逆序数

ACM模版

归并排序求逆序数

/*
 *  也可以用树状数组做
 *  a[0...n-1] cnt=0; call: MergeSort(0, n)
 */
const int N = 1010;
int a[N];
int c[N];
int cnt = 0;

void MergeSort(int l, int r)
{
    int mid, i, j, tmp;
    if (r > l + 1)
    {
        mid = (l + r) / 2;
        MergeSort(l, mid);
        MergeSort(mid, r);
        tmp = l;
        for (i = l, j = mid; i < mid && j < r;)
        {
            if (a[i] > a[j])
            {
                c[tmp++] = a[j++];
                cnt += mid - i;
            }
            else
            {
                c[tmp++] = a[i++];
            }
        }
        if (j < r)
        {
            for (; j < r; ++j)
            {
                c[tmp++] = a[j];
            }
        }
        else
        {
            for (; i < mid; ++i)
            {
                c[tmp++]=a[i];
            }
        }
        for (i = l; i < r; ++i)
        {
            a[i] = c[i];
        }
    }
    return ;
}

逆序数推排列数

/*
 *  动态规划:f(n,m)表示逆序数为m的n元排列的个数,则
 *  f(n + 1, m) = f(n, m) + f(n, m - 1) + ... + f(n, m - n)(当 b<0 时,f(a,b) = 0) 
 *  优化又考虑到如果直接利用上式计算时间复杂度为O(n^3),我们分析上 
 *  式不难发现f(n + 1, m) = f(n, m) + f(n + 1, m - 1)
 *  end = node[index].e,
 *  if(m-n-1 >= 0) f(n+1, m) -= f(n, m-n-1).
 */
const int N = 1001;
const int C = 10001;
const long MOD = 1000000007;
long arr[N][C];
long long temp;

int main()
{
    int i, j;
    arr[1][0] = arr[2][0] = arr[2][1] = 1;
    for (i = 3; i < N; ++i)
    {
        arr[i][0] = 1;
        long h = i * (i + 1) / 2 + 1;
        if (h > C)
        {
            h = C;
        }
        for (j = 1; j < h; ++j)
        {
            temp = arr[i - 1][j] + arr[i][j - 1];
            arr[i][j] = temp % MOD;
            if (j - i >= 0)
            {
                arr[i][j] -= arr[i - 1][j - i];
                if (arr[i][j] < 0)
                {   //  注意:由于arr[i][j]和arr[i - 1][j - i]都是模过的,所以可能会得到负数
                    arr[i][j] += MOD;
                }
            }
        }
    }
    while (scanf("%d %d", &i, &j) != EOF)
    {
        printf("%ld\n", arr[i][j]);
    }
    return 0;
}