思路:类似归并排序算法,在合并已经有序的相邻子数组的时候,计算前面数组相对于后面数组的逆序对数,整个递归过程可以算出所有逆序对
#include <stdio.h>
void merge(int A[], int front, int middle, int end, int &count)
{
if (front >= end) return;
int i = front;
int j = middle + 1;
int k = 0;
int *p = new int[end - front + 1];
while (i <= middle && j <= end) {
if (A[i] < A[j]) {
p[k++] = A[i++];
} else {
count += middle - i + 1;
p[k++] = A[j++];
}
}
if (j <= end) while (j <= end) {
p[k++] = A[j++];
count += middle - i + 1;
}
if (i <= middle) while (i <= middle) p[k++] = A[i++];
printf("front:%d end:%d :", front, end);
for (int i = 0; i < k; ++i) {
printf("%d ", p[i]);
A[front + i] = p[i];
}
printf("
");
delete [] p;
}
void merge_sort(int A[], int front, int end, int &count)
{
if (front >= end) return;
int middle = (front + end) / 2;
merge_sort(A, front, middle, count);
merge_sort(A, middle + 1, end, count);
merge(A, front, middle, end, count);
}
int main()
{
int count = 0;
int rand[] = {3,1,5,23,2,32,56,76,65,34,2,2,333,1,0};