我有一个nxm个由非负整数组成的矩阵。例如:
2 3 4 7 1
1 5 2 6 2
4 3 4 2 1
2 1 2 4 1
3 1 3 4 1
2 1 4 3 2
6 9 1 6 4
“投下炸弹”会使目标细胞及其所有八个邻居的数量减少一个，直到最小值为零。
x x x
x X x
x x x
什么样的算法可以确定将所有细胞减少到零所需的最少炸弹数量?
B选项(因为我不是一个细心的读者)
事实上，问题的第一个版本并不是我要寻找的答案。我没有仔细阅读整个任务，有额外的约束条件，让我们说:
那么简单的问题是，当行中的序列必须是非递增的:
8 7 6 6 5是可能的输入序列
7 8 5 5 2是不可能的，因为7 -> 8在一个序列中增长。
也许为“简单”的问题找到答案会有助于为更难的问题找到解决方案。
PS:我相信当我们有几个相同的情况需要最少的炸弹来清除上面的线时，我们会选择在“左侧”使用最多炸弹的一个。还有什么证据是正确的吗?
当前回答
评价函数，总和:
int f (int ** matrix, int width, int height, int x, int y)
{
int m[3][3] = { 0 };
m[1][1] = matrix[x][y];
if (x > 0) m[0][1] = matrix[x-1][y];
if (x < width-1) m[2][1] = matrix[x+1][y];
if (y > 0)
{
m[1][0] = matrix[x][y-1];
if (x > 0) m[0][0] = matrix[x-1][y-1];
if (x < width-1) m[2][0] = matrix[x+1][y-1];
}
if (y < height-1)
{
m[1][2] = matrix[x][y+1];
if (x > 0) m[0][2] = matrix[x-1][y+1];
if (x < width-1) m[2][2] = matrix[x+1][y+1];
}
return m[0][0]+m[0][1]+m[0][2]+m[1][0]+m[1][1]+m[1][2]+m[2][0]+m[2][1]+m[2][2];
}
目标函数:
Point bestState (int ** matrix, int width, int height)
{
Point p = new Point(0,0);
int bestScore = 0;
int b = 0;
for (int i=0; i<width; i++)
for (int j=0; j<height; j++)
{
b = f(matrix,width,height,i,j);
if (b > bestScore)
{
bestScore = best;
p = new Point(i,j);
}
}
retunr p;
}
破坏功能:
void destroy (int ** matrix, int width, int height, Point p)
{
int x = p.x;
int y = p.y;
if(matrix[x][y] > 0) matrix[x][y]--;
if (x > 0) if(matrix[x-1][y] > 0) matrix[x-1][y]--;
if (x < width-1) if(matrix[x+1][y] > 0) matrix[x+1][y]--;
if (y > 0)
{
if(matrix[x][y-1] > 0) matrix[x][y-1]--;
if (x > 0) if(matrix[x-1][y-1] > 0) matrix[x-1][y-1]--;
if (x < width-1) if(matrix[x+1][y-1] > 0) matrix[x+1][y-1]--;
}
if (y < height-1)
{
if(matrix[x][y] > 0) matrix[x][y+1]--;
if (x > 0) if(matrix[x-1][y+1] > 0) matrix[x-1][y+1]--;
if (x < width-1) if(matrix[x+1][y+1] > 0) matrix[x+1][y+1]--;
}
}
目标函数:
bool isGoal (int ** matrix, int width, int height)
{
for (int i=0; i<width; i++)
for (int j=0; j<height; j++)
if (matrix[i][j] > 0)
return false;
return true;
}
线性最大化函数:
void solve (int ** matrix, int width, int height)
{
while (!isGoal(matrix,width,height))
{
destroy(matrix,width,height, bestState(matrix,width,height));
}
}
这不是最优的，但可以通过找到更好的评价函数来优化。
. .但是考虑到这个问题，我在想一个主要的问题是在0中间的某个点上得到废弃的数字，所以我要采取另一种方法。这是支配最小值为零，然后试图转义零，这导致一般的最小现有值(s)或这样
其他回答
这可以用深度为O(3^(n))的树来求解。其中n是所有平方和。
首先考虑用O(9^n)树来解决问题是很简单的，只需考虑所有可能的爆炸位置。有关示例，请参阅Alfe的实现。
接下来我们意识到，我们可以从下往上轰炸，仍然得到一个最小的轰炸模式。
Start from the bottom left corner. Bomb it to oblivion with the only plays that make sense (up and to the right). Move one square to the right. While the target has a value greater than zero, consider each of the 2 plays that make sense (straight up or up and to the right), reduce the value of the target by one, and make a new branch for each possibility. Move another to the right. While the target has a value greater than zero, consider each of the 3 plays that make sense (up left, up, and up right), reduce the value of the target by one, and make a new branch for each possibility. Repeat steps 5 and 6 until the row is eliminated. Move up a row and repeat steps 1 to 7 until the puzzle is solved.
