def run_terminal(grid_start, agent1=ep2.RandomAgent, agent2=ep2.RandomAgent):
""" Runs a entire game in terminal mode """
tank = util.TANK_CAPACITY
problem = util.USPber(grid=grid_start,
multi_agent=True,
tank_capacity=tank)
pl1 = agent1(player_number=1,
number_of_agents=2,
tank_capacity=tank,
max_depth=util.MAX_DEPTH)
pl2 = agent2(player_number=2,
number_of_agents=2,
tank_capacity=tank,
max_depth=util.MAX_DEPTH)
if util.SHOW_INFO:
print("********************")
print(" Starting game ")
print("********************")
print("Initial board")
for line in grid_start:
print(line)
ctrl = GameControlller(problem, pl1, pl2)
score = None
while score is None:
score = ctrl.step_terminal(show_info=util.SHOW_INFO)
def run_view(grid_start, agent1=ep2.RandomAgent, agent2=ep2.RandomAgent):
""" Runs a game in the graphical mode """
tank = util.TANK_CAPACITY
problem = util.USPber(grid=grid_start,
multi_agent=True,
tank_capacity=tank)
pl1 = agent1(player_number=1,
number_of_agents=2,
tank_capacity=tank,
max_depth=util.MAX_DEPTH)
pl2 = agent2(player_number=2,
number_of_agents=2,
tank_capacity=tank,
max_depth=util.MAX_DEPTH)
if util.SHOW_INFO:
print("********************")
print(" Starting game ")
print("********************")
print("Initial board")
for line in grid_start:
print(line)
ctrl = GameControlller(problem, pl1, pl2)
game_view = view.View(controller=ctrl, gridboard=grid_start)
game_view.root.mainloop()
def run_tests():
""" Simple function to perform a small subset of tests for EP2 """
if os.path.exists("final_result.txt"):
file_flag = "a"
else:
file_flag = "w"
file_results = open("final_result.txt", file_flag)
final_grades = [] # Grade list (code, p01_01, p01_02, p02, p03)
lint_result = pylint.lint.Run(['ep2.py'], exit=False)
for k in lint_result.linter.stats["by_module"]:
mod_name = k
msg_types = lint_result.linter.stats["by_module"][mod_name]
msg_names = lint_result.linter.stats['by_msg']
if 'global_note' in lint_result.linter.stats:
lint_score = lint_result.linter.stats['global_note']
else:
lint_score = -1000.0
file_results.write("******************************\n")
file_results.write("Code quality:\n")
file_results.write("------------------------------\n")
file_results.write("Code grade: {0:.2f}/10.0".format(lint_score))
if lint_score < 0:
file_results.write("\t(Code has serious problems)\n")
elif lint_score < 5.0:
file_results.write("\t(Code need some adjustments)\n")
elif lint_score < 7.0:
file_results.write("\t(Code with reasonably quality)\n")
else:
file_results.write("\t(Great code quality)\n")
file_results.write("------------------------------\n")
file_results.write("Standards deviation:\n")
for k1, v1 in msg_types.items():
file_results.write("{0}:\t{1}\n".format(k1, v1))
file_results.write("------------------------------\n")
file_results.write("Deviation description:\n")
for k2, v2 in msg_names.items():
file_results.write("{0}:\t{1}\n".format(k2, v2))
file_results.write("******************************\n")
file_results.write("Tests:\n")
final_grades.append(lint_score)
#global_results = 0 # Student's result in all tests for all parts
#total_tests_global = 0 # Number of tests performed in all parts
#total_tests = 0 # Number of tests in a specific part
test_results = 0 # Student's result in that specific
#part_tests = 0 # Keep total tests in each part for weighted score
#part_correct = 0 # Keep correct in each part for weighted score
try:
print("Finished pylint analysis")
print("*************************")
print("Starting tests:")
print("Starting Part 00: Test 01")
file_results.write("------------------------------\n")
