Fill the Cisterns!

题目链接:

http://acm.hust.edu.cn/vjudge/contest/129783#problem/F

Description


During the next century certain regions on earth will experience severe water shortages. The old town of Uqbar has already started to prepare itself for the worst. Recently they created a network of pipes connecting the cisterns that distribute water in each neighbourhood, making it easier to fill them at once from a single source of water. But in case of water shortage the cisterns above a certain level will be empty since the water will to the cisterns below.
You have been asked to write a program to compute the level to which cisterns will be lled with a certain volume of water, given the dimensions and position of each cistern. To simplify we will neglect the volume of water in the pipes.
Task
Write a program which for each data set:
reads the description of cisterns and the volume of water,
computes the level to which the cisterns will be filled with the given amount of water,
writes the result.

Input


The first line of the input contains the number of data sets k, 1

Output


The output should consist of exactly d lines, one line for each data set.
Line i, 1

Sample Input


```
3
2
0 1 1 1
2 1 1 1
1
4
11 7 5 1
15 6 2 2
5 8 5 1
19 4 8 1
132
4
11 7 5 1
15 6 2 2
5 8 5 1
19 4 8 1
78
```

Sample Output


```
1.00
OVERFLOW
17.00
```

Source


2016-HUST-线下组队赛-3


##题意:

给出n个长方体水箱,从下往上依次注入V升水. 求最后结果的高度.


##题解:

对长方体的上下底面排序后直接模拟即可,维护当前高度时长方体的截面面积之和. 每次枚举到下底时把面积加入,枚举到下底时减去面积.
当枚举到某个上底时,水不够注满到这个高度,那么用剩余体积除以当前截面,就是剩下的高度.
还有种做法是二分最终高度,并遍历所有长方体计算是否能够用.

这道水题成为了今天的败笔. 2个多小时才过.
一开始就想的是二分,然后我特意处理都乘了个100来避免浮点数. 结果WA. 应该是因为可能不能恰好用完导致的.
然后改成double的还是WA. 后来重写了一份模拟,还是WA.
错点在于,一开始把rounded up错误理解成为向上取整,结果每次输出我都加了个0.005导致GG.
模拟过了之后以为二分的精度不够,这题不能用二分. 回来补题才发现输出一定只能用 %f , 不能用 %lf. (否者WA,至于精度,各种姿势处理都能过).

double输出再用%lf就吃键盘!!!!!