这个算法是正确的，因为
有必要在某一时刻完成每一行。 完成一行总是需要一个游戏，一个在上面，一个在下面，或者在这一行内。 选择在未清除的最低行之上的玩法总是比选择在该行之上或该行之下的玩法更好。
在实践中，这个算法通常会比它的理论最大值做得更好，因为它会定期轰炸邻居并减少搜索的大小。如果我们假设每次轰炸都会减少4个额外目标的价值，那么我们的算法将运行在O(3^(n/4))或大约O(1.3^n)。
Because this algorithm is still exponential, it would be wise to limit the depth of the search. We might limit the number of branches allowed to some number, X, and once we are this deep we force the algorithm to choose the best path it has identified so far (the one that has the minimum total board sum in one of its terminal leaves). Then our algorithm is guaranteed to run in O(3^X) time, but it is not guaranteed to get the correct answer. However, we can always increase X and test empirically if the trade off between increased computation and better answers is worthwhile.
我想不出一个计算实际数字的方法除非用我最好的启发式方法计算轰炸行动并希望得到一个合理的结果。
So my method is to compute a bombing efficiency metric for each cell, bomb the cell with the highest value, .... iterate the process until I've flattened everything. Some have advocated using simple potential damage (i.e. score from 0 to 9) as a metric, but that falls short by pounding high value cells and not making use of damage overlap. I'd calculate cell value - sum of all neighbouring cells, reset any positive to 0 and use the absolute value of anything negative. Intuitively this metric should make a selection that help maximise damage overlap on cells with high counts instead of pounding those directly.
下面的代码在28个炸弹中达到了测试场的完全破坏(注意，使用潜在伤害作为度量，结果是31!)
using System;
using System.Collections.Generic;
using System.Linq;
namespace StackOverflow
{
internal class Program
{
// store the battle field as flat array + dimensions
private static int _width = 5;
private static int _length = 7;
private static int[] _field = new int[] {
2, 3, 4, 7, 1,
1, 5, 2, 6, 2,
4, 3, 4, 2, 1,
2, 1, 2, 4, 1,
3, 1, 3, 4, 1,
2, 1, 4, 3, 2,
6, 9, 1, 6, 4
};
// this will store the devastation metric
private static int[] _metric;
// do the work
private static void Main(string[] args)
{
int count = 0;
while (_field.Sum() > 0)
{
Console.Out.WriteLine("Round {0}:", ++count);
GetBlastPotential();
int cell_to_bomb = FindBestBombingSite();
PrintField(cell_to_bomb);
Bomb(cell_to_bomb);
}
Console.Out.WriteLine("Done in {0} rounds", count);
}
// convert 2D position to 1D index
private static int Get1DCoord(int x, int y)
{
if ((x < 0) || (y < 0) || (x >= _width) || (y >= _length)) return -1;
else
{
return (y * _width) + x;
}
}
// Convert 1D index to 2D position
private static void Get2DCoord(int n, out int x, out int y)
{
if ((n < 0) || (n >= _field.Length))
{
x = -1;
y = -1;
}
else
{
x = n % _width;
y = n / _width;
}
}
// Compute a list of 1D indices for a cell neighbours
