I created a video editorial talking about how to use Atcoder Library's Segment Tree to optimize CF2019D: Speedbreaker
# | User | Rating |
---|---|---|
1 | tourist | 4009 |
2 | jiangly | 3839 |
3 | Radewoosh | 3646 |
4 | jqdai0815 | 3620 |
4 | Benq | 3620 |
6 | orzdevinwang | 3612 |
7 | Geothermal | 3569 |
8 | ecnerwala | 3494 |
9 | Um_nik | 3396 |
10 | gamegame | 3386 |
# | User | Contrib. |
---|---|---|
1 | maomao90 | 164 |
2 | Um_nik | 163 |
3 | atcoder_official | 158 |
3 | cry | 158 |
3 | -is-this-fft- | 158 |
6 | awoo | 156 |
7 | adamant | 154 |
7 | nor | 154 |
9 | TheScrasse | 153 |
10 | Dominater069 | 152 |
I created a video editorial talking about how to use Atcoder Library's Segment Tree to optimize CF2019D: Speedbreaker
Name |
---|
Very hard problems
I dont get it, D was extremely simple. You just have to track the first and last occursnce of each character, and then keep intersecting the segments.
A 20 line implementation would reach $$$O(n)$$$ without segment trees on this problem.
Yes, I agree Segment Tree might be an overkill for this problem. However, I arrived at the segtree solution without a lot of efforts, hence wanted to share an alternate solution from the editorial.
Also, looking at submission count of Problem C (6529) vs Problem D (1373), a lot of people did not submit the overkill segtree solution, possibly because they were not aware of such data structures or thought process of reducing the problem to a standard domain, hence the video.
Yes it's always useful to learn different ways to solve a problem.
Can you please explain your code and thought process of this problem ?