Codeforces Round 950 (Div. 3)

5 месяцев назад, # ^ |

← Rev. 4 →

+25

(I think that if you send 2 or more accepted submissions then the last one is the penalty giving one).

Actually, every submission after the first accepted one does not count, so the penalty is (the number of wrong submissions before the first accepted submission) * 10 + (the time in minute the first accepted submission was made). This is why you can freely submit multiple solutions if you have any doubt in your previous accepted solution without worrying about the penalty in Educational / Div. 3&4 rounds (as opposed to normal Div. 1 or 2 rounds where the earlier solutions become WA penalty).

→ Ответить

Im_Gonna_Be_King_Of_Reds

5 месяцев назад, # ^ |

Thanks everyone for clearing my doubt.

→ Ответить

aryan23147

5 месяцев назад, # ^ |

If we had submitted 2 wrong answers for a problem and are unable to get an accept verdict for it then how much pentaly would be added.

→ Ответить

Prince2618

5 месяцев назад, # ^ |

Then there will be no penalty

→ Ответить

hewhocodes

5 месяцев назад, # |

Hopefully last rated div 3 for me

→ Ответить

5 месяцев назад, # ^ |

same

→ Ответить

nootnoot1729

5 месяцев назад, # ^ |

isn't it already unrated?

→ Ответить

5 месяцев назад, # ^ |

nah I signed up as a specialist

→ Ответить

mathmath33

5 месяцев назад, # ^ |

Well most probably you will become expert brother, congrats !

→ Ответить

Whitedevil08

5 месяцев назад, # |

-9

Please fix the ratings for Educational CF Round 166 as soon as possible, the ratings were changed till 14,000 (approx), what about others? Thank you for your efforts!

→ Ответить

Emil_Aliyevv

5 месяцев назад, # |

-8

Hi. MikeMirzayanov, please reply my comment.

→ Ответить

Emil_Aliyevv

5 месяцев назад, # |

-38

Hi, MikeMirzayanov. Please reply my comment.

→ Ответить

Avik_AB17

5 месяцев назад, # |

Excited for the milestone Codeforces Round 950! A big milestone for the community! :)

→ Ответить

5 месяцев назад, # ^ |

you're 50 rounds too early

→ Ответить

Avik_AB17

5 месяцев назад, # ^ |

Setting problems for each of the 950 rounds is not an easy task. It's worth celebrating every milestone. Big thanks to all the problem setters, authors, testers and MikeMirzayanov!

→ Ответить

5 месяцев назад, # ^ |

-7

bro is teaching an expert about problem setting (aura -1000)

→ Ответить

Nxxlt

5 месяцев назад, # ^ |

+17

Anyways, he is right. Problem setting is a hard task

→ Ответить

5 месяцев назад, # ^ |

ik he is right, i have done problem-setting in few clg contests, but what's the sense of celebrating 950 ? has he been celebrating each round if that's the case

→ Ответить

Nxxlt

5 месяцев назад, # ^ |

You are right, but you wrote about problem setting, so I commented about that.

→ Ответить

5 месяцев назад, # ^ |

agreed sir

→ Ответить

Stuck_On_Gray

5 месяцев назад, # |

+70

I like when FBI tests the round ! I feel more secure

→ Ответить

uchiha_itachi_1308

5 месяцев назад, # |

-13

Problems are getting more difficult these days. I can solve at most B in div-3, div-2. Are testers creating more difficult questions? Previous questions were easier than recent ones, and question difficulty was according to Div, but now the scenario has changed.

→ Ответить

Codertang

5 месяцев назад, # ^ |

← Rev. 2 →

+25

As the problems change and adapt, we too must become stronger and evolve into more superior problem solvers

             (\__/)
              (UwU)
          ＿ノ ヽ ノ＼＿ 
        /　`/ ⌒Ｙ⌒ Ｙ　 \
        ( 　(三ヽ人　 /　　|
        |　ﾉ⌒＼ ￣￣ヽ　 ノ
         ヽ＿＿＿＞､＿＿／
            ｜( 王 ﾉ〈 
             /ﾐ`ー―彡\ 
            / ╰    ╯ \

POST CONTEST UPDATE: i did not in fact get any stronger ;-;

→ Ответить

mostafaabdelaal_03

5 месяцев назад, # |

-12

احلى مسا عليك ياعم vladosiya يلي مصر كلها بتستناك تنزل بالديفاية

→ Ответить

My post contest discussion stream

5 месяцев назад, # |

Will surely try to cross and reach pupil in this round!<3

→ Ответить

RAJAB

5 месяцев назад, # |

-17

Hope to reach pupil in this round ^^ .

→ Ответить

daud04

5 месяцев назад, # |

← Rev. 2 →

-15

Please pray for my friend to be blue in this contest. Best of luck Mahi vai

→ Ответить

Murad134

5 месяцев назад, # |

-11

hoping to return back the blue handle

→ Ответить

Alikoshikus

5 месяцев назад, # |

-12

Good luck guys

→ Ответить

aryanc403

5 месяцев назад, # |

+13

→ Ответить

burstingfire355

5 месяцев назад, # |

another round, another reality check for me

→ Ответить

iamshihab100

5 месяцев назад, # |

Aim to reach pupil

→ Ответить

iamshihab100

5 месяцев назад, # |

Aim to reach pupil.

→ Ответить

ad1tya_kumar

5 месяцев назад, # |

-38

To be honest in Div 3 its more difficult to increase rating than Div2 as in Div 2 you solve 2-3 problem is equivalent to solve 4-5 problems in Div3 and I think its also the case that many high rating people give it with their fake accounts.

→ Ответить

Dominater069

5 месяцев назад, # ^ |

+79

have you considered that maybe the problems are easier?

→ Ответить

5 месяцев назад, # ^ |

lol

→ Ответить

5 месяцев назад, # ^ |

+11

Statistically, and for reasons it's the opposite; lower divisions are easier to gain rating due to new accounts usually being overrated (note that the performance is calculated based on effective rating, not displayed rating).

→ Ответить

ad1tya_kumar

5 месяцев назад, # ^ |

-39 :(

→ Ответить

emam_hasan_himu

5 месяцев назад, # |

so if someone didn't attend at least 5 rated contests and solved at least one problem on each contest can't register on this contest ?

→ Ответить

synthborne

5 месяцев назад, # |

People are often excited about their first unrated Div3. But I'm excited about my first rated Div3 as an expert. Good luck to all the participants!

→ Ответить

synthborne

5 месяцев назад, # |

← Rev. 2 →

+10

People are often excited about their first unrated Div3. But I'm excited for my first rated Div3 as an expert. Good luck to all the participants!

Edit : They probably did me unrated haha, all good!

→ Ответить

5 месяцев назад, # ^ |

they unrated me too lol. tfw you can't div 3 abuse as an expert

→ Ответить

Mr_FM

5 месяцев назад, # |

Good luck to everyone!

→ Ответить

kinitic013

5 месяцев назад, # |

Hurray!!! my 100th contest. lets see how it goes fingers crossed

→ Ответить

ShivamJNU

5 месяцев назад, # ^ |

Expert in 100th contest :) All the best !!

