程序员面试金典——18.9实时中位数

程序员面试金典——18.9实时中位数

Solution1：我的答案。利用排序，比较弱智。。

class Middle {
public:
    vector<int> getMiddle(vector<int> A, int n) {
        // write code here
        vector<int> res;
        if (n <= 0)
            return res;
        auto iter1 = A.begin(), iter2 = ++A.begin();
        for (; iter2 <= A.end(); iter2++) {
            sort(iter1, iter2);
            res.push_back(A[(iter2 - iter1 - 1) / 2]);//注意准确计算下标
        }
        return res;
    }
};

Solution2：利用优先队列priority_queue，说实话还是第一次用呢~~
此题思路如下图所示：

利用priority_queue自动排序了
关于priority_queue介绍的博客：
[1]https://blog.csdn.net/pzhu_cg_csdn/article/details/79166858
[2]http://www.cplusplus.com/reference/queue/priority_queue/

class Middle {
public:
    vector<int> getMiddle(vector<int> A, int n) {
        // write code here
        priority_queue<int, vector<int>, less<int> > small;
        priority_queue<int, vector<int>, greater<int> > big;
        vector<int> ans;
        for (int i = 0; i < A.size(); ++i) {
            if (small.empty() || A[i] <= small.top())
                small.push(A[i]);
            else
                big.push(A[i]);
            if (small.size() == big.size() + 2) {
                big.push(small.top());
                small.pop();
            }
            else if (small.size() == big.size() - 1) {
                small.push(big.top());
                big.pop();
            }
            ans.push_back(small.top());
        }
        return ans;
    }
};

Solution3：
利用最大堆和最小堆实现，其思想和用priority_queue差不多，但是代码却麻烦了好多。。。。还是优先队列好啊
关于最大堆和最小堆的介绍：
[1]https://blog.csdn.net/guoweimelon/article/details/50904346
关于push_heap和pop_heap的介绍
[1]https://blog.csdn.net/fuyufjh/article/details/47144869

class Middle {
public:
    vector<int> getMiddle(vector<int> A, int n) {
        // write code here
        vector<int> maxheap;
        vector<int> minheap;
        vector<int> re;
        for (int i = 0; i < n; ++i) {
            if (maxheap.size() == minheap.size()) {
                if (minheap.size() > 0 && minheap.front() < A[i]) {
                    pop_heap(minheap.begin(), minheap.end(), greater<int>());
                    maxheap.push_back(minheap.back());
                    push_heap(maxheap.begin(), maxheap.end(), less<int>());
                    minheap.pop_back();
                    minheap.push_back(A[i]);
                    push_heap(minheap.begin(), minheap.end(), greater<int>());
                }
                else {
                    maxheap.push_back(A[i]);
                    push_heap(maxheap.begin(), maxheap.end(), less<int>());
                }
            }
            else {
                if(maxheap.front() > A[i]) {
                    pop_heap(maxheap.begin(), maxheap.end(), less<int>());
                    minheap.push_back(maxheap.back());
                    push_heap(minheap.begin(), minheap.end(), greater<int>());
                    maxheap.pop_back();
                    maxheap.push_back(A[i]);
                    push_heap(maxheap.begin(), maxheap.end(), less<int>());
                }
                else {
                    minheap.push_back(A[i]);
                    push_heap(minheap.begin(), minheap.end(), greater<int>());
                }
            }
            re.push_back(maxheap.front());
        }
        return re;
    }
};