I seriously want that Mike's T-shirt...

Please don't be another SpeedForces Edu Round, otherwise excited about this.

its taking too much time to be in queue is it problem on my end or happening with yall

WHY ARE THERE NO TESTERS FOR ANY EDUCATIONAL ROUNDS?

this is pure stupidity from the organizers

https://mirror.codeforces.com/contest/1837/problems never gets updated! I waited 10 minutes and then finally got to know it's WA on test 1 after 10 minutes when solutions were rejudged. queue is also long, taking so much time for the verdict

Still in queue for B. Should i resubmit?

https://mirror.codeforces.com/contest/1837/submission/207170399 была на 10 минуте, перетестирование позже. Уберите её, пожалуйста!

stringforce

"problems and examples will be in announcements"forces

(()deforces

I dont think i will participate in edu rounds anymore. I asked if my submission is going to be judged after it was in queue for 10 minutes. While I got the answer it was judged and the answer was "No comments". There is no need to be arrogant and in my opinion that was. Also, this would not happen if they tested the round.

This contest is really bad. What is this problem D? The solution is pretty dumb for a D. I think ABCD were really silly and this round is really bad. I liked problem E though.

Why time constraint was so small on F? my nlog^2n solution with segment tree could not pass, had to remake it into a priority_queue solution with same nlog^2n. (smaller constant with priority_queue)

the worst contest

another really bad educational round

B completely ruined my round,TEST YOUR PROBLEMS.

For the problem D is there any chance that we need to use more than 2 color because i tried many sequence ans only i found only 2 color is enought to paint.

Worst contest I have ever seen!..

The sample test case was wrong and it delayed me a lot. Also, because of rejudges, there was a queue so I could not see my verdict for a while.

Also, I could not submit my code in the last 30 seconds maybe it was about my internet but giving the sample wrong was a big mistake :(

may anyone help me ,plz? i dont know why does it give me WA (problem B:) https://mirror.codeforces.com/contest/1837/submission/207236832

Last tasks DEF look like CDE to be honest.

It was 1st Div. 2 round that i almost have solved full. (but wa2 in F on last minutes)

Unrated would be better than this situation

Nice E, solved it 2 minutes after contest:)

The last 2 rounds made me realise that expert is not for me ;(.

i'm gonna cry because of the problem E....

Really bad Contest . Solutions can be easily guessed .

This is really the most unexperienced game, abcd is violent, e is too big, there is no room to learn, total useless。

Again, a shit educational round. Problem A,B,C,D can be done in 25-30 minutes and rating depends on whether or not you incorrectly submit B. Last educational round for me,you learn nothing and only have headache.

Worst Contest !!! Didn't liked the problems at all

Very Cool F except the TL

Very bad experience, all the time was wasted on understanding the meaning of the question, and I once doubted whether I was from Earth

I don't know why I wasted time waiting in the queue. Could have solved other Problems earlier. LMAO

Solved A-E and passed pretest of F in 3899ms (which will be TLE in system test).

A: If x%k!=0, we just need to move by x, otherwise since k>=2 we need to move by x-1, and then move by 1.

B: Consider the longest maximal block of same symbol, let it's length be k. Then If we put 1 at every local minimum and k+1 at every local maximum, we always have enough numbers from 2 to k to fill other positions.

C: We can see that a consecutive block of same numbers (0 or 1) plays the same role as a single number, and more blocks takes more operations to sort the string. Therefore, we can fill '0' into a block of '?' between '0' and '0', fill '1' between '1' and '1', and fill '0' or '1' between '0' and '1'. For these blocks at the front or the back of the string, we can assume that there's an additional '0' at front and '1' at back of the string, which won't affect the answer.

D: For any beautiful sequence the number of occurences of '(' and ')' will be the same, so if they are not the same we can simply output -1. Otherwise, there must be a solution with k<=2. In fact, we can record the prefix-sum of the string (assume '('=1, ')'=-1) and change color whenever the sign of the prefix-sum changes, then brackets with non-negative prefix-sum will form a regular sequence, and other brackets will form a reversed regular sequence.