→ Ответить

kinitic013

5 месяцев назад, # ^ |

Thank you :D, again fingers crossed

→ Ответить

rangerscowboys

5 месяцев назад, # |

Good luck everyone!

→ Ответить

bhataktaBhai

5 месяцев назад, # |

I am not sure I understand the participation rules. I recently hit 1600, but I do meet both criteria for being a trusted participant. However, when I enter the contest, it shows me "Out of competition".

→ Ответить

5 месяцев назад, # ^ |

I also wonder why the 'official standings table' only seems to show people under 1600.

It even states '[trusted participants only]' when I uncheck 'show unofficial', while not showing contestants ranked [1600,1900), so there seems to be some mismatch between the definition used by the standings and the one in the announcement.

→ Ответить

5 месяцев назад, # ^ |

If I understand correctly, to be in the official standings table you need to currently be in below 1600 and never have been above 1900. This is probably to allow people who briefly used to be expert but not people who were much higher at one point

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

Interesting, thanks!

Doesn't that mean that a contestant with top rating of 1900+ and current rating below 1600 will somehow both be rated and unofficial at the same time?

Edit: Oh, I guess it would not be that weird, since e.g. new accounts would be in the same boat. Still pretty funny to have unofficial participants who are rated.

→ Ответить

5 месяцев назад, # ^ |

It's not unofficial, it's "trusted" which basically means that the person belongs in a div 3 contesr

→ Ответить

5 месяцев назад, # ^ |

I mean unofficial as per the 'show unofficial' checkbox on the standings page. So if you wanted to see only the rated contestants, there's no easy way to do so, right?

→ Ответить

5 месяцев назад, # ^ |

Actually a good question which I unfortunately can't answer

→ Ответить

Matveyko

5 месяцев назад, # |

+30

YES NO forces

→ Ответить

MJSaif

5 месяцев назад, # |

I am too dumb to understand the problem statement of problem C...

→ Ответить

AntiCoreEnthu

5 месяцев назад, # ^ |

I understood the problem and when I build the approach, on checking the test cases I find out I was wrong and then tried again the same thing, similar things happened 4 to 5 times and now I think the test cases are wrong

→ Ответить

MJSaif

5 месяцев назад, # ^ |

i asked the authors,, "what is suitable sequence???"

they replied,, "It is the sequence which satisfy the problem statement"..

&& it was very helpful....

→ Ответить

Rache_Bartmoss

5 месяцев назад, # ^ |

"Hmm Yes the Floor Here Is Made Out of Floor"

→ Ответить

accord

5 месяцев назад, # ^ |

← Rev. 2 →

+18

one thing you learn the hard way: it is almost always you who are wrong, not the testcases

→ Ответить

huanyizhiyuan

5 месяцев назад, # |

← Rev. 2 →

-34

How terrible problems to me!I should rest to think myself.

→ Ответить

Curious69

5 месяцев назад, # ^ |

+18

couldn't solve == terrible problems

→ Ответить

huanyizhiyuan

5 месяцев назад, # ^ |

You are right.Perhaps it's due to losing too many points and having a bad mentality.I should pay attention to my own problems

→ Ответить

NAINAAAA

5 месяцев назад, # |

← Rev. 2 →

MAYBE hashmapForces

→ Ответить

demitronization

5 месяцев назад, # ^ |

+10

ORDERED!!

→ Ответить

Vampire

5 месяцев назад, # |

Excited to see the forest.

→ Ответить

5 месяцев назад, # ^ |

+36

what competitive programmers think outside is like

→ Ответить

this_is_torcher

5 месяцев назад, # |

+17

Firstly ,i loved the problems but there might be too many problems of high difficulty, spent a lot of time solving F2 which was fun in its own way and didn't have any time left to read problem G.

→ Ответить

5 месяцев назад, # |

F2 and G are nice simple to understand and practice some techniques. Overall round feels quite hard for rated participants

→ Ответить

SymphonyOp2-CATPARADISE-

5 месяцев назад, # ^ |

Need to improve aesthetics.

→ Ответить

The_dreamer_

5 месяцев назад, # |

+12

Miss G by just 1 second. Emotional damage

→ Ответить

5 месяцев назад, # |

What was the intuition behind E?

→ Ответить

5 месяцев назад, # ^ |

The rows and the columns of b must be permutations of the rows and the columns of a. If you think about it, when you switch columns, you are just changing the order of the rows and when you switch rows, you are changing the order of columns.

→ Ответить

5 месяцев назад, # ^ |

Thanks

→ Ответить

smit_384

5 месяцев назад, # ^ |

what is time comlpexity to find vector in set if vector is sorted?

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

-19

In general. Time complexity for finding in a set is :

O(log(N) * F(M))

where F(M) is the time needed for comparing two objects of class M.

In the case in an integer, it is just 1.

But in the case of a vector, this is the size of the vector as you compare elements one by one, so you need to iterate over all the elements in the set.

It doesn't matter whether the vector is sorted or not, what only matters is the time needed for the compare you do, which is O(#of vector elements).

This is the case in strings as well.

→ Ответить

garlic_bread_0204

5 месяцев назад, # ^ |

when swapping rows , u will swap the entire rows , when swapping columns , u will swap entire columns...

so elements in same row will tend to remain in same row and elements in same column will tend to remain in same column...that's all..

→ Ответить

5 месяцев назад, # ^ |

So won't taking the sum of each row and column and then checking whether those sum exists in both suffice?

→ Ответить

garlic_bread_0204

5 месяцев назад, # ^ |

you can see my code , it is quite clear ig.. I created two maps , first one for storing rows corresponding to every element in b and second one for storing columns for every element in b...

I just iterated in matrix A twice , first row wise and then column wise

when i iterated row wise , I checked that every element in the same row of a matrix should have the same row in b as well (I already have the rows of every element in b matrix pre-computed)

Did similar for columns as well...

264014966

→ Ответить

5 месяцев назад, # ^ |

Thought of a similar approach first but believed it would give tle.

Thanks

→ Ответить

daud04

5 месяцев назад, # ^ |

fix first row and first column of b.

then check others

→ Ответить

5 месяцев назад, # ^ |

Can you elaborate a bit more on this?

→ Ответить

daud04

5 месяцев назад, # ^ |

If you fix the 1st row. You cant make any column swap later, because any kind of column swap will change your 1st row.
Do same for 1st column.
Then you can't swap anything else.
You can just check weather you got the sequence or not.

→ Ответить

5 месяцев назад, # ^ |

Thanks for explaining.

→ Ответить

ag._.

