线段树是一种二叉搜索树,与区间树相似,它将一个区间划分成一些单元区间,每个单元区间对应线段树中的一个叶结点。
使用线段树可以快速的查找某一个节点在若干条线段中出现的次数,时间复杂度为O(logN)。而未优化的空间复杂度为2N,实际应用时一般还要开4N的数组以免越界,因此有时需要离散化让空间压缩。
区间更新(查询)的时间复杂度是O(logn),使用懒惰法只会影响以下四类节点,每类节点数量都不超过logn。一,左边界及其祖先。二,右边界及其祖先。三,第一类的兄弟节点。四,第二类的兄弟节点。
本封装类,都是使用数组模拟,如果不是连续的节点,需要离散化。如果是流,无法离散化。则需要用哈希映射或树。
封装类
单点初始化、更新
template<class TSave,class TRecord>
class CSingUpdateLineTree
{
public:CSingUpdateLineTree(int iEleSize):m_iEleSize(iEleSize), m_vSave(iEleSize*4){}void Update(int index, TRecord update) {Update(1, 1, m_iEleSize, index + 1, update);}void Query(int leftIndex, int leftRight) {Query(1, 1, m_iEleSize, leftIndex + 1, leftRight + 1);}void Init() {Init(1, 1, m_iEleSize);}const int m_iEleSize;
protected:void Init(int iNodeNO, int iSaveLeft, int iSaveRight){if (iSaveLeft == iSaveRight) {OnInit(m_vSave[iNodeNO], iSaveLeft);return;}const int mid = iSaveLeft + (iSaveRight - iSaveLeft) / 2;Init(iNodeNO * 2, iSaveLeft, mid);Init(iNodeNO * 2+1, mid+1, iSaveRight);OnUpdateParent(m_vSave[iNodeNO], m_vSave[iNodeNO * 2], m_vSave[iNodeNO * 2 + 1], iSaveLeft, iSaveRight);}void Query(int iNodeNO, int iSaveLeft, int iSaveRight, int iQueryLeft,int iQueryRight) {if (( iSaveLeft >= iQueryLeft) && (iSaveRight <= iQueryRight )) {OnQuery(m_vSave[iNodeNO]);return;}if (iSaveLeft == iSaveRight) {//没有子节点return;}const int mid = iSaveLeft + (iSaveRight - iSaveLeft) / 2;if (mid >= iQueryLeft) {Query(iNodeNO * 2, iSaveLeft, mid, iQueryLeft, iQueryRight);}if( mid+1 <= iQueryRight ){Query(iNodeNO * 2+1, mid+1, iSaveRight, iQueryLeft, iQueryRight);}}void Update(int iNodeNO,int iSaveLeft,int iSaveRight,int iUpdateNO, TRecord update) {if (iSaveLeft == iSaveRight){OnUpdate(m_vSave[iNodeNO], update);return;}const int mid = iSaveLeft + (iSaveRight - iSaveLeft) / 2;if (iUpdateNO <= mid) {Update(iNodeNO * 2, iSaveLeft, mid, iUpdateNO, update);}else {Update(iNodeNO * 2+1, mid+1, iSaveRight, iUpdateNO, update);}OnUpdateParent(m_vSave[iNodeNO], m_vSave[iNodeNO * 2], m_vSave[iNodeNO * 2+1],iSaveLeft,iSaveRight);}virtual void OnInit(TSave& save,int iSave)=0;virtual void OnQuery(TSave& save) = 0;virtual void OnUpdate(TSave& save, const TRecord& update) = 0;virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r,int iSaveLeft,int iSaveRight) = 0;vector<TSave> m_vSave;
};
区间更新、区间查询
template<class TSave, class TRecord>
class CLineTree
{
public:CLineTree(int iEleSize, TRecord recordNull=0):m_iEleSize(iEleSize), m_vArr(m_iEleSize * 4), m_vRecord(m_iEleSize * 4, recordNull), m_recordNull(recordNull){}void Update(int iLeftIndex, int iRightIndex, TRecord value){Update(1, 1, m_iEleSize, iLeftIndex + 1, iRightIndex + 1, value);}void Query( int iLeftIndex, int iRightIndex){Query( 1, 1, m_iEleSize, iLeftIndex + 1, iRightIndex + 1);}
private:virtual void OnQuery(TSave& save) = 0;virtual void OnUpdateRecord(TRecord& old, const TRecord& newRecord) = 0;virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r) = 0;virtual void OnUpdate(TSave& save, const int& len, const TRecord& update) = 0;void Query( int iNode, int iSaveLeft, int iSaveRight, int iQueryLeft, int iQueryRight){if ((iQueryLeft <= iSaveLeft) && (iQueryRight >= iSaveRight)){OnQuery(m_vArr[iNode]);return;}Fresh(iNode, iSaveLeft, iSaveRight);const int iMid = iSaveLeft + (iSaveRight - iSaveLeft) / 2;if (iMid >= iQueryLeft){Query( iNode * 2, iSaveLeft, iMid, iQueryLeft, iQueryRight);}if (iMid + 1 <= iQueryRight){Query( iNode * 2 + 1, iMid + 1, iSaveRight, iQueryLeft, iQueryRight);}}void Update(int iNode, int iSaveLeft, int iSaveRight, int iOpeLeft, int iOpeRight, TRecord value){if (iNode >= m_vArr.size()){return;}if ((iOpeLeft <= iSaveLeft) && (iOpeRight >= iSaveRight)){OnUpdate(m_vArr[iNode], min(iSaveRight, iOpeRight) - max(iSaveLeft, iOpeLeft) + 1, value);OnUpdateRecord(m_vRecord[iNode], value);return;}Fresh(iNode, iSaveLeft, iSaveRight);const int iMid = iSaveLeft + (iSaveRight - iSaveLeft) / 2;if (iMid >= iOpeLeft){Update(iNode * 2, iSaveLeft, iMid, iOpeLeft, iOpeRight, value);}if (iMid + 1 <= iOpeRight){Update(iNode * 2 + 1, iMid + 1, iSaveRight, iOpeLeft, iOpeRight, value);}// 如果有后代,至少两个后代OnUpdateParent(m_vArr[iNode], m_vArr[iNode * 2], m_vArr[iNode * 2 + 1]);}void Fresh(int iNode, int iDataLeft, int iDataRight){if (m_recordNull == m_vRecord[iNode]){return;}const int iMid = iDataLeft + (iDataRight - iDataLeft) / 2;Update(iNode * 2, iDataLeft, iMid, iDataLeft, iMid, m_vRecord[iNode]);Update(iNode * 2 + 1, iMid + 1, iDataRight, iMid + 1, iDataRight, m_vRecord[iNode]);m_vRecord[iNode] = m_recordNull;}const int m_iEleSize;vector<TSave> m_vArr;vector<TRecord> m_vRecord;const TRecord m_recordNull;
};
样例汇总
【线段树】【前缀和】:1687从仓库到码头运输箱子
template<class TSave = int , class TRecord = int>
class CMinLineTree : public CLineTree<TSave, TRecord>
{
public:using CLineTree<TSave, TRecord>::CLineTree;int m_iQueryValue = INT_MAX;
protected:virtual void OnQuery(TSave& save) override{m_iQueryValue = min(m_iQueryValue, save);}virtual void OnUpdateRecord(TRecord& old, const TRecord& newRecord) override{old += newRecord;}virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r) override{par = min(left, r);}virtual void OnUpdate(TSave& save, const int& len, const TRecord& update) override{save += update;}};
【线段树】1622. 奇妙序列
template<class TSave = C1097Int<>, class TRecord = pair<C1097Int<>,C1097Int<>> >class CMyLineTree : public CLineTree<TSave, TRecord>{public:using CLineTree<TSave, TRecord>::CLineTree; protected:virtual void OnUpdateRecord(TRecord& old, const TRecord& newRecord) override{old.first *= newRecord.first;old.second = old.second * newRecord.first + newRecord.second;}virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r) override{}virtual void OnUpdate(TSave& save, const int& len, const TRecord& iUpdate) override{save = save * iUpdate.first + iUpdate.second;}};
【最大值线段树】【二分查找】2286. 以组为单位订音乐会的门票
template<class TSave, class TRecord, TRecord RecordNull = 0>
class CMaxLineTree : public CLineTree<TSave, TRecord, RecordNull>
{using CLineTree< TSave, TRecord, RecordNull>::CLineTree;virtual void OnUpdateRecord(TRecord& old, const TRecord& newRecord) override{old = newRecord;}virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r) override{par = max(left, r);}virtual void OnUpdate(TSave& save, const int& len, const TRecord& iUpdate) override{save = iUpdate;}
};
【线段树】【区间更新】2916. 子数组不同元素数目的平方和 II
class CPOW2LineTree : public CLineTree<pair<C1097Int<>, C1097Int<>>,int>
{
