The Parallel Challenge Ballgame

Time Limit: 2000/1000 MS (Java/Others) Memory Limit: 65536/32768 K (Java/Others)
Total Submission(s): 547 Accepted Submission(s): 26

Problem Description
Before the ACM/ICPC world final 2005, there is a competition called “The Parallel Challenge Ballgame”. The Parallel Challenge ballgame gives each team a chance to pit their programming skills against those of other teams in a fast-moving parallel-programming game of skill, treachery, and hazard avoidance. Each team will write a single C++ class named MyPlayer which defines the characteristics of a “player”. The MyPlayer class will be instantiated five times in the environment, making up a five-player team, which will then compete in a series of Parallel Challenge ballgames running on an IBM Power Architecture Blue Gene supercomputer – the world’s fastest computer.

A Parallel Challenge ballgame is played on a rectangular filed. The filed is surrounded by a wall; balls will bounce off the walls if they run into it. The rule of bouncing is the same as light (In figure 1,angle 1 equals angle 2). Near the edges of the fields are a number of goals where points can be scored. Goals are rectangular areas lying near the edges of the field but within the field boundaries. When the game starts there are a number of balls placed at random locations on the field. A player can move to a ball, pick it up, and throw (of course, it is not football, why not use hand?) it. At the start of each game there are also a number of “nets” distributed at various locations on the edges of the field. A player can move to and pick up one of these nets, and can then use them to “trap” players on other teams by throwing the net on top of them. Once a player is trapped beneath a net, that player cannot do anything more in the game until a teammate comes and lifts the net from the trapped player. A player may “tackle” another player, normally in an attempt to dislodge a ball being carried by the other player (although it is also legal to tackle a player who is not carrying a ball).

The objective of each team is to write their MyPlayer class so that their players (the five instances of the class) operate in a coordinated fashion, taking advantage of the various ways to score points while at the same time avoiding both hazards on the game filed and impediments thrown at them by players from other teams. The winner of a game is the team whose players score the largest total number of points.
There are many ways to score points:
(1) Successful Tackle (tackle not caught by Referees)
(2) Opponent’s Failed Tackle
(3) Throwing a net on one or more opponents
(4) Lifting a net off a teammate
And of course, the normal approach: (5) Carry or throw the balls into goals. This is also the easiest way to score points. In order to get more chance to make a ball into the goal, it is better to throw a ball to a goal once you get it. The ball may be blocked by other player, but the probability is low, because a thrown ball moves at the maximum speed allowed in the game, and the player can not catch up with it. So the only way it is blocked is that some player is just on the direction, which the ball moves. And because the ball will bounce off the wall when it hits the wall, it is not necessary to throw a ball straight to a goal (see figure 2). But the more times the ball bounces off the wall, the higher probability that other player will head off it. So we only consider the ball bounces off the wall no more than once.

Here is our problem. Given the range of the field, the position of the ball and the goals, the size of the goals, your task is to calculate how many percents of the direction that the team can score points through method (5).

Input
In the first line of input, there is an integer t, which is the number of test cases and followed by the data for the test cases. The first line of each test case contains four integers: x1, y1, x2, y2, and (x1, y1) and (x2, y2) (-1000 <= x1, y1, x2, y2 <= 1000) are the coordinates of the diagonally points of the field. In the next line, there are two integers x and y, and (x, y) is the coordinate of the ball. The third line of each test case contains an integer n (0 <= n <= 100), which is the number of the goals. In next n lines, each line contains four integers: xi1, yi1, xi2, yi2, and (xi1, yi1) and (xi2, yi2) are the coordinates of the diagonally points of the i-th goal. You may assume the goals and the ball are inside the field. And if a ball move into or on the boundaries of a goal, the team scores points.

Output
The output contains one line per test case containing a number, which is described above and followed a “%”. The number should be rounded up to two decimal digits. See the Sample Output to know the exact format.

Sample Input
1
100 100 -100 -100
0 0
1
10 10 -10 20

Sample Output
28.34%

  当这道代表了我与ACM在四年里多次分分合合的题目终于被AC时,我的心情是激动而茫然的。

  

  题意:一个长方形球场,其中有一些长方形区域,和一个球。把球从初始位置沿直线打出,在与球场边缘碰撞最多一次后若能经过某个长方形区域则可以得分。求可以得分的发射角度占总的角度2π的百分比,触壁反弹的规则同光反射。

  首先,把每个小长方形根据球场4条边做轴对称,得到一共5个小长方形。这样操作后,完全可以假定不允许反弹。

  然后对每个小长方形,算出可以打中的角度:

    如果球就在小长方形里,可以不用计算,直接输出100%。

    其他情况球和小长方形有8中位置关系和相应的角度范围:

