Merge branch 'main' of https://github.com/Farama-Foundation/MO-Gymnasium

mo_gymnasium/envs/highway/__init__.py

@@ -1,6 +1,6 @@
 from gymnasium.envs.registration import register
 
 
-register(id="mo-highway-v0", entry_point="mo_gymnasium.envs.highway.highway:MOHighwayEnv")
+register(id="mo-highway-v0", entry_point="mo_gymnasium.envs.highway.highway:MOHighwayEnv", nondeterministic=True)
 
-register(id="mo-highway-fast-v0", entry_point="mo_gymnasium.envs.highway.highway:MOHighwayEnvFast")
+register(id="mo-highway-fast-v0", entry_point="mo_gymnasium.envs.highway.highway:MOHighwayEnvFast", nondeterministic=True)

mo_gymnasium/envs/minecart/minecart.py

@@ -177,7 +177,7 @@ def __init__(
                 dtype=np.uint8,
             )
         else:
-            self.observation_space = Box(np.zeros(7), np.ones(7), dtype=np.float32)
+            self.observation_space = Box(-np.ones(7), np.ones(7), dtype=np.float32)
 
         self.action_space = Discrete(6)
         self.reward_space = Box(low=-1, high=self.capacity, shape=(self.ore_cnt + 1,))

mo_gymnasium/envs/water_reservoir/dam_env.py

@@ -94,9 +94,9 @@ def reset(self, *, seed: Optional[int] = None, options: Optional[dict] = None):
         if not self.penalize:
             state = self.np_random.choice(DamEnv.s_init, size=1)
         else:
-            state = self.np_random.randint(0, 160, size=1).astype(np.float32)
+            state = self.np_random.randint(0, 160, size=1)
 
-        self.state = np.array(state)
+        self.state = np.array(state, dtype=np.float32)
         return self.state, {}
 
     def step(self, action):
@@ -112,7 +112,7 @@ def step(self, action):
         action = bounded_action
         dam_inflow = self.np_random.normal(DamEnv.DAM_INFLOW_MEAN, DamEnv.DAM_INFLOW_STD, len(self.state))
         # small chance dam_inflow < 0
-        n_state = np.clip(self.state + dam_inflow - action, 0, None)
+        n_state = np.clip(self.state + dam_inflow - action, 0, None).astype(np.float32)
 
         # cost due to excess level wrt a flooding threshold (upstream)
         r0 = -np.clip(n_state / DamEnv.S - DamEnv.H_FLO_U, 0, None) + penalty

tests/test_envs.py

@@ -1,142 +1,134 @@
 import pickle
 
 import gymnasium as gym
+import numpy as np
+import pytest
+from gymnasium.envs.registration import EnvSpec
 from gymnasium.utils.env_checker import check_env, data_equivalence
 
 import mo_gymnasium as mo_gym
-from mo_gymnasium import LinearReward
 
 
-def _test_pickle_env(env: gym.Env):
-    pickled_env = pickle.loads(pickle.dumps(env))
-
-    data_equivalence(env.reset(), pickled_env.reset())
-
-    action = env.action_space.sample()
-    data_equivalence(env.step(action), pickled_env.step(action))
-    env.close()
-    pickled_env.close()
-
-
-def test_deep_sea_treasure():
-    env = mo_gym.make("deep-sea-treasure-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
-
-
-def test_fishwood():
-    env = mo_gym.make("fishwood-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
-
-
-def test_four_room():
-    env = mo_gym.make("four-room-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
-
-
-def test_minecart():
-    env = mo_gym.make("minecart-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
-
-
-def test_mountaincar():
-    env = mo_gym.make("mo-mountaincar-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
-
-
-def test_continuous_mountaincar():
-    env = mo_gym.make("mo-mountaincarcontinuous-v0")
-    env = LinearReward(env)
+all_testing_env_specs = []
+for env_spec in gym.envs.registry.values():
+    if type(env_spec.entry_point) is not str:
+        continue
+    # collect MO Gymnasium envs
+    if env_spec.entry_point.split(".")[0] == "mo_gymnasium":
+        # Ignore highway as they do not deal with the random seed appropriately
+        if not env_spec.id.startswith("mo-highway"):
+            all_testing_env_specs.append(env_spec)
+
+
+@pytest.mark.parametrize(
+    "spec",
+    all_testing_env_specs,
+    ids=[spec.id for spec in all_testing_env_specs],
+)
+def test_all_env_api(spec):
+    """Check that all environments pass the environment checker."""
+    env = mo_gym.make(spec.id)
+    env = mo_gym.LinearReward(env)
     check_env(env, skip_render_check=True)
     _test_pickle_env(env)
 
 
-def test_resource_gathering():
-    env = mo_gym.make("resource-gathering-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+@pytest.mark.parametrize("spec", all_testing_env_specs, ids=[spec.id for spec in all_testing_env_specs])
+def test_all_env_passive_env_checker(spec):
+    env = mo_gym.make(spec.id)
+    env.reset()
+    env.step(env.action_space.sample())
+    env.close()
 
