def knight(algorithm_type, starting_x, starting_y, goal_x, goal_y):
# Create a board of size 8
b = board(8)
# Set starting position and goal position
p_start = position(starting_x, starting_y)
p_goal = position(goal_x, goal_y)
# Check if the positions are available
if not p_start.is_available_pos(b) or not p_goal.is_available_pos(b):
print("position out of board range")
return -1
# Create agent
a = agent(p_start, p_goal)
# Classify the type of agent
if algorithm_type == 0:
a = bfs(p_start, p_goal)
elif algorithm_type == 1:
a = dfs(p_start, p_goal)
elif algorithm_type == 2:
a = ids(p_start, p_goal)
elif algorithm_type == 3:
a = astar(p_start, p_goal)
elif algorithm_type == 4:
a = idastar(p_start, p_goal)
else:
usage()
return -1
# Search path
path_list = a.search(b)
# Print path
for path in path_list:
print(path, end="")
print()
# Return the number of expanded nodes
return a.expanded_node_count
def board_size_test():
print("======== Board size comparison ========")
p_start = position(0, 0)
p_goal = position(2, 2)
agents = [
bfs(p_start, p_goal),
dfs(p_start, p_goal),
ids(p_start, p_goal),
astar(p_start, p_goal),
idastar(p_start, p_goal)
]
for a in agents:
print(a)
print("Board size\tSearch time (ms)\tSteps\t\tExpanded nodes")
bm = [-1, -1, -1, -1]
for b_size in range(3, 17):
b = board(b_size)
start_time = time.time()
path = a.search(b)
search_time = (time.time() - start_time) * 100
steps = len(path) - 1
bm[0] = b_size if bm[0] == -1 else bm[0]
bm[1] = search_time if bm[1] == -1 else bm[1]
bm[2] = steps if bm[2] == -1 and steps != 0 else bm[2]
bm[3] = a.expanded_node_count != 0 if bm[
3] == -1 and a.expanded_node_count else bm[3]
print("{} ({:.3f})\t{:.6f} ({:.3f})\t{} ({:.3f})\t{} ({:.3f})".
format(b_size, b_size / bm[0], search_time,
search_time / bm[1], steps, steps / bm[2],
a.expanded_node_count, a.expanded_node_count / bm[3]))
print()
i = 0
while i < 5:
command = input()
if command == 'usi':
print('id name suisui')
print('usiok')
if command == 'isready':
print('readyok')
if re.match('position', command):
sfen = command
if re.match('go', command):
#初期局面にする
Bboard.shoki()
Wboard.shoki()
#startposを読み込み、現局面を表示する
board.position(sfen)
board.synth()
board.kyokumen()
print(f'{board.turn = }')
#以下合法手生成
if board.turn == 1:
Bmoves.move1()
print(f'{Bmoves.depth1=}')
moves = Bmoves.depth1
#合法手がなければ投了
if Bmoves.depth1 == []:
print('bestmove resign')
#合法手があればランダムで選択
else:
Beval.eval(moves)
if Wkikimoves1.depth1 == []:
def algorithm_type_test():
print("======== Algorithm type comparison ========")
b = board(8)
pairs = [(position(2, 2), position(4, 4)), (position(4, 4), position(2,
2)),
(position(0, 0), position(7, 7)), (position(7, 7), position(0,
0)),
(position(0, 0), position(0, 1)), (position(0, 1), position(0,
0)),
(position(0, 0), position(9, 10))]
for pair in pairs:
p_start = pair[0]
p_goal = pair[1]
agents = [
bfs(p_start, p_goal),
dfs(p_start, p_goal),
ids(p_start, p_goal),
astar(p_start, p_goal),
idastar(p_start, p_goal)
]
bm = [-1, -1, -1]
print("From", p_start, "to", p_goal)
print("Algorithm\tSearch time (ms)\tSteps\t\tExpanded nodes")
for a in agents:
start_time = time.time()
path = a.search(b)
search_time = (time.time() - start_time) * 100
steps = len(path) - 1
bm[0] = search_time if bm[0] == -1 else bm[0]
bm[1] = steps if bm[1] == -1 and steps != 0 else bm[1]
bm[2] = a.expanded_node_count if bm[
2] == -1 and a.expanded_node_count else bm[2]
print("{}\t\t{:.6f} ({:.3f})\t{} ({:.3f})\t{} ({:.3f})".format(
a, search_time, search_time / bm[0], steps, steps / bm[1],
a.expanded_node_count, a.expanded_node_count / bm[2]))
print()
cur_node.add_child(child)
# Set frontier
pair = (estimated_cost(child.depth, child, self.goal), child)
bisect.insort(frontier, pair)
if len(explorered_set) >= math.pow(b.size, 2):
return []
return []
if __name__ == '__main__':
# Examples
b = board(8)
p_start = position(0, 0)
p_goal = position(2, 2)
a = bfs(p_start, p_goal)
path = a.search(b)
b.print_pathway(path)
print(a.expanded_node_count)
a = dfs(p_start, p_goal)
path = a.search(b)
b.print_pathway(path)
print(a.expanded_node_count)
a = ids(p_start, p_goal)
path = a.search(b)
b.print_pathway(path)