public:typedef pair<C1097Int<>, C1097Int<>> TSave;typedef int TRecord;const TRecord RecordNull = 0 ;using CLineTree::CLineTree;// 通过 CLineTree 继承virtual void OnUpdateRecord(TRecord& old, const TRecord& newRecord) override{old += newRecord;}// 通过 CLineTree 继承virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r) override{par.first = left.first + r.first;par.second = left.second + r.second;}virtual void OnUpdate(TSave& save, const int& len, const TRecord& iUpdate) override{save.second += save.first * 2 * iUpdate + C1097Int<>(len) * iUpdate * iUpdate;save.first += C1097Int<>(iUpdate) * len;}
};
单点初始化
【线段树】【众数】1157数组中占绝大多数的元素
template<class TSave = std::pair<int,int>, class TRecord = int >
class CMyLineTree : public CSingUpdateLineTree<TSave, TRecord>
{
public:CMyLineTree(const vector<int>& arr):m_moreNum(arr),CSingUpdateLineTree<TSave,TRecord>(arr.size()){m_arr = arr; CSingUpdateLineTree<TSave, TRecord>::Init();}int Query(int left, int r, int threshold){m_vCan.clear();CSingUpdateLineTree<TSave, TRecord>::Query(left,r);auto [i1, i2] = m_moreNum.Query(left, r, m_vCan);return (i2 >= threshold) ? i1 : -1;}
protected:vector<int> m_vCan;virtual void OnQuery(TSave& save) override {m_vCan.emplace_back(save.first);}virtual void OnUpdate(TSave& save, const TRecord& update) override{};virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r, int iSaveLeft, int iSaveRight) override {vector<int> vCan = { left.first,r.first };par = m_moreNum.Query(iSaveLeft - 1, iSaveRight - 1, vCan);} vector<int> m_arr;CMoreNum m_moreNum;virtual void OnInit(TSave& save, int iSave) override {save = { m_arr[iSave - 1],1 };}
};
【线段树】2213. 由单个字符重复的最长子字符串
template<class TSave = std::tuple<int,int,int>, class TRecord = char, class TSaveCon = CUnorderMapSave<TSave> >
class CMyLineTree :public CSingUpdateLineTree<TSave,TRecord, TSaveCon>
{
public:CMyLineTree(const string& s) :m_s(s), CSingUpdateLineTree<TSave, TRecord, TSaveCon>(s.length() ,{ 0,0,0 }) {}void Update(int index, TRecord update) {m_s[index] = update;CSingUpdateLineTree<TSave, TRecord, TSaveCon>::Update(index, update);}
protected:virtual void OnInit(TSave& save, int iSave) override{save = { 1,1,1 };}virtual void OnQuery(TSave& save) override{}virtual void OnUpdate(TSave& save, int iSaveLeft, const TRecord& update) override{save = { 1,1,1 };}virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r, int iSaveLeft, int iSaveRight) override{int i1 = get<0>(left);//最长前缀int i2 = max(get<1>(left), get<1>(r));//最长字符串int i3 = get<2>(r);//最长后缀const int mid = iSaveLeft + (iSaveRight - iSaveLeft) / 2;if (m_s[mid] == m_s[mid + 1]){//拼接i2 = max(i2, get<2>(left) + get<0>(r));if (mid - iSaveLeft + 1 == i1) {i1 += get<0>(r);}if (iSaveRight - mid == i3) {i3 += get<2>(left);}}par = { i1,i2,i3 };}string m_s;
};
【线段树】2276. 统计区间中的整数数目
template<class TSave=int, class TRecord =int >
class CMyTreeRangeLineTree : public CTreeRangeLineTree<TSave, TRecord>
{
public:using CTreeRangeLineTree<TSave, TRecord>::CTreeRangeLineTree;
protected:virtual void OnQuery(TSave& save) override{}virtual void OnUpdate(TSave& save, int iSaveLeft, int iSaveRight, const TRecord& update) override{ save = update*(iSaveRight-iSaveLeft+1);}virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r, int iSaveLeft, int iSaveRight) override{par = left + r;}virtual void OnUpdateRecord(TRecord& old, const TRecord& newRecord) override{old = newRecord;}
};
【线段树 有序映射】715. Range 模块
