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

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

其他回答

这是一个广度搜索,通过这个“迷宫”的位置寻找最短路径(一系列轰炸)。不,我不能证明没有更快的算法,抱歉。

#!/usr/bin/env python

M = ((1,2,3,4),
     (2,3,4,5),
     (5,2,7,4),
     (2,3,5,8))

def eachPossibleMove(m):
  for y in range(1, len(m)-1):
    for x in range(1, len(m[0])-1):
      if (0 == m[y-1][x-1] == m[y-1][x] == m[y-1][x+1] ==
               m[y][x-1]   == m[y][x]   == m[y][x+1] ==
               m[y+1][x-1] == m[y+1][x] == m[y+1][x+1]):
        continue
      yield x, y

def bomb(m, (mx, my)):
  return tuple(tuple(max(0, m[y][x]-1)
      if mx-1 <= x <= mx+1 and my-1 <= y <= my+1
      else m[y][x]
      for x in range(len(m[y])))
    for y in range(len(m)))

def findFirstSolution(m, path=[]):
#  print path
#  print m
  if sum(map(sum, m)) == 0:  # empty?
    return path
  for move in eachPossibleMove(m):
    return findFirstSolution(bomb(m, move), path + [ move ])

def findShortestSolution(m):
  black = {}
  nextWhite = { m: [] }
  while nextWhite:
    white = nextWhite
    nextWhite = {}
    for position, path in white.iteritems():
      for move in eachPossibleMove(position):
        nextPosition = bomb(position, move)
        nextPath = path + [ move ]
        if sum(map(sum, nextPosition)) == 0:  # empty?
          return nextPath
        if nextPosition in black or nextPosition in white:
          continue  # ignore, found that one before
        nextWhite[nextPosition] = nextPath

def main(argv):
  if argv[1] == 'first':
    print findFirstSolution(M)
  elif argv[1] == 'shortest':
    print findShortestSolution(M)
  else:
    raise NotImplementedError(argv[1])

if __name__ == '__main__':
  import sys
  sys.exit(main(sys.argv))

有一种方法可以把这个问题简化为一个简单的子问题。

解释分为两部分,算法和算法的原因 提供最优解决方案。没有第二个,第一个就说不通了,所以我 从为什么开始。

如果你想轰炸矩形(假设一个大矩形-还没有边缘情况) 你可以看到,只有这样才能减少空心矩形上的正方形 周长到0的意思是炸毁周长或者炸毁的空心矩形 就在外围的方块里。我称周长为图层1,其中的矩形为图层2。

一个重要的观点是,没有点轰炸层1,因为 你这样做得到的“爆炸半径”总是包含在爆炸半径内 另一个来自第2层的正方形。你应该很容易就能说服自己。

所以,我们可以把问题简化为找到一个最优的方法来炸开周长,然后我们可以重复这个过程,直到所有的平方都为0。

但当然了,如果有爆炸的可能,并不总能找到最优解 以一种不太理想的方式远离周边,但通过使用X个额外的炸弹制造 用>X炸弹减少内层的问题。如果我们调用 第一层,如果我们在第二层的某个地方放置一个额外的X炸弹(只是 在第1层内,我们可以减少之后轰炸第2层的努力吗 X ?换句话说,我们必须证明我们可以贪心化简外部 周长。

但是,我们知道我们可以贪婪。因为第2层的炸弹永远不会更多 有效减少第2层到0比战略上放置炸弹在第3层。和 因为和之前一样的原因-总有一个炸弹我们可以放在第3层 将影响第2层的每一个方块,炸弹放在第2层可以。所以,它可以 永远不要伤害我们的贪婪(在这个意义上的贪婪)。

所以,我们要做的就是找到最优的方法,通过轰炸将许可值降为0 下一个内层。

我们永远不会因为先把角落炸到0而受伤,因为只有内层的角落可以到达,所以我们真的没有选择(并且,任何可以到达角落的周长炸弹的爆炸半径都包含在内层角落的爆炸半径中)。

一旦我们这样做了,与0角相邻的周长上的正方形只能由内层的2个正方形到达:

0       A       B

C       X       Y

D       Z

在这一点上,周长实际上是一个封闭的1维环,因为任何炸弹都会减少3个相邻的正方形。除了角落附近的一些奇怪之处——X可以“击中”A、B、C和D。

Now we can't use any blast radius tricks - the situation of each square is symmetric, except for the weird corners, and even there no blast radius is a subset of another. Note that if this were a line (as Colonel Panic discusses) instead of a closed loop the solution is trivial. The end points must be reduced to 0, and it never harms you to bomb the points adjacent to the end points, again because the blast radius is a superset. Once you have made your endpoint 0, you still have a new endpoint, so repeat (until the line is all 0).

所以,如果我们可以优化地将层中的单个正方形减少到0,我们就有了一个算法(因为我们已经切断了循环,现在有了一条带有端点的直线)。我相信轰炸与最小值相邻的正方形(给你2个选项),这样在最小值的2个正方形内的最大值就是最小值(你可能不得不分割你的轰炸来管理这一点)将是最优的,但我还没有证明。

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

你的新问题,有跨行不递减的值,很容易解决。

Observe that the left column contains the highest numbers. Therefore, any optimal solution must first reduce this column to zero. Thus, we can perform a 1-D bombing run over this column, reducing every element in it to zero. We let the bombs fall on the second column so they do maximum damage. There are many posts here dealing with the 1D case, I think, so I feel safe in skipping that case. (If you want me to describe it, I can.). Because of the decreasing property, the three leftmost columns will all be reduced to zero. But, we will provably use a minimum number of bombs here because the left column must be zeroed.

现在,一旦左边的列归零,我们只要剪掉最左边的三列现在归零,然后对现在化简的矩阵重复这一步骤。这必须给我们一个最优的解决方案,因为在每个阶段我们使用可证明的最少数量的炸弹。

使用分支和定界的数学整数线性规划

As it has already been mentioned, this problem can be solved using integer linear programming (which is NP-Hard). Mathematica already has ILP built in. "To solve an integer linear programming problem Mathematica first solves the equational constraints, reducing the problem to one containing inequality constraints only. Then it uses lattice reduction techniques to put the inequality system in a simpler form. Finally, it solves the simplified optimization problem using a branch-and-bound method." [see Constrained Optimization Tutorial in Mathematica.. ]

我写了下面的代码,利用ILP库的Mathematica。它的速度快得惊人。

solveMatrixBombProblem[problem_, r_, c_] := 
 Module[{}, 
  bombEffect[x_, y_, m_, n_] := 
   Table[If[(i == x || i == x - 1 || i == x + 1) && (j == y || 
        j == y - 1 || j == y + 1), 1, 0], {i, 1, m}, {j, 1, n}];
  bombMatrix[m_, n_] := 
   Transpose[
    Table[Table[
      Part[bombEffect[(i - Mod[i, n])/n + 1, Mod[i, n] + 1, m, 
        n], (j - Mod[j, n])/n + 1, Mod[j, n] + 1], {j, 0, 
       m*n - 1}], {i, 0, m*n - 1}]];
  X := x /@ Range[c*r];
  sol = Minimize[{Total[X], 
     And @@ Thread[bombMatrix[r, c].X >= problem] && 
      And @@ Thread[X >= 0] && Total[X] <= 10^100 && 
      Element[X, Integers]}, X];
  Print["Minimum required bombs = ", sol[[1]]];
  Print["A possible solution = ", 
   MatrixForm[
    Table[x[c*i + j + 1] /. sol[[2]], {i, 0, r - 1}, {j, 0, 
      c - 1}]]];]

对于问题中提供的示例:

solveMatrixBombProblem[{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}, 7, 5]

输出

对于那些用贪婪算法读这篇文章的人

在下面这个10x10的问题上试试你的代码:

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

这里用逗号分隔:

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

对于这个问题,我的解决方案包含208个炸弹。这里有一个可能的解决方案(我能够在大约12秒内解决这个问题)。

作为一种测试Mathematica产生结果的方法,看看你的贪婪算法是否能做得更好。