Merge pull request #58 from PettingZoo-Team/utilities_fixes

requirements.txt

@@ -1,3 +1,3 @@
-pettingzoo[butterfly]>=1.4.0
+pettingzoo[butterfly]>=1.6.0
 opencv-python
 cloudpickle

setup.py

@@ -27,7 +27,7 @@ def get_version():
     keywords=["Reinforcement Learning", "gym"],
     packages=setuptools.find_packages(),
     python_requires=">=3.6",
-    install_requires=["pettingzoo>=1.4.0", "opencv-python~=3.4.0", "cloudpickle"],
+    install_requires=["pettingzoo>=1.6.0", "opencv-python~=3.4.0", "cloudpickle"],
     classifiers=[
         "Programming Language :: Python :: 3.6",
         "Programming Language :: Python :: 3.7",

supersuit/__init__.py

@@ -1,13 +1,13 @@
 import gym
-from pettingzoo.utils.to_parallel import to_parallel, ParallelEnv, from_parallel
+from pettingzoo.utils.conversions import to_parallel, ParallelEnv, from_parallel
 from pettingzoo.utils.env import AECEnv
 from . import aec_wrappers
 from . import gym_wrappers
 from . import parallel_wrappers
 from . import vector_constructors
 from . import aec_vector
 
-__version__ = "2.5.1"
+__version__ = "2.6.0"
 
 
 class WrapperFactory:

supersuit/aec_wrappers.py

@@ -180,6 +180,7 @@ def _update_items(self):
             self._cumulative_rewards[agent] = self.env._cumulative_rewards.get(agent, 0)
 
     def step(self, action):
+        self._has_updated = True
         if self.dones[self.agent_selection]:
             return self._was_done_step(action)
 
@@ -327,6 +328,7 @@ def observe(self, agent):
         return fin_observe if fin_observe is not None else super().observe(agent)
 
     def step(self, action):
+        self._has_updated = True
         if self.dones[self.agent_selection]:
             if self.env.agents and self.agent_selection == self.env.agent_selection:
                 self.env.step(None)

supersuit/parallel_wrappers.py

@@ -1,4 +1,4 @@
-from pettingzoo.utils.to_parallel import ParallelEnv
+from pettingzoo.utils.conversions import ParallelEnv
 import gym
 from gym.spaces import Box, Discrete
 from .adv_transforms.frame_stack import stack_obs_space, stack_init, stack_obs

test/parallel_env_test.py

@@ -1,4 +1,4 @@
-from pettingzoo.utils.to_parallel import ParallelEnv
+from pettingzoo.utils import ParallelEnv
 from gym.spaces import Box, Discrete
 import numpy as np
 import supersuit

test/pettingzoo_api_test.py

@@ -1,7 +1,7 @@
 import numpy as np
 from pettingzoo.test import api_test, seed_test, parallel_test
 from pettingzoo.mpe import simple_push_v2, simple_world_comm_v2
-from pettingzoo.butterfly import knights_archers_zombies_v6, prison_v2
+from pettingzoo.butterfly import knights_archers_zombies_v7, prison_v3
 
 import supersuit
 import pytest
@@ -42,30 +42,30 @@ def test_pettingzoo_pad_action_space():
 def test_pettingzoo_parallel_env():
     _env = simple_world_comm_v2.parallel_env()
     wrapped_env = pad_action_space_v0(_env)
-    parallel_test.parallel_play_test(wrapped_env)
+    parallel_test.parallel_api_test(wrapped_env)
 
