def test_ac3_two(self):
all_values = set()
for var in self.const_problem2.get_variables():
for val in var.domain:
all_values.add(val)
self.assertEqual(all_values, {2, 4, 5})
res2 = csp.ac3(self.const_problem2)
self.assertTrue(res2)
reduced_all_values = set()
for var in self.const_problem2.get_variables():
for val in var.domain:
reduced_all_values.add(val)
self.assertEqual(reduced_all_values, {2, 4})
"constraints_weighting",
10000,
with_history=True)
measure_performance(2, "n_queens_problem", n_queens_problem,
"simulated_annealing", 100000, 0.5, 0.99999)
measure_performance(2, "n_queens_problem", n_queens_problem,
"random_restart_first_choice_hill_climbing", 100, 100, 10)
general_genetic_n_queens_problem = csp.GeneralGeneticConstraintProblem(
n_queens_problem, 0.1)
measure_performance(2, "n_queens_problem", general_genetic_n_queens_problem,
"genetic_local_search", 100, 100, 0.1)
ac3_n_queens_problem = copy.deepcopy(n_queens_problem)
ac3_n_queens_problem.unassign_all_variables()
ac3_start_time = time.process_time()
ac3_is_arc_consistent = csp.ac3(ac3_n_queens_problem)
ac3_end_time = time.process_time()
if ac3_is_arc_consistent:
print()
print()
print("-" * 145)
print("using ac3 as a preprocessing stage which took",
ac3_end_time - ac3_start_time, "seconds")
print("-" * 145)
measure_performance(1,
"ac3_n_queens_problem",
ac3_n_queens_problem,
"backtracking_search",
with_history=True)
measure_performance(1,
"ac3_n_queens_problem",
measure_performance(2, "sudoku_problem", sudoku_problem,
"simulated_annealing", 100000, 0.5, 0.99999, sudoku_start_state_generator,
sudoku_successor_generator, sudoku_score_calculator, read_only_variables=read_only_variables)
measure_performance(2, "sudoku_problem", sudoku_problem,
"random_restart_first_choice_hill_climbing", 100, 10, 10, sudoku_start_state_generator,
sudoku_successor_generator, sudoku_score_calculator, read_only_variables=read_only_variables)
general_genetic_sudoku_problem = GeneticSudokuConstraintProblem(sudoku_problem, read_only_names, 0.1)
measure_performance(2, "general_genetic_sudoku_problem", general_genetic_sudoku_problem,
"genetic_local_search", 100, 100, 0.1, read_only_variables=read_only_variables)
genetic_sudoku_problem = GeneticSudokuConstraintProblem(sudoku_problem, read_only_names, 0.1)
measure_performance(2, "sudoku_genetic_sudoku_problem", genetic_sudoku_problem,
"genetic_local_search", 100, 100, 0.1, read_only_variables=read_only_variables)
ac3_start_time = time.process_time()
ac3_is_arc_consistent = csp.ac3(ac3_sudoku_problem)
ac3_end_time = time.process_time()
if ac3_is_arc_consistent:
print()
print()
print("-" * 145)
print("using ac3 as a preprocessing stage which took", ac3_end_time - ac3_start_time, "seconds")
print("-" * 145)
measure_performance(1, "ac3_sudoku_problem", ac3_sudoku_problem,
"backtracking_search", with_history=True, read_only_variables=read_only_variables)
measure_performance(1, "ac3_sudoku_problem", ac3_sudoku_problem,
"backtracking_search", inference=csp.forward_check, with_history=True,
read_only_variables=read_only_variables)
measure_performance(1, "ac3_sudoku_problem", ac3_sudoku_problem,
"heuristic_backtracking_search", with_history=True,
read_only_variables=read_only_variables)
10000,
with_history=True)
measure_performance(2, "verbal_arithmetic_problem", verbal_arithmetic_problem,
"simulated_annealing", 100000, 0.5, 0.99999)
measure_performance(2, "verbal_arithmetic_problem", verbal_arithmetic_problem,
"random_restart_first_choice_hill_climbing", 100, 100, 10)
general_genetic_verbal_arithmetic_problem = csp.GeneralGeneticConstraintProblem(
verbal_arithmetic_problem, 0.1)
measure_performance(2, "verbal_arithmetic_problem",
general_genetic_verbal_arithmetic_problem,
"genetic_local_search", 100, 100, 0.1)
ac3_verbal_arithmetic_problem = copy.deepcopy(verbal_arithmetic_problem)
ac3_verbal_arithmetic_problem.unassign_all_variables()
ac3_start_time = time.process_time()
ac3_is_arc_consistent = csp.ac3(ac3_verbal_arithmetic_problem)
ac3_end_time = time.process_time()
if ac3_is_arc_consistent:
print()
print()
print("-" * 145)
print("using ac3 as a preprocessing stage which took",
ac3_end_time - ac3_start_time, "seconds")
print("-" * 145)
measure_performance(1,
"ac3_verbal_arithmetic_problem",
ac3_verbal_arithmetic_problem,
"backtracking_search",
with_history=True)
measure_performance(1,
"ac3_verbal_arithmetic_problem",
100000,
with_history=True)
measure_performance(2, "car_assembly_problem", car_assembly_problem,
"simulated_annealing", 100000, 0.5, 0.99999)
measure_performance(2, "car_assembly_problem", car_assembly_problem,
"random_restart_first_choice_hill_climbing", 10, 10, 10)
general_genetic_car_assembly_problem = csp.GeneralGeneticConstraintProblem(
car_assembly_problem, 0.1)
measure_performance(2, "car_assembly_problem",