E: We can see that teams of number [2^(k-1)+1 , 2^k] will lose in the first round, teams of number [2^(k-2)+1 , 2^(k-1)] will lose in the second round, and so on. So in the first round, we check 2^(k-1) pairs of adjacent positions, and each pair must contain a team with number larger than 2^(k-1). If any of such a pair contains 2 numbers > 2^(k-1) or 2 numbers <= 2^(k-1), there's no solution. Otherwise we have 2^a*(b!) ways to arrange teams of number [2^(k-1)+1 , 2^k], where a=(how many pairs of positions are both undefined), b=(how many pairs of positions contains no teams with number > 2^(k-1) ). Then we can eliminate teams with larger number in every pairs, and solve the problem recursively.

F: I tried to solve it by segment tree and ternary search in O(n*(log(n))^2), but it takes too long times to pass the system test. Maybe there will be solution with smaller constant.

10k ACs on C and 4k ACs on D, wow! Never seen anything this extreme before.

Nice contest. Incase you are stuck on Problem A, Problem B, Problem C, Problem D. You can use check the following video tutorial- video link

The Worst Education Round I ever seen.

Reasons:

1.Big changes on problem B:Wrong sample,too long queue(after change),etc.This is terrible.

2.Speedy Edu.Problem A,B,C,D is too easy and E is hard.4,600 people passed D,but only 367 pass E.This mean the rank between ~400 and ~4000 is FULLY depend on the Penalty.

More I know:the B is similar as atcoder contests

My solution to B got WA on test 1 at first, but once the samples were changed (and before the announcement that the previously solutions would be rejudged) I resubmitted the same code with a no-op to see if it would pass. Then both solutions gave a WA time penalty :(.

I didn't know that a priority queue is faster than a multiset, and if the intended solution has a complexity of O(n * log^2n), then, in my opinion, should have also considered the multiset solution.

So many errors in the statements... Contests of such scale should be thoroughly tested!

worst contest I have ever seen!

this round should be unrated due to long queue time(~1 min for every submission) and wrong examples in B.

dude python was not working for question c.It was giving tle at testcase 6

Problem B has a quite similar approach with this problem.

I find it a bit funny tho because problem B is about <>, problem C is about 01 and problem D is about () lol

Using discrete to convert time complexity from nlog(1e18)log(1e9) to nlog(1e18)log(3e5) then pass pretest on Problem F. Time limit is toooooooooo tight.

One tip if you got WA on test 8 of E: Please notice the input format.

Can someone please assist me with submission :207266027 ? I am unable to determine why my submission is incorrect. Thank you very much.

800-Forces

Can anyone please explain why the answer to this test case is 4 for question 'E':

3 -1 7 -1 8 2 -1 -1 5

What I understood from the question is illustrated below: _ & 5th | _ & _ | 8th & _ | 2nd & 4th So according to this arrangement, 4th would have rank '5' which is not desired. So shouldn't the answer be 0?

Problem B ruined my contest. Please at least take a glance and make sure that the examples are correct (it was so obvious that the examples were wrong ...).

Hello, I enjoyed this contest. thanks to the problem setters.

During the contest, I solved problems $$$A - E$$$. Just after the contest, I solved problem $$$F$$$.

Problem $$$A$$$: if $$$N \mod K == 0$$$ it's $$$N-1$$$, $$$1$$$ otherwise it's $$$N$$$.

Problem $$$B$$$: the answer is the longest increasing(or decreasing) subarray.

Problem $$$C$$$: how many disjoint $$$1$$$ s are there ($$$-1$$$ if the last number is $$$1$$$).

Problem $$$D$$$: if the number of opening brackets isn't the same as the number of closing brackets it's $$$-1$$$. otherwise, if it's a beautiful bracket sequence the answer is $$$1$$$ otherwise $$$2$$$. We can color the biggest beautiful bracket sequence with the color $$$1$$$ and the remaining brackets make a beautiful bracket sequence.

Problem $$$E$$$: We know which players will be eliminated in the first round. Let's say there is a $$$P$$$ number of $$$(-1, -1)$$$ matchup, and $$$Q$$$ number of players we need to eliminate on that round. On round $$$k$$$ the number of ways we can place the players that should be eliminated is $$$(2^{P_k}\cdot Q_k!)$$$. Then, we can assume those players are eliminated on that round and solve the problem for $$$k+1$$$-th round. Thus, we can recursively find the answer.