 
 wrappers = [
-    supersuit.color_reduction_v0(knights_archers_zombies_v6.env(), "R"),
-    supersuit.resize_v0(dtype_v0(knights_archers_zombies_v6.env(), np.uint8), x_size=5, y_size=10),
-    supersuit.resize_v0(dtype_v0(knights_archers_zombies_v6.env(), np.uint8), x_size=5, y_size=10, linear_interp=True),
-    supersuit.dtype_v0(knights_archers_zombies_v6.env(), np.int32),
-    supersuit.flatten_v0(knights_archers_zombies_v6.env()),
-    supersuit.reshape_v0(knights_archers_zombies_v6.env(), (512 * 512, 3)),
-    supersuit.normalize_obs_v0(dtype_v0(knights_archers_zombies_v6.env(), np.float32), env_min=-1, env_max=5.0),
-    supersuit.frame_stack_v1(knights_archers_zombies_v6.env(), 8),
-    supersuit.pad_observations_v0(knights_archers_zombies_v6.env()),
-    supersuit.pad_action_space_v0(knights_archers_zombies_v6.env()),
-    supersuit.black_death_v1(knights_archers_zombies_v6.env()),
-    supersuit.agent_indicator_v0(knights_archers_zombies_v6.env(), True),
-    supersuit.agent_indicator_v0(knights_archers_zombies_v6.env(), False),
-    supersuit.reward_lambda_v0(knights_archers_zombies_v6.env(), lambda x: x / 10),
-    supersuit.clip_reward_v0(knights_archers_zombies_v6.env()),
-    supersuit.clip_actions_v0(prison_v2.env(continuous=True)),
-    supersuit.frame_skip_v0(knights_archers_zombies_v6.env(), 4),
-    supersuit.sticky_actions_v0(knights_archers_zombies_v6.env(), 0.75),
-    supersuit.delay_observations_v0(knights_archers_zombies_v6.env(), 3),
-    supersuit.max_observation_v0(knights_archers_zombies_v6.env(), 3),
+    supersuit.color_reduction_v0(knights_archers_zombies_v7.env(), "R"),
+    supersuit.resize_v0(dtype_v0(knights_archers_zombies_v7.env(), np.uint8), x_size=5, y_size=10),
+    supersuit.resize_v0(dtype_v0(knights_archers_zombies_v7.env(), np.uint8), x_size=5, y_size=10, linear_interp=True),
+    supersuit.dtype_v0(knights_archers_zombies_v7.env(), np.int32),
+    supersuit.flatten_v0(knights_archers_zombies_v7.env()),
+    supersuit.reshape_v0(knights_archers_zombies_v7.env(), (512 * 512, 3)),
+    supersuit.normalize_obs_v0(dtype_v0(knights_archers_zombies_v7.env(), np.float32), env_min=-1, env_max=5.0),
+    supersuit.frame_stack_v1(knights_archers_zombies_v7.env(), 8),
+    supersuit.pad_observations_v0(knights_archers_zombies_v7.env()),
+    supersuit.pad_action_space_v0(knights_archers_zombies_v7.env()),
+    supersuit.black_death_v1(knights_archers_zombies_v7.env()),
+    supersuit.agent_indicator_v0(knights_archers_zombies_v7.env(), True),
+    supersuit.agent_indicator_v0(knights_archers_zombies_v7.env(), False),
+    supersuit.reward_lambda_v0(knights_archers_zombies_v7.env(), lambda x: x / 10),
+    supersuit.clip_reward_v0(knights_archers_zombies_v7.env()),
+    supersuit.clip_actions_v0(prison_v3.env(continuous=True)),
+    supersuit.frame_skip_v0(knights_archers_zombies_v7.env(), 4),
+    supersuit.sticky_actions_v0(knights_archers_zombies_v7.env(), 0.75),
+    supersuit.delay_observations_v0(knights_archers_zombies_v7.env(), 3),
+    supersuit.max_observation_v0(knights_archers_zombies_v7.env(), 3),
 ]
 
