def test_find_closest_randomized():
RANGES = (-1000, 1000)
DATA_SIZE = 6000
data = [(uniform(*RANGES), uniform(*RANGES)) for _ in xrange(DATA_SIZE)]
tree = KDTree(data)
new_point = (uniform(*RANGES), uniform(*RANGES))
KDTree.find_closest = counted(KDTree.find_closest)
result = tree.find_closest(new_point)
result_distance = KDTree._get_distance(result, new_point)
distances = [KDTree._get_distance(point, new_point) for point in data]
assert_equals(result_distance, min(distances))
# Number of calls will be inconsistent, but should not check all points
print "Number of calls: ", tree.find_closest.calls
assert tree.find_closest.calls < len(data)
def test_find_closest_identical_point():
tree = KDTree([(10, 12), (70, 70), (35, 45)])
new_point = (10, 12)
KDTree.find_closest = counted(KDTree.find_closest)
result = tree.find_closest(new_point)
result_distance = KDTree._get_distance(result, new_point)
assert_equals(result, (10, 12))
assert_equals(result_distance, 0)
assert_equals(tree.find_closest.calls, 2)
