当前位置:网站首页>Game mathematical derivation AC code (high precision and low precision multiplication and division comparison) + 60 code (long long) + 20 point code (Full Permutation + deep search DFS)
Game mathematical derivation AC code (high precision and low precision multiplication and division comparison) + 60 code (long long) + 20 point code (Full Permutation + deep search DFS)
2020-11-08 07:14:00 【osc_56801rv0】
NOIP 2012 Improvement group The rematch The first day The second question is King's game game Mathematical derivation AC Code ( High precision Low precision ride except Compare )+60 Code (long long)+20 Code division ( Full Permutation + Deep search dfs)
For the general catalogue, please refer to :NOIP Improvement group The rematch test questions Catalog Xinao Calendar year
Online evaluation address :https://www.luogu.com.cn/problem/P1080
The mathematical derivation is as follows :
We analyze the two points behind the king
The queue might look like this :
* | Left | Right |
---|---|---|
king: | a0 | b0 |
p1 | a1 | b1 |
p2 | a2 | b2 |
So we can calculate that ans1=max(a0/b1,a0*a1/b2)
The queue could be like this
* | Left | Right |
---|---|---|
king: | a0 | b0 |
p2 | a2 | b2 |
p1 | a1 | b1 |
So we can calculate that ans2=max(a0/b2,a0*a2/b1)
consider ans1 The optimal , that ans1<ans2,
Please note that :
(ans1 Medium )a0/b1<(ans2 Medium )a0*a2/b1
(ans1 Medium )a0*a1/b2>(ans2 Medium )a0/b2
Make sure that ans1<ans2 It must be true , must (ans1 Medium )a0*a1/b2<(ans2 Medium )a0*a2/b1
namely a1/b2<a2/b1
namely a1*b1<a2*b2
namely , Press a*b Sort from small to large , It can be guaranteed that it is the optimal solution to the problem .
In terms of success rate , It's still highly recommended 60 Code division .
Provide a set of test data
Input :
5
9999 1
9999 1
9999 2
9999 3
9999 4
9999 5
Output :
19990001999800009999
1.AC Code ( High precision Low precision ride except Compare )
#include <bits/stdc++.h>
#define maxn 1010
#define BASE 10000
using namespace std;
int n,aa[maxn],an,bb[maxn],bn,cc[maxn],cn;
struct node{
int a,b,c;
}q[maxn];
int cmp(node a,node b){
return a.c<b.c;
}
void mul(int x){// High precision * Low precision
int i;
for(i=1;i<=an;i++)
aa[i]=aa[i]*x;
for(i=1;i<=an;i++){
aa[i+1]+=aa[i]/BASE;
aa[i]%=BASE;
}
//if(aa[an+1])an++; You can also write this sentence , But I don't think it's right
while(aa[an+1])an++,aa[an+1]+=aa[an]/BASE,aa[an]%=BASE;
}
void div(int x){// High precision / Low precision
int i,yu;
yu=0;
for(i=an;i>=1;i--){
bb[i]=(aa[i]+yu*BASE)/x;
yu=(aa[i]+yu*BASE)%x;
}
bn=an;
while(!bb[bn])bn--;// Remove the lead 0
}
int compare(){
if(bn>cn)return 1;//bb>cc
if(bn<cn)return -1;//bb<cc
for(int i=bn;i>=1;i--)//bb==cc
if(bb[i]>cc[i])return 1;//bb>cc
else if(bb[i]<cc[i])return -1;//bb<cc
return 0;//bb==cc
}
int main(){
int i,j;
scanf("%d",&n);
for(i=0;i<=n;i++)scanf("%d%d",&q[i].a,&q[i].b),q[i].c=q[i].a*q[i].b;
sort(q+1,q+1+n,cmp);
aa[1]=1,an=1;
cc[1]=q[0].a/q[1].b,cn=1;
for(i=1;i<=n;i++){
mul(q[i-1].a);
div(q[i].b);
if(compare()==1){
cn=bn;
for(j=bn;j>=1;j--)cc[j]=bb[j];