##代码:
####模拟:
``` cpp
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#define LL long long
#define eps 1e-8
#define maxn 50010
#define mod 100000007
#define inf 0x3f3f3f3f3f3f3f3f
#define mid(a,b) ((a+b)>>1)
#define IN freopen("in.txt","r",stdin);
using namespace std;

int n;

LL V;

struct node {

LL b,h,w,d;

}p[maxn];

bool vis[maxn];

typedef pair<LL,int> pii;

pii edge[maxn*2];

int main(int argc, char const *argv[])

{

//IN;

int t; cin >> t;
while(t--)
{
scanf("%d", &n);
int cnt = 0;
for(int i=1; i<=n; i++) {
scanf("%I64d %I64d %I64d %I64d", &p[i].b, &p[i].h, &p[i].w, &p[i].d);
edge[++cnt] = make_pair(p[i].b, i);
edge[++cnt] = make_pair(p[i].b + p[i].h, i);
}
scanf("%I64d", &V); sort(edge+1, edge+1+2*n);
memset(vis, 0, sizeof(vis));
LL area = 0;
LL ans = edge[1].first;
double last = inf;
for(int i=1; i<=2*n; i++) {
LL top = edge[i].first;
int num = edge[i].second; if((top-ans)*area >= V) {
last = (double)V / (double)area;
V = 0;
break;
} if(vis[num]) {
V -= area * (top-ans);
area -= p[num].w*p[num].d;
ans = top;
} else {
V -= area * (top-ans);
area += p[num].w*p[num].d;
vis[num] = 1;
ans = top;
}
} if(V > 0) printf("OVERFLOW\n");
else {
printf("%.2f\n", last+(double)ans);
}
} return 0;

}

####二分:

include

include

include

include

include

include

include

include

include

include

include

define LL long long

define eps 1e-8

define maxn 50010

define mod 100000007

define inf 0x3f3f3f3f

define mid(a,b) ((a+b)>>1)

define IN freopen("in.txt","r",stdin);

using namespace std;

int n;

double V;

struct node {

double b,h,w,d;

bool operator < (const node& B) const {

return b < B.b;

}

}p[maxn];

double cal(double dep) {

double ret = V;

for(int i=1; i<=n && dep>=p[i].b; i++) {

if(dep >= p[i].b + p[i].h) ret -= p[i].h * p[i].w * p[i].d;

else ret -= (dep - p[i].b) * p[i].w * p[i].d;

if(ret < 0) break;

}

return ret;

}

int main(int argc, char const *argv[])

{

//IN;

int t; cin >> t;
while(t--)
{
scanf("%d", &n);
double L = inf, R = -inf;
for(int i=1; i<=n; i++) {
scanf("%lf %lf %lf %lf", &p[i].b, &p[i].h, &p[i].w, &p[i].d);
L = min(L, p[i].b);
R = max(R, p[i].b + p[i].h);
}
scanf("%lf", &V); double tmp = V;
for(int i=1; i<=n; i++) {
tmp -= p[i].h*p[i].w*p[i].d;
if(tmp < 0) break;
} if(tmp > 0) {
printf("OVERFLOW\n");
continue;
} sort(p+1, p+1+n); double mid;
double ans = inf;
while(L <= R) {
mid = (L + R) / 2.0;
double cur = cal(mid);
if(cur <= 0) {
if(fabs(cur) < eps) ans = min(ans, mid);
R = mid - 0.001;
}
else L = mid + 0.001;
}
ans = min(ans, mid); printf("%.2f\n", ans);
} return 0;

}

POJ 1434 Fill the Cisterns! (模拟 or 二分)的更多相关文章

  1. poj1434 Fill the Cisterns!

    地址:http://poj.org/problem?id=1434 题目:Fill the Cisterns! Fill the Cisterns! Time Limit: 5000MS   Memo ...

  2. POJ 2723 Get Luffy Out(2-SAT+二分答案)

    Get Luffy Out Time Limit: 2000MS   Memory Limit: 65536K Total Submissions: 8851   Accepted: 3441 Des ...

  3. POJ 3414 Pots【bfs模拟倒水问题】

    链接: http://poj.org/problem?id=3414 http://acm.hust.edu.cn/vjudge/contest/view.action?cid=22009#probl ...

  4. poj 1008:Maya Calendar(模拟题,玛雅日历转换)

    Maya Calendar Time Limit: 1000MS   Memory Limit: 10000K Total Submissions: 64795   Accepted: 19978 D ...

  5. Poj 3233 Matrix Power Series(矩阵二分快速幂)

    题目链接:http://poj.org/problem?id=3233 解题报告:输入一个边长为n的矩阵A,然后输入一个k,要你求A + A^2 + A^3 + A^4 + A^5.......A^k ...

  6. POJ 1027 The Same Game(模拟)

    题目链接 题意 : 一个10×15的格子,有三种颜色的球,颜色相同且在同一片内的球叫做cluster(具体解释就是,两个球颜色相同且一个球可以通过上下左右到达另一个球,则这两个球属于同一个cluste ...

  7. poj 1247 The Perfect Stall 裸的二分匹配,但可以用最大流来水一下

    The Perfect Stall Time Limit: 1000MS   Memory Limit: 10000K Total Submissions: 16396   Accepted: 750 ...

  8. POJ 2010 Moo University - Financial Aid( 优先队列+二分查找)

    POJ 2010 Moo University - Financial Aid 题目大意,从C头申请读书的牛中选出N头,这N头牛的需要的额外学费之和不能超过F,并且要使得这N头牛的中位数最大.若不存在 ...

  9. POJ 3233 Matrix Power Series (矩阵+二分+二分)

    题目地址:http://poj.org/problem?id=3233 题意:给你一个矩阵A,让你求A+A^2+……+A^k模p的矩阵值 题解:我们知道求A^n我们可以用二分-矩阵快速幂来求,而 当k ...

随机推荐

  1. 深入理解Linux-hostname

    当我觉得对Linux系统下修改hostname已经非常熟悉的时候,今天碰到了几个个问题,这几个问题给我好好上了一课,很多知识点,当你觉得你已经掌握的时候,其实你了解的还只是皮毛.技术活,切勿浅尝则止! ...

  2. JAVA基础面向对象分析

    面向对象内存的分析: 一:内存的分类 1:栈(tack) 2:堆(heop) 3: 静态区 4:代码区 二:引用数据类型内存特点 三:引用数据类型传值的特点 四:引用数据类型在作为参数时的特点 面向对 ...

  3. 006 Notepad++ 运行 C/C++

    目录 0. 前言 1. 准备 2. 开工 setp 1 step 2 step 3 step 4 step 5 step 6 3. 修改与删除 3.1修改名称.快捷键 3.2 删除 4. 运行 5. ...

  4. 关于E980

    1. 浪商官网上面的内容貌似有点问题 来源: https://www.inspurpower.com/product/others.php?f=E980 但是wiki 里面的东西: 其实只有12cor ...

  5. 【监控笔记】【2.4】SQL Server中的 Ring Buffer 诊断各种系统资源压力情况

    SQL Server 操作系统(SQLOS)负责管理特定于SQL Server的操作系统资源. 其中相关的动态管理试图sys.dm_os_ring_buffers将被标识为仅供参考.不提供支持.不保证 ...

  6. spring - 第N篇 一些笔记

    1.properties文件的引入 <bean id="propertyConfigurer" class="org.springframework.beans.f ...

  7. MySQL-快速入门(7)索引

    1.什么是索引 索引是对数据库表中一列或者多列的值进行排序的一种结构.索引是在存储引擎中实现的,每种存储引擎中的索引不一定完全相同. MySQL中索引的存储类型有两种:btree和hash.MyISA ...

  8. 连连看(简单搜索)bfs

    连连看Time Limit: 20000/10000 MS (Java/Others)    Memory Limit: 65536/32768 K (Java/Others)Total Submis ...

  9. egg框架实现表单验证及获取验证的错误信息

    最近再看egg框架 是阿里今年开源的一款基于nodejs的mvc架构的web框架 感兴趣的同学可以看看 因为之前一直做php的关系 在tp,yii,laravel这些框架中对于前台发送的数据,模型里都 ...

  10. Hive常用数据库操作

    1.创建表的三种姿势 第一种 //员工表 create table if not exists default.emp( empno int, ename string, job string, mg ...