file_results.write("Part00: RandomAgent Improved\n")
# Part00 - 5 tests
# Default grid: [[1,3,0,3],[7,5,4,2],[3,0,6,0]] Check if no faults
# Blocking grid [[1,5,3],[2,4,5],[3,0,0]]
# Blocked and gas [[5,0,5],[5,8,5],[3,9,5]]
# Big grid 2000x2000 car 1 at (1998,1998), person (1,1), car2 (0,0)
# Changes in perception
# Test 01
# Default grid: [[1,3,0,3],[7,5,4,2],[3,0,6,0]] Check if no faults
grid1 = [[1, 3, 0, 3], [7, 5, 4, 2], [3, 0, 6, 0]]
problem = util.USPber(grid=grid1,
multi_agent=True,
tank_capacity=5,
max_turns=20)
pl1 = ep2.RandomAgent(player_number=1,
number_of_agents=2,
tank_capacity=5,
max_depth=4)
moves = 0
is_goal = False
state = problem.initial_state()
n_grid, _, nt01g, _, _, _ = state
state1 = (n_grid, nt01g)
while moves < 15 and not is_goal:
with Timeout(10):
act = pl1.get_action(state1)
new_state = problem.next_state(state, act)
new_state = problem.next_state(new_state, 'STOP')
if problem.is_goal_state(new_state):
final_info = problem.get_game_info(new_state)
is_goal = True
else:
n_grid, _, nt01g, _, _, _ = new_state
state1 = (n_grid, nt01g)
final_info = problem.get_game_info(new_state)
state = copy.deepcopy(new_state)
moves += 1
if final_info['agent1_faults'] == 0:
test_results += 1
print("Finished Part 00: Test 01")
print("*************************")
print("Starting Part 00: Test 02")
# Test 02
# Blocking grid [[1,5,3],[2,4,5],[3,0,0]]
grid2 = [[1, 5, 3], [2, 4, 5], [3, 0, 0]]
problem = util.USPber(grid=grid2,
multi_agent=True,
tank_capacity=10,
max_turns=20)
pl1 = ep2.RandomAgent(player_number=1,
number_of_agents=2,
tank_capacity=10,
max_depth=4)
moves = 0
state = problem.initial_state()
n_grid, _, nt01g, _, _, _ = state
state1 = (n_grid, nt01g)
while moves < 15:
with Timeout(10):
act = pl1.get_action(state1)
new_state = problem.next_state(state, act)
new_state = problem.next_state(new_state, 'STOP')
n_grid, _, nt01g, _, _, _ = new_state
state1 = (n_grid, nt01g)
final_info = problem.get_game_info(new_state)
state = copy.deepcopy(new_state)
moves += 1
if final_info['agent1_faults'] == 0:
test_results += 1
print("Finished Part 00: Test 02")
print("*************************")
print("Starting Part 00: Test 03")
# Test 03
# Blocked gas [[5,0,5],[5,8,5],[3,9,5]]
grid3 = [[5, 0, 5], [5, 8, 5], [3, 9, 5]]
problem = util.USPber(grid=grid3,
multi_agent=True,
tank_capacity=10,
max_turns=201)
pl1 = ep2.RandomAgent(player_number=1,
number_of_agents=2,
tank_capacity=10,
max_depth=4)
moves = 0
state = problem.initial_state()
n_grid, _, nt01g, _, _, _ = state
state1 = (n_grid, nt01g)
list_of_actions = []
while moves < 200: # To check if list of actions contains all
with Timeout(10):
act = pl1.get_action(state1)
list_of_actions.append(act)
new_state = problem.next_state(state, act)
new_state = problem.next_state(new_state, 'STOP')
n_grid, _, nt01g, _, _, _ = new_state
state1 = (n_grid, nt01g)
final_info = problem.get_game_info(new_state)
state = copy.deepcopy(new_state)
moves += 1
if final_info['agent1_faults'] == 0:
for action in ['STOP', 'REFILL', 'UP', 'DOWN']:
if action in list_of_actions:
test_results += 0.25 # Very low chance to skip one action
print("Finished Part 00: Test 03")
print("*************************")
