bound, but you can still have corner cases that submitter may have not handle.
I don't think that's a good idea, some guy can just create new account, lock, look at other guy solution, and re-code it for his main account and you know what happen next. So I think it will increase the number of cheater since it's really easy to cheat.
Good point, the fact that hacks can be made during coding time is the problem, indeed.
I solved A in 8 minutes, but was stuck with B because of a bug, which I manage to find only after an hour. :(
многие ошиблись по поводу деления на ноль :))) я взломал 5 решении
First time top 5 :)
Будет ли разбор задач?
разбор/tutorial
are the tutorials posted ???
yes, here they are.