|
|
|
|
package class37;
|
|
|
|
|
|
|
|
|
|
// 来自字节
|
|
|
|
|
// 扑克牌中的红桃J和梅花Q找不到了,为了利用剩下的牌做游戏,小明设计了新的游戏规则
|
|
|
|
|
// 1) A,2,3,4....10,J,Q,K分别对应1到13这些数字,大小王对应0
|
|
|
|
|
// 2) 游戏人数为2人,轮流从牌堆里摸牌,每次摸到的牌只有“保留”和“使用”两个选项,且当前轮必须做出选择
|
|
|
|
|
// 3) 如果选择“保留”当前牌,那么当前牌的分数加到总分里,并且可以一直持续到游戏结束
|
|
|
|
|
// 4) 如果选择“使用”当前牌,那么当前牌的分数*3,加到总分上去,但是只有当前轮,下一轮,下下轮生效,之后轮效果消失。
|
|
|
|
|
// 5) 每一轮总分大的人获胜
|
|
|
|
|
// 假设小明知道每一轮对手做出选择之后的总分,返回小明在每一轮都赢的情况下,最终的最大分是多少
|
|
|
|
|
// 如果小明怎么都无法保证每一轮都赢,返回-1
|
|
|
|
|
public class Code02_GameForEveryStepWin {
|
|
|
|
|
|
|
|
|
|
// public static max(int[] cands, int[] sroces) {
|
|
|
|
|
// return f(cands, sroces, 0, 0, 0, 0);
|
|
|
|
|
// }
|
|
|
|
|
|
|
|
|
|
// 当前来到index位置,牌是cands[index]值
|
|
|
|
|
// 对手第i轮的得分,sroces[i]
|
|
|
|
|
// int hold : i之前保留的牌的总分
|
|
|
|
|
// int cur : 当前轮得到的,之前的牌只算上使用的效果,加成是多少
|
|
|
|
|
// int next : 之前的牌,对index的下一轮,使用效果加成是多少
|
|
|
|
|
// 返回值:如果i...最后,不能全赢,返回-1
|
|
|
|
|
// 如果i...最后,能全赢,返回最后一轮的最大值
|
|
|
|
|
|
|
|
|
|
// index -> 26种
|
|
|
|
|
// hold -> (1+2+3+..13) -> 91 -> 91 * 4 - (11 + 12) -> 341
|
|
|
|
|
// cur -> 26
|
|
|
|
|
// next -> 13
|
|
|
|
|
// 26 * 341 * 26 * 13 -> ? * (10 ^ 5)
|
|
|
|
|
public static int f(int[] cands, int[] sroces, int index, int hold, int cur, int next) {
|
|
|
|
|
if (index == 25) { // 最后一张
|
|
|
|
|
int all = hold + cur + cands[index] * 3;
|
|
|
|
|
if (all <= sroces[index]) {
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
return all;
|
|
|
|
|
}
|
|
|
|
|
// 不仅最后一张
|
|
|
|
|
// 保留
|
|
|
|
|
int all1 = hold + cur + cands[index];
|
|
|
|
|
int p1 = -1;
|
|
|
|
|
if (all1 > sroces[index]) {
|
|
|
|
|
p1 = f(cands, sroces, index + 1, hold + cands[index], next, 0);
|
|
|
|
|
}
|
|
|
|
|
// 爆发
|
|
|
|
|
int all2 = hold + cur + cands[index] * 3;
|
|
|
|
|
int p2 = -1;
|
|
|
|
|
if (all2 > sroces[index]) {
|
|
|
|
|
p2 = f(cands, sroces, index + 1, hold, next + cands[index] * 3, cands[index] * 3);
|
|
|
|
|
}
|
|
|
|
|
return Math.max(p1, p2);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// 26 * 341 * 78 * 39 = 2 * (10 ^ 7)
|
|
|
|
|
public static int process(int[] cards, int[] scores, int index, int hold, int cur, int next) {
|
|
|
|
|
if (index == 25) {
|
|
|
|
|
int all = hold + cur + cards[index] * 3;
|
|
|
|
|
if (all > scores[index]) {
|
|
|
|
|
return all;
|
|
|
|
|
} else {
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
int d1 = hold + cur + cards[index];
|
|
|
|
|
int p1 = -1;
|
|
|
|
|
if (d1 > scores[index]) {
|
|
|
|
|
p1 = process(cards, scores, index + 1, hold + cards[index], next, 0);
|
|
|
|
|
}
|
|
|
|
|
int d2 = hold + cur + cards[index] * 3;
|
|
|
|
|
int p2 = -1;
|
|
|
|
|
if (d2 > scores[index]) {
|
|
|
|
|
p2 = process(cards, scores, index + 1, hold, next + cards[index] * 3, cards[index] * 3);
|
|
|
|
|
}
|
|
|
|
|
return Math.max(p1, p2);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
// cur -> 牌点数 -> * 3 之后是效果
|
|
|
|
|
// next -> 牌点数 -> * 3之后是效果
|
|
|
|
|
public static int p(int[] cands, int[] sroces, int index, int hold, int cur, int next) {
|
|
|
|
|
if (index == 25) { // 最后一张
|
|
|
|
|
int all = hold + cur * 3 + cands[index] * 3;
|
|
|
|
|
if (all <= sroces[index]) {
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
return all;
|
|
|
|
|
}
|
|
|
|
|
// 不仅最后一张
|
|
|
|
|
// 保留
|
|
|
|
|
int all1 = hold + cur * 3 + cands[index];
|
|
|
|
|
int p1 = -1;
|
|
|
|
|
if (all1 > sroces[index]) {
|
|
|
|
|
p1 = f(cands, sroces, index + 1, hold + cands[index], next, 0);
|
|
|
|
|
}
|
|
|
|
|
// 爆发
|
|
|
|
|
int all2 = hold + cur * 3 + cands[index] * 3;
|
|
|
|
|
int p2 = -1;
|
|
|
|
|
if (all2 > sroces[index]) {
|
|
|
|
|
p2 = f(cands, sroces, index + 1, hold, next + cands[index], cands[index]);
|
|
|
|
|
}
|
|
|
|
|
return Math.max(p1, p2);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// 改出动态规划,记忆化搜索!
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
}
|