def test_sample_only_preload(self):
action_space = ActionSpace(self.push_groups, disable_push=True)
action_space.seed(10000)
all_actions = []
num_iters = 150
for _ in range(num_iters):
action_id = action_space.sample()
action_type, push_action, preload_action = (action_id[0], tuple(action_id[1:4]), tuple(action_id[4:]))
if action_type == 0:
assert action_id == NOOP_ACTION_ID, "should be a noop"
all_actions.append((0,))
if 1 <= action_type <= 4:
assert push_action == NOOP_PUSH_ACTION_ID, "push should be a noop"
assert preload_action != NOOP_PRELOAD_ACTION_ID, "preload should not be noop"
assert action_space.push_space.contains(push_action)
all_actions.append((1,))
if 5 <= action_type <= 8:
assert False, "action_type == 2 should not be possible"
# Should return approximately 10-40 proportion of each type of action
action_types = Counter(a for (a, *_) in all_actions)
assert (num_iters * 0.2 * 0.8) <= action_types[0] <= (num_iters * 0.2 * 1.2)
assert (num_iters * 0.8 * 0.9) <= action_types[1] <= (num_iters * 0.8 * 1.1)
assert sum(action_types.values()) == num_iters == len(all_actions)
def test_sample_returns_infinitely_when_actions_not_used(self):
action_space = ActionSpace(self.push_groups)
action_space.seed(20000)
all_actions = []
num_iters = 300
for _ in range(num_iters):
action_id = action_space.sample()
action_type, push_action, preload_action = (action_id[0], tuple(action_id[1:4]), tuple(action_id[4:]))
if action_type == 0:
assert action_id == NOOP_ACTION_ID, "should be a noop"
all_actions.append((0,))
if 1 <= action_type <= 4:
assert push_action != NOOP_PUSH_ACTION_ID, "push should not be a noop"
assert preload_action == NOOP_PRELOAD_ACTION_ID, "preload should be noop"
assert action_space.push_space.contains(push_action)
all_actions.append((1,))
if 5 <= action_type <= 6:
assert push_action == NOOP_PUSH_ACTION_ID, "preload should not be noop"
assert preload_action != NOOP_PRELOAD_ACTION_ID, "push should be a noop"
assert action_space.preload_space.contains(preload_action)
all_actions.append((2,))
# Should return approximately 1-4-1 proportion of each type of action
action_types = Counter(a for (a, *_) in all_actions)
assert (num_iters * (1.0 / 6.0) * 0.75) <= action_types[0] <= (num_iters * (1.0 / 6.0) * 1.25)
assert (num_iters * (4.0 / 6.0) * 0.75) <= action_types[1] <= (num_iters * (4.0 / 6.0) * 1.25)
assert (num_iters * (1.0 / 6.0) * 0.75) <= action_types[2] <= (num_iters * (1.0 / 6.0) * 1.25)
assert sum(action_types.values()) == num_iters == len(all_actions)
def test_init_disabled_preload(self):
action_space = ActionSpace(self.push_groups, disable_preload=True)
assert not action_space.disable_push
assert action_space.disable_preload
assert action_space.num_action_types == 5
assert len(action_space.spaces) == 6
assert action_space.spaces[0].n == 5
assert action_space.spaces[1].n == action_space.push_space.spaces[0].n
assert action_space.spaces[2].n == action_space.push_space.spaces[1].n
assert action_space.spaces[3].n == action_space.push_space.spaces[2].n
assert action_space.spaces[4].n == action_space.preload_space.spaces[0].n
assert action_space.spaces[5].n == action_space.preload_space.spaces[1].n
def test_sample_returns_all_actions_uniquely_when_used(self):
action_space = ActionSpace(self.push_groups)
action_space.seed(10000)
num_iters = 150
num_non_noop = 0
assert not action_space.empty()
all_actions = set()
for i in range(num_iters):
action_id = action_space.sample()
action_type, *_ = action_id
if action_type == 0 and action_space.empty():
break
if action_type != 0:
assert action_id not in all_actions
num_non_noop += 1
all_actions.add(action_id)
action_space.use_action(action_space.decode_action(action_id))
else:
# If there was no break, sample did not stop
print(len(all_actions))
assert False, "sample returned too many items"
assert len(all_actions) == num_non_noop
assert action_space.empty()
def test_init_raises_when_both_push_and_preload_disabled(self):
with pytest.raises(AssertionError):
ActionSpace(self.push_groups, disable_preload=True, disable_push=True)