我有一个任务是编写一个程序,在国际象棋棋盘上设置8位主教占据整个董事会。它应该在找到第一个解决方案时结束并打印出所有内容。这是我用Java编写的代码,我很难用回溯来完成它(这个地方在代码中被注释)。
/*
* 0 - not occupied square
* 1 - bishop standing square
* 2 - occupied square (diagonal)
*/
public class BishopsBT {
public int [][] solution;
final int N = 8; // number of squares in column and row (chess board)
final int solved = 120; //Sum of 1's and 2's in case of all occupied board
int sum; //current sum of board
public BishopsBT(){
solution = new int [N][N] ;
}
public void solve() {
if(placeBishops(0)){
//print the result
clear(); // clears all 2's
for (int i = 0; i < N; i++) {
for (int j = 0; j < N; j++) {
System.out.print(" " + solution[i][j]);
}
System.out.println();
}
} else{
System.out.println("NO SOLUTION EXISTS");
}
}
public boolean placeBishops (int bishop){
for (int row = 0; row < N; row++) {
// check if bishop can be placed
if (canPlace(solution, row, bishop)) {
// place the bishop
solution[row][bishop] = 1;
}
}
if (allSpaceOccupied()) {
return true;
} else {
// SOME BACKTRACKING CODE HERE
return false;
}
}
// check if bishop can be placed at matrix[row][column]
public boolean canPlace(int[][] matrix, int row, int column) {
// we need to check all diagonals
// whether no bishop is standing there
for (int i = row, j = column; i >= 0 && j >= 0; i--, j--) {
if (matrix[i][j] == 1) {
return false;
}
}
for (int i = row, j = column; i >= 0 && j < matrix.length; i--, j++) {
if (matrix[i][j] == 1) {
return false;
}
}
for (int i = row, j = column; i < matrix.length && j >= 0; i++, j--) {
if (matrix[i][j] == 1) {
return false;
}
}
for (int i = row, j = column; i < matrix.length && j < matrix.length; i++, j++) {
if (matrix[i][j] == 1) {
return false;
}
}
// if we are here that means we are safe to place Bishop
return true;
}
public boolean allSpaceOccupied() {
// clears previously occupied space
clear();
// occupies new space
for (int i = 0; i < solution.length; i++) {
for ( int j = 0; j < solution.length; j++) {
if (solution[i][j] == 1) diagonalOccupy(i,j);
}
}
sum = 0;
// counts sum of occupied space
for (int i = 0; i < solution.length; i++) {
for ( int j = 0; j < solution.length; j++) {
sum += solution [i][j];
}
}
if (sum == solved) return true;
// else
return false;
}
public void diagonalOccupy(int row, int column) {
// writes 2 in each bishop's occupied square
for (int i = row, j = column; i >= 0 && j >= 0; i--, j--) {
if (solution[i][j] == 0) {
solution[i][j] = 2;
}
}
for (int i = row, j = column; i >= 0 && j < solution.length; i--, j++) {
if (solution[i][j] == 0) {
solution[i][j] = 2;
}
}
for (int i = row, j = column; i < solution.length && j >= 0; i++, j--) {
if (solution[i][j] == 0) {
solution[i][j] = 2;
}
}
for (int i = row, j = column; i < solution.length && j < solution.length; i++, j++) {
if (solution[i][j] == 0) {
solution[i][j] = 2;
}
}
}
// clears all 2's on the board
public void clear() {
for (int i = 0; i < solution.length; i++) {
for ( int j = 0; j < solution.length; j++) {
if (solution[i][j] == 2) solution[i][j] = 0;
}
}
}
public static void main(String[] args) {
BishopsBT q = new BishopsBT();
q.solve();
}
}
事情是,目前我的程序将主教放在第一列,这种布局不占用所有空间。当然,我可以简单地将所有内容放在第三列,问题就解决了。但是,我必须使用回溯并且不知道如何。如果您有任何想法或提示,我会很高兴听到它们。
答案 0 :(得分:2)
您的解决方案假定所有主教必须放在不同的行中。并非所有解决方案都适用。 (有一个解决方案,所有主教都在第三或第四列。你不是在寻找所有的解决方案,但如果你是这样的话,你会错过这个假设的许多解决方案。)
你也不需要canPlace
检查:主教不能相互威胁是没有限制的。 (这可能是加速搜索的有效技术,但同样,当你应用它时,你会错过一些解决方案。如果你想使用它,就没有必要检查已经放置过的主教的所有对角线单元格;它足以检查当前单元格是否已被标记为“已占用”或受到威胁。)
如果你打算使用蛮力方法进行回溯,你可以测试所有可能的主教组合。这是C(64,8)或4,426,165,368种组合。
你可以大幅削减可能性,但不要假设主教必须在不同的行。相反,请注意您的解决方案包含两个独立的解决方案。白色广场上的主教只能威胁到白色方块,黑色方块上的主教只能威胁到黑色方块。因此,找到一个解决方案,在板上放置四个主教,威胁所有的白色方块。然后
(如果您想找到所有解决方案,请找到所有 k 子解决方案,并将它们合并到 k ²完整解决方案中。)
这种案件分离减少了对C(32,8)或35,960进行测试的可能安排。你的策略只考虑每行有一个主教的配置,检查8 ^ 8(约1600万)的可能性。它错过了一些解决方案,并检查了多个配置,其中没有四个主教在白色方块上,四个在黑色方块上。
回溯的原则在另一个答案中给出。如果您将32个白色方块标记为:
01 02 03 04
05 06 07 08
09 10 11 12
13 14 15 16
17 18 19 20
21 22 23 24
25 26 27 28
29 30 31 32
你可以使用像这样的递归方法(在伪Java中):
bool place(int i, int start) {
if (i == 8) {
if (allOccupied()) {
print();
return true;
}
} else {
for (int j = start, j < 32; j++) {
int row = j / 4;
int col = 2 * (j % 4) + row % 2;
// add bishop at (col, row)
// save occupancy matrix
// add threat by (col, row) to matrix
if (place(i + 1, j + 1)) return true;
// revert matrix to saved matrix
// remove bishop from (col, row)
}
}
return false;
}
并以
开头place(0, 0);
答案 1 :(得分:0)
你应该这样做:
public boolean placeBishops (int bishop){
if(bishop == 8){
if(allSpaceOccupied()){
//Print the board here, i.e found the solution
//also check the indexing of the bishop,
//i have assumed that they start from 0
return true;
}else{
return false;
}
}
for (int row = 0; row < N; row++) {
// check if bishop can be placed
if (canPlace(solution, row, bishop)) {
// place the bishop
solution[row][bishop] = 1;
boolean found = placeBishops(bishop+1);
if(found == true) return true;
solution[row][bishop] = 0;
}
}
return false;
}
我们可以检查一个地方是否适合特定的主教,并相应地增加主教数量,如果我们没有找到解决方案沿着那条路走下去,我们重置当前主教索引的solution array
以及该主教的相应行,以便我们可以寻找另一种可能的解决方案。