5 месяцев назад, # |

I spent 85 minutes on Problem E and cannot find the error: **** int main(){ ios::sync_with_stdio(false); cin.tie(NULL); cout.tie(NULL); ll t=1; cin >> t; while(t--) { ll n,m; cin >> n >> m; ll r1[n],r2[n],c1[m],c2[m]; for (ll i=0;i<n;i++) { r1[i]=0; for (ll j=0;j<m;j++) { if (i==0) c1[j]=0; ll x; cin>>x; r1[i]+=x; c1[j]+=x; } } for (ll i=0;i<n;i++) { r2[i]=0; for (ll j=0;j<m;j++) { if (i==0) c2[j]=0; ll x; cin>>x; r2[i]+=x; c2[j]+=x; } } sort(r1,r1+n); sort(r2,r2+n); sort(c1,c1+m); sort(c2,c2+m); bool flag=true; for (ll i=0; i<n;i++) { if (r1[i]!=r2[i]) { flag=false; break; } } for (ll i=0; i<m;i++) { if (c1[i]!=c2[i]) { flag=false; break; } } if (flag) { cout<<"YES\n"; } else { cout<<"NO\n"; } } return 0;

→ Ответить

Rev

5 месяцев назад, # ^ |

Your Answer: Yes

Correct Answer: No

→ Ответить

ag._.

5 месяцев назад, # ^ |

Thank you

→ Ответить

senguptasushruta

5 месяцев назад, # |

+19

CODEFORCES wants us to communicate only in YES and NO

→ Ответить

accord

5 месяцев назад, # ^ |

+22

And sometimes MAYBE (problem B)

→ Ответить

N00bCode

5 месяцев назад, # ^ |

Maybe

→ Ответить

ad1tya_kumar

5 месяцев назад, # ^ |

YES

→ Ответить

5 месяцев назад, # |

I spent whole time figuring why there was RUNTIME in F , test case 3

→ Ответить

Stuck_On_Gray

5 месяцев назад, # |

E<C<D

→ Ответить

sachin22423

5 месяцев назад, # |

this was my first contest im unrated I did solve 1 problem and I'm still unrated will I get rated after this

→ Ответить

RonTheprogrammer

5 месяцев назад, # ^ |

you will

It will take around 12 more hours

→ Ответить

sachin22423

5 месяцев назад, # ^ |

Thanks for replying

→ Ответить

Wasif_Shahzad

5 месяцев назад, # ^ |

you can expert roughly 20 hours or something because system testing usually takes 3-4 hrs or something and it starts after 2-3 hrs after hacking phase too

→ Ответить

sachin22423

5 месяцев назад, # ^ |

Thanks for the reply

→ Ответить

exquisite27

5 месяцев назад, # ^ |

this was my first contest too, I am still unrated, is it just me? or are the ratings not out yet?

→ Ответить

Wasif_Shahzad

5 месяцев назад, # ^ |

the submissions are still being judged. you may expect ratings to come out in 3-4 hrs.

→ Ответить

Newbie-1

5 месяцев назад, # |

Can someone explain why the answer to problem C in test case N5 is 'yes'? As I understand it, modifications should be applied in the same order, so all the modifications will be used on the first number, leaving no more modifications for the rest of the array.

→ Ответить

5 месяцев назад, # ^ |

It can be used on any number. Just last element in $$$D$$$ should be present in $$$B$$$

→ Ответить

Newbie-1

5 месяцев назад, # ^ |

I thought all the previous modifications should be used before applying the last one. But thank you anyway.

→ Ответить

NunoLealF

5 месяцев назад, # ^ |

Yeah, this is what tripped me up too; from my perspective, either the order doesn't matter and test case 4 should also be YES, or order does matter and both test cases should be NO

I spent more time trying to figure out what was actually going on than working on the initial solution, only to not get it right anyways .-.

→ Ответить

jayasting98

5 месяцев назад, # ^ |

You can use the following $$$c$$$.

[3, 3, 5, 1, 1, 4, 2, 1]

If you apply that and $$$d$$$ to $$$a$$$, you should get $$$b$$$. Try it.

→ Ответить

Newbie-1

5 месяцев назад, # ^ |

The announcement said that all operations must be applied to the array in the given order. I thought that C should be 1, 1, 1, 1, 1, 1, 1, 1 to make the first number 4, but that way I couldn't make any other modifications to the array. Anyway, I understand it now and I got accepted after knowing how the modifications work.

→ Ответить

osource

5 месяцев назад, # |

I think E was too easy to be a Div 3 E. Loved solving F1. Thought F2 was on similar lines of F1 but it was actually way more difficult. Overall a nice problem set!

→ Ответить

DuongForeverAlone

5 месяцев назад, # |

If people didn't go for sequential solving, then E would be hugely accepted in this round.

→ Ответить

Stuck_On_Gray

5 месяцев назад, # ^ |

I agree It was clearer than C and easier than D

→ Ответить

smit_384

5 месяцев назад, # |

← Rev. 2 →

are you kidding me , this happend with me again, solved F just after 15 mins contest ends.(same happend in round 949)

Spoiler

→ Ответить

rasesh_shetty

5 месяцев назад, # ^ |

++ same here

→ Ответить

stefdasca

5 месяцев назад, # |

-13

Here is a video Solution for Problem F

→ Ответить

WizzyRocky

5 месяцев назад, # |

-6

Is E that easy as comment suggests i think it's way easier than C. C is just an observation then use a greedy approach

→ Ответить

physics0523

5 месяцев назад, # |

Very cool problems for education, and according to today's DEFG and some recent Div3s, I think the problems are hard enough to make the round rated for Experts.

→ Ответить

5 месяцев назад, # ^ |

Agreed. But had experts knew it would be rated they'd approach round slightly differently

→ Ответить

NunoLealF

5 месяцев назад, # |

I don't have much experience with Codeforces, does anybody know what rating we could expect for problem C?

(I know that it won't be released for a few days, I just want to have an idea; I don't know if it's closer to 1000 or to 1800)

→ Ответить

ShivamJNU

5 месяцев назад, # ^ |

+12

Maybe around 1100.

→ Ответить

Can somebody please tell me why my solution is wrong? Thanks in advance!

5 месяцев назад, # ^ |

-18

I think it would be in range 1100 — 1300. This is fair from my perspective

→ Ответить

Taha_adeel

5 месяцев назад, # ^ |

Not more than 1200 I would say

→ Ответить

GermanGuy

5 месяцев назад, # ^ |

I feel somewhere around 1200, as the problem was a bit heavy in terms of implementation and had 1-2 edge cases to take care of.

→ Ответить

okkookko

5 месяцев назад, # ^ |

You can check on clist

→ Ответить

NunoLealF

5 месяцев назад, # ^ |

Oh bruh, I didn't know it came out there already

→ Ответить

cwz2024

5 месяцев назад, # |

+14

Maybe too noob today, I failed to solve D, and it wasted me too much time and I failed to solve F1 in time:(

Luckily I'm out of competition.

→ Ответить

Astroflexx

5 месяцев назад, # |

← Rev. 2 →

First, I reorder the columns according to the first row, and then check if all rows are same by converting them to string.

→ Ответить

5 месяцев назад, # ^ |

You do not add separators when concatenating numbers into a string, so 1 11 and 11 1 are the same to you.

Example test:

Spoiler

→ Ответить

Astroflexx

5 месяцев назад, # ^ |

Thank you so much man!