private static List<int> GetNeighbours(int cell)
{
List<int> neighbours = new List<int>();
int x, y;
Get2DCoord(cell, out x, out y);
if ((x >= 0) && (y >= 0))
{
List<int> tmp = new List<int>();
tmp.Add(Get1DCoord(x - 1, y - 1));
tmp.Add(Get1DCoord(x - 1, y));
tmp.Add(Get1DCoord(x - 1, y + 1));
tmp.Add(Get1DCoord(x, y - 1));
tmp.Add(Get1DCoord(x, y + 1));
tmp.Add(Get1DCoord(x + 1, y - 1));
tmp.Add(Get1DCoord(x + 1, y));
tmp.Add(Get1DCoord(x + 1, y + 1));
// eliminate invalid coords - i.e. stuff past the edges
foreach (int c in tmp) if (c >= 0) neighbours.Add(c);
}
return neighbours;
}
// Compute the devastation metric for each cell
// Represent the Value of the cell minus the sum of all its neighbours
private static void GetBlastPotential()
{
_metric = new int[_field.Length];
for (int i = 0; i < _field.Length; i++)
{
_metric[i] = _field[i];
List<int> neighbours = GetNeighbours(i);
if (neighbours != null)
{
foreach (int j in neighbours) _metric[i] -= _field[j];
}
}
for (int i = 0; i < _metric.Length; i++)
{
_metric[i] = (_metric[i] < 0) ? Math.Abs(_metric[i]) : 0;
}
}
//// Compute the simple expected damage a bomb would score
//private static void GetBlastPotential()
//{
// _metric = new int[_field.Length];
// for (int i = 0; i < _field.Length; i++)
// {
// _metric[i] = (_field[i] > 0) ? 1 : 0;
// List<int> neighbours = GetNeighbours(i);
// if (neighbours != null)
// {
// foreach (int j in neighbours) _metric[i] += (_field[j] > 0) ? 1 : 0;
// }
// }
//}
// Update the battle field upon dropping a bomb
private static void Bomb(int cell)
{
List<int> neighbours = GetNeighbours(cell);
foreach (int i in neighbours)
{
if (_field[i] > 0) _field[i]--;
}
}
// Find the best bombing site - just return index of local maxima
private static int FindBestBombingSite()
{
int max_idx = 0;
int max_val = int.MinValue;
for (int i = 0; i < _metric.Length; i++)
{
if (_metric[i] > max_val)
{
max_val = _metric[i];
max_idx = i;
}
}
return max_idx;
}
// Display the battle field on the console
private static void PrintField(int cell)
{
for (int x = 0; x < _width; x++)
{
for (int y = 0; y < _length; y++)
{
int c = Get1DCoord(x, y);
if (c == cell)
Console.Out.Write(string.Format("[{0}]", _field[c]).PadLeft(4));
else
Console.Out.Write(string.Format(" {0} ", _field[c]).PadLeft(4));
}
Console.Out.Write(" || ");
for (int y = 0; y < _length; y++)
{
int c = Get1DCoord(x, y);
if (c == cell)
Console.Out.Write(string.Format("[{0}]", _metric[c]).PadLeft(4));
else
Console.Out.Write(string.Format(" {0} ", _metric[c]).PadLeft(4));
}
Console.Out.WriteLine();
}
Console.Out.WriteLine();
}
}
}
产生的轰炸模式输出如下(左边是字段值，右边是度量值)
Round 1:
2 1 4 2 3 2 6 || 7 16 8 10 4 18 6
3 5 3 1 1 1 9 || 11 18 18 21 17 28 5
