結果
問題 | No.2161 Black Market |
ユーザー | koba-e964 |
提出日時 | 2023-07-05 02:13:47 |
言語 | Rust (1.77.0 + proconio) |
結果 |
AC
|
実行時間 | 423 ms / 7,000 ms |
コード長 | 5,283 bytes |
コンパイル時間 | 13,893 ms |
コンパイル使用メモリ | 379,744 KB |
実行使用メモリ | 13,492 KB |
最終ジャッジ日時 | 2024-07-18 20:36:25 |
合計ジャッジ時間 | 15,997 ms |
ジャッジサーバーID (参考情報) |
judge4 / judge1 |
(要ログイン)
テストケース
テストケース表示入力 | 結果 | 実行時間 実行使用メモリ |
---|---|---|
testcase_00 | AC | 1 ms
5,248 KB |
testcase_01 | AC | 1 ms
5,376 KB |
testcase_02 | AC | 2 ms
5,376 KB |
testcase_03 | AC | 0 ms
5,376 KB |
testcase_04 | AC | 0 ms
5,376 KB |
testcase_05 | AC | 1 ms
5,376 KB |
testcase_06 | AC | 1 ms
5,376 KB |
testcase_07 | AC | 1 ms
5,376 KB |
testcase_08 | AC | 1 ms
5,376 KB |
testcase_09 | AC | 1 ms
5,376 KB |
testcase_10 | AC | 1 ms
5,376 KB |
testcase_11 | AC | 1 ms
5,376 KB |
testcase_12 | AC | 1 ms
5,376 KB |
testcase_13 | AC | 0 ms
5,376 KB |
testcase_14 | AC | 1 ms
5,376 KB |
testcase_15 | AC | 1 ms
5,376 KB |
testcase_16 | AC | 1 ms
5,376 KB |
testcase_17 | AC | 1 ms
5,376 KB |
testcase_18 | AC | 1 ms
5,376 KB |
testcase_19 | AC | 1 ms
5,376 KB |
testcase_20 | AC | 85 ms
11,924 KB |
testcase_21 | AC | 84 ms
11,924 KB |
testcase_22 | AC | 316 ms
11,540 KB |
testcase_23 | AC | 309 ms
12,176 KB |
testcase_24 | AC | 423 ms
13,460 KB |
testcase_25 | AC | 101 ms
9,772 KB |
testcase_26 | AC | 374 ms
13,492 KB |
testcase_27 | AC | 25 ms
6,592 KB |
testcase_28 | AC | 28 ms
6,592 KB |
testcase_29 | AC | 11 ms
5,376 KB |
testcase_30 | AC | 10 ms
5,376 KB |
testcase_31 | AC | 151 ms
9,332 KB |
testcase_32 | AC | 22 ms
6,336 KB |
testcase_33 | AC | 21 ms
5,376 KB |
testcase_34 | AC | 7 ms
5,376 KB |
testcase_35 | AC | 7 ms
5,376 KB |
testcase_36 | AC | 24 ms
5,376 KB |
testcase_37 | AC | 10 ms
5,376 KB |
testcase_38 | AC | 84 ms
5,684 KB |
testcase_39 | AC | 6 ms
5,376 KB |
ソースコード
// https://qiita.com/tanakh/items/0ba42c7ca36cd29d0ac8 macro_rules! input { ($($r:tt)*) => { let stdin = std::io::stdin(); let mut bytes = std::io::Read::bytes(std::io::BufReader::new(stdin.lock())); let mut next = move || -> String{ bytes.by_ref().map(|r|r.unwrap() as char) .skip_while(|c|c.is_whitespace()) .take_while(|c|!c.is_whitespace()) .collect() }; input_inner!{next, $($r)*} }; } macro_rules! input_inner { ($next:expr) => {}; ($next:expr,) => {}; ($next:expr, $var:ident : $t:tt $($r:tt)*) => { let $var = read_value!($next, $t); input_inner!{$next $($r)*} }; } macro_rules! read_value { ($next:expr, ( $($t:tt),* )) => { ($(read_value!($next, $t)),*) }; ($next:expr, [ $t:tt ; $len:expr ]) => { (0..$len).map(|_| read_value!