# | User | Rating |
---|---|---|
1 | tourist | 4009 |
2 | jiangly | 3823 |
3 | Benq | 3738 |
4 | Radewoosh | 3633 |
5 | jqdai0815 | 3620 |
6 | orzdevinwang | 3529 |
7 | ecnerwala | 3446 |
8 | Um_nik | 3396 |
9 | ksun48 | 3390 |
10 | gamegame | 3386 |
# | User | Contrib. |
---|---|---|
1 | cry | 166 |
2 | maomao90 | 163 |
2 | Um_nik | 163 |
4 | atcoder_official | 161 |
5 | adamant | 160 |
6 | -is-this-fft- | 158 |
7 | awoo | 157 |
8 | TheScrasse | 154 |
9 | nor | 153 |
9 | Dominater069 | 153 |
Name |
---|
I am just able to come up with O(n^2 log n). First we will sort the array . We are finding all the pair possible for (i,j) ---> (i<j) such that (A[i] + A[j])<= thresold ,now let the remaining rem = thresold — (A[i] + A[j] ) now if (rem > A[j]), then we will find the larges (value<= rem) in array of remaining elements using binary search and no of valid triplets is (k — j), where 'k' is index of that value.
We will find this for all the valid pairs.
So O(n^2) for valid pairs and for each pair finding the A[k] is (logn). So O(n^2 log n)
You can achieve $$$O(n^2)$$$ if you replace binary search with two pointers.
There are no known significantly better upper bounds and the lower bound is widely known open problem. More details here.
Just curious, does Citadel hire from India?