Brython: Navigate a Maze

How to use

Your task: implement the decision logic that makes the agent follow the right wall until it reaches the goal.

Your decision function

Edit only inside decide_next_move(state). Return either a single move ('forward', 'left', 'right', 'back') or a list/tuple of moves. The engine will queue your moves and execute one per time step until the queue is empty, then call your function again. Use Reset to restart the run while keeping your algorithm’s internal memory; use Clear Memory to wipe any persisted state in your Python logic.

def decide_next_move(state): """ Decide the next action(s) based on sensor state. state dictionary contains: - facing: one of 'N','E','S','W' - at_goal: True/False - sensors: dict with booleans for walls: {'front': bool, 'right': bool, 'left': bool} - pos: (row, col) current grid position - start: (row, col) start location - goal: (row, col) goal location - rows: int number of rows in the maze grid - cols: int number of columns in the maze grid - debug: a DOM element for live messages; e.g., state['debug'].text = 'Scanning...' Return either: - a single move: 'forward' | 'left' | 'right' | 'back' - or a list/tuple of moves, e.g., ['right', 'forward'] The engine will execute one move per time step until your list is done. """ # Implement right-wall following. # Heuristic when not at goal: # 1) If no wall on the right, turn right and move forward (queue two steps). # 2) Else if no wall in front, go forward. # 3) Else if no wall on the left, turn left. # 4) Else, turn back. s = state['sensors'] if not s['right']: return ['right', 'forward'] if not s['front']: return 'forward' if not s['left']: return 'left' return 'back'

Tip: If the agent bumps into walls, your function may be suggesting invalid moves; the engine prevents walking through walls and will not move when the requested direction is blocked.