def test_measure_probability_clustering():
# Assert that the measured resource clustering of each known layout is correct
for layout_data in known_probability_clustering_layouts:
assert measure_probability_clustering(
Board(
terrain_types=layout_data[0]['terrain_types'],
tile_numbers=layout_data[0]['tile_numbers'])) == layout_data[1]
def test_measure_probability_distribution_per_resources():
# Assert that the measured resource clustering of each known layout is correct
for layout_data in known_probability_distribution_layouts:
assert round(
measure_probability_distribution_per_resources(
Board(terrain_types=layout_data[0]['terrain_types'],
tile_numbers=layout_data[0]['tile_numbers'])),
4) == round(layout_data[1], 4)
def generate_random_board():
# Uses random.shuffle() on each required component of the beginner_layout (the terrain_types and tile_numbers arrays
# ) to build and return a randomized board. We shuffle copies of the arrays to leave beginner_layout unaffected
terrain_types = beginner_layout['terrain_types'].copy()
tile_numbers = beginner_layout['tile_numbers'].copy()
random.shuffle(terrain_types)
random.shuffle(tile_numbers)
return Board(terrain_types=terrain_types, tile_numbers=tile_numbers)
def find_best_board(num_random, is_seeking_balanced):
# The current best (either highest or lowest, as determined by is_seeking_balanced) returned score
current_best_score = 1 if is_seeking_balanced else 0
# The Board that resulted in the current_best_score - defaults to the beginner layout so that a valid board will
# always be returned, regardless of the value of num_random
current_best_board = Board(terrain_types=beginner_layout['terrain_types'],
tile_numbers=beginner_layout['tile_numbers'])
for board_num in range(num_random if num_random > 0 else 0):
board_to_test = generate_random_board()
balance_score = calculate_balance(board_to_test)
# Update the current_best_score and current_best board accordingly, based on is_seeking_balanced
if (is_seeking_balanced and balance_score <= current_best_score) or (
not is_seeking_balanced and balance_score >= current_best_score):
current_best_board = board_to_test
current_best_score = balance_score
return current_best_board, current_best_score
def test_measure_resource_distribution():
# Assert that the measured resource distribution of the perfectly distributed layout is 0
assert measure_resource_distribution(
Board(terrain_types=perfectly_distributed_layout['terrain_types'],
tile_numbers=perfectly_distributed_layout['tile_numbers'])) == 0