6044. 花期内花的数目（点被区间覆盖）

6044. 花期内花的数目

题目

---

样例

---

数据

1 <= flowers.length <= 5 * 104
1 <= starti <= endi <= 109
1 <= persons.length <= 5 * 104
1 <= persons[i] <= 109

---

题解

map差分 + 双指针

思路：

1. $[star{t}_i,en{d}_i]区间每个时间点都多一朵花，计算perso{n}_i时间点的花数$。
2. 利用map差分思路，$start增加1，end+1减少1$。
3. iota(id)是$person$排序的对应下标数组。
4. 从小到大遍历$mp$，指针同从小到大移动到$perso{n}_i$时间点。

class Solution {
public:
    vector<int> fullBloomFlowers(vector<vector<int>>& flowers, vector<int>& persons) {
        map<int, int> mp;
        for (auto &x: flowers) mp[x[0]]++, mp[x[1] + 1]--;
        vector<int> id(persons.size());
        iota(id.begin(), id.end(), 0);
        sort(id.begin(), id.end(), [&](int i, int j){ return persons[i] < persons[j]; });

        vector<int> ans(persons.size());
        auto it = mp.begin();
        int sum = 0;
        for (int i = 0; i < id.size(); i++) {
            while(it != mp.end() && it->first <= persons[id[i]]) sum += it++->second;
            ans[id[i]] = sum;
        }
        return ans;
    }
};

---

二分（清晰透彻）

思路：

• 正在开的花的个数 = 所有在此时间节点及之前开放的花的个数 - 在时间节点之前凋谢的花的个数

class Solution {
public:
    vector<int> fullBloomFlowers(vector<vector<int>>& flowers, vector<int>& persons) {
        int n = flowers.size();
        vector<int> open(n), close(n);
        for (int i = 0; i < n; i++) open[i] = flowers[i][0], close[i] = flowers[i][1];
        sort(open.begin(), open.end());
        sort(close.begin(), close.end());

        vector<int> ans;
        for (int i = 0; i < persons.size(); i++) {
            int now = persons[i];
            int op = upper_bound(open.begin(), open.end(), now) - open.begin();
            int cl = upper_bound(close.begin(), close.end(), now - 1) - close.begin();
            ans.push_back(op - cl);
        }
        return ans;
    }
};

---

差分 + 树状数组 + 哈希

哈希树状数组

• 注释正常思路，对开花，谢花，人到的时间点都标记，排序。
• 但数据太大，$tr$数组维护不了1e9的长度

时间开销大

class Solution {
public:
    static constexpr int N = 1e9 + 7;
    // int tr[N];
    unordered_map<int, int> tr;
    void add(int x, int c = 1) {
        for (; x <= N; x += x & -x) tr[x] += c;
    }
    int sum(int x) {
        int res = 0;
        for (; x; x -= x & -x) res += tr[x];
        return res;
    }
    // #define INF 0x3f3f3f3f
    vector<int> fullBloomFlowers(vector<vector<int>>& flowers, vector<int>& persons) {
        int n = flowers.size();
        // vector<pair<int, int>> v;
        // for (int i = 0; i < n; i++) v.push_back({flowers[i][0], -INF}), v.push_back({flowers[i][1], INF});
        // for (int i = 0; i < persons.size(); i++) v.push_back({persons[i], i});
        // sort(v.begin(), v.end());
        for (auto &x: flowers) add(x[0], 1), add(x[1] + 1, -1); // 从这之后少1，差分嘛
        vector<int> ans(persons.size());
        for (int i = 0; i < persons.size(); i++) {
            // auto x = v[i];
            // if(x.second == -INF) add(x.first, 1);
            // else if(x.second == INF) add(x.first + 1, -1);
            // else ans[x.second] = sum(x.first);
            ans[i] = sum(persons[i]);
        }
        return ans;
    }
};

---

离散化 + 前缀和 （简洁明了）

思路：

• 离散化（map）后，$cnt<=3\ast 5e4$
• 直接前缀和求得答案。

class Solution {
public:
    map<int, int> a, id;
    int s[150010];
    int cnt = 0;
    vector<int> fullBloomFlowers(vector<vector<int>>& flowers, vector<int>& persons) {
        for (auto &x: flowers) a[x[0]] = 1, a[x[1] + 1] = 1;
        for (auto &x: persons) a[x] = 1;
        for (auto &x: a) id[x.first] = ++cnt;
        for (auto &x: flowers) s[id[x[0]]]++, s[id[x[1] + 1]]--;
        for (int i = 1; i <= cnt; i++) s[i] += s[i - 1];
        vector<int> res;
        for (auto &x: persons) res.push_back(s[id[x]]);
        return res;
    }
};