→ Ответить

ChaosIsALadder

5 месяцев назад, # |

Really nice problemset! Kudos to the authors and coordinators for creating such an interesting round.

→ Ответить

m_nature

5 месяцев назад, # |

← Rev. 2 →

[Sorry, I was dumb and realized that my solution is actually correct, ignore what I said]

→ Ответить

Zahad_2706

5 месяцев назад, # ^ |

if the condition is already satisfied without removing any element, then i think you can just remove the last element.

→ Ответить

m_nature

5 месяцев назад, # ^ |

Huh, but what if I need to remove more than one element?

→ Ответить

m_nature

5 месяцев назад, # ^ |

Oh wait, I'm dumb lmao. Sorry all. After I remove it, I check the array after removing immediately to see if any of my removal is good, so there is no way that I will get to the point of removing more than one element

→ Ответить

https://codeforces.net/contest/1980/submission/264011923

5 месяцев назад, # |

← Rev. 5 →

Can someone please help me in figuring out what's wrong with my solution to problem C?

I am calculating elements in b array which aren't there in a array. Now I check if the elements in this diff are there in d array too (also ensuring that the count of each element in diff is less than its corresponding count in d array). Finally I am checking that there should be at least one element in b which is equal to d[m-1].

Thanks in advance!

→ Ответить

kaustubhkaty

5 месяцев назад, # ^ |

erase erases all occurrences , you should find the val first and then erase 264052322

→ Ответить

5 месяцев назад, # ^ |

Oh fricc. Thanks a lot man!

→ Ответить

kaustubhkaty

5 месяцев назад, # ^ |

It's fine Tanka

→ Ответить

5 месяцев назад, # ^ |

lmao

→ Ответить

5 месяцев назад, # ^ |

There is also a cool alternative since C++17 called extract which removes only 1 instance of an element from multiset

→ Ответить

https://codeforces.net/contest/1980/submission/264046924

5 месяцев назад, # ^ |

That's neat. Thanks!

→ Ответить

TOMAS_SHELBY

5 месяцев назад, # |

← Rev. 2 →

E < D for me, lost 1.5h on D unsuccessfully.

→ Ответить

chayankumawat2004

5 месяцев назад, # |

YES NO YES NO

→ Ответить

gholyo

5 месяцев назад, # ^ |

MAYBE?

→ Ответить

nxsang1611

5 месяцев назад, # |

nah pray for expect

→ Ответить

giorgi_pkhaladze

5 месяцев назад, # |

can someone help me on D problem? I have wrong answer test N2(testcase 56). I am not sure but maybe my whole idea is incorrect, my idea is to solve this using dinamic programming, starting from the last.

https://codeforces.net/contest/1980/submission/264027336 this is my code

→ Ответить

alpha1215

5 месяцев назад, # |

-14

problem G is nice, too bad i started contest late :(

→ Ответить

Shoeib

5 месяцев назад, # |

unordered hacks again

→ Ответить

__NO__NAME__

5 месяцев назад, # |

+13

i was so happy that i got +85 delta. Now 3rd got hacked of unordered :(

→ Ответить

ashishk.it18

5 месяцев назад, # ^ |

how do u know of delta ?

→ Ответить

__NO__NAME__

5 месяцев назад, # ^ |

+21

carrot extension

→ Ответить

Lakshay_2021059

5 месяцев назад, # ^ |

Same

→ Ответить

krishna_08

5 месяцев назад, # |

-8

C > D > E > B >A

→ Ответить

giorgi_pkhaladze

5 месяцев назад, # |

← Rev. 2 →

I have wrong answer on problem D, and can someone help? I'm tring to solve dinamicaly, maybe my idea is wrong, I don't know for sure. https://codeforces.net/contest/1980/submission/264027336 this is my code. pls help

→ Ответить

rishiraj1905

5 месяцев назад, # |

← Rev. 2 →

→ Ответить

Marvell0us

5 месяцев назад, # ^ |

you can check with i +- 5 use one for loop :) 263959259

→ Ответить

rishiraj1905

5 месяцев назад, # ^ |

is it not sufficient to check only a[i],a[i+1],a[i-1]

→ Ответить

dhruv0902

5 месяцев назад, # ^ |

checking only for i-1 , i , i+1 is sufficient , AC submission : https://codeforces.net/contest/1980/submission/264071477

can someone help me with another submission of same problem , idea is same just a little change in implementation. please find the comment where the doubt is elaborated https://codeforces.net/blog/entry/130032#comment-1155519

→ Ответить

rishiraj1905

5 месяцев назад, # ^ |

can anyone point out where i am going wrong?

→ Ответить

srabons_thinkLab

5 месяцев назад, # |

+13

Solved 5 problems in Div-3 for the 1st time

→ Ответить

is_this_greedy

5 месяцев назад, # |

← Rev. 3 →

G using dfs and trie. Don't know why it is getting tle for n*(log10^9) . Maybe because I used java.

→ Ответить

ttirth1234518

5 месяцев назад, # |

-7

Failed in this contest desperately (-1000 aura)

It was unrated for me (+100 aura)

Still what did I do :(

→ Ответить

eshaansodhi

5 месяцев назад, # ^ |

rated for me but still failed

→ Ответить

ttirth1234518

5 месяцев назад, # ^ |

I did worse :(

→ Ответить

karna001

5 месяцев назад, # |

what was the idea in D

→ Ответить

ttirth1234518

5 месяцев назад, # ^ |

You traverse the array, from 2 to n. The first index where gcd(a[i], a[i+1)<gcd(a[i], a[i-1]), you can either remove i-1, i, or i+1. You do all of it, and check if any of them gives you a favourable outcome.

→ Ответить

Kazake

5 месяцев назад, # ^ |

bruh, it's just brute forcing all possible ways, and I thought gcd calculation would be a heavy operation and there should be a clever way to solve the problem

→ Ответить

Noluck_167

5 месяцев назад, # ^ |

Well there may be a more mathematical side to the problem, but i guess in DIV3 its about logic and implementation ,however its not true everytime.

→ Ответить

oishi_161

5 месяцев назад, # ^ |

can you please give me one example when I have to remove i. I am checking i-1 and i+1 only and after changing my implementation it gives correct answer in test 1.

→ Ответить

shreyk

5 месяцев назад, # |

YES_NOForces!

→ Ответить

shreyk

5 месяцев назад, # |

After a long time, a contest that went well for me :)

→ Ответить

AdityaTakkar

5 месяцев назад, # |

YesNoForces Round.