4 [2] 4 2 3 4 1 || 19 [32] 21 20 17 24 22
7 6 2 4 4 3 6 || 8 17 20 14 16 22 8
1 2 1 1 1 2 4 || 14 15 14 11 13 16 7
Round 2:
2 1 4 2 3 2 6 || 5 13 6 9 4 18 6
2 4 2 1 1 [1] 9 || 10 15 17 19 17 [28] 5
3 2 3 2 3 4 1 || 16 24 18 17 17 24 22
6 5 1 4 4 3 6 || 7 14 19 12 16 22 8
1 2 1 1 1 2 4 || 12 12 12 10 13 16 7
Round 3:
2 1 4 2 2 1 5 || 5 13 6 7 3 15 5
2 4 2 1 0 1 8 || 10 15 17 16 14 20 2
3 [2] 3 2 2 3 0 || 16 [24] 18 15 16 21 21
6 5 1 4 4 3 6 || 7 14 19 11 14 19 6
1 2 1 1 1 2 4 || 12 12 12 10 13 16 7
Round 4:
2 1 4 2 2 1 5 || 3 10 4 6 3 15 5
1 3 1 1 0 1 8 || 9 12 16 14 14 20 2
2 2 2 2 2 [3] 0 || 13 16 15 12 16 [21] 21
5 4 0 4 4 3 6 || 6 11 18 9 14 19 6
1 2 1 1 1 2 4 || 10 9 10 9 13 16 7
Round 5:
2 1 4 2 2 1 5 || 3 10 4 6 2 13 3
1 3 1 1 0 [0] 7 || 9 12 16 13 12 [19] 2
2 2 2 2 1 3 0 || 13 16 15 10 14 15 17
5 4 0 4 3 2 5 || 6 11 18 7 13 17 6
1 2 1 1 1 2 4 || 10 9 10 8 11 13 5
Round 6:
2 1 4 2 1 0 4 || 3 10 4 5 2 11 2
1 3 1 1 0 0 6 || 9 12 16 11 8 13 0
2 2 2 2 0 2 0 || 13 16 15 9 14 14 15
5 4 [0] 4 3 2 5 || 6 11 [18] 6 11 15 5
1 2 1 1 1 2 4 || 10 9 10 8 11 13 5
Round 7:
2 1 4 2 1 0 4 || 3 10 4 5 2 11 2
1 3 1 1 0 0 6 || 8 10 13 9 7 13 0
2 [1] 1 1 0 2 0 || 11 [15] 12 8 12 14 15
5 3 0 3 3 2 5 || 3 8 10 3 8 15 5
1 1 0 0 1 2 4 || 8 8 7 7 9 13 5
Round 8:
2 1 4 2 1 0 4 || 1 7 2 4 2 11 2
0 2 0 1 0 0 6 || 7 7 12 7 7 13 0
1 1 0 1 0 2 0 || 8 8 10 6 12 14 15
4 2 0 3 3 [2] 5 || 2 6 8 2 8 [15] 5
1 1 0 0 1 2 4 || 6 6 6 7 9 13 5
Round 9:
2 1 4 2 1 0 4 || 1 7 2 4 2 11 2
0 2 0 1 0 0 6 || 7 7 12 7 6 12 0
1 1 0 1 0 [1] 0 || 8 8 10 5 10 [13] 13
4 2 0 3 2 2 4 || 2 6 8 0 6 9 3
1 1 0 0 0 1 3 || 6 6 6 5 8 10 4
Round 10:
2 1 4 2 1 0 4 || 1 7 2 4 2 10 1
0 2 [0] 1 0 0 5 || 7 7 [12] 7 6 11 0
1 1 0 1 0 1 0 || 8 8 10 4 8 9 10
4 2 0 3 1 1 3 || 2 6 8 0 6 8 3
1 1 0 0 0 1 3 || 6 6 6 4 6 7 2
Round 11:
2 0 3 1 1 0 4 || 0 6 0 3 0 10 1
0 1 0 0 0 [0] 5 || 4 5 5 5 3 [11] 0
1 0 0 0 0 1 0 || 6 8 6 4 6 9 10
4 2 0 3 1 1 3 || 1 5 6 0 5 8 3
1 1 0 0 0 1 3 || 6 6 6 4 6 7 2
Round 12:
2 0 3 1 0 0 3 || 0 6 0 2 1 7 1
0 1 0 0 0 0 4 || 4 5 5 4 1 7 0
1 0 0 0 0 [0] 0 || 6 8 6 4 5 [9] 8
4 2 0 3 1 1 3 || 1 5 6 0 4 7 2
1 1 0 0 0 1 3 || 6 6 6 4 6 7 2
Round 13:
2 0 3 1 0 0 3 || 0 6 0 2 1 6 0
0 1 0 0 0 0 3 || 4 5 5 4 1 6 0
1 [0] 0 0 0 0 0 || 6 [8] 6 3 3 5 5
4 2 0 3 0 0 2 || 1 5 6 0 4 6 2
1 1 0 0 0 1 3 || 6 6 6 3 4 4 0
Round 14:
2 0 3 1 0 [0] 3 || 0 5 0 2 1 [6] 0
0 0 0 0 0 0 3 || 2 5 4 4 1 6 0
0 0 0 0 0 0 0 || 4 4 4 3 3 5 5
3 1 0 3 0 0 2 || 0 4 5 0 4 6 2
1 1 0 0 0 1 3 || 4 4 5 3 4 4 0
Round 15:
2 0 3 1 0 0 2 || 0 5 0 2 1 4 0
0 0 0 0 0 0 2 || 2 5 4 4 1 4 0
0 0 0 0 0 0 0 || 4 4 4 3 3 4 4
3 1 0 3 0 [0] 2 || 0 4 5 0 4 [6] 2
1 1 0 0 0 1 3 || 4 4 5 3 4 4 0
Round 16:
2 [0] 3 1 0 0 2 || 0 [5] 0 2 1 4 0
0 0 0 0 0 0 2 || 2 5 4 4 1 4 0
