Weak test cases for TKCONVEX.

The test cases for the problem TKCONVEX are very weak. Even though the problem was rejudged during the contest, still I am able to find many solutions which fail very simple test case.

8 4
1 2 3 6 6 6 6 13

It can be seen that 1 2 6 6 form a convex polygon, and 3 6 6 13 form the second one.

16 Likes

same is the case with me! even I can show a solution which shows “No” on this simple case
6 3
4 5 10 12 13 17
well, weaker test cases and wrong solutions getting passed.
i think this would be taken care of now onwards.

4 Likes

O(n) solutions passed whereas the limit should had been O(n^2k)

if a problem can be rejudged 3 hours before the close of the contest, i dont see why it cant be rejudged at the end of the contest as well!

9 Likes

A better option would be to enforce a system like codeforces in which some pretests are done just after the submissions.
And rest of the cases are checked after the contest is over.
This will also prevent the server from overload during the end of the contest.

1 Like

a good problem made easy by weak cases probably.

My solution was O(N) + Factorial(k) and it’s correct on all these examples.

@logic_max and the others, be more understanding. The problems are written by people, people make mistakes. I’m sure the setters will be a lot more careful from here on but don’t criticize so much. Codechef is a great platform and a lot of users grew a lot here thanks to the problems and the dedication the setters and testers put behind it. There’s no reason for you to be “disappointed”. I’m sure it must be frustrating for you to see those kind of solutions pass. Email to feedback@codechef.com, they are really quick answering the queries even though I believe this situation is already being discussed.

I guess the main problem here is the time it took for the problem to be noticed. If they change the status of the solutions, the users that didn’t pass the corner cases will get mad because they didn’t have time to try another approach, if they don’t change a lot of people will get mad because users with the wrong approach scored with this problem. They can’t just do something without pondering about it. I’m ashamed to say that my solution also fails the corner cases but I personally can deal with a lower score, I’m not going to be happy about it but what can I do, should’ve had more attention, but what about the others? People who solved the problems days ago and could have had the time to correct it if the problem was announced earlier? Have patience, they will say something about it.

2 Likes

Though it’s unfair to those who got AC instead of WA, I think it is more unfair to those whose codes pass all the corner cases too. Because ultimately gaining that 1 point actually means nothing.

4 Likes

I really think that the solutions should be rejudged with better test cases.

Yes , even the best can make minor mistakes and therefore these solutions can be rejudged . There is no big deal if a lot of solutions go wrong because in the end codeforces and topcoder do these things and everyone is adept to it . On the other hand, later it might create confusions for those who would look for the wrong solutions to clear their doubts when the problem is put up in practice sets.And no one criticizes the mistake . Psetting is one of the toughest things to do everyone knows that .

1 Like

@karan173 : The solutions were not rejudged 3 hours before the contest ended, they were rejudged a night before the contest ended. :slight_smile:

So, the people here are completely missing the point. I am not blaming anyone and I didn’t mean so say that you rejudge the solutions after the contest or even within the contest. What I was doing here was to bring to notice the people concerned that a mistake has been committed and better precautions would be taken in future. I think it would be an awful, awful decision, (to rejudge the solutions now) as it goes against the spirit of programming competitions: predefined judge and procedural justice.
It would be like I give you 10 tries to solve a certain problem. When you gave the answer first time I told you it;s correct, but after the exam I tell you, “no, your solution was wrong, I won’t give you marks.” Now that’s what no one wants to be happening here.
Procedural justice should be considered. Adding a test IN CONTEST shouldn’t be there. The contest organizers should consider every aspect of problems properly BEFORE THE CONTEST.

8 Likes

I did’nt submit the simple solution because I knew that it would fail for simplest of simple test case

1 2 3 3 4 5

But i was really disappointed to see that people used the same algo that i rejected(after thinking of the above test case). Could have got one more AC :frowning:

That one point means something @flaminrage. But seeing your comment I have to agree with you. Those who were able to figure out the corner cases deserve that one point way more than the rest of us.

And what would you say to people who figured out the test case in the contest, and did not submit it because they knew that their solution is incorrect.And keep on trying something different. Like in this case http://discuss.codechef.com/answer_link/13690/. IS this justice? And to people who did not know that the solutions have been rejudged at the last moment of the contest. And got it wrong last moment. What about those.

To me, it would be fair to rejudge even now. Because if you didnt get the correct solution ( or think of some test case).And even when you did not get right in the first attempt, there is no gurantee that you would have done it right in 10 tries as well. So in my view, either cancel this question for all/ or rejudge it/ or make the entire contest unrated.
PS: You in all context means the people who got the problem AC but with incorrect solutions and not to @darkshadows.

2 Likes

I agree with @kavishrox . Even some of the top 20 in India have wrong solutions for this problem, they fail the test cases given here. This implies that people looking out for some great solutions would in fact be glaring at wrong approaches or missing corner cases to this great problem. Also, as it involves prizes and pride, I think that it would be unfair to those who thought of correct solutions. I think that the solutions of this problem should be rejudged as it happens on topcoder and codeforces.

2 Likes

I can understand that anyone can make a mistake and its not a very big problem that they missed some test cases but here is the thing that surprised me. Even when the user fake_coder gave the missing test case in the comments(which I think was the reason for rejudging)Why wasn’t that test case included while rejudging which was done 3 hours after the comment was posted. Some solutions have been judged correct even after rejudging which fail the test case given in the comments.
6 3
1 5 6 6 8 9
What is the reason for this?

//