→ Ответить

Lakshay_2021059

5 месяцев назад, # |

Can someone explain how was i hacked? https://codeforces.net/contest/1980/submission/263943355

→ Ответить

namanbhardwaj1454

5 месяцев назад, # ^ |

https://codeforces.net/blog/entry/62393

→ Ответить

Rakesh_898

5 месяцев назад, # ^ |

Can you tell me why this code gives invalid input during hacking.

spoiler


#include <bits/stdc++.h>
using namespace std;

int main()
{
#ifndef ONLINE_JUDJE
    freopen("ip2.txt", "r", stdin);
    freopen("output.txt", "w", stdout);
#endif
    srand(time(0));
    srand(time(0));
    int n = 100000 - 1;
    int a[n];
    int b[n];
    int c[n];
    for (int i = 0; i < 1e4; ++i)
    {
        c[i] = i + 1;
    }
    int val = 1000010;
    for (int i = 10000; i < n; ++i)
    {
        c[i] = val;
        val++;
    }
    int j = 0;
    for (int i = n - 1; i >= 0; --i)
    {
        b[j] = c[i];
        j++;
    }
    for (int i = 0; i < n; ++i)
    {
        if (b[i] == 1)
        {
            a[i] = b[i] + 1;
        }
        else
        {
            a[i] = b[i] - 1;
        }
    }
    cout << n << endl;
    for (int i = 0; i < n; ++i)
    {
        cout << a[i] << " ";
    }
    cout << endl;
    for (int i = 0; i < n; ++i)
    {
        cout << b[i] << " ";
    }
    cout << endl;
    cout << n << endl;
    for (int i = 0; i < n; ++i)
    {
        cout << c[i] << " ";
    }

    return 0;
}

→ Ответить

5 месяцев назад, # ^ |

Is there any other blog answering same question on python as Using any data structures such as Counter of collections ,sets or dictionaries,all get hacked to show TLE on system testing.

→ Ответить

5 месяцев назад, # |

How to G(I'm not gonna be making hacks but if ure concerned just dm me instead of replying)

→ Ответить

FahimR

5 месяцев назад, # |

← Rev. 3 →

I can able to to solve problems till E after the contest with first submission but I can't do that during the contest time.Moreover I feel very shaky and make silly mistakes and blunder during contests. It this my fate / bad luck only or is there any lack of my skills on it?

It there, anyone please help me to overcome this frustrating conduct, share me some tips and strategies if you used to face these conditions before.

Advance thanks to you. May almighty bless you.

→ Ответить

5 месяцев назад, # ^ |

The solution is actually to take it slow. Even though there is a time limit, obviously, you aren't getting much from being fast if you make mistakes. Whereas if you focus on being slow and steady you make a lot less mistakes while gradually getting faster to the point where you are doing things fast and correct after a lot of practice.

→ Ответить

FahimR

5 месяцев назад, # ^ |

So, recommend me the practice way. Do I practice from problemset section or participate in virtual contests? Which will be better for me?

→ Ответить

nilay_agarwal

5 месяцев назад, # ^ |

You should give virtual contests, since you don't have the fear of making wrong submissions, once you are a little sure you will listen to your gut, that would surely build up your confidence!

→ Ответить

5 месяцев назад, # ^ |

You need to step it down a bit in regular contests and even if it slightly negatively affects your rating, you will learn a lot. Virtual contests are not a good solution because they do not normally cause the same pressure as regular ones

→ Ответить

dhruv0902

5 месяцев назад, # |

← Rev. 2 →

hello guys , would really appreciate if someone helps me through this , in round 950(Div 3) problem D , my approach was to brute force the implementation of how one element should be removed and then the gcd should be checked . The logic seems right because today I tried to submit the same logic with a slight change of implementation and it got AC . But I'm still unable to understand why was first approach failing.

First approach : to check if the currGcd is smaller than the lastGcd if so remove either ith element or i-1 or i-2 from the original array , did this by marking it -1 and then restoring the ith element that was marked as -1 after the recursive call where I check the gcd of the new array. I'm pretty sure that the mistake is in the -1 marking and restoring the element since its the only part that is changed in AC solution, but I'm unable to think of a case.

Link to the approach : https://codeforces.net/contest/1980/submission/263995703

Link to the AC approach with slightly different implementation : https://codeforces.net/contest/1980/submission/264071477

EDIT: Got the mistake it is if the first element has to be market -1 , in that case lst will be 0 according to the first approach , but the gcd of (0,a[1]) would always be a[1] , instead it should be such that I completely ignore the above mentioned gcd since i have removed the element (marked it as -1) so either start the loop in that case from i = 2 of else lst = 1e9 + 7 , prime number grater than any of the possible elements , hence the gcd of them being 1 always which would not hinder gcds

→ Ответить

Astroflexx

5 месяцев назад, # |

Amazing contest! Problems were high quality and fun to solve. Thanks a lot for the round myav Gornak40 senjougaharin Vladosiya and all testers!

→ Ответить

5 месяцев назад, # |

Can anyone please explain why Counter of collections in python(pypy3-64)often leads to Time limit exceeded on some questions even though the logic and time complexity is completely fine.(I am really annoyed by this problem as this has happened to me that questions were accepted initially but on larger test cases later,they were hacked or Faced TlE,and if I submitted them on different language like python3.8.10 or submit the same logic in another data structure ,the get accepted.) Ps:I found that it is not the problem only with Counter,but also sets . like in this question 1927C - Choose the Different Ones! ,[submission:245138579]is not accepted ,whereas [submission:245382629]is accepted as sets used are removed to lists.

→ Ответить

LastTryFr

5 месяцев назад, # |

can someone explain why does O(n) solution in C using python dict get hacked?

→ Ответить

5 месяцев назад, # ^ |

Same issue with me as I posted above,I think that python sucks in data structures as it has happened to me multiple times .I sincerely wanted to switch to C++ for coding but I do not have the time for it as i am proficient to write code only in python .I urge Vladosiya to either please clarify or give our solutions fair judgment.

→ Ответить

LastTryFr

5 месяцев назад, # ^ |

same here i only enjoy coding in python and i think our solutions should be getting a fair judgement same as everyone else's, Vladosiya please look into this.

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

It's actually Python's fault: first of all, dict is not O(n) it's just average complexity of O(n) because it's a hashmap. But as others mentioned it can get as bad as O(n^2) which can be hacked. C++ has regular map (nlogn red-black-tree) and unordered_map (a hashmap) where the latter can be hacked as well... There is no discrimination against python here and I assume that if you implemented your own red-black-tree in python, even mediocre, you'd not get hacked. What happened here is just Python's problem that it doesn't offer a built-in nlogn stable map by default.

Edit: you can use OrderedDict in python I think

→ Ответить

elseecay

5 месяцев назад, # ^ |

← Rev. 3 →

OrderedDict is just list of pairs, not a tree map

UPD: Mistake

class OrderedDict(dict):
    'Dictionary that remembers insertion order'
    # An inherited dict maps keys to values.
    # The inherited dict provides __getitem__, __len__, __contains__, and get.
    # The remaining methods are order-aware.
    # Big-O running times for all methods are the same as regular dictionaries.

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

oh, my bad then, thanks

edit: it's still a hashmap just preserves order so it's not what I thought it to be anyways

→ Ответить

5 месяцев назад, # ^ |

dict uses hashing, and has worst case time complexity of $$$\mathcal{O}(n^2)$$$ when you insert $$$\mathcal{O}(n)$$$ elements. Counter is a subclass of dict so it's essentially the same. The way to make the worst case is well-known and is almost surely used by the hackers during open hack phase.