0 0 0 0 0 0 0 || 4 4 4 3 3 3 3
3 1 0 3 0 0 1 || 0 4 5 0 3 3 1
1 1 0 0 0 0 2 || 4 4 5 3 3 3 0
Round 17:
1 0 2 1 0 0 2 || 0 3 0 1 1 4 0
0 0 0 0 0 0 2 || 1 3 3 3 1 4 0
0 0 0 0 0 0 0 || 4 4 4 3 3 3 3
3 1 [0] 3 0 0 1 || 0 4 [5] 0 3 3 1
1 1 0 0 0 0 2 || 4 4 5 3 3 3 0
Round 18:
1 0 2 1 0 0 2 || 0 3 0 1 1 4 0
0 0 0 0 0 0 2 || 1 3 3 3 1 4 0
0 0 0 0 0 0 0 || 3 3 2 2 2 3 3
3 [0] 0 2 0 0 1 || 0 [4] 2 0 2 3 1
1 0 0 0 0 0 2 || 2 4 2 2 2 3 0
Round 19:
1 0 2 1 0 [0] 2 || 0 3 0 1 1 [4] 0
0 0 0 0 0 0 2 || 1 3 3 3 1 4 0
0 0 0 0 0 0 0 || 2 2 2 2 2 3 3
2 0 0 2 0 0 1 || 0 2 2 0 2 3 1
0 0 0 0 0 0 2 || 2 2 2 2 2 3 0
Round 20:
1 [0] 2 1 0 0 1 || 0 [3] 0 1 1 2 0
0 0 0 0 0 0 1 || 1 3 3 3 1 2 0
0 0 0 0 0 0 0 || 2 2 2 2 2 2 2
2 0 0 2 0 0 1 || 0 2 2 0 2 3 1
0 0 0 0 0 0 2 || 2 2 2 2 2 3 0
Round 21:
0 0 1 1 0 0 1 || 0 1 0 0 1 2 0
0 0 0 0 0 0 1 || 0 1 2 2 1 2 0
0 0 0 0 0 0 0 || 2 2 2 2 2 2 2
2 0 0 2 0 [0] 1 || 0 2 2 0 2 [3] 1
0 0 0 0 0 0 2 || 2 2 2 2 2 3 0
Round 22:
0 0 1 1 0 0 1 || 0 1 0 0 1 2 0
0 0 0 0 0 0 1 || 0 1 2 2 1 2 0
[0] 0 0 0 0 0 0 || [2] 2 2 2 2 1 1
2 0 0 2 0 0 0 || 0 2 2 0 2 1 1
0 0 0 0 0 0 1 || 2 2 2 2 2 1 0
Round 23:
0 0 1 1 0 0 1 || 0 1 0 0 1 2 0
0 0 [0] 0 0 0 1 || 0 1 [2] 2 1 2 0
0 0 0 0 0 0 0 || 1 1 2 2 2 1 1
1 0 0 2 0 0 0 || 0 1 2 0 2 1 1
0 0 0 0 0 0 1 || 1 1 2 2 2 1 0
Round 24:
0 0 0 0 0 0 1 || 0 0 0 0 0 2 0
0 0 0 0 0 0 1 || 0 0 0 0 0 2 0
0 0 [0] 0 0 0 0 || 1 1 [2] 2 2 1 1
1 0 0 2 0 0 0 || 0 1 2 0 2 1 1
0 0 0 0 0 0 1 || 1 1 2 2 2 1 0
Round 25:
0 0 0 0 0 [0] 1 || 0 0 0 0 0 [2] 0
0 0 0 0 0 0 1 || 0 0 0 0 0 2 0
0 0 0 0 0 0 0 || 1 1 1 1 1 1 1
1 0 0 1 0 0 0 || 0 1 1 0 1 1 1
0 0 0 0 0 0 1 || 1 1 1 1 1 1 0
Round 26:
0 0 0 0 0 0 0 || 0 0 0 0 0 0 0
0 0 0 0 0 0 0 || 0 0 0 0 0 0 0
[0] 0 0 0 0 0 0 || [1] 1 1 1 1 0 0
1 0 0 1 0 0 0 || 0 1 1 0 1 1 1
0 0 0 0 0 0 1 || 1 1 1 1 1 1 0
Round 27:
0 0 0 0 0 0 0 || 0 0 0 0 0 0 0
0 0 0 0 0 0 0 || 0 0 0 0 0 0 0
0 0 [0] 0 0 0 0 || 0 0 [1] 1 1 0 0
0 0 0 1 0 0 0 || 0 0 1 0 1 1 1
0 0 0 0 0 0 1 || 0 0 1 1 1 1 0
Round 28:
0 0 0 0 0 0 0 || 0 0 0 0 0 0 0
0 0 0 0 0 0 0 || 0 0 0 0 0 0 0
0 0 0 0 0 0 0 || 0 0 0 0 0 0 0
0 0 0 0 0 [0] 0 || 0 0 0 0 0 [1] 1
0 0 0 0 0 0 1 || 0 0 0 0 0 1 0
Done in 28 rounds
由于时间不够，我不得不停留在部分解决方案上，但希望即使是这个部分解决方案也能提供解决这个问题的潜在方法的一些见解。
当面对一个困难的问题时，我喜欢想出一些简单的问题来培养对问题空间的直觉。这里，我采取的第一步是将这个二维问题简化为一维问题。考虑一行字:
0 4 2 1 3 0 1
不管怎样，你知道你需要在4点附近炸4次才能把它降到0。因为左边是一个较低的数字，所以轰炸0或4比轰炸2没有任何好处。事实上，我相信(但缺乏严格的证明)轰炸2，直到4点降到0，至少和任何其他策略一样好，让4点降到0。从左到右，我们可以采用如下策略:
index = 1
while index < line_length
while number_at_index(index - 1) > 0
bomb(index)
end
index++
end
# take care of the end of the line
while number_at_index(index - 1) > 0
bomb(index - 1)
end
几个轰炸命令示例:
0 4[2]1 3 0 1
0 3[1]0 3 0 1
0 2[0]0 3 0 1
0 1[0]0 3 0 1
0 0 0 0 3[0]1
0 0 0 0 2[0]0
