def execute_current_algorithm():
Adj, source, target = Tile.neighborsDict, Config.source, Config.target
W = {} # weight dictionary that maps each tile to its cost
for tile in Tile.tilesDict.values():
W[tile.id] = tile.W
algDict = {
"BFS" : algorithms.BFS(Adj, source, target),
"DFS" : algorithms.DFS(Adj, source, target),
"Dijkstra" : algorithms.Dijkstra(Adj, W, source, target),
"B_FS" : algorithms.B_FS(Adj, W, source, target),
"A*" : algorithms.A_star(Adj, W, source, target),
}
alg = algDict[Config.currentAlgorithm]
if Config.currentAlgorithm in ["BFS", "DFS"]:
Tile.pathToTargetList, Tile.idToLevelDict, Tile.levelToIdList = alg.search()
Tile.explored_tiles = Tile.idToLevelDict.keys() # to draw explored tiles
Tile.idToLevelAux = Tile.idToLevelDict
elif Config.currentAlgorithm in ["Dijkstra", "B_FS", "A*"]:
Tile.pathToTargetList, Tile.idToCostDict, Tile.levelToIdList, Tile.explored_tiles, Tile.levelToCostList = alg.search()
Tile.idToCostAux = Tile.idToCostDict
GameController.currentAlg = alg
def main( ) : if validInput( sys.argv ) : fileName = sys.argv[1] board = helper.loadState( fileName ) print "Puzzle: " + fileName time0 = time.clock() if (sys.argv[2] == "bfs" ) : metrics = algorithms.BFS( board ) elif (sys.argv[2] == "dfs" ) : metrics = algorithms.DFS( board ) elif (sys.argv[2] == "id" ) : metrics = algorithms.IDS( board ) elif (sys.argv[2] == "astar") : metrics = algorithms.aStar( board ) time1 = time.clock() delTime = time1-time0 print "Nodes Explored: " + str(metrics[0]) print "Solution Length: " + str(metrics[1]) print "Time Elapsed: " + str(delTime) + " sec" print "" else : print "invalid input" print ""
def dfs_python():
data = request.json
startNode = data["startNode"]
nodes = data["nodes"]
edges = data["edges"]
dfs = algo.DFS(nodes, startNode, None)
trace = dfs.run()
return jsonify(trace)
def plotSearchAlgorithms(dim, p):
start = (0, 0)
goal = (dim - 1, dim - 1)
grid = gd.generateGrid(dim, p)
solved = [False for _ in range(5)]
solved[0], DFSPath = algos.DFS(grid, start, goal)
solved[1], BFSPath, _ = algos.BFS(grid, start, goal)
solved[2], BDBFSPath = algos.BDBFS(grid, start, goal)
solved[3], AStarPathEuclidean, _ = algos.aStar(grid, start, goal, h.Euc)
solved[4], AStarPathManhattan, _ = algos.aStar(grid, start, goal,
h.Manhattan)
paths = {
"DFS": DFSPath,
"BFS": BFSPath,
"BDBFS": BDBFSPath,
"A* Euclidean": AStarPathEuclidean,
"A* Manhattan": AStarPathManhattan
}
# if solvable, then all search algorithms should work, and so plot all of them one at a time
if all(solved):
print("Close window after each search algorithm to see next one")
for i, path in enumerate(paths):
print(path)
for y, x in paths.get(path):
grid[y][x] = 3
pyplot.matshow(grid)
pyplot.title(path)
pyplot.show()
for y, x in paths.get(path):
grid[y][x] = 0
# otherwise just show the unsolvable grid
else:
pyplot.matshow(grid)
pyplot.title("Unsolvable")
pyplot.show()
def update(self):
self.sprite_group_all.update()
# catch inputs
keystate = pg.key.get_pressed()
if keystate[pg.K_ESCAPE]:
pg.event.post(pg.event.Event(pg.QUIT))
if keystate[pg.K_TAB]:
self.sprite_group_all.empty()
self.tilemap.randomize_start_goal(self)
if keystate[pg.K_KP0]:
self.neural.train()
if keystate[pg.K_KP1]:
self.neural.save(assets.neural_network_model_folder)
if keystate[pg.K_KP2]:
self.neural.load(assets.neural_network_model_folder)
if keystate[pg.K_q]: # BFS
self.pathqueue = None
self.path = alg.BFS(alg.SquareGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
if keystate[pg.K_a]: # Visual BFS
self.path = alg.BFS(alg.SquareGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
self.visual_helper()
if keystate[pg.K_w]: # DFS
self.pathqueue = None
self.path = alg.DFS(alg.SquareGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
