代码对我来说似乎很好,但是输出太短了,答案是肯定的,因为它应该是肯定的(从0开始,骑士应该能够游遍整个棋盘)
顺便说一句,我的positionsVisted
数组是[9] [9]的原因是因为我希望值为1-8,以匹配输出。
public class KnightChessRiddle {
static // given a chess board and a 0,0 starting point, can a knight pass through
// all the the squares without passing
// a square twice
int[][] positionsVisited = new int[9][9];
static int positionX = 1;
static int positionY = 1;
boolean stop = false;
static boolean continUe = false;
static int moveCounter = -1;
public static void main(String[] args) {
if (recursive(1, 1, 0)) {
System.out.println("yes");
}
else System.out.println("no");
}
public static boolean recursive(int x, int y, int moveType){
if (x>8||x<=0||y>8||y<=0) return false;
if (positionsVisited[x][y]==1) {
return false;
}
positionX = x;
positionY = y;
positionsVisited[positionX][positionY]++;
char c;
c='a';
switch (positionX) {
case 1:
c='a';
break;
case 2:
c='b';
break;
case 3:
c='c';
break;
case 4:
c='d';
break;
case 5:
c='e';
break;
case 6:
c='f';
break;
case 7:
c='g';
break;
case 8:
c='h';
break;
default:
break;
}
moveCounter++;
System.out.println("doing move "+moveType+" move count: "+moveCounter);
System.out.println("Knight is in "+ c +","+positionY);
try {
Thread.sleep(100);
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
if (recursive(positionX+2, positionY+1, 1)) {
return true;
}
else if (recursive(positionX+1, positionY+2, 2) ) {
return true;
}
else if (recursive(positionX+2, positionY-1, 3) ) {
return true;
}
else if (recursive(positionX+1, positionY-2, 4)) {
return true;
}
else if (recursive(positionX-2, positionY+1, 5)) {
return true;
}
else if (recursive(positionX-1, positionY+2, 6)) {
return true;
}
else if (recursive(positionX-2, positionY-1, 7)) {
return true;
}
else if (recursive(positionX-1, positionY-2, 8)) {
return true;
}
else return false;
}
这是该计划的输出:
doing move 0 move count: 0
Knight is in a,1
doing move 1 move count: 1
Knight is in c,2
doing move 1 move count: 2
Knight is in e,3
doing move 1 move count: 3
Knight is in g,4
doing move 2 move count: 4
Knight is in h,6
doing move 5 move count: 5
Knight is in f,7
doing move 1 move count: 6
Knight is in h,8
doing move 8 move count: 7
Knight is in g,6
doing move 4 move count: 8
Knight is in h,4
doing move 5 move count: 9
Knight is in f,5
doing move 2 move count: 10
Knight is in g,7
doing move 4 move count: 11
Knight is in h,5
doing move 5 move count: 12
Knight is in f,6
doing move 1 move count: 13
Knight is in h,7
doing move 5 move count: 14
Knight is in f,8
doing move 7 move count: 15
Knight is in d,7
doing move 4 move count: 16
Knight is in e,5
doing move 4 move count: 17
Knight is in f,3
doing move 2 move count: 18
Knight is in g,5
doing move 4 move count: 19
Knight is in h,3
doing move 5 move count: 20
Knight is in f,4
doing move 4 move count: 21
Knight is in g,2
doing move 7 move count: 22
Knight is in e,1
doing move 6 move count: 23
Knight is in d,3
doing move 3 move count: 24
Knight is in f,2
doing move 3 move count: 25
Knight is in h,1
doing move 6 move count: 26
Knight is in g,3
doing move 5 move count: 27
Knight is in e,4
doing move 5 move count: 28
Knight is in c,5
doing move 1 move count: 29
Knight is in e,6
doing move 5 move count: 30
Knight is in c,7
doing move 1 move count: 31
Knight is in e,8
doing move 8 move count: 32
Knight is in d,6
doing move 5 move count: 33
Knight is in b,7
doing move 1 move count: 34
Knight is in d,8
doing move 8 move count: 35
Knight is in c,6
doing move 1 move count: 36
Knight is in e,7
doing move 1 move count: 37
Knight is in g,8
no
答案 0 :(得分:4)
您实施的算法如下:
按如下方式从正方形命令可能的骑士移动:
在每次移动中,选择仍然允许的编号最小的移动。
(A)如果您覆盖所有方格,请返回
true
。(B)如果您无法再移动,请返回
false
。
您的代码存在两个问题。第一个已经指出的是你错过了支票(A)。第二个更严重的问题是这个算法不起作用。事实上,你最终会得到以下结论:
在这张照片中,黑色骑士代表开始和结束方块,而白色骑士则是所有其他方块。如果你遵循你的算法,你最终会进入一个方格,在那里你无法到达任何其他尚未被覆盖的方格。这并不意味着你不能在棋盘上进行骑士之旅,只是你的算法不起作用。
如果这确实是你想要实现的算法,那么就没有理由使用递归,因为for
循环也可以正常工作。当有效骑士从我们当前所在的方格移动时,你的函数recursive
会返回true。有两个原因,这实际上不是一个递归算法:
recursive
函数不是幂等的 - 它有副作用,即填充positionsVisited
数组的正方形。 recursive
函数调用全局变量 - positionsVisited
(我说&#39;全局变量&#39;,而不是&#39;私有字段&#39;,因为你有什么写的基本上是程序代码)。 相反,函数recursive
应该告诉你一个更为一般的信息:给定棋盘上的特定方块,以及我们不允许访问的特定方块,是否有骑士& #39;参观剩下的广场? (当然,用方形a1调用该函数和一个空的访问位置数组将告诉你是否有一个骑士游览。)该函数还可以返回一个字符串,它会告诉你什么是骑士&# 39; s tour is。
递归函数的结构应该类似于以下内容:
private boolean isKnightsTour(Square currentSquare,
int[9][9] visitedSquares,
KnightsTour tour)
{
// Append the current square to the array of visited squares.
