What?

Perhaps we are not in the same time zone.

wtf

abcde for me ,B is so easy,why you failed on it??OHHH look！！！！！！！ a big error!!!!,this code is absurd but AAAAACCCCCCCC!!!It AC at all random but Wa at simple.so I add some code to AC the simple .ItACed!!——at21:43

Congratulations!>w<

You are so strong!

Suppose we have x < y (initially x = 0, and we assign p1 to y). Iterating i from 2 to n: if pi > y, we must perform operation 2 to keep x as minimal as possible; otherwise, we just move to i + 1. Finally, we use the remaining elements only for x and use DP LIS.

You didn't need to use binary search or string hashing. For n <= 10000, doing a linear scan works fine

$$$N \leqslant 10^4$$$, so you can replace two binary searches with just linear search of the first and second indices in your approach.

I use DP. I set an $$$10000 \times 10000$$$ array,but I haven't got a MLE ($$$340$$$ MB) :)

And no TLE ($$$10^8$$$ in the max size of $$$N$$$).

Use Manacher to solve it.

https://atcoder.jp/contests/abc468/submissions/77868088

About 1 hour.

Maybe you can use C++

Let pos[i] be the position of i, then with s[k] = 'o' we have prefix_max[k]-prefix_min[k]+1 = k, otherwise prefix_max[k]-prefix_min[k]+1 > k. Let dp[k] be the number of permutaions satisfying the first k conditions, but s[k] = 'o', Find the nearest j < i and s[j] = 'o', then subtract the contribution of dp[k] from j+1 to i-1, k is the first position that s[i] must be 'o', consider first k elements as one block and shuffle it with i — k remaing position.See my code

If you solve for the case $$$\text{o}\text{x}^k\text{o}$$$ which I call $$$g_k$$$ then you can multiply $$$g_k$$$ for each gap between o's of length $$$k$$$ in the final string since you can treat each sub-segment as an atomic unit when picking surrounding positions. To calculate $$$g_k$$$ remove all possible prefixes $$$\text{o}\text{x}^i\text{o}$$$ for $$$i \lt k$$$

I am not involved in the banning process, so I do not know anything about it.

Also, appeals regarding bans should be submitted through AtCoder’s contact form, not on the Codeforces Blog: https://atcoder.jp/contact

Think of finding the sum of all subarrays of a particular size, you may need some precomputations for that. Then the task is easy because we can just add that sum divided by that size i for instance with some modular mathematics in hand....

can you please share your submission? and the similar codeforces problems that you mentioned if you have them? Thanks in advance :)

AtCoder allows users to post editorials, but there is a minimum rating requirement. If I remember correctly, you need a rating of at least 2000 to write an editorial for an ABC.

You are also free to post an editorial somewhere other than the AtCoder contest page, including on this Codeforces blog.

Sorry for my mistake. I didn't see it. It seems like we have to do something every time