ConcentrationAI (CI翻译)
Problem Statement
Takahashi is playing a solitaire game similar to the memory-matching game.
There are $2N$ cards. A number is written on the front of each card, and nothing is written on the back.
For each $i$ satisfying $1 \leq i \leq N$, there are exactly two cards on which $A_i$ is written. (The $A_i$ are pairwise distinct.)
Takahashi plays the game with these cards using the following procedure.
- Shuffle the
$2N$cards and lay them out face down. - Set life to
$L$and score to$0$. - Repeat the following until life becomes
$0$or there are no cards left on the table: - Choose one face-down card on the table, turn it face up, and check the number
$X$written on it.- Choose one face-down card on the table, turn it face up, and check the number
$Y$written on it. - If
$X=Y$, remove those two cards from the table, and increase the score by$X$. - If
$X\neq Y$, turn the two cards face down again, and decrease life by$1$.
- Choose one face-down card on the table, turn it face up, and check the number
Find the expected value of the score at the end of the game when Takahashi acts optimally to maximize the score at the end of the game.
Below is a more formal description of the game.
- Takahashi knows the rules of this game described below.
- Takahashi knows the values of
$A_1,\dots,A_N$. - Let
$B$be a sequence obtained by choosing a permutation of the length-$2N$sequence$(A_1,A_1,A_2,A_2,\ldots,A_N,A_N)$uniformly at random. - Initially, Takahashi knows none of the values of
$B$. Once he learns the value of$B_i$, he remembers it completely thereafter. - Let life be
$L$, score be$0$, and$S$be$\{1,2,3,\ldots,2N\}$. Takahashi always knows these values. - Repeat the following until life becomes
$0$or$S$becomes empty: - Based on the information obtained up to this point, Takahashi chooses an element from
$S$, and calls it$i$.- The value of
$B_i$is revealed, and Takahashi learns it. - Based on the information obtained up to this point (including the value of
$B_i$), Takahashi chooses an element from$S\setminus\{i\}$, and calls it$j$. - The value of
$B_j$is revealed, and Takahashi learns it. - If
$B_i=B_j$, remove$i$and$j$from$S$, and add$B_i$to the score. - If
$B_i \neq B_j$, decrease life by$1$.
- The value of
- Takahashi acts optimally to maximize the expected value of the score at the end of the game.
Constraints
$1 \leq N \leq 200$$1 \leq L \leq 200$$1 \leq A_1 < A_2 < \dots < A_N \leq 10^5$- All input values are integers.
Input
The input is given from Standard Input in the following format:
N L
A_1 A_2 \dots A_N
Output
Output the answer.
Your output will be considered correct if its absolute or relative error from the true answer is at most $10^{-5}$.
{"titleZh":"集中力","bodyZh":"## 题目描述\n\n> 高桥正在玩一种类似于记忆配对游戏的单人游戏。\n\n有 $2N$ 张卡片。每张卡片的正面写有一个数字,背面没有任何内容。对于满足 $1 \\leq i \\leq N$ 的每个 $i$,恰好有两张卡片写着 $A_i$。($A_i$ 两两不同。)\n\n高桥按照以下流程用这些卡片进行游戏。\n\n- 将 $2N$ 张卡片洗匀,背面朝上摆放。\n- 将 生命值 设为 $L$,将 得分 设为 $0$。\n- 重复以下操作,直到生命值变为 $0$ 或桌上没有剩余卡片:\n- 选择桌上一张背面朝上的卡片,将其翻开并查看上面写的数字 $X$。\n - 选择桌上另一张背面朝上的卡片,将其翻开并查看上面写的数字 $Y$。\n - 如果 $X=Y$,则将这两张卡片从桌上移除,并将得分增加 $X$。\n - 如果 $X\\neq Y$,则将这两张卡片重新翻回背面朝上,并将生命值减少 $1$。\n\n假设高桥采取最优策略,使游戏结束时的得分最大,求游戏结束时得分的期望值。\n\n下面给出游戏的更形式化描述。\n\n- 高桥知道下述游戏规则。\n- 高桥知道 $A_1,\\dots,A_N$ 的值。\n- 令 $B$ 为序列 $(A_1,A_1,A_2,A_2,\\ldots,A_N,A_N)$ 的一个长度为 $2N$ 的排列,并且所有排列等概率选取。\n- 初始时,高桥不知道 $B$ 中的任何值。一旦得知 $B_i$ 的值,此后便会一直记住它。\n- 令生命值为 $L$,得分为 $0$,且 $S$ 为集合 $\{1,2,3,\\ldots,2N\\}$。高桥始终知道这些值。\n- 重复以下操作,直到生命值变为 $0$ 或 $S$ 变为空集:\n- 根据目前为止获得的信息,高桥从 $S$ 中选择一个元素,并将其记为 $i$。\n - 揭示 $B_i$ 的值,高桥得知该值。\n - 根据目前为止获得的信息(包括 $B_i$ 的值),高桥从 $S\\setminus\\{i\\}$ 中选择一个元素,并将其记为 $j$。\n - 揭示 $B_j$ 的值,高桥得知该值。\n - 如果 $B_i=B_j$,则从 $S$ 中移除 $i$ 和 $j$,并将 $B_i$ 加到得分中。\n - 如果 $B_i \\neq B_j$,则将生命值减少 $1$。\n- 高桥采取最优策略,使游戏结束时得分的期望值最大。\n\n## 约束条件\n\n- $1 \\leq N \\leq 200$\n- $1 \\leq L \\leq 200$\n- $1 \\leq A_1 < A_2 < \\dots < A_N \\leq 10^5$\n- 所有输入值均为整数。\n\n## 输入格式\n\n输入通过标准输入给出,格式如下:\n\n\n\n## 输出格式\n\n输出答案。\n只要答案与正确答案的绝对误差或相对误差不超过 text\nN L\nA_1 A_2 \\dots A_N\n$10^{-5}$,就会被判定为正确。"}AI (CI翻译)
样例
3 2
1 2 3
3.8666666667
5 2
2 3 5 7 101
17.8560846561
20 10
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
770.7122293087
题面与样例版权归 AtCoder 及出题人所有;中文翻译由 AI 生成(gpt-6.1-sol),如有出入请以原文为准。