solvability_algorithm.py

import queue
from train_GAN_4_fl import *


def is_f(maze, row, column, tile):
    return maze[4 * column + 31 * 4 * row + tile] == 1


def data(row, column):
    return 30 - row + 29 - column, (row, column)


def neighbour(x, y):
    neighbours = []
    if x != 0:
        neighbours.append((x-1, y))
    if x != 30:
        neighbours.append((x+1, y))
    if y != 0:
        neighbours.append((x, y-1))
    if y != 30:
        neighbours.append((x, y+1))
    return neighbours


def solvable(maze):
    open_set = queue.PriorityQueue()
    open_set.put(data(0, 1))
    if not is_f(maze, 0, 1, 2):
        return False

    closed_set = []

    while not open_set.empty():
        position = open_set.get()[1]
        closed_set.append(position)
        neighbours = neighbour(position[0], position[1])
        for pos in neighbours:
            if is_f(maze, pos[0], pos[1], 0) and pos not in closed_set:
                open_set.put(data(pos[0], pos[1]))
            elif is_f(maze, pos[0], pos[1], 3) and pos == (30, 29):
                return True

    return False


if __name__=='__main__':
    dataset = MyDataset('levels_corrected.csv')
    loader = DataLoader(
        dataset,
        batch_size=40,
        shuffle=True,
        num_workers=2
    )

    for d in loader:
        break
    #print(d[0])
    maze = d[0]

    print("maze solvable: ", solvable(maze))