We will hold SuntoryProgrammingContest2024（AtCoder Beginner Contest 357）.

- Contest URL: https://atcoder.jp/contests/abc357
- Start Time: http://www.timeanddate.com/worldclock/fixedtime.html?iso=20240608T2100&p1=248
- Duration: 100 minutes
- Writer: Nyaan, math957963, physics0523, MtSaka
- Tester: toam, Aus21
- Rated range: ~ 1999
- The point values: 100-200-250-350-450-550-650

We are looking forward to your participation!

Excited and Hoping for the best ABC Round!

Auto comment: topic has been updated by atcoder_official (previous revision, new revision, compare).only 5 contributions away from the top-contributors.

wu~hu~qi~fei~

glhf

Problem E is similar to this!

E was exactly same as this. Probably that explains so many submissions on E.

what is hand_00.txt in D? only failed that test :/

\begin{equation} \sum_{i=0}^{len(s) — 1}(\frac{10^{n * len(n) + i} — 10^{i}}{10^i — 1}) * d[i] \end{equation} where $$$len(n)$$$ is no of digits in $$$n$$$ and $$$d[i]$$$ is the $$$i$$$ th digit.

need to be careful when calculate the numerator it might overflow: AC

n*len(n) can overflow. Maybe you missed that ?

thanks for trying to help guys. I realized I was using the built-in pow to calculate pow(10, len(n)) instead of the pow in my mod template :(

I changed

`pow(10, num_digits(n))`

to`pow((mlong)10, num_digits(n))`

and got AC nowWA AC

How to change my name?

I spent over 40 mins implementing E only to find out here that it was available on Leetcode :/

I used Condensation Graph for E which is a bit overkill but it worked.

Why there are so many D&C NTT problems in ABC, but almost no problems about suffix array or suffix automaton ??? I think they are educational too :)

how to solve F?

I personally didn't solve it, but my idea was correct. We maintain 3 arrays: $$${a_i}, {b_i}, {v_i}$$$. The first two should be self-explanatory and $$$v_i=a_ib_i$$$. To maintain these arrays, we can use many RURQ data structures (I used Sqrt Decomposition). Maintaining $$${a_i}$$$ and $$${b_i}$$$ should be simple as it is just a classic RURQ problem. To maintain $$${v_i}$$$, we notice that when $$${a_l, a_{l+1}, \dots, a_r}$$$ are each incremented by $$$x$$$, then $$${v_l+v_{l+1}+\dots+v_r}$$$ is incremented by $$$x\times\left(b_l+b_{l+1}+\dots+b_r\right)$$$. This allows us to maintain array $$$v$$$ lazily. You can see the editorials for clearer explanations. Also, if anyone finds the problem in my implementation, I would appreciate it.

Some segment tree magic. I misread statement and solved for $$$\sum_i \sum_j A_iB_j $$$

Segment Tree

Why submission giving WA on test_03.txt :(

you must print # for N = 0 not .

Joined 1 hour late (on accident), but still solved ABD.

WA — https://atcoder.jp/contests/abc357/submissions/54368958 AC — https://atcoder.jp/contests/abc357/submissions/54381026

Can somebody explain why my first submission was not accepted whereas the other one was accepted and the only difference both have is that in the not accepted one, I have define k = 1+(int)log10(n) whereas, in the accepted one I have define k = to_string(n).size()

This may help you.

Reduced problem D to Geometric Progression, and then utilised the sum of first N terms formula. We can split the original number multiply it later and make a GP serious out of multipliers of powers of 10. a = 10^(N*d — d), r = 1/(10^d). Final ans = N * a(1-r^n)/(1-r)

Here d = no. of digits in the original given number. Have taken care of fermat's theorem too by applying (mod-1) for huge numbers in exponents. Why does it fail in 15 and passes 13 test cases. What is wrong in the solution ? https://atcoder.jp/contests/abc357/submissions/54378604

Let $$$M$$$ be the length of the number $$$N$$$.

To calculate this, we have to take

Code

I think there is some issue, with mod being used. I used your approach too with r = 10^d instead of 1/(10^d) which I previously used. Yet it still gives same results(15 fail and 13 pass). for some reason

https://atcoder.jp/contests/abc357/submissions/54385841

Not sure what is the exact issue

Wait that is literally the exact same thing that happened to me

My first D submission had 15WA and 13AC, so I might have had the same error as you

Later in the contest, I found out that, in my method, I was creating an ll overflow by multiplying the original number n by another number. I fixed this by making the original number n=n%998244353. This might work for you, too.

Here's my WA solution:

https://atcoder.jp/contests/abc357/submissions/54371464

And here's my AC solution:

https://atcoder.jp/contests/abc357/submissions/54371922

Thanks! That totally skipped my mind at the time

Anyone please explain me the

logic or Intution of Problem D.Prerequisite: Binary Exponentiation and Modular arithmetic

Could You please tell me why this fails — https://atcoder.jp/contests/abc357/submissions/54764020

https://mirror.codeforces.com/blog/entry/130226?#comment-1157408

Only solved first two, sad, suggest some topic.

Can anyone help me with problem F? I used lazy segment tree. code

Anyone help please

@atcoder_official try to add more general editorial instead of any snippet, sometimes its get hard to understand .