feat(bfs): add bfs on a 2d grid

Author: jessicayung

algorithms/bfs.py

@@ -0,0 +1,72 @@
+"""
+Breadth-first Search on a 2D grid
+
+Note: 'steps' indicates order of expansion, not distance from the source cell.
+
+Jessica Yung
+Jan 2018
+"""
+import numpy as np
+from queue import Queue
+
+class BFS2D:
+
+    def __init__(self, grid_width, grid_height, visited=None):
+        self.grid_width = grid_width
+        self.grid_height = grid_height
+        if visited is None:
+            self.visited = np.zeros((grid_height, grid_width))
+        else:
+            self.visited = visited
+        self.steps_taken = np.zeros((grid_height, grid_width))
+        self.queue = Queue()
+
+
+    def bfs(self, x, y, steps=0):
+        """Breadth-First Search algorithm that visits every cell in a 2D grid, where adjacent cells are defined as up to eight cells next to the current cell (can move diagonally).
+        """
+        if x >= self.grid_width or y >= self.grid_height:
+            return
+        if x < 0 or y < 0:
+            return
+        if self.visited[y][x]:
+            return
+
+        # Mark current cell as visited
+        self.queue.put([x,y])
+        self.visited[y][x] = True
+        self.steps_taken[y][x] = steps
+        steps += 1
+        print("Visited ", x, y, "in ", steps, " steps")
+        print(self.steps_taken)
+        print("Visited is now")
+        print(self.visited)
+
+        # Visit every neighbouring cell
+        while not self.queue.empty():
+            node = self.queue.get()
+            for cell in self.neighbour(node[0], node[1]):
+                cellx = cell[0]
+                celly = cell[1]
+                # print(cell)
+                if cellx < 0 or cellx >= self.grid_width:
+                    continue
+                if celly < 0 or celly >= self.grid_height:
+                    continue
+                if not self.visited[cell[1], cell[0]]:
+                    self.queue.put(cell)
+                    self.visited[cell[1], cell[0]] = True
+                    self.steps_taken[celly, cellx] = steps
+                    steps += 1
+
+    def neighbour(self, x, y):
+        neighbours = []
+        for dx in [-1,0,1]:
+            for dy in [-1,0,1]:
+                neighbours.append([x + dx, y + dy])
+        return neighbours
+
+bfs = BFS2D(4, 3)
+bfs.bfs(1,2)
+print(bfs.steps_taken)
+