Problem $$$F$$$: I didn't solve this problem during the contest. I used multi_set, which I later changed into priority_queue to get accepted. My time complexity is $$$O(N \log^2 N)$$$. I don't know if the problem setters intended to cut submissions that used set. The idea of this problem is to find the answer with a binary search. For a chosen value, we check for each element to be the last problem of the Monocarp, and find the maximum number of editorials we can make in that situation. If it's over or equal to $$$k$$$ we can make it in for the chosen value. I still think the time limit for this problem is a little too tight.

Overall it was a good contest :)

The problems were very easy till D , it must be somewhat tough at least for C and D.

please start writing div3 contests

I passed F in last two minutes! Hope it won't FST.

D can u tell what's wrong with my soln first i eliminate all the regular seq from string and colored them with 1 then i reversed remaining and colored them with 2, if finally the stack is not empty i print -1; then did above process again but this time i reversed first then printed min of above 2 processes

Seeing even experts getting WA for first time in B is such a relief..

unrate it please.

Problem B is bad :( Promblem D is easy but I made many stupid mistakes... bad experience.

"How many greedy questions have you prepared for this contest?" Problemsetters: Yes

This was not a good contest for div2. All problems were very easy compared to their expected hardness.

Can anyone please explain why my solution is incorrect?

207239600

https://mirror.codeforces.com/contest/1837/submission/207239600

Was this the easiest div 2D ever in terms of problem rating and solve count?

queue is so long!

maybe change the problemsetters for some time

Can anyone give a counterexample where my code gives a wrong answer?

207246737

What is this: not accepted / accepted
Only difference that the first one may only print out 2s instead of ones, but there is nowhere specified that colors must start with 1...

Wow, I am little surprised to see the number of submissions of B. Was the solution leaked somewhere ?

Hi this is the first time I join in a competition. But I only resolve the first three problems (A, B, and C). Will I get a rating score? Because I saw that my score is still 0 in my profile.

Я сегодня впервые участвовал в контесте и с интересом изучаю чужие взломы.

При этом мне показалось, что в некоторых случаях отдельные участники соревнования кое-где у нас порой специально публикуют решения с изъяном от имени клонов и потом их с гордостью взламывают.

Например, четыре успешных взлома задачи А: https://mirror.codeforces.com/contest/1837/hacks?verdictName=CHALLENGE_SUCCESSFUL&chosenProblemIndex=A — удивляют схожестью идеи и явным сходством в никах автора решения и взломщика. Или мне только кажется?

Возможно, есть какой-то принятый способ сообщать о подобных хитростях?

very high time limit for F

gordan ramsay after giving todays contest.

"My gran could do better!"

F is the most messed up problem I've ever seen. I can't believe it's that easy yet I can't solve :(

In problem B, if I start with the number X initially, if I encounter '<' I increment X else decrement X . Here is the code for it

vi ans;
ans.pb(5);
ll val = 5;

f(i, 0, n)
{
    if (s[i] == '>') {
        val--;
        ans.pb(val);

    }
    else {
        val++;
        ans.pb(val);
    }
}

deb(ans);

set<ll>st(all(ans));
cout << st.size() << endl;

}

Why is this approach wrong?

someone pls explain how to solve problem F

this contest should be unrated because of the problem B.

My approach for Problem F:

Let us say we already have the k elements which are picked out from the original array, we can create a prefix and a suffix sum array and then iterate over the two picking the minimum of max(prefix[i], suffix[i]).

So now all we have to do is to find the k elements, we can find the k smallest elements in the original array and then take the maximum of that array, then check if we can/have to pick more than one occurrences of the max element from the original element, if yes we somehow have to find which occurrences of the max element do we have to pick in order to reach the minimum later on when we calculate the answer, I understand that the position of the max element matters but I cannot figure out how to optimize it in such a way so that it leads to the minimum ans.

Auto comment: topic has been updated by awoo (previous revision, new revision, compare).

Trash round.

My rating was updated to expert few days ago and today it is showing back to specialist, it has reduced by some points in this contest. Have anyone else faced this?