($next, $t)).collect::<Vec<_>>() }; ($next:expr, $t:ty) => ($next().parse::<$t>().expect("Parse error")); } // Segment Tree. This data structure is useful for fast folding on intervals of an array // whose elements are elements of monoid I. Note that constructing this tree requires the identity // element of I and the operation of I. // Verified by: yukicoder No. 2220 (https://yukicoder.me/submissions/841554) struct SegTree<I, BiOp> { n: usize, orign: usize, dat: Vec<I>, op: BiOp, e: I, } impl<I, BiOp> SegTree<I, BiOp> where BiOp: Fn(I, I) -> I, I: Copy { pub fn new(n_: usize, op: BiOp, e: I) -> Self { let mut n = 1; while n < n_ { n *= 2; } // n is a power of 2 SegTree {n: n, orign: n_, dat: vec![e; 2 * n - 1], op: op, e: e} } // ary[k] <- v pub fn update(&mut self, idx: usize, v: I) { debug_assert!(idx < self.orign); let mut k = idx + self.n - 1; self.dat[k] = v; while k > 0 { k = (k - 1) / 2; self.dat[k] = (self.op)(self.dat[2 * k + 1], self.dat[2 * k + 2]); } } // [a, b) (half-inclusive) // http://proc-cpuinfo.fixstars.com/2017/07/optimize-segment-tree/ #[allow(unused)] pub fn query(&self, rng: std::ops::Range<usize>) -> I { let (mut a, mut b) = (rng.start, rng.end); debug_assert!(a <= b); debug_assert!(b <= self.orign); let mut left = self.e; let mut right = self.e; a += self.n - 1; b += self.n - 1; while a < b { if (a & 1) == 0 { left = (self.op)(left, self.dat[a]); } if (b & 1) == 0 { right = (self.op)(self.dat[b - 1], right); } a = a / 2; b = (b - 1) / 2; } (self.op)(left, right) } } trait Bisect<T> { fn lower_bound(&self, val: &T) -> usize; fn upper_bound(&self, val: &T) -> usize; } impl<T: Ord> Bisect<T> for [T] { fn lower_bound(&self, val: &T) -> usize { let mut pass = self.len() + 1; let mut fail = 0; while pass - fail > 1 { let mid = (pass + fail) / 2; if &self[mid - 1] >= val { pass = mid; } else { fail = mid; } } pass - 1 } fn upper_bound(&self, val: &T) -> usize { let mut pass = self.len() + 1; let mut fail = 0; while pass - fail > 1 { let mid = (pass + fail) / 2; if &self[mid - 1] > val { pass = mid; } else { fail = mid; } } pass - 1 } } fn calc(ab: &[(i64, i64)]) -> Vec<Vec<(i64, i64)>> { let n = ab.len(); let mut ans = vec![vec![]; n + 1]; for bits in 0usize..1 << n { let mut x = 0; let mut y = 0; for i in 0..n { if (bits & 1 << i) != 0 { x += ab[i].0; y += ab[i].1; } } let bc = bits.count_ones() as usize; ans[bc].push((x, y)); } ans } // https://yukicoder.me/problems/no/2161 (3) // 半分全列挙 + 前半 x 個、後半 y 個 (x + y <= k) であるものを平面走査で数える fn main() { input! { n: usize, k: usize, l: i64, p: i64, ab: [(i64, i64); n], } let fst = calc(&ab[..n / 2]); let snd = calc(&ab[n / 2..]); let mut ans = 0i64; for i in 0..k + 1 { if k - i > n - n / 2 { continue; } let mut coo = vec![]; for &(_, y) in &snd[k - i] { coo.push(y); } coo.sort(); coo.dedup(); let mut ev = vec![]; for j in 0..std::cmp::min(i, n / 2) + 1 { for &(x, y) in &fst[j] { ev.push((l - x, 1, coo.lower_bound(&(p - y)))); } } for &(x, y) in &snd[k - i] { ev.push((x, 0, coo.lower_bound(&y))); } ev.sort(); let m = coo.len(); let mut st = SegTree::new(m, |x, y| x + y, 0i64); for &(_, ty, idx) in &ev { if ty == 1 { ans += st.query(idx..m); } else { let old = st.query(idx..idx + 1); st.update(idx, old + 1); } } } println!("{}", ans); }