int[9][9] newVisitedSquares = visitedSquares;
newVisitedSquares[currentSquare.getX()][currentSquare.getY()] = 1;
// If we have visited all the squares, there is a knight's tour.
// Add some code here to check for that.
if (allSquaresVisited()) {
tour = new KnightsTour(currentSquare);
return true;
}
// Test all squares a knight's move away. If you get a knight's tour,
// append the current square to the start and return that.
KnightsTour newTour;
if (isKnightsTour(currentSquare.doMove1(), newVisitedSquares, newTour) {
newTour.appendStart(currentSquare);
tour = newTour;
return true;
}
// Repeat for the other knight's moves.
else {
tour = null;
return false;
}
}
或者,用文字:
递归检查骑士从当前方块移开的所有方格,传递新方块和通过添加当前方块形成的新访问方块阵列。如果从其中一个方格中有骑士之旅,请将当前方格附加到其开头,以便从当前方块获得骑士之旅。
而不是你所写的:
通过在广场上开始并且(相当随意地)选择合法骑士在每一步的移动来递归地建立一个骑士之旅。如果我们到达一个我们无法再进行骑士行动的位置,请返回
false
。
你能看出为什么第一个(不可否认的是更复杂的)递归算法起作用而你的不是吗?
答案 1 :(得分:0)
除了“所有被访问”的问题之外,我看到至少还有一个问题,即在类字段中。当算法通过一个递归分支时,它会将一些正方形标记为已访问,并且由于此信息是类字段,当它失败当前分支并启动另一个分支时,它会看到之前尝试的所有无效信息。
如果您尝试将positionsVisited
,positionX
和positionY
作为方法参数传递并将其从类字段中删除,那么每个方法调用都会拥有自己的实际副本,该怎么办?
最终版
public class KnightChessRiddle {
private final static Map<Integer, Character> letters = new HashMap<>();
static {
letters.put(0, 'a');
letters.put(1, 'b');
letters.put(2, 'c');
letters.put(3, 'd');
letters.put(4, 'e');
letters.put(5, 'f');
}
public static void main(String[] args) {
if (recursive(0, 0, 0, new boolean[6][6], 1, "")) {
System.out.println("yes");
} else {
System.out.println("no");
}
}
private static boolean allVisited(boolean[][] positionsVisited) {
for (int i = 0; i < positionsVisited.length; i++) {
for (int j = 0; j < positionsVisited.length; j++) {
if (!positionsVisited[i][j]) {
return false;
}
}
}
return true;
}
private static boolean recursive(int positionX, int positionY, int moveType,
boolean[][] positionsVisited, int moveCounter, String currentMoves) {
// checks
if (allVisited(positionsVisited)) {
System.out.println(currentMoves);
return true;
}
if (positionX > 5 || positionX < 0 || positionY > 5 || positionY < 0) {
return false;
}
if (positionsVisited[positionX][positionY]) {
return false;
}
// make move
positionsVisited[positionX][positionY] = true;
char c = letters.get(positionX);
currentMoves += "" + c + (positionY + 1) + " (move type: " + (moveType + 1) + ")\r\n";
if (recursive(positionX + 2, positionY + 1, 1, cloneArray(positionsVisited), moveCounter + 1, currentMoves)) {
return true;
} else if (recursive(positionX + 1, positionY + 2, 2, cloneArray(positionsVisited), moveCounter + 1, currentMoves)) {
return true;
} else if (recursive(positionX + 2, positionY - 1, 3, cloneArray(positionsVisited), moveCounter + 1, currentMoves)) {
return true;
} else if (recursive(positionX + 1, positionY - 2, 4, cloneArray(positionsVisited), moveCounter + 1, currentMoves)) {
return true;
} else if (recursive(positionX - 2, positionY + 1, 5, cloneArray(positionsVisited), moveCounter + 1, currentMoves)) {
return true;
} else if (recursive(positionX - 1, positionY + 2, 6, cloneArray(positionsVisited), moveCounter + 1, currentMoves)) {
return true;
} else if (recursive(positionX - 2, positionY - 1, 7, cloneArray(positionsVisited), moveCounter + 1, currentMoves)) {
return true;
} else if (recursive(positionX - 1, positionY - 2, 8, cloneArray(positionsVisited), moveCounter + 1, currentMoves)) {
return true;
} else {
return false;
}
}
private static boolean[][] cloneArray(boolean[][] src) {
boolean[][] newPositions = new boolean[src.length][src.length];
for (int i = 0; i < src.length; i++) {
System.arraycopy(src[i], 0, newPositions[i], 0, src[0].length);
}
return newPositions;
}
}
以下是6x6板的工作变化。使用8x8电路板计算会占用我的机器上太多时间。如果随机选择移动选择,也许可以更快。