  将5n个小长方形的角度全部算出来之后,求出它们的并的大小即可。

  其实这道题真的不难,配不上这样的通过率。

  理解题意点有困难。

  然后主要就是精度问题吧。三角函数是肯定要用的,但是一定要少用,推荐这个atan2。对double类型进行任何逻辑判断都要用eps=1e-8来代替0。

  真的不难,被这样一个纸老虎虐了,惭愧。

#include <math.h>
#include <stdio.h>
#define eps (1e-8)
#define pi 3.1415926536
struct rectangle {
double x1, x2, y1, y2;
} b, g[];
double x, y, l[], r[];
void adjust(rectangle & r) {
double t;
if (r.x1 - r.x2 > eps) {
t = r.x1;
r.x1 = r.x2;
r.x2 = t;
}
if (r.y1 - r.y2 > eps) {
t = r.y1;
r.y1 = r.y2;
r.y2 = t;
}
}
int pos(rectangle a) {
if ((a.x1 - x <= eps) && (x - a.x2 <= eps) && (a.y1 - y <= eps) && (y - a.y2 <= eps)) {
return ;
}
if (x - a.x1 < eps) {
if (y > a.y2) {
return ;
} else if (y - a.y1 < eps) {
return ;
} else {
return ;
}
} else if (x - a.x2 > eps) {
if (y > a.y2) {
return ;
} else if (y - a.y1 < eps) {
return ;
} else {
return ;
}
} else {
if (y - a.y2 > eps) {
return ;
} else {
return ;
}
}
}
int main() {
int T, n;
scanf("%d", &T);
while (T--) {
scanf("%lf%lf%lf%lf", &b.x1, &b.y1, &b.x2, &b.y2);
adjust(b);
scanf("%lf%lf%d", &x, &y, &n);
for (int i = ; i <= n; i++) {
scanf("%lf%lf%lf%lf", &g[i].x1, &g[i].y1, &g[i].x2, &g[i].y2);
adjust(g[i]);
g[i + n].x1 = g[i].x1;
g[i + n].y1 = * b.y1 - g[i].y1;
g[i + n].x2 = g[i].x2;
g[i + n].y2 = * b.y1 - g[i].y2;
adjust(g[i + n]);
g[i + * n].x1 = g[i].x1;
g[i + * n].y1 = * b.y2 - g[i].y1;
g[i + * n].x2 = g[i].x2;
g[i + * n].y2 = * b.y2 - g[i].y2;
adjust(g[i + * n]);
g[i + * n].x1 = * b.x1 - g[i].x1;
g[i + * n].y1 = g[i].y1;
g[i + * n].x2 = * b.x1 - g[i].x2;
g[i + * n].y2 = g[i].y2;
adjust(g[i + * n]);
g[i + * n].x1 = * b.x2 - g[i].x1;
g[i + * n].y1 = g[i].y1;
g[i + * n].x2 = * b.x2 - g[i].x2;
g[i + * n].y2 = g[i].y2;
adjust(g[i + * n]);
}
n *= ;
int m = ;
bool flag = false;
double a1, a2;
for (int i = ; i <= n; i++) {
int tmp = pos(g[i]);
if (tmp == ) {
flag = true;
break;
} else if (tmp == ) {
m++;
a1 = atan2(y - g[i].y2, x - g[i].x1);
a2 = atan2(y - g[i].y2, g[i].x2 - x);
l[m] = pi + a1;
r[m] = * pi - a2;
} else if (tmp == ) {
m++;
a1 = atan2(y - g[i].y2, x - g[i].x1);
a2 = atan2(x - g[i].x2, y - g[i].y1);
l[m] = pi + a1;
r[m] = 1.5 * pi - a2;
} else if (tmp == ) {
m++;
a1 = atan2(x - g[i].x2, g[i].y2 - y);
a2 = atan2(x - g[i].x2, y - g[i].y1);
l[m] = 0.5 * pi + a1;
r[m] = 1.5 * pi - a2;
} else if (tmp == ) {
m++;
a1 = atan2(x - g[i].x2, g[i].y2 - y);
a2 = atan2(g[i].y1 - y, x - g[i].x1);
l[m] = 0.5 * pi + a1;
r[m] = pi - a2;
} else if (tmp == ) {
m++;
a1 = atan2(g[i].y1 - y, g[i].x2 - x);
a2 = atan2(g[i].y1 - y, x - g[i].x1);
l[m] = a1;
r[m] = pi - a2;
} else if (tmp == ) {
m++;
a1 = atan2(g[i].y1 - y, g[i].x2 - x);
a2 = atan2(g[i].x1 - x, g[i].y2 - y);
l[m] = a1;
r[m] = 0.5 * pi - a2;
} else if (tmp == ) {
m++;
a1 = atan2(g[i].y2 - y, g[i].x1 - x);
l[m] = ;
r[m] = a1;
m++;
a2 = atan2(y - g[i].y1, g[i].x1 - x);
l[m] = * pi - a2;
r[m] = * pi;
} else {
m++;
a1 = atan2(g[i].x1 - x, y - g[i].y1);
a2 = atan2(y - g[i].y2, g[i].x2 - x);
l[m] = 1.5 * pi + a1;
r[m] = * pi - a2;
}
}
if (flag) {
printf("100.00%%\n");
} else {
double tot = , t, ll, rr;
for (int i = ; i < n; i++) {
for (int j = i + ; j <= n; j++) {
if (l[i] - l[j] > eps) {
t = l[i];
l[i] = l[j];
l[j] = t;
t = r[i];
r[i] = r[j];
r[j] = t;
}
}
}
ll = l[];
rr = r[];
m++;
l[m] = ;
r[m] = ;
for (int i = ; i <= m; i++) {
if (l[i] - rr <= eps) {
if (rr - r[i] < eps) {
rr = r[i];
}
} else {
tot += (rr - ll);
ll = l[i];
rr = r[i];
}
}
printf("%.2lf%%\n", tot * / pi);
}
}
return ;
}

