请问为什么线段树+树状数组全部TLE

已解决主要TLE问题：现在求帮卡常 /* * Author: xiaohei_AWM * Date: 5.24 * Mutto: Face to the weakness, expect for the strength. */ #include<cstdio> #include<cstring> #include<algorithm> #include<iostream> #include<cstdlib> #include<ctime> #include<set> #include<queue> #include<utility> #include<functional> #include<cmath> #include<vector> #include<assert.h> using namespace std; #define reg register #define endfile fclose(stdin);fclose(stdout); typedef long long ll; typedef unsigned long long ull; typedef double db; typedef std::pair<int,int> pii; typedef std::pair<ll,ll> pll; namespace IO{ char buf[1<<15],*S,*T; inline char gc(){ if (S==T){ T=(S=buf)+fread(buf,1,1<<15,stdin); if (S==T)return EOF; } return *S++; } inline int read(){ reg int x;reg bool f;reg char c; for(f=0;(c=gc())<'0'||c>'9';f=c=='-'); for(x=c^'0';(c=gc())>='0'&&c<='9';x=(x<<3)+(x<<1)+(c^'0')); return f?-x:x; } inline ll readll(){ reg ll x;reg bool f;reg char c; for(f=0;(c=gc())<'0'||c>'9';f=c=='-'); for(x=c^'0';(c=gc())>='0'&&c<='9';x=(x<<3)+(x<<1)+(c^'0')); return f?-x:x; } } using namespace IO; const int maxn = 1e5 + 10; int n, m, k, s[maxn]; ll ans, sum, c[maxn], temp[maxn], kth, kthid; struct Element{ int id; ll val; bool operator<(const Element &v)const{ if(val == v.val) return id < v.id; return val > v.val; } } a[maxn]; struct Query{ int l, r, id; ll val; } b[maxn]; inline int lowbit(int x){ return x & -x; } void add(int pos, ll val){ for(int i = pos; i <= n; i += lowbit(i)) c[i] += val; } ll queryS(int pos){ ll res = 0; for(int i = pos; i; i -= lowbit(i)) res += c[i]; return res; } bool cmp2(Query u, Query v){ return u.val > v.val; } vector<int>date[maxn*4+5]; //存线段树节点信息 void build(int ro,int l,int r) { if(l == r) { date[ro].push_back(-1e7); //乱塞一个数（0号位置） date[ro].push_back(s[l]); return; } int mid = (l+r)>>1; build(ro<<1,l,mid); build(ro<<1|1,mid+1,r); //这一部分类似于归并排序（应该说就是一样的），保证时间复杂度。 int size1 = date[ro<<1].size()-1; int size2 = date[ro<<1|1].size()-1; int i = 1,j = 1; date[ro].push_back(-1e7); while(i<=size1 && j<=size2) { if(date[ro<<1][i]>date[ro<<1|1][j])date[ro].push_back(date[ro<<1|1][j++]); else date[ro].push_back(date[ro<<1][i++]); } while(i<=size1)date[ro].push_back(date[ro<<1][i++]); while(j<=size2)date[ro].push_back(date[ro<<1|1][j++]); return; } //二分查找完全包含的区间中小于等于R的数的个数 int low_bound(int ro,int target) { int L,R,res; L =1 ; R = date[ro].size()-1;res = 0; while(L<=R) { int mid = (L+R)>>1; if(date[ro][mid]<=target){res = mid; L = mid + 1;} else R = mid - 1; } return res; } //线段树查询区间内小于等于R的数的个数 //tl--target L ; tr--target R; l--now L; r--now R 。 int query(int ro,int l,int r,int tl,int tr,int K) { if(tr<l || tl>r)return 0; //没有交集 else if(tl<=l && r<=tr) //完全包含 { int temp_num = low_bound(ro,K); return temp_num; } else //有交集但不完全包含，递归查找左右子树 { int mid = (l + r)>>1; int temp1 = query(ro<<1,l,mid,tl,tr,K); int temp2 = query(ro<<1|1,mid+1,r,tl,tr,K); return (temp1+temp2); //合并答案 } } //二分查找目标值R，二分出值R后再用线段树进行验证 int search(int l,int r,int k) { int L,R,RES,NUM; L = 0; R = maxn; while(L<=R) { int mid = (L+R)>>1; NUM = query(1,1,n,l,r,mid); //RES不断向下逼近，保证精确性。 if(NUM >= k){RES = mid ; R = mid-1;} else L = mid+1; } return RES; } bool cmpx(Element x, Element y){ if(x.val == y.val) return x.id < y.id; return x.val < y.val; } int main(){ n = read(), m = read(), k = read(); for(int i = 1; i <= n; i++) a[i].val = temp[i] = readll(), a[i].id = i; sort(a+1, a+1+n, cmpx); for(int i = 1; i <= n; i++) s[a[i].id] = i; build(1, 1, n); for(int i = 1; i <= n-m+1; i++){ a[i].val = search(i, i+m-1, k); a[i].id = i; } int LIM = n-m+1; for(int i = 1; i <= n; i++){ b[i].l = max(1, i - m + 1); b[i].r = min(LIM, i); b[i].val = s[i]; b[i].id = i; } sort(a+1, a+1+LIM); sort(b+1, b+1+n, cmp2); int cnt = 1; for(int i = 1; i <= n; i++){ while(cnt <= LIM && a[cnt].val >= b[i].val){ add(a[cnt].id, 1); cnt++; } int num = (queryS(b[i].r) - queryS(b[i].l-1)); ans += (ll)(queryS(b[i].r) - queryS(b[i].l-1)) * temp[b[i].id]; } cout << ans << endl; return 0; }