Problem solution · Python

ABC351 D — Grid and Magnet

ABC351 D — Grid and Magnet: a Python solution using disjoint set union. Learn the idea, check the complexity, and read the full code, with credit to KATO-Hiro AtCoder Solutions.

Technique
Disjoint set union
Source
KATO-Hiro AtCoder Solutions
Length
252 lines
Start with the idea.

Try the problem first. If you get stuck, read the approach below, then write your own solution. The full code is at the bottom.

Approach

Disjoint set union

For ABC351 D — Grid and Magnet, the implementation maintains connected components and merges them as relationships are processed.

  1. Give each element a component representative.
  2. Merge representatives when a connection is accepted.
  3. Answer connectivity or component queries from the compressed representatives.

Code notes

  • 252 lines of Python from the credited upstream file abc351_d.py.
  • The implementation visibly relies on sequence storage, ordered lookup.
  • No explicit loop blocks detected.

Complexity

Account for every find and union operation; with path compression and ranked merging, the amortized cost is nearly constant per operation.

Check the problem constraints before deciding whether this complexity will pass.

Source

Code and credit

This code comes from KATO-Hiro AtCoder Solutions by KATO-Hiro and is used under the CC0-1.0 licence.

Full codeABC351 D — Grid and Magnet · PythonPython
Use this to learn the idea, then write your own version.
# -*- coding: utf-8 -*-  from typing import List  class UnionFind:    """Represents a data structure that tracks a set of elements partitioned       into a number of disjoint (non-overlapping) subsets.     Landau notation: O(α(n)), where α(n) is the inverse Ackermann function.     See:    https://www.youtube.com/watch?v=zV3Ul2pA2Fw    https://en.wikipedia.org/wiki/Disjoint-set_data_structure    https://atcoder.jp/contests/abc120/submissions/4444942    https://atcoder.jp/contests/abc292/submissions/39410075    https://github.com/not522/ac-library-python/blob/master/atcoder/dsu.py    """     def __init__(self, number_count: int) -> None:        """        Args:            number_count: The size of elements (greater than 2).        """        self.number_count = number_count        self.parent_numbers = [-1 for _ in range(number_count)]        self.edge_count = [0 for _ in range(number_count)]        self.group_count = number_count     def find_root(self, number: int) -> int:        """Follows the chain of parent pointers from number up the tree until           it reaches a root element, whose parent is itself.        Args:            number: The trees id (0-index).         Returns:            The index of a root element.        """        if self.parent_numbers[number] < 0:            return number         self.parent_numbers[number] = self.find_root(self.parent_numbers[number])        return self.parent_numbers[number]     def get_group_size(self, number: int) -> int:        """        Args:            number: The trees id (0-index).         Returns:            The size of group.        """        return -self.parent_numbers[self.find_root(number)]     def is_same_group(self, number_x: int, number_y: int) -> bool:        """Represents the roots of tree number_x and number_y are in the same           group.        Args:            number_x: The trees x (0-index).            number_y: The trees y (0-index).        """        return self.find_root(number_x) == self.find_root(number_y)     def merge_if_needs(self, number_x: int, number_y: int) -> bool:        """Uses find_root to determine the roots of the tree number_x and           number_y belong to. If the roots are distinct, the trees are combined           by attaching the roots of one to the root of the other.        Args:            number_x: The trees x (0-index).            number_y: The trees y (0-index).        """        x = self.find_root(number_x)        y = self.find_root(number_y)         self.edge_count[x] += 1         if x == y:            return False         self.group_count -= 1         if self.parent_numbers[x] > self.parent_numbers[y]:            x, y = y, x         self.parent_numbers[x] += self.parent_numbers[y]        self.parent_numbers[y] = x        self.edge_count[x] += self.edge_count[y]        return True     def get_roots(self) -> List[int]:        return [i for i, x in enumerate(self.parent_numbers) if x < 0]     def get_groups(self) -> List[List[int]]:        roots: List[int] = [self.find_root(i) for i in range(self.number_count)]        groups: List[List[int]] = [[] for _ in range(self.number_count)]         for i in range(self.number_count):            groups[roots[i]].append(i)         return list(filter(lambda g: g, groups))     def get_edge_count(self, number: int) -> int:        return self.edge_count[number]     def get_group_count(self) -> int:        return self.group_count  class UnionFind2D:    """Extends UnionFind to two dimensions.     See:    https://atcoder.jp/contests/past202010-open/submissions/21472171    """     def __init__(self, height: int, width: int) -> None:        self.height: int = height        self.width: int = width        self.size: int = height * width        self.uf: UnionFind = UnionFind(self.size)     def find_root(self, x: int, y: int) -> int:        assert 0 <= x < self.width        assert 0 <= y < self.height         return self.uf.find_root(self._to_number(x, y))     def get_group_size(self, x: int, y: int) -> int:        assert 0 <= x < self.width        assert 0 <= y < self.height         return self.uf.get_group_size(self._to_number(x, y))     def is_same_group(self, x1: int, y1: int, x2: int, y2: int) -> bool:        assert 0 <= x1 < self.width        assert 0 <= y1 < self.height        assert 0 <= x2 < self.width        assert 0 <= y2 < self.height         return self.find_root(x1, y1) == self.find_root(x2, y2)     def merge_if_needs(self, x1: int, y1: int, x2: int, y2: int) -> bool:        assert 0 <= x1 < self.width        assert 0 <= y1 < self.height        assert 0 <= x2 < self.width        assert 0 <= y2 < self.height         return self.uf.merge_if_needs(self._to_number(x1, y1), self._to_number(x2, y2))     def get_roots(self) -> List[int]:        return self.uf.get_roots()     def get_groups(self) -> List[List[int]]:        """        Returns:            List of trees id (0-index).        """        return self.uf.get_groups()     def get_edge_count(self, x: int, y: int) -> int:        assert 0 <= x < self.width        assert 0 <= y < self.height         return self.uf.get_edge_count(self._to_number(x, y))     def get_group_count(self) -> int:        return self.uf.get_group_count()     def _to_number(self, x: int, y: int) -> int:        """        Args:            x, y: Coordinates in grid (0-index).         Returns:            The trees id (0-index).        """        return x + self.width * y     def _to_yx(self, number: int) -> tuple[int, int]:        """        Args:            The trees id (0-index).         Returns:            y, x: Coordinates in grid (0-index).        """        return divmod(number, self.width)  def main():    import sys     input = sys.stdin.readline     h, w = map(int, input().split())    s = [list(input().rstrip()) for _ in range(h)]    uf = UnionFind2D(height=h, width=w)    dxy = [(-1, 0), (1, 0), (0, -1), (0, 1), (-1, -1), (1, -1), (-1, 1), (1, 1)]    dxy = dxy[:4]     # 磁石の上下左右のマスを便宜的に!に書き換える    for i in range(h):        for j in range(w):            if s[i][j] != "#":                continue             for dx, dy in dxy:                nx, ny = j + dx, i + dy                 if 0 <= nx < w and 0 <= ny < h and s[ny][nx] == ".":                    s[ny][nx] = "!"     # 何もないマス同士を同じ連結成分として扱う    for i in range(h):        for j in range(w):            if s[i][j] != ".":                continue             for dx, dy in dxy:                nx, ny = j + dx, i + dy                 if 0 <= nx < w and 0 <= ny < h and s[ny][nx] == ".":                    uf.merge_if_needs(j, i, nx, ny)     # 空きマスの連結成分に対して、周囲にある!の個数を調べる    ans = 1     for group in uf.get_groups():        root_yx = set()         for g in group:            y, x = uf._to_yx(g)             if s[y][x] != ".":                continue             for dx, dy in dxy:                nx, ny = x + dx, y + dy                 if 0 <= nx < w and 0 <= ny < h and s[ny][nx] == "!":                    root_yx.add((ny, nx))         # 着目している連結成分のサイズ +  周囲にある!の個数 (重複カウントしないように注意)        ans = max(ans, len(group) + len(root_yx))     print(ans)  if __name__ == "__main__":    main() 

Did this explanation save you time? I'm a Grade 11 student building this free library to make difficult algorithms easier to understand.

Buy me a coffee ↗