I think the tests are weak too, I literally just simulated the $$$k$$$ constraints $$$2 \cdot n$$$ times and checked if the upper bound at least the lower bound, and it passes all test cases. If this is actually correct, I don't have a proof for this, it's just random intuition lmao. Otherwise, the tests are just bad.

omeganot will play hsr!

This is a common mistake in dp, where 0 + (a previous answer) might be used to update your current dp state but that 0 is not supposed to be possible. If it is set to negative infinity, its effects will be neglected.

-0x3f is negative infinity. When you fill the array with 0s, you might add states that are impossible to reach before.

this is cool, especially the test case generation but some linting would be nice

based

PLEASE reschedule it on another weekend :prayge:

Whether you're driving to work

sir it's going to be so thought provoking i'm going to end up crashing

3 months late but orz grindset :o

slayyyyy

:o slay

i agree with your points (except the last statement) though idk if this will do anything. It’s not too hard to just submit a compile error when a contest happens and move on.

Most trolls i’ve seen are even newbie, and it doesn’t seem like they’re too ashamed of their own rating lol.

Has someone tried translating using chatGPT before? I sure do hope it's not as cursed as google translate...

why downvotes? this is cool :o

This is so epic!!

Please tell me there are going to be problems about sausages again :DD

:scare:

hacked uwu

lol someone got too pissed and decided to ddos usaco :/

rip to anyone who has their contest window running rn

I've always had this thought, but what about administering the USACO like the AMC? Schools send proctors to watch the students, like an actual exam in a set time period. Or, like the SAT, where students can go to designated testing sites all over the country?

road to a positive score on plat #1

as a tester, the problems are

btw contest difficulty is around div2 C-D in codeforces terms

never had a coach, never paid for any cp services, and never will

every time i skip a round i end up vcing and doing better than i ever could've imagined :(

though i was in fact, 208 points off of 1016 :(

i learned how to modify segment tree when i was a hardstuck pupil and couldn’t solve div 2 Cs. Checkmate

plat pleaseeeeeeeeeeeee it has been way too long :sob:

would kind of sucky if 16 points away from cutoff again :((((

wow tysm!! I had a slight feeling that only the factors of $$$k$$$ are important but ultimately couldn't prove it in contest :(

May someone explain how to get full credit on Bamboni? Getting more than half credit with $$$n^2k$$$ dp was pretty trivial... Is there some number theory optimization I missed?

I agree — It's mostly about how much you love and how much time you dedicate to problem solving and each activity that makes the difference. Excelling in one does not give you free magic to ace the other :)

youre reading test cases when it's just one test

Not too big of a fan of them, mostly because I feel like if i seen them once i won’t really see them again, especially the “given a binary string let me invent two random operations you can do”. It also comes down too how fast you get the one observation in most cases, which can also be sometimes rng and trial and error for lower rated people while tbh for most of the top rated guys it’s just intuition.

omg this is what i call quality service!!

um tbh i wouldn’t count it on too hard

I thought this contest was easier than last open quite a bit (in terms of getting over 800 points) and last open had a 800 cutoff..

yeah i agree

i always thought of pretests as a system to minimize the load on the judges instead of running the billion systests all at once

At the bottom of the contest page: “Please email the contest director ([email protected]) if you come across any technical problems.”

man really admitted to breaking the rules :sob:

you always want to guarantee youre able to reach the highest value while differing by at most m.

let a[i] be max(s[i], a[i+1] — m) for i from n to 1. This guarantees the least amount you have to assign to reach the peaks that come after.

let ans[i] = max(ans[i-1] — m, a[i]) for i from 1 to n. This represents taking the max of the least amount you have to assign and if you were to subtract m from previous.

what no not exactly but they give you like a 20 day gap between dec 20th to like jan 10th

^^^

no way days 5-10 is 2000 rated worthy and 10-15 is div 1. Most problems are essentially just brute force but the harder part is understanding the statement, so i don’t know if you’re incorporating the language barrier factor as well :v

it’s a combination of readforces and implementationforces based on past years

as a NA east coaster, me when i sacrifice sleep to do aoc and fall asleep in class the next day >_<

lol why candidate master specifically and not anything lower or higher? lgm is pretty cool too!

no cap kinda neato burrito

ummm for that problem since $$$n,k \leq 50$$$ you can literally brute force it (create a separate string and keep adding occurrences of a prefix of $$$t$$$ and see if $$$t$$$ eventually becomes a prefix of that string). This runs in $$$n^2$$$ which is more than enough lol

But if you're a gigachad and want to solve a harder version of this problem on CSES you can try using string hashing (my editorial, definitely not a shameless plug) or use the kmp thing in the editorial (why is that even there in the editorial for a div 3 A)

vector resizing is cozy as well :v

mmm an even more similar problem lol https://mycode.prepbytes.com/problems/sorting/ELMT

Will it be available as a CF gym?

i’m cm and i still cant work out nvim for the life of me

instead of posting it here you should hire a guidance counselor for cp

Problem submission syncing (i.e. if you solve a problem in a mashup or a div 2, your submission will show up in the div 1 version of the problem as well).

how about you post this on your main account mr most likely newbie

Amazing effort! One of the best and most educational video on a cp problem i've seen.

Great idea, a downside would be more annoying to hack. Assuming the headers would be forceably public, sometimes you'd have to go all the way to the person's personal directory to find out what a macro does or search for anything vulnerable in their ds implementations, rather than it being right there at the top of their code.

and 99.7% of those people haven't played the level for themselves.

They have the right to ego inflate all they want, it's not your business lmao

As a problemsetter, I'm a setter of the problems

theres always this one guy scrolling for unordered maps lol

it isn't something new and apparently 400+ people still fell for it

Not necessarily, the problems are similar and you get a lot more time. Qualifying for gold should be kind of trivial for CMs though

i can 100% relate to this when testing the round D:

Sometimes thinking too hard can screw you over :/

as a tester, the problems are pretty neato burrito!!

I think I kinda cheesed C lol (weak tests?) I used a DSU with a simple set. A element can only start its own connected component if it is the greatest in the set so far (if you sweep from left to right). Therefore any element smaller than that will auto join its connected component. In the set, I only kept the biggest values of each connected component and for each element, binary search and manually join the components with max values bigger than that element.

Here is my code that passes in nearly 800ms lol https://codeforces.me/contest/1638/submission/146576523

Edit: there is a smart optimization for this but its still kinda cool my brute force passed

Edit 2: i appreciate the hack :D

hilarious DSU solution i wrote for D. Diverse Garland because i was bored https://codeforces.me/contest/1108/submission/138819168

Because binary search is monotonic.

cout<<fixed<<ans; I also made this mistake when i first started :)

https://codeforces.me/contest/771/submission/137069507

Nevermind about the last edit, I read your code wrong. You can check my code if you'd like, using an adjacency matrix was way more practical. https://codeforces.me/contest/954/submission/135211021

when i was 10 i was playing roblocks not Codeblocks :pensive:

haha this is helpful since i can solve problems of any rating without knowing.. i wont be scared when i see a high rating. thanks!

imagine having negative reputation... guess im a criminal now

im getting "Validator 'testvalidator.exe' returns exit code 3 [FAIL Expected EOF (stdin, line 1)]" Why is that?