LWC 67: 764. Largest Plus Sign

本文介绍了一种算法,用于寻找二维网格中最大的轴对齐加号标志。通过动态规划记录四个方向上连续1的最大数量,进而确定加号标志的大小。

摘要生成于 C知道 ,由 DeepSeek-R1 满血版支持, 前往体验 >

LWC 67: 764. Largest Plus Sign

传送门:764. Largest Plus Sign

Problem:

In a 2D grid from (0, 0) to (N-1, N-1), every cell contains a 1, except those cells in the given list mines which are 0. What is the largest axis-aligned plus sign of 1s contained in the grid? Return the order of the plus sign. If there is none, return 0.

An “axis-aligned plus sign of 1s of order k” has some center grid[x][y] = 1 along with 4 arms of length k-1 going up, down, left, and right, and made of 1s. This is demonstrated in the diagrams below. Note that there could be 0s or 1s beyond the arms of the plus sign, only the relevant area of the plus sign is checked for 1s.

Examples of Axis-Aligned Plus Signs of Order k:

Order 1:
000
010
000

Order 2:
00000
00100
01110
00100
00000

Order 3:
0000000
0001000
0001000
0111110
0001000
0001000
0000000

Example 1:

Input: N = 5, mines = [[4, 2]]
Output: 2
Explanation:
11111
11111
11111
11111
11011
In the above grid, the largest plus sign can only be order 2. One of them is marked in bold.

Example 2:

Input: N = 2, mines = []
Output: 1
Explanation:
There is no plus sign of order 2, but there is of order 1.

Example 3:

Input: N = 1, mines = [[0, 0]]
Output: 0
Explanation:
There is no plus sign, so return 0.

Note:

  • N will be an integer in the range [1, 500].
  • mines will have length at most 5000.
  • mines[i] will be length 2 and consist of integers in the range [0, N-1].
  • (Additionally, programs submitted in C, C++, or C# will be judged with a slightly smaller time limit.)

思路:
动态规划,分别记录4个方向上的最大连续1的个数。比如”1001111”, 每个位置出现的最大连续1的个数分别为:”1001234”,有了4个方向的最长连续1,order就是这四个方向的最小值,遍历每个位置的order,求出最大order即可。

Java版本:

    public int orderOfLargestPlusSign(int N, int[][] mines) {
        int[][][] grid = new int[N][N][4];
        for (int i = 0; i < N; ++i) {
            for (int j = 0; j < N; ++j) {
                for (int k = 0; k < 4; ++k) {
                    grid[i][j][k] = 1;
                }
            }
        }

        for (int[] mine : mines) {
            int r = mine[0];
            int c = mine[1];
            for (int k = 0; k < 4; ++k) grid[r][c][k] = 0;
        }

        for (int i = 0; i < N; ++i) {
            for (int j = 1; j < N; ++j) {
                if (grid[i][j][0] == 1)
                    grid[i][j][0] += grid[i][j - 1][0]; 
            }
            for (int j = N - 2; j >= 0; --j) {
                if (grid[i][j][1] == 1)
                    grid[i][j][1] += grid[i][j + 1][1];
            }
        }

        for (int j = 0; j < N; ++j) {
            for (int i = 1; i < N; ++i) {
                if (grid[i][j][2] == 1)
                    grid[i][j][2] += grid[i - 1][j][2];
            }
            for (int i = N - 2; i >= 0; --i) {
                if (grid[i][j][3] == 1)
                    grid[i][j][3] += grid[i + 1][j][3];
            }
        }

        int ans = 0;
        for (int i = 0; i < N; ++i) {
            for (int j = 0; j < N; ++j) {
                int order = Math.min(Math.min(grid[i][j][0], grid[i][j][1]), Math.min(grid[i][j][2], grid[i][j][3]));
                ans = Math.max(ans, order);
            }
        }
        return ans;
    }

Python版本:

class Solution(object):
    def orderOfLargestPlusSign(self, N, mines):
        """
        :type N: int
        :type mines: List[List[int]]
        :rtype: int
        """
        grid = [[1] * N for _ in range(N)]
        lf = [[0] * N for _ in range(N)]
        dn = [[0] * N for _ in range(N)]
        rt = [[0] * N for _ in range(N)]
        up = [[0] * N for _ in range(N)]

        for i, j in mines: grid[i][j] = 0

        for i in xrange(N):
            for j in xrange(N):
                if grid[i][j] == 1:
                    lf[i][j] = 1 if j == 0 else lf[i][j - 1] + 1
                if grid[j][i] == 1:
                    dn[j][i] = 1 if j == 0 else dn[j - 1][i] + 1
        for i in xrange(N):
            for j in xrange(N - 1, -1, -1):
                if grid[i][j] == 1:
                    rt[i][j] = 1 if j == N - 1 else rt[i][j + 1] + 1
                if grid[j][i] == 1:
                    up[j][i] = 1 if j == N - 1 else up[j + 1][i] + 1

