Very weak test cases for problem IEMCO5D

weak-test-case

#1

I want to report wrong constraints / incorrect test cases in wakeup sid question(IMECO5D).
I have gone through accepted submissions of others and noticed that those who didn’t assert the constraint 0 <= L + R <= N - 1, got their answers accepted.
I ensured that constraint and got Wrong Answer.

I also found that many bruteforce solutions were accepted and some other solutions that didn’t handle integer overflows were also accepted. But according to the given constraints, they should result in Time Limit Exceeded or Wrong Answers.

So either this question must be removed or all the submissions must be rejudged with proper testcases.


#2

Yes, we used assert to check the constraints. Even on telling the organisers about the constraint failing, there was no reply.
https://www.codechef.com/viewsolution/18329302

Such a ridiculous waste of time, for a contest whose organisers don’t even care to check there test cases (even on informing them about the mistake) and whose toughest problem passed with brute solutions.

Time to go back to codeforces. :stuck_out_tongue:


#3

Guys it was a server problem, as the size of the input file was 118 MB. Codechef server wasn’t responding with that much big file size. So bear with the issue

Proof: click


#4

@sid55 No, the problem was with the test data only.

Proof :

This solution of the team the_human1 fails assert :

https://www.codechef.com/viewsolution/18329354

This solution from the same team where they have handled the case where l+r >= m-1 explicitly passes :

https://www.codechef.com/viewsolution/18329428

In the question it was clearly mentioned 0 ≤ L+R ≤ N-1 .


#5

@sid55 I looks like that only your pc shows server problem every website. Instead of admitting mistake only finding out codechef / codeforces server problems. If there was actually a server problem then you should have contacted @admin to resolve issue at earliest.

“So bear with the issue” why ??

We are here to solve questions not regret after contest due to mistake of tester/setter. Poor contest. I’m happy that I have not participated in the contest.

And the worst part of situation is that after reportedly informed to administrators (https://discuss.codechef.com/questions/125933/very-weak-test-cases-for-problem-iemco5d) issue was not resolved.

This is not what codechef is known for.
They also have issues but they always resolved them taking this as their first priority. And make sure nobody is at an unfair advantage.


#6

Hello everyone,

I am the manager of IEMCO. Sorry for the delay in rejudge, I was stuck with some personal work as I have already mentioned. There was a bug in one of the test files of IEMCO5D (Wake Up Sid) which I didn’t catch with my assert statements due to the WA > RTE display preference. The rank list has been updated. The management team will soon be sending the prizes.

Shout out to @achaitanyasai for pointing out the possible bug in the test-cases.

Sorry for the inconvenience and delay and thank you to everyone for participating.


#7

I have seen that @admin has submitted the code: https://www.codechef.com/viewsolution/18333107 to check the constraints in practice. It got Wrong Answer. But your code still gives SIGABRT. However, if I comment the print statements, your code also gives Wrong Answer.

So is the display preference of Wrong Answer > Runtime Error?? (I mean if the same code got Wrong Answer on some test files and RTE on some other files, does verdict displayed is Wrong Answer)?


#8

Yes, I think WA > RE, this is why the setter thought his test data is correct.


#9

Sorry for the confusion: By “noticed that those who didn’t assert the constraint 0 <= L + R <= N - 1, got their answers accepted”,

I meant, those who didn’t take care of that condition got their solution accepted (it’s not those who didn’t put assert() statement in code got their solution accepted).


#10

The setter (and surely admin too) can see the results for individual test files.


#11

Hey @achaitanyasai, @vaibhav138, @meooow, I am the manager of IEMCO… I believe that if there is really an issue, we probably didn’t catch it due to the WA>RE display preference thing. None of my assertions gave an RTE. Our team also received an email from CodeChef admin regarding this today. I will have all the files checked again and rejudged. Just give a bit of time, as I am stuck with some work. I will ensure that the prizes are distributed only after correcting errors and rejudge (if it is to be done). Just asking for a small amount of time due to personal work. Hope you guys understand.


#12

+1, well said.