从a、b、c中随机挑出一个，a出现的几率为20%，b出现的记录为50%，c出现的几率为70%。
求最简洁效率最高的算法，最好用php。
求最简洁效率最高的算法，最好用php。
This topic created in 4840 days ago, the information mentioned may be changed or developed.
Supplement 1 · Aug 3, 2013
把我最终的代码发出来，我遇到的情况更复杂，对象个数未知，对象权重未知。当然原理都是一样的。
$total_weightiness = 0; //符合条件内容的权重之和
$weightiness_arr = array(); //存放每个对象权重的范围
foreach ($final_rules as $key => $value) {
$weightiness_arr[$key] = $total_weightiness + $value->weightiness;
$total_weightiness += $value->weightiness;
}
$rand = rand(1,$total_weightiness);
foreach ($weightiness_arr as $key => $value) {
if($rand < $value)
return $final_rules[$key];
}
$total_weightiness = 0; //符合条件内容的权重之和
$weightiness_arr = array(); //存放每个对象权重的范围
foreach ($final_rules as $key => $value) {
$weightiness_arr[$key] = $total_weightiness + $value->weightiness;
$total_weightiness += $value->weightiness;
}
$rand = rand(1,$total_weightiness);
foreach ($weightiness_arr as $key => $value) {
if($rand < $value)
return $final_rules[$key];
}
Supplement 2 · Aug 5, 2013
又发现一种牛掰的方法：
$weight = 0;
$tempdata = array();
foreach ($data as $one) {
$weight += $one['weight'];
for ($i = 0; $i < $one['weight']; $i ++) {
$tempdata[] = $one;
}
}
$use = rand(0, $weight – 1);
$one = $tempdata[$use];
$weight = 0;
$tempdata = array();
foreach ($data as $one) {
$weight += $one['weight'];
for ($i = 0; $i < $one['weight']; $i ++) {
$tempdata[] = $one;
}
}
$use = rand(0, $weight – 1);
$one = $tempdata[$use];
Supplement 3 · Aug 5, 2013
@Mutoo 根据mutoo的方法进一步精简
$total_weightiness = 0; //符合条件内容的权重之和
foreach ($final_rules as $key => $value) {
$total_weightiness += $value->weightiness;
}
$rand = rand(1,$total_weightiness);
foreach ($final_rules as $value) {
$temp = $rand -= $value->weightiness;
if($temp < 0)
return $value;
}
$total_weightiness = 0; //符合条件内容的权重之和
foreach ($final_rules as $key => $value) {
$total_weightiness += $value->weightiness;
}
$rand = rand(1,$total_weightiness);
foreach ($final_rules as $value) {
$temp = $rand -= $value->weightiness;
if($temp < 0)
return $value;
}
 |
dndxAug 3, 2013 via iPad
1
70 + 50 + 20 > 100
这如何实现？ |
 |
iloahzAug 3, 2013
2
r = rand() % 100;
if r < 20: return a else if r < 70: return b else: return c |
 |
wang2191195Aug 3, 2013 via iPhone
3
这个要求太高了=_=
|
 |
rwxAug 3, 2013
6
自己都说是权重了，那就按权重来办啊
总权重=20+50+70 取一个 1 到总权重的随机数 看随机数落到哪个权重范围就选哪个 |
 |
windywinterAug 3, 2013
11
把2L的% 100改成%140就对了。
|
 |
jjplayAug 3, 2013
15
我觉得你是体彩中心新招聘的程序员！！！！！！！！！
|
 |
MutooAug 3, 2013
16
你需要《代码大全》第12章，表驱动法。
不需要有百分数表示，用整数就好。 1) 先求合sum 2) 产生0~sum之间的随机数 3) sum依次减去各个权重，直到sum为负数，被减的那个就是被抽中的 这个算法要求每次减的顺序是一致的，这样才可以保证概率。 其实就是把古典概型（权数）转化成几何概型（一维长度） |
 |
zurocAug 3, 2013
18
|
 |
justfinduAug 5, 2013
19
http://www.helloweba.com/view-blog-216.html
类似于这种? 表示不会贴代码~ 只是这个示例里面有一个转盘范围 min和max , v表示权重或者叫做概率~ 然后就可以抽奖了~ |
Select Language