general_genetic_car_assembly_problem,
"genetic_local_search", 100, 100, 0.1)
ac3_car_assembly_problem = copy.deepcopy(car_assembly_problem)
ac3_car_assembly_problem.unassign_all_variables()
ac3_start_time = time.process_time()
ac3_is_arc_consistent = csp.ac3(ac3_car_assembly_problem)
ac3_end_time = time.process_time()
if ac3_is_arc_consistent:
print()
print()
print("-" * 145)
print("using ac3 as a preprocessing stage which took",
ac3_end_time - ac3_start_time, "seconds")
print("-" * 145)
measure_performance(1,
"ac3_car_assembly_problem",
ac3_car_assembly_problem,
"backtracking_search",
with_history=True)
measure_performance(1,
"ac3_car_assembly_problem",
10000,
with_history=True)
measure_performance(2, "map_coloring_problem", map_coloring_problem,
"simulated_annealing", 100000, 0.5, 0.99999)
measure_performance(2, "map_coloring_problem", map_coloring_problem,
"random_restart_first_choice_hill_climbing", 100, 100, 10)
general_genetic_map_coloring_problem = csp.GeneralGeneticConstraintProblem(
map_coloring_problem, 0.1)
measure_performance(2, "map_coloring_problem",
general_genetic_map_coloring_problem,
"genetic_local_search", 100, 100, 0.1)
ac3_map_coloring_problem = copy.deepcopy(map_coloring_problem)
ac3_map_coloring_problem.unassign_all_variables()
ac3_start_time = time.process_time()
ac3_is_arc_consistent = csp.ac3(ac3_map_coloring_problem)
ac3_end_time = time.process_time()
if ac3_is_arc_consistent:
print()
print()
print("-" * 145)
print("using ac3 as a preprocessing stage which took",
ac3_end_time - ac3_start_time, "seconds")
print("-" * 145)
measure_performance(1,
"ac3_map_coloring_problem",
ac3_map_coloring_problem,
"backtracking_search",
with_history=True)
measure_performance(1,
"ac3_map_coloring_problem",
def tes_ac3_three(self):
res = csp.ac3(self.const_problem3)
self.assertTrue(res)
wanted_reduced_domains = [[1, 2], [2, 3]]
for var in self.const_problem3.get_variables():
self.assertIn(var.domain, wanted_reduced_domains)
def test_ac3_one(self):
self.const_problem1.unassign_all_variables()
res1 = csp.ac3(self.const_problem1)
self.assertTrue(res1)
"constraints_weighting",
10000,
with_history=True)
measure_performance(2, "einstein_problem", einstein_problem,
"simulated_annealing", 100000, 0.5, 0.99999)
measure_performance(2, "einstein_problem", einstein_problem,
"random_restart_first_choice_hill_climbing", 100, 100, 10)
general_genetic_einstein_problem = csp.GeneralGeneticConstraintProblem(
einstein_problem, 0.1)
measure_performance(2, "einstein_problem", general_genetic_einstein_problem,
"genetic_local_search", 100, 100, 0.1)
ac3_einstein_problem = copy.deepcopy(einstein_problem)
ac3_einstein_problem.unassign_all_variables()
ac3_start_time = time.process_time()
ac3_is_arc_consistent = csp.ac3(ac3_einstein_problem)
ac3_end_time = time.process_time()
if ac3_is_arc_consistent:
print()
print()
print("-" * 145)
print("using ac3 as a preprocessing stage which took",
ac3_end_time - ac3_start_time, "seconds")
print("-" * 145)
measure_performance(1,
"ac3_einstein_problem",
ac3_einstein_problem,
"heuristic_backtracking_search",
with_history=True)
measure_performance(1,
"ac3_einstein_problem",
"min_conflicts", 100000, with_history=True)
measure_performance(2, "magic_square_problem", magic_square_problem,
"constraints_weighting", 10000, with_history=True)
measure_performance(2, "magic_square_problem", magic_square_problem,
"simulated_annealing", 200000, 0.5, 0.99999)
measure_performance(2, "magic_square_problem", magic_square_problem,
"random_restart_first_choice_hill_climbing", 100, 100, 10)
general_genetic_magic_square_problem = csp.GeneralGeneticConstraintProblem(magic_square_problem, 0.1)
measure_performance(2, "general_genetic_magic_square_problem", general_genetic_magic_square_problem,
"genetic_local_search", 1000, 1000, 0.1)
ac3_magic_square_problem = copy.deepcopy(magic_square_problem)
ac3_magic_square_problem.unassign_all_variables()
ac3_start_time = time.process_time()
ac3_is_arc_consistent = csp.ac3(ac3_magic_square_problem)
ac3_end_time = time.process_time()
if ac3_is_arc_consistent:
print()
print()
print("-" * 145)
print("using ac3 as a preprocessing stage which took", ac3_end_time - ac3_start_time, "seconds")
print("-" * 145)
measure_performance(1, "ac3_magic_square_problem", ac3_magic_square_problem,
"backtracking_search", with_history=True)
measure_performance(1, "ac3_magic_square_problem", ac3_magic_square_problem,
"backtracking_search", inference=csp.forward_check, with_history=True)
measure_performance(1, "ac3_magic_square_problem", ac3_magic_square_problem,
"heuristic_backtracking_search", with_history=True)
measure_performance(1, "ac3_magic_square_problem", ac3_magic_square_problem,
"heuristic_backtracking_search", inference=csp.forward_check, with_history=True)