→ Ответить

5 месяцев назад, # ^ |

can you please suggest any alternative for such cases,so that the alternative should do the same logic but in good time complexity?

→ Ответить

5 месяцев назад, # ^ |

The safest way would be finding a solution that does not use hash-based modules like dict, set, Counter, etc. There are many custom libraries for balanced binary search tree that guarantees $$$\mathcal{O}(log(n))$$$ per operation, or something similar to such structures. If you really need to use hashes, using a custom hash function with randomization would work.

→ Ответить

5 месяцев назад, # ^ |

Can you please tell then why konstantin_rokossovsky's solution to 1974C - Beautiful Triple Pairs i.e. 261980102 is accepted?Should it also not face TLE?or the Test cases for testing were not good enough to reject the solution?

→ Ответить

5 месяцев назад, # ^ |

Just a small bit of change in the keys invalidates hacks on another keys. I think that is also hackable, but since people don't have all the time it's probably just neglected.

→ Ответить

5 месяцев назад, # ^ |

I did not understand,can you please elaborate

→ Ответить

5 месяцев назад, # ^ |

Oh wait, didn't notice that it's from a different problem. In that problem's case the range of $$$a_i$$$ prevents hacks, because for sufficient amount of hash collisions you usually need a wide element range, and in that problem the constraint is only $$$10^6$$$, so it's difficult to produce hacks. I think only Java's HashMap was the only one prone to it.

→ Ответить

5 месяцев назад, # ^ |

But the constriant for this problem was 2 * 10 ^ 5 for both n and m. So isnt it less than 10 ^ 6? Why did the hack work for this problem?

→ Ответить

5 месяцев назад, # ^ |

In this problem the constraint on the element's value was $$$10^9$$$, which is far greater than $$$10^6$$$. You need a wide range for the values to make many collisions when the hash's bucket size is large enough to cause significant slowdown.

→ Ответить

https://codeforces.net/blog/entry/101817

5 месяцев назад, # ^ |

Here is example of unhackadle dict: 247436179

Well, veeeery unlikely to be hacked to be precise

→ Ответить

jzavaglia

5 месяцев назад, # ^ |

This was something I learned from this contest. For integer keys in particular, they are easily hackable. Using an XOR on top of your key like this might make it more resilient as mentioned in the linked article.

Someone else suggests str(key) is also a workable transformation.

import random
RANDOM = random.randrange(2**62)
counter = Counter()

counter[key ^ RANDOM] += 1

→ Ответить

5 месяцев назад, # ^ |

Thanks!was helpful indeed

→ Ответить

5 месяцев назад, # ^ |

sortedcontainers is the typical starting point.

→ Ответить

5 месяцев назад, # ^ |

But there is not any in-built library for sorted containers in python(atleast not for CP,I heard that you have to download it from Github pages for usage on your device.)

→ Ответить

5 месяцев назад, # ^ |

Oh, I thought it was available for import on the CF judge, pretty weird that it isn't!

→ Ответить

Salicylismypop

5 месяцев назад, # |

First time attempting a contest, although I had made this account a long while back. I gave this contest yesterday and solved 2 questions, but I am still unrated. Can anyone please help me understand why? (Clicking on my contests link, no contests are displayed, but when I filter the search to unrated, this contest shows up). Can anyone tell me what I have to do to make at least the next contests that I give become rated, things are a bit elusive here.

→ Ответить

5 месяцев назад, # ^ |

Don't worry, ratings are just not in yet. They always take a few hours and in the case of this contest we had a 12 hour open hacking phase after the contest itself. The ratings are probably going to be updated later today

→ Ответить

Salicylismypop

5 месяцев назад, # ^ |

oh okay thanks. Also right now it shows system testing and all my solutions are in red at the moment, I guess I got them all wrong then :/

→ Ответить

5 месяцев назад, # |

← Rev. 2 →

Can anyone please this solution for C :- 263941647

why it got hacked by TLE?

→ Ответить

5 месяцев назад, # ^ |

← Rev. 4 →

-7

You use Counter in your code. Which depends on the hash to store items.
The problem with hash is that it can be hacked easily, so the time for one get would be O(N) instead of O(1) resulting in time limit for sure.
This is also the case in unordered_map in c++, all of the solutions using unordered_map got hacked.

Tip: Use SortedDict in python instead. It uses red-black tree and the time for get is O(logn) instead of O(1)

UPD: It turned out that SortedDict is forbidden in Codeforces :(

Useful Read

→ Ответить

bevigilantheiscomingback

5 месяцев назад, # |

Lesson Learned... Use a prime number to make your Counter/dict/set in python SAFE... The collision problem is very interesting (and honestly quite dirty), but I guess we all (or most of us) have to learn the hard way... It seems quite impractical in real world use to use a prime number (for example in Python I saw someone use p=109008647)... Which will save you from a hack, but again in the real world how useful is it? Probably not much... Unfortunately that means I don't get credit for 1 of my solutions this time around... (Better to learn now than later I guess).

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

-10

Good suggestion. But maybe simply using an ordered data structure is better :)

In python you can use OrderedDict instead of a normal dict, you will get a slightly worse time of O(log(n)) instead of O(1), but you will be much safer overall.

P.S I am not sure but I think I've read once that this prime trick is also hackable, but I can't remember the methodology.

UPD: OrderedDict are no good also, never the less, SortedDict is the python equivalent to std::map in c++.

→ Ответить

5 месяцев назад, # ^ |

I think as OrderedDict is also a hashmap ,it can be hacked as well,and it is also O(1) (and not O(logn)).I do not find any improvements in using OrderedDict instead of Counter

→ Ответить

5 месяцев назад, # ^ |

You are correct, OrderedDict are no good also.
Just made a quick search, found that the one I meant is SortedDict

This one is indeed sorted and takes log(n) per get/delete.

→ Ответить

5 месяцев назад, # ^ |

Problem with this SortedDict is that it cannot be used on codeforces. You can only use it locally on your system.Hence,it is of no use as a competitive programmer.

→ Ответить

5 месяцев назад, # ^ |

So strange. Any ideas why?

→ Ответить

5 месяцев назад, # ^ |

I don't know precisely ,but it is not any standard library that is itself incorporated into python.Instead it was rather a project whose focus was to develop some data structures in python like sorted sets of c++,which are in C++STL,so that they can use it freely during CP.Feels like python is useless where data strucutres are concerned :(

→ Ответить

silvertides

5 месяцев назад, # |

I wish I can be Specialist

→ Ответить

qxq_xqx

5 месяцев назад, # ^ |

You will be

→ Ответить

silvertides

5 месяцев назад, # ^ |

thanks for your wishes

→ Ответить

qxq_xqx

5 месяцев назад, # ^ |

WOW Congrats!

→ Ответить

x1f0cb32d3_

5 месяцев назад, # ^ |

At this delta in your first rated rounds? You'll probably end up way higher.

→ Ответить

5 месяцев назад, # |

Can someone please explain me the in-depth intuition on how to solve problem C, whats concepts are involved?