0 0 0 0 1[0]0
0 0 0 0 0 0 0
4[2]1 3 2 1 5
3[1]0 3 2 1 5
2[0]0 3 2 1 5
1[0]0 3 2 1 5
0 0 0 3[2]1 5
0 0 0 2[1]0 5
0 0 0 1[0]0 5
0 0 0 0 0 0[5]
0 0 0 0 0 0[4]
0 0 0 0 0 0[3]
0 0 0 0 0 0[2]
0 0 0 0 0 0[1]
0 0 0 0 0 0 0
从一个需要以某种方式下降的数字开始是一个很有吸引力的想法，因为它突然变得可以找到一个解，就像一些人声称的那样，至少和所有其他解一样好。
The next step up in complexity where this search of at least as good is still feasible is on the edge of the board. It is clear to me that there is never any strict benefit to bomb the outer edge; you're better off bombing the spot one in and getting three other spaces for free. Given this, we can say that bombing the ring one inside of the edge is at least as good as bombing the edge. Moreover, we can combine this with the intuition that bombing the right one inside of the edge is actually the only way to get edge spaces down to 0. Even more, it is trivially simple to figure out the optimal strategy (in that it is at least as good as any other strategy) to get corner numbers down to 0. We put this all together and can get much closer to a solution in the 2-D space.
根据对角子的观察，我们可以肯定地说，我们知道从任何起始棋盘到所有角子都是0的棋盘的最佳策略。这是一个这样的板的例子(我借用了上面两个线性板的数字)。我用不同的方式标记了一些空间，我会解释为什么。
0 4 2 1 3 0 1 0
4 x x x x x x 4
2 y y y y y y 2
1 y y y y y y 1
3 y y y y y y 3
2 y y y y y y 2
1 y y y y y y 1
5 y y y y y y 5
0 4 2 1 3 0 1 0
你会注意到，最上面一行和我们之前看到的线性例子非常相似。回想一下我们之前的观察，将第一行全部降为0的最佳方法是破坏第二行(x行)。轰炸任何y行都无法清除顶部行，轰炸顶部行也没有比轰炸x行相应空间更多的好处。
我们可以从上面应用线性策略(轰炸x行上的相应空间)，只关注第一行，不关注其他任何内容。大概是这样的:
0 4 2 1 3 0 1 0
4 x[x]x x x x 4
2 y y y y y y 2
1 y y y y y y 1
3 y y y y y y 3
2 y y y y y y 2
1 y y y y y y 1
5 y y y y y y 5
0 4 2 1 3 0 1 0
0 3 1 0 3 0 1 0
4 x[x]x x x x 4
2 y y y y y y 2
1 y y y y y y 1
3 y y y y y y 3
2 y y y y y y 2
1 y y y y y y 1
5 y y y y y y 5
0 4 2 1 3 0 1 0
0 2 0 0 3 0 1 0
4 x[x]x x x x 4
2 y y y y y y 2
1 y y y y y y 1
3 y y y y y y 3
2 y y y y y y 2
1 y y y y y y 1
5 y y y y y y 5
0 4 2 1 3 0 1 0
0 1 0 0 3 0 1 0
4 x[x]x x x x 4
2 y y y y y y 2
1 y y y y y y 1
3 y y y y y y 3
2 y y y y y y 2
1 y y y y y y 1
5 y y y y y y 5
0 4 2 1 3 0 1 0
0 0 0 0 3 0 1 0
4 x x x x x x 4
2 y y y y y y 2
1 y y y y y y 1
3 y y y y y y 3
2 y y y y y y 2
1 y y y y y y 1
5 y y y y y y 5
0 4 2 1 3 0 1 0
The flaw in this approach becomes very obvious in the final two bombings. It is clear, given that the only bomb sites that reduce the 4 figure in the first column in the second row are the first x and the y. The final two bombings are clearly inferior to just bombing the first x, which would have done the exact same (with regard to the first spot in the top row, which we have no other way of clearing). Since we have demonstrated that our current strategy is suboptimal, a modification in strategy is clearly needed.