The Parallel Challenge Ballgame[HDU1101]的更多相关文章

  1. 大规模视觉识别挑战赛ILSVRC2015各团队结果和方法 Large Scale Visual Recognition Challenge 2015

    Large Scale Visual Recognition Challenge 2015 (ILSVRC2015) Legend: Yellow background = winner in thi ...

  2. Introduction to Parallel Computing

    Copied From:https://computing.llnl.gov/tutorials/parallel_comp/ Author: Blaise Barney, Lawrence Live ...

  3. Parallel I/O and Columnar Storage

    Parallel I/O and Columnar Storage We begin with a high level overview of the system while follow up ...

  4. Massively parallel supercomputer

    A novel massively parallel supercomputer of hundreds of teraOPS-scale includes node architectures ba ...

  5. A Pattern Language for Parallel Application Programming

    A Pattern Language for Parallel Application Programming Berna L. Massingill, Timothy G. Mattson, Bev ...

  6. .Net多线程编程—System.Threading.Tasks.Parallel

    System.Threading.Tasks.Parallel类提供了Parallel.Invoke,Parallel.For,Parallel.ForEach这三个静态方法. 1 Parallel. ...

  7. Java 8函数编程轻松入门(五)并行化(parallel)

    1.并发与并行的区别 并发: 一个时间段内有几个程序都处于已启动到运行完毕之间,且这几个程序都是在同一个处理机上运行.但在任一个时刻点只有一个程序在处理机上运行 并行: 在同一个时刻,多核处理多个任务 ...

  8. Parallel并行之乱用

    关于Parallel我也不细说了,一则微软封装的很好用,二来介绍这个的遍地都是. 我要说的是,要想成为一个优秀的标题党,一定要把重点放到别的地方,为了节省大家阅读时间,我先把结论说了,然后再慢慢从头说 ...

  9. 代码的坏味道(12)——平行继承体系(Parallel Inheritance Hierarchies)

    坏味道--平行继承体系(Parallel Inheritance Hierarchies) 平行继承体系(Parallel Inheritance Hierarchies) 其实是 霰弹式修改(Sho ...

随机推荐

  1. oracle学习不错的网站

    http://oracle-base.com/articles/linux/rlwrap.php

  2. 让ASP.NET MVC不使用jsonp也可以跨域访问

    跨域问题仅仅发生在Javascript发起AJAX调用,或者Silverlight发起服务调用时,其根本原因是因为浏览器对于这两种请求,所给予的权限是较低的,通常只允许调用本域中的资源,除非目标服务器 ...

  3. Android中Service 使用详解(LocalService + RemoteService)

    Service 简介: Service分为本地服务(LocalService)和远程服务(RemoteService): 1.本地服务依附在主进程上而不是独立的进程,这样在一定程度上节约了资源,另外L ...

  4. POJ2778 DNA Sequence(AC自动机 矩阵)

    先使用AC自动机求得状态转移关系,再建立矩阵,mat[i][j]表示一步可从i到j且i,j节点均非终止字符的方案数,则此矩阵的n次方表示n步从i,到j的方法数. #include<cstdio& ...

  5. Multiple types were found that match the controller named 'Home'. (weird error)

    found the error, because I changed the namespace and assembly name, then on the bin folder the old d ...

  6. 用康托展开实现全排列(STL、itertools)

    康拓展开: $X=a_n*(n-1)!+a_{n-1}*(n-2)!+\ldots +a_2*1!+a_1*0!$ X=an*(n-1)!+an-1*(n-2)!+...+ai*(i-1)!+...+ ...

  7. windows多线程详解

    转自:http://blog.csdn.net/zhouxuguang236/article/details/7775232 在一个牛人的博客上看到了这篇文章,所以就转过来了,地址是http://bl ...

  8. HR外包系统 - 客户员工 发薪需求/个税需求 设置

    最好,客户公司层面进行设置,如果单一情况,只需要设置公司,如果不是单一情况,设置员工, 另外员工只能从公司设置好的地方选择过来. 增强系统简便设置和设置的灵活性.

  9. 设定自动获得DNS服务器地址

    情况说明:操作系统是Win7 64位, 网络是有线 1 2 3 4 5

  10. C# 对象转换为byte[] ,byte[]还原对象

    /// <summary>  /// 将一个object对象序列化,返回一个byte[]          /// </summary>  /// <param name ...