We will hold AtCoder Regular Contest 134.

- Contest URL: https://atcoder.jp/contests/arc134
- Start Time: http://www.timeanddate.com/worldclock/fixedtime.html?iso=20220129T2100&p1=248
- Duration: 120 minutes
- Number of Tasks: 6
- Writer: camypaper
- Tester: maspy, satashun
- Rated range: — 2799

The point values will be 300-400-500-600-900-1000.

We are looking forward to your participation!

I don't know whether I'm happy there are no stupid meme comments under atcoder announcements, or upset that the contests aren't getting too much attention. Either way, I'm looking forward to being proven yet another time that I know nothing about problemsolving

I only received a reminder email about the upcoming ABC 237, but no reminder email about the upcoming ARC 134, which will happen earlier. I expect to receive the reminder for the contest that will be held earlier first.

Will AtCoder ever consider looking into making the times PST-friendly or varied?

P.S. Not demanding anything, just curious.

Starts in 5 minutes.

GG

When can we expect to see Cpp20 support for atcoder ?

How does this not mean "For all box, the number of ball $$$1$$$ is greater than the number of ball $$$i$$$, for all index $$$i \ge 2$$$"?

I wasted like 1 hour solving the wrong version :(

Really? Problem C means "The number of ball 1 should be greater than the sum of ball 2 to ball N" ?

Yeah I was able to solve after asking for the clarification.

read it the same way for like 10 mins... doesnt help that the small samples were also consistent with this wrong version

I was about to ask for a clarification but then decided just to go with the word "majority" and disregard the "in other words" part.

I think they should have added "combined" to the end, as in "...than that of the other balls combined".

I did so,then I wasted 1 h.(

I was also confused by it.

Luckily, after thinking about the problem for some time, I thought if it was "For all box, the number of ball $$$1$$$ is greater than the number of ball $$$i$$$, for all index $$$i \ge 2$$$", it would be too difficult, so that I decided to go with the word "majority".

And after a while, I saw the clarifications came out.

Us Moment

The problem E... I can't imagine the solution has an exponential complexity until i write it out.

Regarding task A). When a tarp covers the range [x, x+w], both ends included, then the tarp has length w+1, right? Am I stupid? Doesn't that make the answer ceil(l_i/(w+1)) for each segment?

"x is real"

You are thinking in terms of number of integers. If you look at them on a number line, the distance between them is W.

It is written that it is real numbers, so the length is not w+1, but simply w.

Edit: For me the statement of A reads like the setter tried to make it more interesting by not giving a clear statement, but a overformal mathematical sounding one. And actually that style of statement repeats, in example also in C.

Not gonna take another contest by this author :D

I see, I guess reading is not my strength

I would second what _lowerBound said , can anyone elaborate how being defined as "real" makes the length W and not W+1? I too was stuck on this in the contest.

so when you have w = 3 and a tarp starts at 0, then it covers [0.00000, 3.00000], however, 3.00001 is not included. That means to, cover a range from 0 to 7, you need three tarps, one at 0.0000, one at 3.00000, one at 6.0000. (in contrast, if we talk about integers: 0-3, 4-7 would only require 2)

For me, this task was rather reading comprehensions ... :(

I registered 5 minutes before the contest so that I forgot to check the author today.

As a result, I found that the problems are not my style and I got negative delta.

I have no idea about "(4,8) is a corner case".

How to solve problem C? I am not smart enough to understand the editorial.

First, you need to know about https://en.wikipedia.org/wiki/Composition_(combinatorics). Then, for every other-colored ball, pretend that you always stick a ball colored 1 to it. Then, put a ball colored 1 in all of the boxes. Now the requirement that 1 is the majority can be ignored, and the ways to put the balls in the boxes can be solved independently for each color. The problem then becomes, to find the number of weak compositions of the number of remaining balls of each color.

Got your idea. Thanks~

I'm having slight confusion with the formula posted in editorial. I understood the editorial .

if i want to distribute $$$a_2$$$ identical balls into K blocks without any constraint then isn't the formula [](https://en.wikipedia.org/wiki/Stars_and_bars_(combinatorics)) $$$ \binom{a_2 + K -1}{K-1} $$$

why is this value $$$\binom{a_2 + K +1}{K+1} $$$ in the editorial ?

The English translation has a typo. In the Japanese editorial, the formula is correct.

camypaper do correct it when you get time

Problem A I wonder where the mistakes are QwQ

`

you need to set

`int`

to`long long`

and also you have to cast

`w`

to long double while divisionhere is the new AC code

Thank you!But actually,I have already set int into long long (line2) QwQ

Oh, I didn't notice the

`#define int long long`

, it is somewhat weird. But Then, the problem is the`long double`

, I got a WA because of it too.The clarification did not popup? I saw it after the contest ended.

So do I.

https://atcoder.jp/contests/arc134/submissions/28877523 plz see this wa on 3 test case in q2 i tried alot of tc but cant come up with AC solution if anyone can figure out tc or whats wrong plz tell

I tried but couldn't find a counter-example. Looks like you have to run a script which keeps looping and trying random strings as inputs to your solution vs a passing solution.

Sorry, for wrong post.

Can anyone identify what i am missing? WA*4, AC*91

~~~~~

Regarding the last sentence of editorial for F,

"explicit" formulacoefficient of $$$x^N$$$ in $$$\left((1-\frac{1}{W^2})e^{-Wx}+\frac{1-Wx}{W^2}\right)^{-1}$$$