数组中的逆序对

在数组中的两个数字,如果前面一个数字大于后面的数字,则这两个数字组成一个逆序对。输入一个数组,求出这个数组中的逆序对的总数P。并将P对1000000007取模的结果输出。 即输出P%1000000007 
输入描述:

题目保证输入的数组中没有的相同的数字

数据范围:

对于%50的数据,size<=10^4

对于%75的数据,size<=10^5

对于%100的数据,size<=2*10^5



输入例子:
1,2,3,4,5,6,7,0

输出例子:

7

/*归并排序的改进,把数据分成前后两个数组(递归分到每个数组仅有一个数据项),
合并数组,合并时,出现前面的数组值array[i]大于后面数组值array[j]时;则前面
数组array[i]~array[mid]都是大于array[j]的,count += mid+1 - i
参考剑指Offer,但是感觉剑指Offer归并过程少了一步拷贝过程。
还有就是测试用例输出结果比较大,对每次返回的count mod(1000000007)求余
*/
 
public class Solution {
    public int InversePairs(int [] array) {
        if(array==null||array.length==0)
        {
            return 0;
        }
        int[] copy = new int[array.length];
        for(int i=0;i<array.length;i++)
        {
            copy[i] = array[i];
        }
        int count = InversePairsCore(array,copy,0,array.length-1);//数值过大求余
        return count;
         
    }
    private int InversePairsCore(int[] array,int[] copy,int low,int high)
    {
        if(low==high)
        {
            return 0;
        }
        int mid = (low+high)>>1;
        int leftCount = InversePairsCore(array,copy,low,mid)%1000000007;
        int rightCount = InversePairsCore(array,copy,mid+1,high)%1000000007;
        int count = 0;
        int i=mid;
        int j=high;
        int locCopy = high;
        while(i>=low&&j>mid)
        {
            if(array[i]>array[j])
            {
                count += j-mid;
                copy[locCopy--] = array[i--];
                if(count>=1000000007)//数值过大求余
                {
                    count%=1000000007;
                }
            }
            else
            {
                copy[locCopy--] = array[j--];
            }
        }
        for(;i>=low;i--)
        {
            copy[locCopy--]=array[i];
        }
        for(;j>mid;j--)
        {
            copy[locCopy--]=array[j];
        }
        for(int s=low;s<=high;s++)
        {
            array[s] = copy[s];
        }
        return (leftCount+rightCount+count)%1000000007;
    }
}

参考《剑指offer》用归并排序的思想, 时间复杂度O(nlogn)

代码如下:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
int InversePairs(vector<int> data) {
        int length  = data.size();
        return mergeSort(data, 0, length-1);
    }
 
    int mergeSort(vector<int>& data, int start, int end) {
        // 递归终止条件
        if(start >= end) {
            return 0;
        }
 
        // 递归
        int mid = (start + end) / 2;
        int leftCounts = mergeSort(data, start, mid);
        int rightCounts = mergeSort(data, mid+1, end);
 
        // 归并排序,并计算本次逆序对数
        vector<int> copy(data); // 数组副本,用于归并排序
        int foreIdx = mid;      // 前半部分的指标
        int backIdx = end;      // 后半部分的指标
        int counts = 0;         // 记录本次逆序对数
        int idxCopy = end;      // 辅助数组的下标
        while(foreIdx>=start && backIdx >= mid+1) {
            if(data[foreIdx] > data[backIdx]) {
                copy[idxCopy--] = data[foreIdx--];
                counts += backIdx - mid;
            else {
                copy[idxCopy--] = data[backIdx--];
            }
        }
        while(foreIdx >= start) {
            copy[idxCopy--] = data[foreIdx--];
        }
        while(backIdx >= mid+1) {
            copy[idxCopy--] = data[backIdx--];
        }
        for(int i=start; i<=end; i++) {
            data[i] = copy[i];
        }
 
        return (leftCounts+rightCounts+counts);
    }


评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值