# print("Starting Part 00: Test 04")
#
# # Test 04
# # Big grid 2000x2000 car 1 at (1998,1998), person (1,1), car2 (0,0)
# grid4 = [[0]*2000 for _ in range(2000)]
# grid4[1][1] = 3
# grid4[0][0] = 2
# grid4[1998][1998] = 1
# problem = util.USPber(grid=grid4,
# multi_agent=True,
# tank_capacity=100,
# max_turns=11)
# pl1 = ep2.RandomAgent(player_number=1, number_of_agents=2,
# tank_capacity=100, max_depth=4)
# moves = 0
# state = problem.initial_state()
# n_grid, _, nt01g, _, _, _ = state
# state1 = (n_grid, nt01g)
# list_of_actions = []
#
# while moves < 5:
# with Timeout(15):
# act = pl1.get_action(state1)
# list_of_actions.append(act)
# new_state = problem.next_state(state, act)
# new_state = problem.next_state(new_state, 'STOP')
# n_grid, _, nt01g, _, _, _ = new_state
# state1 = (n_grid, nt01g)
# final_info = problem.get_game_info(new_state)
# state = copy.deepcopy(new_state)
# moves += 1
#
# if final_info['agent1_faults'] == 0:
# test_results += 1.0
#
# print("Finished Part 00: Test 04")
print("*************************")
print("Starting Part 00: Test 05")
# Test 05
# Changed perception
grid5 = [[1, 2], [0, 3]]
pl1 = ep2.RandomAgent(player_number=1,
number_of_agents=2,
tank_capacity=10,
max_depth=4)
state1 = (grid5, 10)
with Timeout(10):
act = pl1.get_action(state1)
if act in ['STOP', 'DOWN']:
test_results += 0.25
state1 = ([[0, 0, 2], [0, 8, 5], [3, 5, 0]], 6)
with Timeout(10):
act = pl1.get_action(state1)
if act in ['STOP', 'UP', 'LEFT', 'REFILL']:
test_results += 0.25
state1 = ([[0, 5, 2], [0, 8, 5], [3, 5, 0]], 0)
with Timeout(10):
act = pl1.get_action(state1)
if act in ['STOP', 'REFILL']:
test_results += 0.25
state1 = ([[0, 5, 0], [9, 8, 5], [3, 5, 0]], 3)
with Timeout(10):
act = 'STOP'
while act != 'REFILL': # To see if car chooses refill action
act = pl1.get_action(state1)
if act == 'REFILL':
test_results += 0.25
if test_results > 0:
final_grades.append((10.0 * test_results) / 5)
else:
final_grades.append(0.0)
# End of Part00
print("Finished Part 00: Test 05")
print("*************************")
except IOError as e:
file_results.write("Python error: {0}\n".format(e))
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
except NotImplementedError:
file_results.write("NotImplemented\n")
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
except AssertionError as e:
file_results.write("Python error: {0}\n".format(e))
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
except NameError as e:
file_results.write("Python error: {0}\n".format(e))
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
except Timeout.Timeout:
file_results.write("Test did not concluded in time\n")
file_results.write("Receiving proportional grade in the test\n")
final_grades.append((10.0 * test_results) / 5)
except Exception as e:
file_results.write("Python error: {0}\n".format(e))
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
else:
file_results.write("Test concluded without errors\n")
finally:
if len(final_grades) < 2:
final_grades.append(0.0)
file_results.write("Grade in the Test00: {0:.1f}/10.0\n".format(
final_grades[1]))
try:
test_results = 0
print("*************************")
print("Starting Part 01: Test 01")
file_results.write("------------------------------\n")
file_results.write("Part01:\n")
# Part01 - 4 tests
# Default grid: [[1,3,0,3],[7,5,4,2],[3,0,6,0]] fuel spent(fs)= 9
# Trick straight line [[3,0,1,6,7,3,5,3], [2,0,0,0,0,0,4,0]] fs=11
# Refill check [[3,0,0,0,3],[0,0,0,0,0],[0,0,8,0,0],[0,2,0,0,0],[3,0,0,0,3]] fs=28
# Teleporting person
# Test 01
# Default grid: [[1,3,0,3],[7,5,4,2],[3,0,6,0]] Check if no faults
grid1 = [[1, 3, 0, 3], [7, 5, 4, 2], [3, 0, 6, 0]]
problem = util.USPber(grid=grid1,
multi_agent=True,
tank_capacity=5,
max_turns=20)
plr1 = ep2.CollectAllAgent(player_number=1,
number_of_agents=2,
tank_capacity=5,
max_depth=4)
moves = 0
is_goal = False
state = problem.initial_state()
n_grid, _, nt01g, _, _, _ = state
state1 = (n_grid, nt01g)