        ans = 0
        for i in xrange(N):
            for j in xrange(N):
                order = min(lf[i][j], rt[i][j], up[i][j], dn[i][j])
                ans = max(ans, order)
        return ans
class UniformAffineQuantizer(nn.Module): def __init__( self, n_bits: int = 8, symmetric: bool = False, per_channel_axes=[], metric="minmax", dynamic=False, dynamic_method="per_cluster", group_size=None, shape=None, lwc=False, disable_zero_point=False, ): """ support cluster quantize dynamic_method support per_token and per_cluster """ super().__init__() self.symmetric = symmetric self.disable_zero_point = disable_zero_point assert 2 <= n_bits <= 16, "bitwidth not supported" self.n_bits = n_bits if self.disable_zero_point: self.qmin = -(2 ** (n_bits - 1)) self.qmax = 2 ** (n_bits - 1) - 1 else: self.qmin = 0 self.qmax = 2 ** (n_bits) - 1 self.per_channel_axes = per_channel_axes self.metric = metric self.cluster_counts = None self.cluster_dim = None self.scale = None self.zero_point = None self.round_zero_point = None self.cached_xmin = None self.cached_xmax = None self.dynamic = dynamic self.dynamic_method = dynamic_method self.deficiency = 0 self.lwc = lwc init_value = 4. # inti value of learnable weight clipping if lwc: if group_size: dim1 = int(shape[0]*math.ceil(shape[1]/group_size)) self.deficiency = shape[-1]%group_size if self.deficiency > 0: self.deficiency = group_size - self.deficiency assert self.symmetric # support for mlc-llm symmetric quantization else: dim1 = shape[0] self.upbound_factor = nn.Parameter(torch.ones((dim1,1))*init_value) self.lowbound_factor = nn.Parameter(torch.ones((dim1,1))*init_value) self.sigmoid = nn.Sigmoid() self.enable = True self.group_size = group_size def change_n_bits(self, n_bits): self.n_bits = n_bits if self.disable_zero_point: self.qmin = -(2 ** (n_bits - 1)) self.qmax = 2 ** (n_bits - 1) - 1 else: self.qmin = 0 self.qmax = 2 ** (n_bits) - 1 def fake_quant(self, x, scale, round_zero_point): if self.deficiency > 0: pad_zeros = torch.zeros((x.shape[0],self.deficiency),dtype=x.dtype,device=x.device) x = torch.cat((x,pad_zeros),dim=1) if self.group_size: assert len(x.shape)==2, "only support linear layer now" dim1, dim2 = x.shape x = x.reshape(-1, self.group_size) x_int = round_ste(x / scale) if round_zero_point is not None: x_int = x_int.add(round_zero_point) x_int = x_int.clamp(self.qmin, self.qmax) x_dequant = x_int if round_zero_point is not None: x_dequant = x_dequant.sub(round_zero_point) x_dequant = x_dequant.mul(scale) if self.group_size: x_dequant = x_dequant.reshape(dim1, dim2) if self.deficiency > 0: x_dequant = x_dequant[:,:-self.deficiency] return x_dequant def forward(self, x: torch.Tensor): if self.n_bits >= 16 or not self.enable: return x if self.metric == "fix0to1": return x.mul_(2**self.n_bits-1).round_().div_(2**self.n_bits-1) if self.dynamic_method == "per_token" or self.dynamic_method == "per_channel": self.per_token_dynamic_calibration(x) else: raise NotImplementedError() x_dequant = self.fake_quant(x, self.scale, self.round_zero_point) return x_dequant def per_token_dynamic_calibration(self, x): if self.group_size: if self.deficiency == 0: x = x.reshape(-1,self.group_size) else: pad_zeros = torch.zeros((x.shape[0],self.deficiency),dtype=x.dtype,device=x.device) x = torch.cat((x,pad_zeros),dim=1) x = x.reshape(-1,self.group_size) reduce_shape = [-1] xmin = x.amin(reduce_shape, keepdim=True) xmax = x.amax(reduce_shape, keepdim=True) if self.lwc: xmax = self.sigmoid(self.upbound_factor)*xmax xmin = self.sigmoid(self.lowbound_factor)*xmin if self.symmetric: abs_max = torch.max(xmax.abs(),xmin.abs()) scale = abs_max / (2**(self.n_bits-1)-1) self.scale = scale.clamp(min=CLIPMIN, max=1e4) zero_point = (2**(self.n_bits-1)-1)*torch.ones_like(self.scale) else: range = xmax - xmin scale = range / (2**self.n_bits-1) self.scale = scale.clamp(min=CLIPMIN, max=1e4) zero_point = -(xmin) / (self.scale) if self.disable_zero_point: self.round_zero_point = None else: self.round_zero_point = zero_point.clamp(min=-1e4, max=1e4).round() def register_scales_and_zeros(self): self.register_buffer('scales', self.scale) self.register_buffer('zeros', self.round_zero_point) del self.scale del self.round_zero_point
最新发布
07-24
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值