if keystate[pg.K_s]: # Visual DFS
self.path = alg.DFS(alg.SquareGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
self.visual_helper()
if keystate[pg.K_e]: # Dijkstra
self.pathqueue = None
self.path = alg.Dijkstra(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
if keystate[pg.K_d]: # Visual Dijkstra
self.path = alg.Dijkstra(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
self.visual_helper()
if keystate[pg.K_z]: # Dijkstra - only path
self.pathqueue = None
d_path = alg.Dijkstra(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
self.path = alg.ReconstructPath(d_path, self.tilemap.custom_start, self.tilemap.custom_goal)
if keystate[pg.K_x]: # Visual Dijkstra - only path
d_path = alg.Dijkstra(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
self.path = alg.ReconstructPath(d_path, self.tilemap.custom_start, self.tilemap.custom_goal)
self.visual_helper(True)
if keystate[pg.K_r]: # Astar
self.pathqueue = None
self.path, cost = alg.Astar(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
if keystate[pg.K_f]: # Visual Astar
self.path, cost = alg.Astar(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
self.visual_helper()
if keystate[pg.K_c]: # Astar - only path
self.pathqueue = None
d_path, cost = alg.Astar(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
self.path = alg.ReconstructPath(d_path, self.tilemap.custom_start, self.tilemap.custom_goal)
if keystate[pg.K_v]: # Visual Astar - only path
d_path, cost = alg.Astar(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
self.path = alg.ReconstructPath(d_path, self.tilemap.custom_start, self.tilemap.custom_goal)
self.visual_helper(True)
if keystate[pg.K_t]: # Astar NN
self.pathqueue = None
self.path, cost = alg.Astar(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal, self.neural)
if keystate[pg.K_g]: # Visual Astar NN
self.path, cost = alg.Astar(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal, self.neural)
self.visual_helper()
if keystate[pg.K_p]: # test
num_tests=50
def bfs_test():
self.sprite_group_all.empty()
self.tilemap.randomize_start_goal(self)
self.pathqueue = None
self.path = alg.BFS(alg.SquareGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
def dfs_test():
self.sprite_group_all.empty()
self.tilemap.randomize_start_goal(self)
self.pathqueue = None
self.path = alg.DFS(alg.SquareGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
def dijkstra_test():
self.sprite_group_all.empty()
self.tilemap.randomize_start_goal(self)
self.pathqueue = None
self.path = alg.Dijkstra(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
def astar_test():
self.sprite_group_all.empty()
self.tilemap.randomize_start_goal(self)
self.pathqueue = None
self.path, cost = alg.Astar(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)
def neural_test():
self.sprite_group_all.empty()
self.tilemap.randomize_start_goal(self)
self.pathqueue = None
self.path, cost = alg.Astar(alg.WeightedGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal, self.neural)
bfs_time = timeit.timeit(bfs_test, number=num_tests) / num_tests
print("BFS: " + str(bfs_time))
dfs_time = timeit.timeit(dfs_test, number=num_tests) / num_tests
print("DFS: " + str(dfs_time))
# dijkstra_time = timeit.timeit(dijkstra_test, number=num_tests) / num_tests
# print("Dijkstra: " + str(dijkstra_time))
astar_time = timeit.timeit(astar_test, number=num_tests) / num_tests
print("Astar: " + str(astar_time))
neural_time = timeit.timeit(neural_test, number=num_tests) / num_tests
print("Neural: " + str(neural_time) + "\n\n")
def dfs_test():
self.sprite_group_all.empty()
self.tilemap.randomize_start_goal(self)
self.pathqueue = None
self.path = alg.DFS(alg.SquareGraph(self.tilemap), self.tilemap.custom_start, self.tilemap.custom_goal)