→ Ответить

silverstorm

5 месяцев назад, # ^ |

For the array d, we can observe that some elements are useful while others are not. For those elements that are not useful, we wish they could be overwritten. That is, for the current element, if it is useful, then the useless elements before it can be moved here and subsequently overwritten by the useful element. Ultimately, we need to check whether array a has transformed into array b, and at the same time, there are no useless elements left.

→ Ответить

5 месяцев назад, # ^ |

thanks

→ Ответить

ChuanSWTtest

5 месяцев назад, # ^ |

The essence is that you can customize the first index for each pair of the "modification operations". Your job is to Restore the array as much as possible...by customizing wisely. Used Google Translate, my English is not good

→ Ответить

5 месяцев назад, # ^ |

thanks

→ Ответить

5 месяцев назад, # ^ |

can you share your solution? and explain it too, if you don't mind?

→ Ответить

5 месяцев назад, # ^ |

Most importantly we need to transform the original into the target, so for every position where they differ we need one operation with value specified by the target array. If there aren't enough, then it's impossible.

If there are enough operations, we still need to make sure all the other operations don't destroy our progress. We know we can do multiple operations on a single index, which allows us to overwrite unnecessary/bad operations. So we look at the last operation. If the value doesn't exist in target, then we for sure cannot end up with target since it will be written at some index. Otherwise, all the unnecessary/bad operations can be overwritten by the last operation and we're happy.

→ Ответить

5 месяцев назад, # |

Hi, i failed to solve problem C. Can someone help me what am i doing wrong? Its giving tle for hidden tc 7

(Python User) here is my code

→ Ответить

5 месяцев назад, # ^ |

it's correct code. You have TLE. On this task problems with TLE on python

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

i still cant understand why because similar codes got accepted but my code is failing.

like for this submission code is ac and for my code where i also made some changes thinking that it will be ac code is still having tle on tc13. Can you tell why so

→ Ответить

5 месяцев назад, # ^ |

system test is not already end

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

on this problem some people use dict or Counter which the time gets for it O(1) instead or OrderedDict with O(log(n)). Solution with OrderedDict are not hacked

→ Ответить

logankeede

5 месяцев назад, # ^ |

try pypy, that being said even my pypy solution was hacked so idk.

→ Ответить

5 месяцев назад, # ^ |

use OrderedDict instead of Counter

→ Ответить

5 месяцев назад, # ^ |

will OrderedDict make code any better as it maybe having some overhead as compared to Counter and this case imo both the codes i have linked in the above comment are almost identical and i dont see anyway why my code failed there.

→ Ответить

5 месяцев назад, # ^ |

I agree with u, i have the same trouble, but users Counter and simple dict got TLE for this reason((

→ Ответить

5 месяцев назад, # ^ |

when are those tc added which are successfully hacked? Is it during hacking time or after system testing?

→ Ответить

5 месяцев назад, # ^ |

after system testing

→ Ответить

smilence

5 месяцев назад, # ^ |

This is common problem with hash collisions in Python sets/dicts. Because in Python int == hash(int), so it could easily be hacked. Check this out: https://codeforces.net/blog/entry/101817

→ Ответить

5 месяцев назад, # |

some issues on task C with python(TLE)

→ Ответить

deeprajj774

5 месяцев назад, # |

so i don't what happened , i am getting better or my day was good i solved A , B , C around 1 hour and i was like i have done way beyond my limit so i just keep on admiring my accepted code . then now after seeing E today m felling bad that i could have done that too . anyway was fun contest regardless of my silliness .

→ Ответить

logankeede

5 месяцев назад, # |

I think the test case from hacking are not added yet because my solution was hacked yet when i run the same code it is accepted .it shows the contest status as finished and standings as final. Please, fix this so that contestant can review and improve their code easily.

Thanks

→ Ответить

nithinraj04

5 месяцев назад, # ^ |

System testing hasn't started yet.

→ Ответить

logankeede

5 месяцев назад, # ^ |

i see that now, standings were shown as final standings, which seems misleading if system testing is pending.

→ Ответить

akshatchaudhary

5 месяцев назад, # |

+10

Please do something about the cheating Vladosiya, MikeMirzayanov. Its on a massive level, people are getting negatives just because of these, and it really demotivates the cause of such a platform. Please try plagcheckers, or anything which might be suitable.

→ Ответить

5 месяцев назад, # |

why is the problem c hacked a lot? can someone explain?

→ Ответить

5 месяцев назад, # ^ |

Because many people used Counter of collections as data structure for maintaining occurences of elements,but it is in worst case O(n*n).It is pretty easy to find the worst case for it as it is quite standard.I myself got hacked as I ignored that fact.only those solutions passed which used their own hash function.(This was all for python submissions,I do not know about those in other languages).

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

what kind of hash function would work for thus problem then? i am newbie and first time something like this has happened to me. I am still very confused lol.

→ Ответить

invincibleunderthesun

5 месяцев назад, # ^ |

check this blog post out.

→ Ответить

https://codeforces.net/blog/entry/101817

5 месяцев назад, # ^ |

you can also see this for python :-

→ Ответить

5 месяцев назад, # ^ |

thank you so much guys! I really really appreciate the help.

→ Ответить

rythmkumar232

5 месяцев назад, # |

← Rev. 2 →

-13

Why you put a question like E which is directly available on geek for geeks https://www.geeksforgeeks.org/check-if-a-given-matrix-can-be-converted-to-another-given-matrix-by-row-and-column-exchanges/ Here is the link for the solution . If possible please remove this question from the contest Vladosiya pls look into this. As people shared this blog with their friends too and no plagiarism can be applied on it

→ Ответить

cwz2024

5 месяцев назад, # |

What happened to my problem C solution? 263930829 I didn't use unordered_map or that sort of thing, I used multiset, and I think this is a O(nlogn) or O((n+m)logn) solution. Why it got TLE?

→ Ответить

qmk

5 месяцев назад, # ^ |

It's because of multiset.count, related blog: https://codeforces.net/blog/entry/122960

→ Ответить

cwz2024

5 месяцев назад, # ^ |

OMG, luckily I learn this trap in an unrated round, thank you orz

→ Ответить

DaBest

5 месяцев назад, # ^ |

Are there any similar pitfalls?

[I already learnt how unordered_maps/sets and multiset.count give TLE and FST, the hard way.]

→ Ответить

mazenamr07

5 месяцев назад, # |

It's my first contest, when does my rating get updated?

→ Ответить

eklavya_kumar

5 месяцев назад, # |

How to do G? I've looking at others solutions they used trie to find the maximum or, please do explain

→ Ответить

5 месяцев назад, # ^ |

it's a bit advanced for ur level, maybe solve other problems first

→ Ответить

eklavya_kumar

5 месяцев назад, # ^ |

my bad, I misread bitwise 'exclusive' OR as bitwise OR, now I solved it. Try solving for bitwise OR, it's probably much more difficult problem. it's well within in my level bro I have 7 solves

→ Ответить

5 месяцев назад, # ^ |

Bro, with cheating one can full a div3(not saying u cheate), and I also read it as OR, that's why I said it's too advanced, thanks for the pointing at the xor

→ Ответить

edu92337

5 месяцев назад, # |

My submission to problem B got Accepted but now It os showing “in queue “ what happend?