while moves < 20 and not is_goal:
with Timeout(15):
act = plr1.get_action(state1)
new_state = problem.next_state(state, act)
new_state = problem.next_state(new_state, 'STOP')
if problem.is_goal_state(new_state):
final_info = problem.get_game_info(new_state)
is_goal = True
else:
n_grid, _, nt01g, _, _, _ = new_state
state1 = (n_grid, nt01g)
final_info = problem.get_game_info(new_state)
state = copy.deepcopy(new_state)
moves += 1
if final_info['agent1_gas_spent'] <= 9:
if final_info['agent1_people'] == 18:
test_results += 1
print("Finished Part 01: Test 01")
print("*************************")
print("Starting Part 01: Test 02")
# Test 02
# Trick straight line [[3,0,1,6,7,3,5,3], [2,0,0,0,0,0,4,0]] fs=11
grid2 = [[3, 0, 1, 6, 7, 3, 5, 3], [2, 0, 0, 0, 0, 0, 4, 0]]
problem = util.USPber(grid=grid2,
multi_agent=True,
tank_capacity=10,
max_turns=30)
plr1 = ep2.CollectAllAgent(player_number=1,
number_of_agents=2,
tank_capacity=10,
max_depth=4)
moves = 0
is_goal = False
state = problem.initial_state()
n_grid, _, nt01g, _, _, _ = state
state1 = (n_grid, nt01g)
while moves < 30 and not is_goal:
with Timeout(15):
act = plr1.get_action(state1)
new_state = problem.next_state(state, act)
new_state = problem.next_state(new_state, 'STOP')
if problem.is_goal_state(new_state):
final_info = problem.get_game_info(new_state)
is_goal = True
else:
n_grid, _, nt01g, _, _, _ = new_state
state1 = (n_grid, nt01g)
final_info = problem.get_game_info(new_state)
state = copy.deepcopy(new_state)
moves += 1
if final_info['agent1_gas_spent'] <= 11:
if final_info['agent1_people'] == 18:
test_results += 1
print("Finished Part 01: Test 02")
print("*************************")
print("Starting Part 01: Test 03")
# Test 03 fs=28
# Refill check [[3,0,0,0,3],[0,0,0,0,0],[0,0,8,0,0],[0,2,0,0,0],[3,0,0,0,3]]
grid3 = [[3, 0, 0, 0, 3], [0, 0, 0, 0, 0], [0, 0, 8, 0, 0],
[0, 2, 0, 0, 0], [3, 0, 0, 0, 3]]
problem = util.USPber(grid=grid3,
multi_agent=True,
tank_capacity=10,
max_turns=70)
plr1 = ep2.CollectAllAgent(player_number=1,
number_of_agents=2,
tank_capacity=10,
max_depth=4)
moves = 0
is_goal = False
state = problem.initial_state()
n_grid, _, nt01g, _, _, _ = state
state1 = (n_grid, nt01g)
while moves < 60 and not is_goal:
with Timeout(20):
act = plr1.get_action(state1)
new_state = problem.next_state(state, act)
new_state = problem.next_state(new_state, 'STOP')
if problem.is_goal_state(new_state):
final_info = problem.get_game_info(new_state)
is_goal = True
else:
n_grid, _, nt01g, _, _, _ = new_state
state1 = (n_grid, nt01g)
final_info = problem.get_game_info(new_state)
state = copy.deepcopy(new_state)
moves += 1
if final_info['agent1_gas_spent'] <= 28:
if final_info['agent1_people'] == 4:
test_results += 1
print("Finished Part 01: Test 03")
print("*************************")
print("Starting Part 01: Test 04")
# Test 04
# Teleporting with no solution on some instances
grid4 = [[3, 0, 0, 1, 0, 0], [2, 0, 0, 4, 0, 0], [0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0]]
plr1 = ep2.CollectAllAgent(player_number=1,
number_of_agents=2,
tank_capacity=8,
max_depth=4)
moves = 0
state1 = (grid4, 8)
list_of_actions = []
still_correct = True
if still_correct:
with Timeout(10):
act = plr1.get_action(state1)
if act != 'LEFT':
still_correct = False
else:
test_results += 0.1
if still_correct:
grid4 = [[3, 0, 1, 0, 0, 0], [2, 0, 0, 4, 0, 0],
[0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0]]
state1 = (grid4, 7)
with Timeout(10):
act = plr1.get_action(state1)
if act != 'LEFT':
still_correct = False
else:
test_results += 0.1
if still_correct:
grid4 = [[0, 1, 0, 0, 0, 0], [2, 0, 0, 4, 0, 0],
[0, 0, 0, 0, 0, 0], [0, 3, 0, 0, 0, 0]]
state1 = (grid4, 6)
with Timeout(10):
act = plr1.get_action(state1)
if act != 'DOWN':
still_correct = False
else:
test_results += 0.1
if still_correct:
grid4 = [[0, 0, 0, 0, 0, 0], [2, 1, 0, 4, 0, 0],
[0, 0, 0, 0, 0, 0], [0, 3, 0, 0, 0, 0]]
state1 = (grid4, 5)
with Timeout(10):
act = plr1.get_action(state1)
if act != 'DOWN':
still_correct = False
else:
test_results += 0.1
if still_correct:
grid4 = [[0, 0, 0, 0, 0, 0], [2, 0, 0, 4, 0, 0],
[0, 1, 0, 0, 0, 3], [0, 0, 0, 0, 0, 0]]
state1 = (grid4, 4)
with Timeout(10):
act = plr1.get_action(state1)
if act != 'RIGHT':
still_correct = False
else:
test_results += 0.1
if still_correct:
grid4 = [[0, 0, 0, 0, 0, 0], [2, 0, 0, 4, 0, 0],
[0, 0, 1, 0, 0, 3], [0, 0, 0, 0, 0, 0]]
state1 = (grid4, 3)
with Timeout(10):
act = plr1.get_action(state1)
if act != 'RIGHT':
still_correct = False
else:
test_results += 0.1
if still_correct:
grid4 = [[0, 0, 0, 0, 0, 3], [2, 0, 0, 4, 0, 0],
[0, 0, 0, 1, 0, 0], [0, 0, 0, 0, 0, 0]]
state1 = (grid4, 2)
with Timeout(10):
act = plr1.get_action(state1)
if act != 'UP':
still_correct = False
else:
test_results += 0.1
if still_correct:
grid4 = [[0, 0, 0, 0, 0, 3], [2, 0, 0, 8, 0, 0],
[0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0]]
state1 = (grid4, 1)
with Timeout(10):
act = plr1.get_action(state1)
if act != 'REFILL':
still_correct = False
else:
test_results += 0.3
if test_results > 0:
final_grades.append((10.0 * test_results) / 4)
else:
final_grades.append(0.0)
print("Finished Part 01: Test 04")
print("*************************")
print("*************************")
print("This autograder does not test Part02")
print("Check final_results.txt for info")
except IOError as e:
file_results.write("Python error: {0}\n".format(e))
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
except NotImplementedError:
file_results.write("NotImplemented\n")
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
except AssertionError as e:
file_results.write("Python error: {0}\n".format(e))
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
except NameError as e:
file_results.write("Python error: {0}\n".format(e))
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
except Timeout.Timeout:
file_results.write("Test did not concluded in time\n")
file_results.write("Receiving proportional grade in the test\n")
final_grades.append((10.0 * test_results) / 116.0)
except Exception as e:
file_results.write("Python error: {0}\n".format(e))
file_results.write("Test did not concluded\n")
file_results.write("Receiving grade 0.0 in the test\n")
else:
file_results.write("Test concluded without errors\n")
finally:
if len(final_grades) < 3:
final_grades.append(0.0)
file_results.write("Grade in the Test01: {0:.1f}/10.0\n".format(
final_grades[2]))
file_results.write("------------------------------\n")
file_results.write("Part02:\n")
file_results.write("This autograder does not test Part02\n")
file_results.write(
"because we need two instances of AlphaBetaAgent:\n")
file_results.write(" -One using evaluation_function\n")
file_results.write(" -One using my_better_evaluation_function\n")
file_results.write("In order to test your implementation\n")
file_results.write(
"you should run both instances against GetClosest\n")
file_results.write("using different game grids, like the ones\n")
file_results.write("presented in interesting_grids.txt.\n")
file_results.write("The agent using my_better should be able to\n")
file_results.write(
"get equal or higher score in most of the examples.")