题目地址:
https://leetcode.com/problems/reverse-pairs/
给定一个长 n n n的数组 A A A,求 ∣ { ( i , j ) : i < j ∧ A [ i ] > 2 A [ j ] } ∣ |\{(i,j):i<j\land A[i]>2A[j]\}| ∣{(i,j):i<j∧A[i]>2A[j]}∣。
可以用树状数组。先将 A A A连同 2 A 2A 2A一起做保序离散化(用哈希表),接着可以用经典求法。遍历 A A A,对于 A [ i ] A[i] A[i],我们求一下 A [ 0 : i − 1 ] A[0:i-1] A[0:i−1]里满足小于等于 2 A [ i ] 2A[i] 2A[i]的数有多少个,可以在树状数组里求前缀和,如果有 t t t个,则大于 2 A [ i ] 2A[i] 2A[i]的就有 i − t i-t i−t个;然后再将 A [ i ] A[i] A[i]离散化之后,将这个位置的数加 1 1 1即可。代码如下:
class Solution {
public:
#define lowbit(x) (x & -x)
vector<int> tr;
int n;
void add(int k, int x) {
for (; k <= n; k += lowbit(k)) tr[k] += x;
}
int sum(int k) {
int res = 0;
for (; k; k -= lowbit(k)) res += tr[k];
return res;
}
int reversePairs(vector<int>& A) {
vector<long> tmp(A.begin(), A.end());
for (auto &x : A) tmp.push_back(2L * x);
sort(tmp.begin(), tmp.end());
n = unique(tmp.begin(), tmp.end()) - tmp.begin();
unordered_map<long, int> mp;
for (int i = 0; i < n; i++) mp[tmp[i]] = i + 1;
tr = vector<int>(n + 1);
int res = 0;
for (int i = 0, x; i < A.size(); i++) {
x = A[i];
res += i - sum(mp[x * 2L]);
add(mp[x], 1);
}
return res;
}
};
时间复杂度 O ( n log n ) O(n\log n) O(nlogn),空间 O ( n ) O(n) O(n)。
Java:
import java.util.Arrays;
import java.util.HashMap;
import java.util.Map;
public class Solution {
int[] tr;
int n;
int lowbit(int x) {
return x & -x;
}
void add(int k, int x) {
for (; k <= n; k += lowbit(k)) {
tr[k] += x;
}
}
int sum(int k) {
int res = 0;
for (; k > 0; k -= lowbit(k)) {
res += tr[k];
}
return res;
}
public int reversePairs(int[] A) {
long[] tmp = new long[A.length << 1];
for (int i = 0; i < A.length; i++) {
tmp[2 * i] = A[i];
tmp[2 * i + 1] = A[i] * 2L;
}
Arrays.sort(tmp);
int m = 0;
for (int i = 0; i < tmp.length; i++) {
if (m == 0 || tmp[i] > tmp[m - 1]) {
tmp[m++] = tmp[i];
}
}
Map<Long, Integer> map = new HashMap<>();
for (int i = 0; i < m; i++) {
map.put(tmp[i], i + 1);
}
n = m;
tr = new int[m + 1];
int res = 0;
for (int i = 0, x; i < A.length; i++) {
x = A[i];
res += i - sum(map.get(x * 2L));
add(map.get((long) x), 1);
}
return res;
}
}
时空复杂度一样。