→ Ответить

5 месяцев назад, # ^ |

System Testing is still going on......what for few hours verdict will come!<3

→ Ответить

Xiao_2710

5 месяцев назад, # |

Why is the system-testing incredibly slow for this round?

→ Ответить

5 месяцев назад, # ^ |

most probably because of problem c. they added a lot more test cases for it. more than 100.

→ Ответить

nithinraj04

5 месяцев назад, # ^ |

Yes. 150+ in fact.

→ Ответить

devansh_khandelwal

5 месяцев назад, # ^ |

How do you know that?

→ Ответить

5 месяцев назад, # ^ |

saw it in status where a problem c was getting checked at test case 137.

→ Ответить

5 месяцев назад, # ^ |

I have seen that there are 155 Test Cases by applying filter!

→ Ответить

devansh_khandelwal

5 месяцев назад, # ^ |

Ohh, okay!

→ Ответить

AHMADUL

5 месяцев назад, # |

Could anyone please give a technical brief, why it takes a long time system testing in div-3,4 rounds and few time for div-1,2 rounds?

→ Ответить

5 месяцев назад, # ^ |

-12

I think because of the big difference in the number of participants. Lots of people join div 3, 4 but much less join div 1, 2

→ Ответить

AHMADUL

5 месяцев назад, # ^ |

Bro, it takes a simple query to filter all the participants having rating more than 1599.

→ Ответить

5 месяцев назад, # ^ |

-14

I am talking about the official participants only, others are not taken into considerations.
Have a look at last 2~3 div2 contests and compare them with div3 contest. Around 10K+ participant difference. All of these extra submissions take time :)

One more thing, I think lots of tests are added to problem C, as there were lots and lots of hacks, that could be a reason as well.

→ Ответить

DonSilvio

5 месяцев назад, # |

Does anybody know why Edu/Div.3/4 System Testing takes a lot longer than Div.1/2 System Testing?

→ Ответить

elseecay

5 месяцев назад, # ^ |

+16

1) more participants 2) easier tasks -> more submissions

→ Ответить

gogo2301

5 месяцев назад, # |

isn't system testing taking longer than usual today

→ Ответить

5 месяцев назад, # ^ |

relax bro it will take approximately 2-3 hrs more!

→ Ответить

aditya2227

5 месяцев назад, # ^ |

lagta hai ab kal he final rating pta chalegi

→ Ответить

Prax27

5 месяцев назад, # |

I gave the contest yesterday. It was my first contest. I solved 2 questions A and B and they were accepted also. I am still unrated? Why is that so?

→ Ответить

w_izard

5 месяцев назад, # ^ |

Wait till system testing is completed

→ Ответить

Prax27

5 месяцев назад, # ^ |

I am new here. What do you mean by system testing?

→ Ответить

5 месяцев назад, # ^ |

after the contest, codeforces runs another set of tests to double check the codes, which will take a bit more, wait till tmrw and definitely will be updated

→ Ответить

5 месяцев назад, # |

-8

How to G

→ Ответить

anantpawar782

5 месяцев назад, # |

On my profile this contest show as unrated and i solved 2 questions in contest. Can anyone tell me why it is happening.

→ Ответить

5 месяцев назад, # ^ |

Ratings are not in yet because there were TONS of hacks for problem C (because people were using hashmaps) and if the hack is successful, the test gets added to the test pool. What that means is that it took a lot of time to rejudge all submissions with all those tests which takes a lot longer than usual since you sometimes have 10 tests but here you probably had way over 100 which means that judging problem C took longer than an entire contest usually does. Tests were only finished less than an hour ago so it will probs take a few more to update ratings but you will get your delta :)

→ Ответить

Divyanshiitbhu

5 месяцев назад, # |

I want to know why my rating is increasing by very few points. Like a 800+ rating candidate with rank around 22k in this contest got +72 in his ratings. But my current rating of 722 only went up to 727 , even though my rank in this contest was 21.5k around. This similar kind of issue has happened to me in previous few contests also.

→ Ответить

Karim__2

5 месяцев назад, # ^ |

The higher your rating is, the harder it is to get a higher rating .

→ Ответить

5 месяцев назад, # ^ |

← Rev. 2 →

Each account has both the displayed rating (shown in user profile) and the effective rating (hidden; used for rating calculation), where the former starts much less than the latter and rapidly catches up during the first 6 rounds of the account, so a new account always 'looks like' that it gains a lot of rating at the beginning.

Tl;dr: https://codeforces.net/blog/entry/77890

→ Ответить

PradyunT

5 месяцев назад, # |

← Rev. 4 →

I have some questions about how the hacking system works.

Are successful hacks added to the tests that are ran during system testing? Or are hacks and system tests unrelated?

I'm asking this because my solution for C got TLE after system testing. I learned about implementing a "safe hashmap" after I saw a lot of people get their solutions for C hacked. Then I resubmitted my solution with the new hash function after the contest, which was accepted.

Do Div 1 and Div 2 contests typically not have a hacking phase? Are hacking phases just a part of Div 3 and 4?
Finally, can anyone submit hacks to solutions? Even if the hacker is not participating in the contest or isn't rated/trusted for the contest?

→ Ответить

_Kee

5 месяцев назад, # ^ |

The successful hacks are added to the final set of system tests.
Div 3/4 and educational rounds have 12-hour open hacking phase after the coding phase. In div 1/2 rounds, you can hack the solutions of other contestants in your room during the contest time, i.e. coding and hacking phases aren't separated. To hack a solution, you first need to solve that problem (passing pretests), and lock the problem (you'll no longer be able to resubmit for the problem). In div 1/2 rounds, you will gain (or lose) points for each successful (or unsuccessful) hack.
For contests with open hacking phase, anybody can submit hacks (I think) during the hacking phase. For div1/2 rounds, only contestants can submit hacks. Additionally, candidate masters and above can submit hacks any time after the contest, but those tests may not be integrated to the final system tests.

→ Ответить

ShikharS

5 месяцев назад, # |

Submission

Can anyone tell why this solution got TLE?

→ Ответить

yadavDipesh

5 месяцев назад, # ^ |

you used unordered map

→ Ответить

ShikharS

5 месяцев назад, # ^ |

Thanks, got it.

This blog might also help.

→ Ответить

Arjit_29

5 месяцев назад, # |

Can anyone plz tell that in this contest i submitted my solution of problem B that is choosing cubes and it got accepted. Now after the contest it is showing wrong answer on test2 but it got accepted in the contest itself. Not only this when i now submitted exactly same solution it got accepted on test 2 als0. How is it possible?

→ Ответить

5 месяцев назад, # ^ |

Your code has an undefined behavior. In if(a[k] == num), k can go out of bounds of a and therefore anything unexpected can happen.

→ Ответить

Arjit_29

5 месяцев назад, # ^ |

Ok thanks for replying but now it is accepted so is it that in contest it detected but now it not.

→ Ответить