Time Limit: 5000MS | Memory Limit: 131072K | |
Total Submissions: 48231 | Accepted: 14226 | |
Case Time Limit: 2000MS |
Description
You have N integers, A1, A2, ... , AN. You need to deal with two kinds of operations. One type of operation is to add some given number to each number in a given interval. The other is to ask for the sum of numbers in a given interval.
Input
The first line contains two numbers N and Q. 1 ≤ N,Q ≤ 100000.
The second line contains N numbers, the initial values of A1, A2, ... , AN. -1000000000 ≤ Ai ≤ 1000000000.
Each of the next Q lines represents an operation.
"C a b c" means adding c to each of Aa, Aa+1, ... , Ab. -10000 ≤ c ≤ 10000.
"Q a b" means querying the sum of Aa, Aa+1, ... , Ab.
Output
You need to answer all Q commands in order. One answer in a line.
Sample Input
10 5 1 2 3 4 5 6 7 8 9 10 Q 4 4 Q 1 10 Q 2 4 C 3 6 3 Q 2 4
Sample Output
4 55 9 15
Hint
The sums may exceed the range of 32-bit integers.
题意:
有N(1到100000)个数和M(1到100000)个操作,给出这个N个数的每个数的值Ai(-100000000到1000000000)。操作有两类,第一类Q a b,代表将[a,b]这区间内的数相加输出结果;第二类C a b c,代表将[a,b]区间内的每个数都+c处理。输出每当Q询问时候的结果。
思路:
线段树。延迟标记。将1到N进行线段树处理,用ad域来标记。
AC:
#include<stdio.h> #define MAX 250000+5 typedef struct { __int64 l; //区间起点 __int64 r; //区间终点 __int64 sum; //区间总和 __int64 ad; //标记增加量 }node; node no[MAX]; __int64 num[MAX]; __int64 s; void build(__int64 from,__int64 to,__int64 i) { __int64 mid=(from+to)/2; no[i].l=from; no[i].r=to; no[i].ad=0; //一开始增加量都为0 if(from==to) { no[i].sum=num[from]; return; } build(from,mid,2*i); build(mid+1,to,2*i+1); no[i].sum=no[i*2].sum+no[i*2+1].sum; } void find(__int64 from,__int64 to,__int64 i) { __int64 mid=(no[i].l+no[i].r)/2; if(from==no[i].l&&to==no[i].r) { if(!no[i].ad) { s+=no[i].sum; return; } else { s+=no[i].sum+(no[i].r-no[i].l+1)*no[i].ad; //这两天一直纠结这个问题 //想着如果在这里往下传标记行不行 //后来发现是忘了维护sum值了,就这点纠结了两天 //RE的原因是如果[1,1]的时候往下传就没有意义了,但是为了确保能下传,数组要开大点 //还有是+=,而不是直接= //no[i].sum+=(no[i].r-no[i].l+1)*no[i].ad; //no[2*i].ad+=no[i].ad; //no[2*i+1].as+=no[i].ad; //no[i].ad=0; //其实这里可以不需要作标记下传操作的,仅仅只是好奇一下 return; } } else { if(no[i].ad) { no[i].sum+=(no[i].r-no[i].l+1)*no[i].ad; no[2*i].ad+=no[i].ad; no[2*i+1].ad+=no[i].ad; no[i].ad=0; //传给左右儿子之后才能清零 } if(from>=mid+1) find(from,to,2*i+1); if(to<=mid) find(from,to,2*i); if(from<=mid&&to>=mid+1) { find(from,mid,2*i); find(mid+1,to,2*i+1); } } } void add(__int64 from,__int64 to,__int64 i,__int64 w) { __int64 mid=(no[i].l+no[i].r)/2; if(no[i].l==from&&no[i].r==to) { no[i].ad+=w; //当找到这个区间之后,增量标记域增加 //而不是直接总和增加 return; } else { if(no[i].ad!=0) { no[i].sum+=(no[i].r-no[i].l+1)*no[i].ad; //当下一次查询的时候,标记域的值ad使总和sum增加 no[2*i].ad+=no[i].ad; no[2*i+1].ad+=no[i].ad; //增加完后则往下传 no[i].ad=0; //传完之后本身标记的值已经使总和sum增加完毕,故变为0 } no[i].sum+=(to-from+1)*w; //这里的相加不是标记域所导致的,是所输入的区间所导致的 if(from>=mid+1) add(from,to,2*i+1,w); if(to<=mid) add(from,to,2*i,w); if(from<=mid&&to>=mid+1) { add(from,mid,2*i,w); add(mid+1,to,2*i+1,w); } //继续查找线段树,找到相同的区间则标记 } } int main() { __int64 n,m; scanf("%I64d%I64d",&n,&m); for(__int64 i=1;i<=n;i++) scanf("%I64d",&num[i]); build(1,n,1); while(m--) { char c; scanf(" %c",&c); if(c=='Q') { __int64 from,to; scanf("%I64d%I64d",&from,&to); s=0; find(from,to,1); printf("%I64d\n",s); } if(c=='C') { __int64 from,to; __int64 w; scanf("%I64d%I64d%I64d",&from,&to,&w); add(from,to,1,w); } } return 0; }
总结:
1.维护ad的同时也要维护sum值,记住;
2.注意数据范围;