 
-def test_fruit_tree():
-    env = mo_gym.make("fruit-tree-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+# Note that this precludes running this test in multiple threads.
+# However, we probably already can't do multithreading due to some environments.
+SEED = 0
+NUM_STEPS = 50
 
 
-def test_mario():
-    env = mo_gym.make("mo-supermario-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+@pytest.mark.parametrize(
+    "env_spec",
+    all_testing_env_specs,
+    ids=[env.id for env in all_testing_env_specs],
+)
+def test_env_determinism_rollout(env_spec: EnvSpec):
+    """Run a rollout with two environments and assert equality.
+    This test run a rollout of NUM_STEPS steps with two environments
+    initialized with the same seed and assert that:
+    - observation after first reset are the same
+    - same actions are sampled by the two envs
+    - observations are contained in the observation space
+    - obs, rew, done and info are equals between the two envs
+    """
+    # Don't check rollout equality if it's a nondeterministic environment.
+    if env_spec.nondeterministic is True:
+        return
 
+    env_1 = env_spec.make(disable_env_checker=True)
+    env_2 = env_spec.make(disable_env_checker=True)
+    env_1 = mo_gym.LinearReward(env_1)
+    env_2 = mo_gym.LinearReward(env_2)
 
-def test_reacher_pybullet():
-    env = mo_gym.make("mo-reacher-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+    initial_obs_1, initial_info_1 = env_1.reset(seed=SEED)
+    initial_obs_2, initial_info_2 = env_2.reset(seed=SEED)
+    assert_equals(initial_obs_1, initial_obs_2)
 
+    env_1.action_space.seed(SEED)
 
-def test_reacher_mujoco():
-    env = mo_gym.make("mo-reacher-v4")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+    for time_step in range(NUM_STEPS):
+        # We don't evaluate the determinism of actions
+        action = env_1.action_space.sample()
 
+        obs_1, rew_1, terminated_1, truncated_1, info_1 = env_1.step(action)
+        obs_2, rew_2, terminated_2, truncated_2, info_2 = env_2.step(action)
 
-# TODO: failing because highway_env is not deterministic given a random seed
-""" def test_highway():
-    env = mo_gym.make('mo-highway-v0')
-    env = LinearReward(env)
-    check_env(env)
+        assert_equals(obs_1, obs_2, f"[{time_step}] ")
+        assert env_1.observation_space.contains(obs_1)  # obs_2 verified by previous assertion
 
-def test_highway_fast():
-    env = mo_gym.make('mo-highway-fast-v0')
-    env = LinearReward(env)
-    check_env(env) """
-
-
-def test_halfcheetah():
-    env = mo_gym.make("mo-halfcheetah-v4")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+        assert rew_1 == rew_2, f"[{time_step}] reward 1={rew_1}, reward 2={rew_2}"
+        assert terminated_1 == terminated_2, f"[{time_step}] done 1={terminated_1}, done 2={terminated_2}"
+        assert truncated_1 == truncated_2, f"[{time_step}] done 1={truncated_1}, done 2={truncated_2}"
+        assert_equals(info_1, info_2, f"[{time_step}] ")
 
+        if terminated_1 or truncated_1:  # terminated_2, truncated_2 verified by previous assertion
+            env_1.reset(seed=SEED)
+            env_2.reset(seed=SEED)
 
-def test_hopper():
-    env = mo_gym.make("mo-hopper-v4")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+    env_1.close()
+    env_2.close()
 
 
-def test_breakable_bottles():
-    env = mo_gym.make("breakable-bottles-v0")
-    env = LinearReward(env)
-    check_env(env)
-    _test_pickle_env(env)
+def _test_pickle_env(env: gym.Env):
+    pickled_env = pickle.loads(pickle.dumps(env))
 
+    data_equivalence(env.reset(), pickled_env.reset())
 
-def test_water_reservoir():
-    env = mo_gym.make("water-reservoir-v0")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+    action = env.action_space.sample()
+    data_equivalence(env.step(action), pickled_env.step(action))
+    env.close()
+    pickled_env.close()
 
 
-def test_lunar_lander():
-    env = mo_gym.make("mo-lunar-lander-v2")
-    env = LinearReward(env)
-    check_env(env, skip_render_check=True)
-    _test_pickle_env(env)
+def assert_equals(a, b, prefix=None):
+    """Assert equality of data structures `a` and `b`.
+    Args:
+        a: first data structure
+        b: second data structure
+        prefix: prefix for failed assertion message for types and dicts
+    """
+    assert type(a) == type(b), f"{prefix}Differing types: {a} and {b}"
+    if isinstance(a, dict):
+        assert list(a.keys()) == list(b.keys()), f"{prefix}Key sets differ: {a} and {b}"
+
+        for k in a.keys():
+            v_a = a[k]
+            v_b = b[k]
+            assert_equals(v_a, v_b)
+    elif isinstance(a, np.ndarray):
+        np.testing.assert_array_equal(a, b)
+    elif isinstance(a, tuple):
+        for elem_from_a, elem_from_b in zip(a, b):
+            assert_equals(elem_from_a, elem_from_b)
+    else:
+        assert a == b