在这一点上，我可以退一步，只关注一个角落。让我们考虑一下这个问题:
0 4 2 1
4 x y a
2 z . .
1 b . .
It is clear the only way to get the spaces with 4 down to zero are to bomb some combination of x, y, and z. With some acrobatics in my mind, I'm fairly sure the optimal solution is to bomb x three times and then a then b. Now it's a matter of figuring out how I reached that solution and if it reveals any intuition we can use to even solve this local problem. I notice that there's no bombing of y and z spaces. Attempting to find a corner where bombing those spaces makes sense yields a corner that looks like this:
0 4 2 5 0
4 x y a .
2 z . . .
5 b . . .
0 . . . .
对于这个问题，我很清楚，最优解决方案是轰炸y 5次，z 5次。让我们更进一步。
0 4 2 5 6 0 0
4 x y a . . .
2 z . . . . .
5 b . . . . .
6 . . . . . .
0 . . . . . .
0 . . . . . .
这里，最优解决方案是轰炸a和b 6次，然后x 4次。
现在它变成了一个如何将这些直觉转化为我们可以建立的原则的游戏。
希望能继续!
这里有一个解决方案，推广良好的性质的角。
让我们假设我们可以为给定的字段找到一个完美的落点，也就是说，一个减少其中值的最佳方法。然后，为了找到最少的炸弹数量，一个算法的初稿可能是(代码是从ruby实现中复制粘贴的):
dropped_bomb_count = 0
while there_are_cells_with_non_zero_count_left
coordinates = choose_a_perfect_drop_point
drop_bomb(coordinates)
dropped_bomb_count += 1
end
return dropped_bomb_count
挑战是choose_a_perfect_drop_point。首先，让我们定义一个完美的落点是什么。
(x, y)的放置点会减少(x, y)中的值。它也可能会减少其他单元格中的值。 (x, y)的放置点A比(x, y)的放置点b更好，如果它减少了b所减少的单元格的适当超集中的值。 如果没有其他更好的投放点，投放点是最大的。 (x, y)的两个放置点是等效的，如果它们减少了同一组单元格。 如果(x, y)的放置点等价于(x, y)的所有最大放置点，那么它就是完美的。
如果(x, y)存在一个完美的投放点，那么您不能比在(x, y)的一个完美投放点上投放炸弹更有效地降低(x, y)处的值。
给定字段的完美放置点是其任何单元格的完美放置点。
以下是一些例子:
1 0 1 0 0
0 0 0 0 0
1 0 0 0 0
0 0 0 0 0
0 0 0 0 0
单元格(0,0)(从零开始的索引)的完美放置点是(1,1)。(1,1)的所有其他放置点，即(0,0)、(0,1)和(1,0)，减少的单元格较少。
0 0 0 0 0
0 0 0 0 0
0 0 1 0 0
0 0 0 0 0
0 0 0 0 0
单元格(2,2)(从零开始的索引)的完美落点是(2,2)，以及所有周围的单元格(1,1)、(1,2)、(1,3)、(2,1)、(2,3)、(3,1)、(3,2)和(3,3)。
0 0 0 0 1
0 0 0 0 0
0 0 1 0 0
0 0 0 0 0
0 0 0 0 0
单元格(2,2)的完美放置点是(3,1):它减少了(2,2)中的值和(4,0)中的值。(2,2)的所有其他放置点都不是最大的，因为它们减少了一个单元格。(2,2)的完美下拉点也是(4,0)的完美下拉点，它是字段的唯一完美下拉点。它为这个领域带来了完美的解决方案(一颗炸弹)。
1 0 0 0 0
0 0 0 0 0
0 0 1 0 0
0 0 0 0 0
1 0 0 0 0
(2,2)没有完美的落点:(1,1)和(1,3)都减少(2,2)和另一个单元格(它们是(2,2)的最大落点)，但它们不相等。然而，(1,1)是(0,0)的完美落点，(1,3)是(0,4)的完美落点。
根据完美落点的定义和一定的检查顺序，我得到了以下问题示例的结果:
Drop bomb on 1, 1
Drop bomb on 1, 1
Drop bomb on 1, 5
Drop bomb on 1, 5
Drop bomb on 1, 5
Drop bomb on 1, 6
Drop bomb on 1, 2
Drop bomb on 1, 2
Drop bomb on 0, 6
Drop bomb on 0, 6
Drop bomb on 2, 1
Drop bomb on 2, 5
Drop bomb on 2, 5
Drop bomb on 2, 5
Drop bomb on 3, 1
Drop bomb on 3, 0
Drop bomb on 3, 0
Drop bomb on 3, 0
Drop bomb on 3, 0
Drop bomb on 3, 0
Drop bomb on 3, 4
Drop bomb on 3, 4
Drop bomb on 3, 3
Drop bomb on 3, 3
Drop bomb on 3, 6
Drop bomb on 3, 6
Drop bomb on 3, 6
