Different GCD Subarray Query
Time Limit: 6000/3000 MS (Java/Others) Memory Limit: 65536/65536 K (Java/Others)
Total Submission(s): 114 Accepted Submission(s): 29
Total Submission(s): 114 Accepted Submission(s): 29
Problem Description
This is a simple problem. The teacher gives Bob a list of problems about GCD (Greatest Common Divisor). After studying some of them, Bob thinks that GCD is so interesting. One day, he comes up with a new problem about GCD. Easy as it looks, Bob cannot figure it out himself. Now he turns to you for help, and here is the problem:
Given an array a of N positive integers a1,a2,⋯aN−1,aN ; a subarray of a is defined as a continuous interval between a1 and aN . In other words, ai,ai+1,⋯,aj−1,aj is a subarray of a , for 1≤i≤j≤N . For a query in the form (L,R) , tell the number of different GCDs contributed by all subarrays of the interval [L,R] .
Given an array a of N positive integers a1,a2,⋯aN−1,aN ; a subarray of a is defined as a continuous interval between a1 and aN . In other words, ai,ai+1,⋯,aj−1,aj is a subarray of a , for 1≤i≤j≤N . For a query in the form (L,R) , tell the number of different GCDs contributed by all subarrays of the interval [L,R] .
Input
There are several tests, process till the end of input.
For each test, the first line consists of two integers N and Q , denoting the length of the array and the number of queries, respectively. N positive integers are listed in the second line, followed by Q lines each containing two integers L,R for a query.
You can assume that
1≤N,Q≤100000
1≤ai≤1000000
For each test, the first line consists of two integers N and Q , denoting the length of the array and the number of queries, respectively. N positive integers are listed in the second line, followed by Q lines each containing two integers L,R for a query.
You can assume that
1≤N,Q≤100000
1≤ai≤1000000
Output
For each query, output the answer in one line.
Sample Input
5 3 1 3 4 6 9 3 5 2 5 1 5
Sample Output
6 6 6
Source
2016 ACM/ICPC Asia Regional Dalian Online
题意:长度n的序列, m个询问区间[L, R], 问区间内的所有子段的不同GCD值有多少种.
题解:考虑固定左端点的不同GCD值,只有不超过logA种, 所以事件点只有nlogA个. 那么离线处理, 按照区间右端点排序从小到大处理询问,
用一个树状数组维护每个GCD值的最大左端点位置即可. 复杂度是O(nlogAlogn).
这份题解里有两个难点:
1、如何快速的离线化处理出固定的左端点的gcd;
2 如何用树状数组维护最大左端点,又如何求出答案??
第一个问题:
很显然,若是平常的离散处理,时间复杂度为n^2;
for(int i=1;i<=n;i++)
{
int x=num[i],y=i;
for(int j=0;j<gg[i-1].size();j++)
{
int cc=__gcd(gg[i-1][j].first,x);
if(x!=cc)
{
gg[i].push_back(make_pair(x,y));
x=cc,y=gg[i-1][j].second;
}
}
gg[i].push_back(make_pair(x,y));
}
这个方式我也是第一次见,将固定的左端点和得出的gcd同存在右端点,这样的确可以做到nlongA的时间离线处理。
而且这种方式存储的gcd值是从小到大的,所以不会有重存的情况出现。
主要是我不太习惯在vector 数组里用pair ,真是汗颜。。。
如何维护最大左端点呢??
首先对查询进行右端点排序。
从左到右遍历一遍节点的gcd值。
并在树状数组里面更新相同gcd的最由端点。
代码:
#include<bits/stdc++.h>
using namespace std;
const int maxn=100100;
int c[maxn];
vector< pair<int,int> >gg[maxn];
int num[maxn],vis[maxn*10],res[maxn];
struct node
{
int l,r,id;
bool operator<(const node&p)const
{
return r<p.r;
}
}A[maxn];
int lowbit(int x)
{
return x&(-x);
}
void add(int x,int value)
{
while(x<maxn)
{
c[x]+=value;
x+=lowbit(x);
}
}
int sum(int x)
{
int ans=0;
while(x)
{
ans+=c[x];
x-=lowbit(x);
}
return ans;
}
int main()
{
int n,q;
while(scanf("%d%d",&n,&q)!=EOF)
{
for(int i=1;i<=n;i++)
{
scanf("%d",&num[i]);
gg[i].clear();
}
memset(c,0,sizeof(c));
for(int i=1;i<=n;i++)
{
int x=num[i],y=i;
for(int j=0;j<gg[i-1].size();j++)
{
int cc=__gcd(gg[i-1][j].first,x);
if(x!=cc)
{
gg[i].push_back(make_pair(x,y));
x=cc,y=gg[i-1][j].second;
}
}
gg[i].push_back(make_pair(x,y));
}
for(int i=1;i<=q;i++)
{
int l,r;
scanf("%d%d",&l,&r);
A[i].l=l,A[i].r=r;
A[i].id=i;
}
sort(A+1,A+q+1);
int len=1;
memset(vis,0,sizeof(vis));
for(int i=1;i<=n;i++)
{
for(int j=0;j<gg[i].size();j++)
{
int c1=gg[i][j].first;
int c2=gg[i][j].second;
if(vis[c1])
add(vis[c1],-1);
vis[c1]=c2;
add(c2,1);
}
while(A[len].r==i)
{
res[A[len].id]=sum(A[len].r)-sum(A[len].l-1);
len++;
}
}
for(int i=1;i<=q;i++)
printf("%d\n",res[i]);
}
return 0;
}