def test_reward(behavior_spec: BehaviorSpec, reward: float) -> None:
buffer = create_agent_buffer(behavior_spec, 1000, reward)
settings = RewardSignalSettings()
extrinsic_rp = ExtrinsicRewardProvider(behavior_spec, settings)
generated_rewards = extrinsic_rp.evaluate(buffer)
assert (generated_rewards == reward).all()
# Test group rewards. Rewards should be double of the environment rewards, but shouldn't count
# the groupmate rewards.
buffer[BufferKey.GROUP_REWARD] = buffer[BufferKey.ENVIRONMENT_REWARDS]
# 2 agents with identical rewards
buffer[BufferKey.GROUPMATE_REWARDS].set(
[np.ones(1, dtype=np.float32) * reward] * 2
for _ in range(buffer.num_experiences))
generated_rewards = extrinsic_rp.evaluate(buffer)
assert (generated_rewards == 2 * reward).all()
# Test groupmate rewards. Total reward should be indiv_reward + 2 * teammate_reward + group_reward
extrinsic_rp = ExtrinsicRewardProvider(behavior_spec, settings)
extrinsic_rp.add_groupmate_rewards = True
generated_rewards = extrinsic_rp.evaluate(buffer)
assert (generated_rewards == 4 * reward).all()
def test_construction(behavior_spec: BehaviorSpec) -> None:
settings = RewardSignalSettings()
settings.gamma = 0.2
extrinsic_rp = ExtrinsicRewardProvider(behavior_spec, settings)
assert extrinsic_rp.gamma == 0.2
assert extrinsic_rp.name == "Extrinsic"
def test_reward(behavior_spec: BehaviorSpec, reward: float) -> None:
buffer = create_agent_buffer(behavior_spec, 1000, reward)
settings = RewardSignalSettings()
extrinsic_rp = ExtrinsicRewardProvider(behavior_spec, settings)
generated_rewards = extrinsic_rp.evaluate(buffer)
assert (generated_rewards == reward).all()