I couldn't find a discussion blog for this contest so here's one.
Does anyone know how to solve Ex?
№ | Пользователь | Рейтинг |
---|---|---|
1 | tourist | 4009 |
2 | jiangly | 3821 |
3 | Benq | 3736 |
4 | Radewoosh | 3631 |
5 | jqdai0815 | 3620 |
6 | orzdevinwang | 3529 |
7 | ecnerwala | 3446 |
8 | Um_nik | 3396 |
9 | ksun48 | 3388 |
10 | gamegame | 3386 |
Страны | Города | Организации | Всё → |
№ | Пользователь | Вклад |
---|---|---|
1 | cry | 164 |
1 | maomao90 | 164 |
3 | Um_nik | 163 |
4 | atcoder_official | 161 |
5 | -is-this-fft- | 158 |
6 | awoo | 157 |
7 | adamant | 156 |
8 | TheScrasse | 154 |
8 | nor | 154 |
10 | Dominater069 | 153 |
I couldn't find a discussion blog for this contest so here's one.
Does anyone know how to solve Ex?
Название |
---|
For problem D, my solution with Binary Search + BFS works fine except for 01_random_03.txt. Any hint ?
Did you binary search till 4e9 becauese I was facing same verdict till i realized cost could be 4e9.
My solution also fails on the same test case. Is there anyone who could look into my code.
..
I have used #define int long long
Overflow. P can be 10^9 10^11 * 10^9 = 10^20 which is too big even for long long.
explain your solution
rude
With 4e9 it works fine. thanks.
Submission In problem G why picking the longest transition (by prefix function) works?
Greedily removing the longest suffix of T that is a prefix of S seems to work as well.
Yeah did so after the contest and it passed but I am unable to prove it why does it work.
Because ans[i]<=ans[j] for i<=j because you can always construct a solution for index i from the constructive words of solution j by either deleting some characters of the last word (giving the same ans) or deleting some words then some characters (giving smaller ans).
D was a nice variation of Floyd Warshall algorithm
BFS is a nice variation of Floyd Warshall?
You can also solve it using a slight modification of floyd warshall by defining
$$$ w[i][j] = \frac{|x_i-x_j|+|y_i-y_j|}{p[i]}$$$
and changing + operation with max
code
Interesting solution
I used this idea as well, but I think that binary search solution is also very nice.
I have failed on F because of a little mistake in my BFS :)
I got WA in one test case in F smh.
Problem C? Can anyone mention the approach? Thanks!
We can multiply all w[i]=w[i]*2, then check all w[i]+1 and w[i]-1 to be the possible best answer.
Each check can be done in O(logn) by sorting childs and adulds separate, then binary search the number of correct guessed ones for fixed x.
Edit: Or use prefix sums. Atfer sorting w[], we can split w[] at any position, then the correct guessed number is the number of childs left of the split position plus the number of adults right of the split position.
Keep track of
score0
andscore1
. Use map to iterate the weights and move the border line of guessing 0 vs guessing 1 from left to right. I submitted a simplified version of the code at: https://atcoder.jp/contests/abc257/submissions/32757902I was unlucky to only solved it one minute after the contest ended.
check for all w[i], then also check for smallest weight-1 and largest weight+1(Solution)
just sort the children array and parent array and apply binary to get the ans. also check if we have only children or only adults
How to solve E?
can any one help me with D , I think I miss understood the question . They are saying that : "Takahashi's objective is to become able to choose a starting trampoline such that he can reach any trampoline from the chosen one with some jumps."
some jumps means we can assume that he will be able to jump for each consecutive pair right ?
No. Read sample test case explanation
Hi Everyone, Can anyone help to find the mistake in my code of F? Link to Solution
Consider the case
3 2
0 1
0 3
Just add a line ans = min(ans, st + en + 2); and it should AC.
Thanks!!!
For problem F, my solution is giving runtime error on 27 hidden cases. Why? https://atcoder.jp/contests/abc257/submissions/32791294
UPD: I have done the same thing as editorial except DFS instead of BFS, which I don't think should make any difference
It makes all the difference, because the graph isn't necessarily a tree/forest, so there may be more than one simple path between nodes of different lengths, and DFS doesn't necessarily find the shortest of them