Description
You are given a tuple generator f ( k ) = ( f 1 ( k ) , f 2 ( k ) , ⋯   , f n ( k ) ) f^{(k)}=(f^{(k)}_1,f^{(k)}_2,\cdots,f^{(k)}_n) f(k)=(f1(k),f2(k),⋯,fn(k)), where f i ( k ) = ( a i ⋅ f i ( k − 1 ) + b i ) m o d    p i f^{(k)}_i=(a_i⋅f^{(k−1)}_i+b_i)\mod p_i fi(k)=(ai⋅fi(k−1)+bi)modpi and f ( 0 ) = ( x 1 , x 2 , ⋯   , x n ) f^{(0)}=(x_1,x_2,\cdots,x_n) f(0)=(x1,x2,⋯,xn). Here x m o d    y x\mod y xmody denotes the remainder of x x x when divided by y y y. All p i p_i pi are primes.
One can see that with fixed sequences x i , y i , a i x_i, y_i, a_i xi,yi,ai the tuples $f^{(k)} starting from some index will repeat tuples with smaller indices. Calculate the maximum number of different tuples (from all f ( k ) f^{(k)} f(k) for k ≥ 0 k≥0 k≥0) that can be produced by this generator, if x i , a i , b i x_i, a_i, b_i xi,ai,bi are integers in the range [ 0 , p i − 1 ] [0,p_i−1] [0,pi−1] and can be chosen arbitrary. The answer can be large, so print the remainder it gives when divided by 1 0 9 + 7 10^9+7 109+7
Input
The first line contains one integer n n n ( 1 ≤ n ≤ 2 × 105 ) (1≤n≤2\times105) (1≤n≤2×105) — the number of elements in the tuple.
The second line contains n n n space separated prime numbers — the modules p 1 , p 2 , ⋯   , p n p_1,p_2,\cdots,p_n p1,p2,⋯,pn ( 2 ≤ p i ≤ 2 × 1 0 6 ) (2≤p_i≤2\times10^6) (2≤pi≤2×106).
Output
Print one integer — the maximum number of different tuples modulo 1 0 9 + 7 10^9+7 109+7.
Examples
Input
4
2 3 5 7
Output
210
Input
3
5 3 3
Output
30
Note
In the first example we can choose next parameters: a = [ 1 , 1 , 1 , 1 ] , b = [ 1 , 1 , 1 , 1 ] , x = [ 0 , 0 , 0 , 0 ] a=[1,1,1,1], b=[1,1,1,1], x=[0,0,0,0] a=[1,1,1,1],b=[1,1,1,1],x=[0,0,0,0], then f i ( k ) = k m o d    p i f^{(k)}_i=k\mod p_i fi(k)=kmodpi.
In the second example we can choose next parameters: a = [ 1 , 1 , 2 ] , b = [ 1 , 1 , 0 ] , x = [ 0 , 0 , 1 ] a=[1,1,2], b=[1,1,0], x=[0,0,1] a=[1,1,2],b=[1,1,0],x=[0,0,1].
感觉整道题非常玄学。在围观大佬题解和官方题解后,大概明白了啥意思。这个函数图像是一个圆圈还有圈外的一部分。最后答案是
l
c
m
i
=
1
⋯
n
(
c
i
)
lcm_{i=1\cdots n}(c_i)
lcmi=1⋯n(ci)+
max
i
=
1
⋯
n
(
p
p
i
)
\max\limits_{i=1\cdots n}(pp_i)
i=1⋯nmax(ppi)。中间噼里啪啦一堆公式看的我酸爽,反正就证明了
max
i
=
1
⋯
n
(
p
p
i
)
=
1
\max\limits_{i=1\cdots n}(pp_i)=1%
i=1⋯nmax(ppi)=1以及我们需要
p
i
−
1
p_i-1
pi−1 的因子。然后就是求一个
l
c
m
lcm
lcm,如果
p
i
p_i
pi 加过了就加
p
i
−
1
p_i-1
pi−1,然后就稀里糊涂写(抄)一遍就过了。
#include<cstdio>
#include<algorithm>
using namespace std;
const int N=2e6+10;
const int mod=1e9+7;
int prime[N],pp,pri[N],ans=1,p[N],n,cnt[N],mx[N];
bool iscomp[N];
void primetable()
{
for(int i=2;i<N;i++)
{
if(!iscomp[i])prime[pp++]=i,pri[i]=i;
for(int j=0;j<pp&&1ll*i*prime[j]<N;j++)
{
iscomp[i*prime[j]]=1;
pri[i*prime[j]]=prime[j];
if(i%prime[j]==0)break;
}
}
}
int main()
{
//freopen("in.txt","r",stdin);
primetable();
scanf("%d\n",&n);
for(int i=1;i<=n;i++)scanf("%d",&p[i]);
sort(p+1,p+n+1);
for(int i=n;i;i--)
{
if(!cnt[p[i]])cnt[p[i]]++,mx[p[i]]=1,ans=(1ll*ans*p[i])%mod;
else
{
int v=--p[i];
while(v!=1)
{
int u=pri[v],ci=0;
while(v%u==0)v/=u,ci++;
if(cnt[u]<ci)
{
for(int j=cnt[u]+1;j<=ci;j++)ans=(1ll*ans*u)%mod;
cnt[u]=ci;mx[u]=1;
}
else mx[u]+=(cnt[u]==ci);
}
}
}
for(int i=1;i<=n;i++)
{
int v=p[i],flag=0;
while(v!=1)
{
int u=pri[v],ci=0;
while(v%u==0)v/=u,ci++;
if(cnt[u]==ci&&mx[u]==1)flag=1;
}
if(!flag)
{
ans++;break;
}
}
printf("%d\n",ans);
return 0;
}
总结
玄学数论,还有那个啥拉格朗日定理是啥……