template<class TSave=int, class TRecord =int >
class CMyTreeRangeLineTree : public CTreeRangeLineTree<TSave, TRecord>
{
public:using CTreeRangeLineTree<TSave, TRecord>::CTreeRangeLineTree;bool queryRange(int left, int right) {m_bHas = true;CTreeRangeLineTree<TSave, TRecord>::Query(left, right);return m_bHas;}
protected:bool m_bHas = true;virtual void OnQuery(const TSave& save, const int& iSaveLeft, const int& iSaveRight) override{m_bHas &= (save == (iSaveRight - iSaveLeft + 1));}virtual void OnUpdate(TSave& save, const int& iSaveLeft, const int& iSaveRight, const TRecord& update) override{ save = update*(iSaveRight-iSaveLeft+1);}virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r, const int& iSaveLeft, const int& iSaveRight) override{par = left + r;}virtual void OnUpdateRecord(TRecord& old, const TRecord& newRecord) override{old = newRecord;}
};
其它有题解的题目
【键值皆有序map 线段树 数学 】100240. 最小化曼哈顿距离|最小线段树,单点更新,区间查询
template<class TSave=int, class TRecord =int >
class CMyTreeRangeLineTree : public CTreeRangeLineTree<TSave, TRecord>
{
public:using CTreeRangeLineTree<TSave, TRecord>::CTreeRangeLineTree;bool queryRange(int left, int right) {m_bHas = true;CTreeRangeLineTree<TSave, TRecord>::Query(left, right);return m_bHas;}
protected:bool m_bHas = true;virtual void OnQuery(TSave& save, int iSaveLeft, int iSaveRight) override{m_bHas &= (save == (iSaveRight - iSaveLeft + 1));}virtual void OnUpdate(TSave& save, int iSaveLeft, int iSaveRight, const TRecord& update) override{ save = update*(iSaveRight-iSaveLeft+1);}virtual void OnUpdateParent(TSave& par, const TSave& left, const TSave& r, int iSaveLeft, int iSaveRight) override{par = left + r;}virtual void OnUpdateRecord(TRecord& old, const TRecord& newRecord) override{old = newRecord;}
};
无题解的题目
LeetCode | |
---|---|
LeetCode 218 际线问题 | 最大值 离散化后 线段树区间修改,单点查询 |
LeetCode 315. 计算右侧小于当前元素的个数 | 求和,单点修改,区间查询 |
LeetCode 327. 区间和的个数 | 求和,前缀和是已知的,所以离散化。单点修改,区间查询 |
LeetCode 493. 翻转对 | 求和,离散化后,单点修改,区间查询 |
LeetCode 699. 掉落的方块 | 最大值,区间修改区间查询 |
LeetCode 715. Range 模块 | 求和,无法离散化,麻烦。直接模拟或差分数组+树状数组。 区间修改,区间查询。 |
LeetCode 732. 我的日程安排表 III | 最大值,区间更新,区间查询 |
LeetCode 850. 矩形面积 II | 求和,离散化+扫描线。线段树实现维护当前x,各y是否被覆盖。可以覆盖多次,也可以解除覆盖。有覆盖时,求和时为1,没覆盖为0。 |
LeetCode 1505. 最多 K 次交换相邻数位后得到的最小整数 | 求和,单点更新,单点查询 |
1521. 找到最接近目标值的函数值 | 与和+二分查找。除了练习,没有任何必要使用线段树。 |
1649. 通过指令创建有序数组 | 求和,单点更新,区间修改 |
2179. 统计数组中好三元组数目 | 等价转换(重新编码)+求和,单点更新,区间求和 |
2213. 由单个字符重复的最长子字符串 | 最长,单点修改,区间查询 |
2276. 统计区间中的整数数目 | 无法离散化,用哈希。区间修改、区间查询 |
2407. 最长递增子序列 II | 最大值,单点修改,区间查询 |
2426. 满足不等式的数对数目 | 离散化,求和,单点修改 |
2569. 更新数组后处理求和查询 | 01反转,区间修改 |
2736. 最大和查询 | 离线查询,最大值。单点更新,区间查找 |
2926. 平衡子序列的最大和 | 最大值,单点更新,区间查询 |
2940. 找到 Alice 和 Bob 可以相遇的建筑 | 二分查找+最大值线段树 |
3072. 将元素分配到两个数组中 II | 求和,单点修改,区间查询 |
LCP 05. 发 LeetCoin | DFS时间序,求和线段树。区间修改、区间查询 |
LCP 09. 最小跳跃次数 | 最小值,单点更新,区间查询 |
扩展阅读
视频课程
有效学习:明确的目标 及时的反馈 拉伸区(难度合适),可以先学简单的课程,请移步CSDN学院,听白银讲师(也就是鄙人)的讲解。
https://edu.csdn.net/course/detail/38771
如何你想快
速形成战斗了,为老板分忧,请学习C#入职培训、C++入职培训等课程
https://edu.csdn.net/lecturer/6176
相关
下载
想高屋建瓴的学习算法,请下载《喜缺全书算法册》doc版
https://download.csdn.net/download/he_zhidan/88348653
我想对大家说的话 |
---|
闻缺陷则喜是一个美好的愿望,早发现问题,早修改问题,给老板节约钱。 |
子墨子言之:事无终始,无务多业。也就是我们常说的专业的人做专业的事。 |
如果程序是一条龙,那算法就是他的是睛 |
测试环境
操作系统:win7 开发环境: VS2019 C++17
或者 操作系统:win10 开发环境: VS2022 C++17
如无特殊说明,本算法用**C++**实现。