把 n 个骰子扔在地上,求点数和为 s 的概率。
Input: n = 1
Output: [[1, 0.17], [2, 0.17], [3, 0.17], [4, 0.17], [5, 0.17], [6, 0.17]]
Explanation: Throw a dice, the sum of the numbers facing up may be 1, 2, 3, 4, 5, 6, and the probability of each result is 0.17.
Input: n = 2
Output: [[2,0.03],[3,0.06],[4,0.08],[5,0.11],[6,0.14],[7,0.17],[8,0.14],[9,0.11],[10,0.08],[11,0.06],[12,0.03]]
Explanation: Throw two dices, the sum of the numbers facing up may be in [2,12], and the probability of each result is different.
使用一个二维数组 dp 存储点数出现的次数,其中 dp[i][j] 表示前 i 个骰子产生点数 j 的次数。空间复杂度:O(N2)
一、动态规划:
//f(n)=f(n-1)+f(n-2)+f(n-3)+f(n-4)+f(n-5)+f(n-6)
参考:https://blog.csdn.net/K346K346/article/details/50988681
解题思路:
第一步,确定问题解的表达式。可将f(n, s) 表示n个骰子点数的和为s的排列情况总数。
第二步,确定状态转移方程。n个骰子点数和为s的种类数只与n-1个骰子的和有关。因为一个骰子有六个点数,那么第n个骰子可能出现1到6的点数。所以第n个骰子点数为1的话,f(n,s)=f(n-1,s-1),当第n个骰子点数为2的话,f(n,s)=f(n-1,s-2),…,依次类推。在n-1个骰子的基础上,再增加一个骰子出现点数和为s的结果只有这6种情况!那么有:
f(n,s)=f(n-1,s-1)+f(n-1,s-2)+f(n-1,s-3)+f(n-1,s-4)+f(n-1,s-5)+f(n-1,s-6) ,0< n<=6n
f(n,s)=0, s< n or s>6n
上面就是状态转移方程,已知初始阶段的解为:
当n=1时, f(1,1)=f(1,2)=f(1,3)=f(1,4)=f(1,5)=f(1,6)=1。
public List<Map.Entry<Integer, Double>> dicesSum(int n) {
final int face = 6;
//可以出现的总点数,6*n
final int pointNum = face * n;
//dp[i][j] 表示前 i 个骰子产生点数 j 的次数。
long[][] dp = new long[n + 1][pointNum + 1];
//初始值,1个骰子每面出现的次数是1
for (int i = 1; i <= face; i++)
dp[1][i] = 1;
for (int i = 2; i <= n; i++)
for (int j = i; j <= pointNum; j++) /* 使用 i 个骰子最小点数为 i */
for (int k = 1; k <= face && k <= j; k++)
//f(n)=f(n-1)+f(n-2)+f(n-3)+f(n-4)+f(n-5)+f(n-6)
dp[i][j] += dp[i - 1][j - k];
//出现次数的所有可能性
final double totalNum = Math.pow(6, n);
List<Map.Entry<Integer, Double>> ret = new ArrayList<>();
for (int i = n; i <= pointNum; i++)
ret.add(new AbstractMap.SimpleEntry<>(i, dp[n][i] / totalNum));
return ret;
}
二、动态规划 + 旋转数组
空间复杂度:O(N)
用一个flag数组,来存储点数之和的次数,利用1-flag来交换数组,再重复计算
public List<Map.Entry<Integer, Double>> dicesSum(int n) {
final int face = 6;
final int pointNum = face * n;
long[][] dp = new long[2][pointNum + 1];
for (int i = 1; i <= face; i++)
dp[0][i] = 1;
int flag = 1; /* 旋转标记 */
for (int i = 2; i <= n; i++, flag = 1 - flag) {
for (int j = 0; j <= pointNum; j++)
dp[flag][j] = 0; /* 旋转数组清零 */
for (int j = i; j <= pointNum; j++)
for (int k = 1; k <= face && k <= j; k++)
dp[flag][j] += dp[1 - flag][j - k];
}
final double totalNum = Math.pow(6, n);
List<Map.Entry<Integer, Double>> ret = new ArrayList<>();
for (int i = n; i <= pointNum; i++)
ret.add(new AbstractMap.SimpleEntry<>(i, dp[1 - flag][i] / totalNum));
return ret;
}
二维数组有清零,不保存之前的结果,节约空间。
其中用到了Map.Entry<Integer, Double>,Map.Entry里有相应的getKey和getValue方法,即JavaBean,让我们能够从一个项中取出Key和Value。
ret.add(new AbstractMap.SimpleEntry<>(i, dp[1 - flag][i] / totalNum));
这里面用了new AbstractMap.SimpleEntry,AbstractMap 是 Map 接口的的实现类之一,也是 HashMap, TreeMap, ConcurrentHashMap 等类的父类。
AbstractMap 提供了 Map 的基本实现,使得我们以后要实现一个 Map 不用从头开始,只需要继承 AbstractMap, 然后按需求实现/重写对应方法即可。public AbstractMap.SimpleEntry(Map.Entry<? extends K,? extends V> entry)
Creates an entry representing the same mapping as the specified entry.
it may be convenient to return arrays of SimpleEntry instances in method Map.entrySet().toArray.
https://docs.oracle.com/javase/7/docs/api/java/util/AbstractMap.SimpleEntry.html