}
}
printf("%d",cc[cn]);
for(i=cn-1;i>=1;i--)printf("%04d",cc[i]);
return 0;
}
2.60 Code (long long)
#include <bits/stdc++.h>
#define LL long long
#define maxn 1100
using namespace std;
struct node{
LL a,b,c;
}q[maxn];
int cmp(node a,node b){
return a.c<b.c;
}
LL mx,a=1;
int main(){
int n,i;
scanf("%d",&n);
for(i=0;i<=n;i++){
scanf("%d%d",&q[i].a,&q[i].b);
q[i].c=q[i].a*q[i].b;
}
sort(q+1,q+1+n,cmp);
for(i=1;i<=n;i++)
a*=q[i-1].a,mx=max(mx,a/q[i].b);
printf("%lld\n",mx);
return 0;
}
3.20 Code division ( Full Permutation + Deep search dfs)
20 Code division ( Full Permutation + Deep search dfs) as follows :
#include <bits/stdc++.h>
using namespace std;
int a[15],b[15],c[15],vis[15],n,mn=2000000000;
void dfs(int step){
int i;
if(step==n+1){
int j,aa=1,mx=0;
for(j=1;j<=n;j++){
aa=aa*a[c[j-1]];
mx=max(mx,aa/b[c[j]]);
}
mn=min(mx,mn);
return;
}
for(i=1;i<=n;i++)
if(!vis[i]){
vis[i]=1;
c[step]=i;
dfs(step+1);
vis[i]=0;
}
}
int main(){
int i;
scanf("%d",&n);
for(i=0;i<=n;i++)scanf("%d%d",&a[i],&b[i]);
dfs(1);
printf("%d\n",mn);
return 0;
}
版权声明
本文为[osc_56801rv0]所创,转载请带上原文链接,感谢
边栏推荐
- QT hybrid Python development technology: Python introduction, hybrid process and demo
- Unparseable date: 'Mon Aug 15 11:24:39 CST 2016',时间格式转换异常
- VC6兼容性及打开文件崩溃问题解决
- QT hybrid Python development technology: Python introduction, hybrid process and demo
- Simple use of future in Scala
- 16.文件传输协议、vsftpd服务
- Using subprocess residue in supervisor and python multiprocessing
- Insight -- the application of sanet in arbitrary style transfer
- Ladongo open source full platform penetration scanner framework
- CPP (2) creating CPP project
猜你喜欢
Insight -- the application of sanet in arbitrary style transfer
Hand tearing algorithm - handwritten singleton mode
[original] about the abnormal situation of high version poi autosizecolumn method
ts流中的pcr与pts计算与逆运算
The real-time display of CPU and memory utilization rate by Ubuntu
The most detailed usage guide for perconaxtradbcluster8.0
FORTRAN77从文件中读入若干数据并用heron迭代公式开方
解决RabbitMQ消息丢失与重复消费问题
Brief history of computer
More than 50 object detection datasets from different industries
随机推荐
Writing method of field and field comparison condition in where condition in thinkphpp6
学习Scala IF…ELSE 语句
CPP (2) creating CPP project
SQL Server 2008R2 18456错误解决方案
Web Security (3) -- CSRF attack
Oschina plays on Sunday - before that, I always thought I was a
Judging whether paths intersect or not by leetcode
Adobe Prelude / PL 2020 software installation package (with installation tutorial)
ROS learning: remote start ROS node
接口
Goland 编写含有template的程序
leetcode之判断路径是否相交
Data transmission of asynchronous serial communication controlled by group bus communication
Insight -- the application of sanet in arbitrary style transfer
Windows下子系统Ubuntu安装
16. File transfer protocol, vsftpd service
ulab 1.0.0发布
Adobe Lightroom / LR 2021 software installation package (with installation tutorial)
Littlest jupyterhub| 02 using nbgitpuller to distribute shared files
个人短网址生成平台 自定义域名、开启防红、统计访问量