Drop bomb on 4, 6
28
然而，该算法只有在每一步之后至少有一个完美落点时才能工作。可以在没有完美落点的情况下构建例子:
0 1 1 0
1 0 0 1
1 0 0 1
0 1 1 0
对于这些情况，我们可以修改算法，这样我们就不会选择完美的落点，而是选择一个具有最大落点的最小选择的坐标，然后计算每个选择的最小值。在上面的例子中，所有有值的单元格都有两个最大落点。例如，(0,1)有最大落点(1,1)和(1,2)。选择其中任何一个，然后计算最小值，会得到这样的结果:
Drop bomb on 1, 1
Drop bomb on 2, 2
Drop bomb on 1, 2
Drop bomb on 2, 1
2
这是对第一个问题的回答。我没有注意到他改变了参数。
创建一个所有目标的列表。根据掉落物品(掉落物品本身和所有邻居)影响的正数值的数量为目标分配一个值。最高值是9。
根据受影响目标的数量(降序)对目标进行排序，对每个受影响目标的和进行二次降序排序。
向排名最高的目标投掷炸弹，然后重新计算目标，直到所有目标值都为零。
同意，这并不总是最优的。例如,
100011
011100
011100
011100
000000
100011
这种方法需要5枚炸弹才能清除。最理想的情况是，你可以在4分钟内完成。不过,很 非常接近，没有回头路。在大多数情况下，这将是最优的，或者非常接近。
使用原来的问题数，该方法解决28个炸弹。
添加代码来演示这种方法(使用带有按钮的表单):
private void button1_Click(object sender, EventArgs e)
{
int[,] matrix = new int[10, 10] {{5, 20, 7, 1, 9, 8, 19, 16, 11, 3},
{17, 8, 15, 17, 12, 4, 5, 16, 8, 18},
{ 4, 19, 12, 11, 9, 7, 4, 15, 14, 6},
{ 17, 20, 4, 9, 19, 8, 17, 2, 10, 8},
{ 3, 9, 10, 13, 8, 9, 12, 12, 6, 18},
{16, 16, 2, 10, 7, 12, 17, 11, 4, 15},
{ 11, 1, 15, 1, 5, 11, 3, 12, 8, 3},
{ 7, 11, 16, 19, 17, 11, 20, 2, 5, 19},
{ 5, 18, 2, 17, 7, 14, 19, 11, 1, 6},
{ 13, 20, 8, 4, 15, 10, 19, 5, 11, 12}};
int value = 0;
List<Target> Targets = GetTargets(matrix);
while (Targets.Count > 0)
{
BombTarget(ref matrix, Targets[0]);
value += 1;
Targets = GetTargets(matrix);
}
Console.WriteLine( value);
MessageBox.Show("done: " + value);
}
private static void BombTarget(ref int[,] matrix, Target t)
{
for (int a = t.x - 1; a <= t.x + 1; a++)
{
for (int b = t.y - 1; b <= t.y + 1; b++)
{
if (a >= 0 && a <= matrix.GetUpperBound(0))
{
if (b >= 0 && b <= matrix.GetUpperBound(1))
{
if (matrix[a, b] > 0)
{
matrix[a, b] -= 1;
}
}
}
}
}
Console.WriteLine("Dropped bomb on " + t.x + "," + t.y);
}
private static List<Target> GetTargets(int[,] matrix)
{
List<Target> Targets = new List<Target>();
int width = matrix.GetUpperBound(0);
int height = matrix.GetUpperBound(1);
for (int x = 0; x <= width; x++)
{
for (int y = 0; y <= height; y++)
{
Target t = new Target();
t.x = x;
t.y = y;
SetTargetValue(matrix, ref t);
if (t.value > 0) Targets.Add(t);
}
}
Targets = Targets.OrderByDescending(x => x.value).ThenByDescending( x => x.sum).ToList();
return Targets;
}
private static void SetTargetValue(int[,] matrix, ref Target t)
{
for (int a = t.x - 1; a <= t.x + 1; a++)
{
for (int b = t.y - 1; b <= t.y + 1; b++)
{
if (a >= 0 && a <= matrix.GetUpperBound(0))
{
if (b >= 0 && b <= matrix.GetUpperBound(1))
{
if (matrix[ a, b] > 0)
{
t.value += 1;
t.sum += matrix[a,b];
}
}
}
}
}
}
你需要的一个类:
class Target
{
public int value;
public int sum;
public int x;
public int y;
}
