def test_encode_with_treasure_position_layer(self):
state = GridWorld(4, 4)
state.add_object(Treasure(2, 3))
encoder = OneHotEncoder(4, 4, agent_position=False, treasure_position=True)
expected = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0]
assert_array_equal(expected, encoder.encode(state))
def define_agent(self, width, height, num_actions):
return NStepDQNAgent(
config=Config(num_actions=num_actions,
encoder=OneHotEncoder(width, height),
optimizer=AdamOptimizer(0.01),
network=MLP(),
policy=EpsilonGreedyPolicy(1, 0.01, 1000),
discount=0.95,
n_step=8))
def define_agent(self, width, height, num_actions):
return DQNAgent(config=Config(num_actions=num_actions,
encoder=OneHotEncoder(width, height),
optimizer=AdamOptimizer(0.01),
network=MLP(),
policy=EpsilonGreedyPolicy(1, 0.01, 500),
discount=0.95,
capacity=100,
batch_size=16))
def test_shape_with_agent_position_layer(self):
encoder = OneHotEncoder(4, 4, agent_position=True, treasure_position=False)
self.assertEqual(16, encoder.shape())
def test_encode_with_no_layers(self):
state = GridWorld(4, 4)
encoder = OneHotEncoder(4, 4, agent_position=False, treasure_position=False)
assert_array_equal(np.empty(0), encoder.encode(state))
def test_shape_with_multiple_layers(self):
encoder = OneHotEncoder(4, 4, agent_position=True, treasure_position=True)
self.assertEqual(32, encoder.shape())