 
@@ -75,16 +75,16 @@ def test_pettingzoo_aec_api(env):
 
 
 parallel_wrappers = [
-    supersuit.frame_stack_v1(knights_archers_zombies_v6.parallel_env(), 8),
-    supersuit.reward_lambda_v0(knights_archers_zombies_v6.parallel_env(), lambda x: x / 10),
-    supersuit.delay_observations_v0(knights_archers_zombies_v6.parallel_env(), 3),
-    supersuit.color_reduction_v0(knights_archers_zombies_v6.parallel_env(), "R"),
-    supersuit.frame_skip_v0(knights_archers_zombies_v6.parallel_env(), 4),
-    supersuit.max_observation_v0(knights_archers_zombies_v6.parallel_env(), 4),
-    supersuit.black_death_v1(knights_archers_zombies_v6.parallel_env()),
+    supersuit.frame_stack_v1(knights_archers_zombies_v7.parallel_env(), 8),
+    supersuit.reward_lambda_v0(knights_archers_zombies_v7.parallel_env(), lambda x: x / 10),
+    supersuit.delay_observations_v0(knights_archers_zombies_v7.parallel_env(), 3),
+    supersuit.color_reduction_v0(knights_archers_zombies_v7.parallel_env(), "R"),
+    supersuit.frame_skip_v0(knights_archers_zombies_v7.parallel_env(), 4),
+    supersuit.max_observation_v0(knights_archers_zombies_v7.parallel_env(), 4),
+    supersuit.black_death_v1(knights_archers_zombies_v7.parallel_env()),
 ]
 
 
 @pytest.mark.parametrize("env", parallel_wrappers)
 def test_pettingzoo_parallel_api(env):
-    parallel_test.parallel_play_test(env)
+    parallel_test.parallel_api_test(env)

test/test_vector/test_aec_vector_identity_env.py

@@ -1,4 +1,4 @@
-from pettingzoo.butterfly import knights_archers_zombies_v6
+from pettingzoo.butterfly import knights_archers_zombies_v7
 from pettingzoo.mpe import simple_push_v2
 from pettingzoo.sisl import waterworld_v3
 from supersuit import frame_skip_v0, vectorize_aec_env_v0
@@ -7,12 +7,12 @@
 
 def test_identical():
     def env_fn():
-        return knights_archers_zombies_v6.env()  # ,20)
+        return knights_archers_zombies_v7.env()  # ,20)
 
     n_envs = 2
     # single threaded
-    env1 = vectorize_aec_env_v0(knights_archers_zombies_v6.env(), n_envs)
-    env2 = vectorize_aec_env_v0(knights_archers_zombies_v6.env(), n_envs, num_cpus=1)
+    env1 = vectorize_aec_env_v0(knights_archers_zombies_v7.env(), n_envs)
+    env2 = vectorize_aec_env_v0(knights_archers_zombies_v7.env(), n_envs, num_cpus=1)
     env1.seed(42)
     env2.seed(42)
     env1.reset()

test/test_vector/test_aec_vector_values.py

@@ -1,7 +1,7 @@
 from supersuit import vectorize_aec_env_v0
 from pettingzoo.classic import rps_v1
 from pettingzoo.classic import mahjong_v2, hanabi_v3
-from pettingzoo.butterfly import knights_archers_zombies_v6
+from pettingzoo.butterfly import knights_archers_zombies_v7
 from pettingzoo.mpe import simple_world_comm_v2
 from pettingzoo.classic import chess_v0
 import numpy as np
@@ -67,5 +67,5 @@ def select_action(vec_env, passes, i):
         test_vec_env(vectorize_aec_env_v0(mahjong_maker(), NUM_ENVS, num_cpus=num_cpus))
         test_infos(vectorize_aec_env_v0(hanabi_maker(), NUM_ENVS, num_cpus=num_cpus))
         test_some_done(vectorize_aec_env_v0(mahjong_maker(), NUM_ENVS, num_cpus=num_cpus))
-        test_vec_env(vectorize_aec_env_v0(knights_archers_zombies_v6.env(), NUM_ENVS, num_cpus=num_cpus))
+        test_vec_env(vectorize_aec_env_v0(knights_archers_zombies_v7.env(), NUM_ENVS, num_cpus=num_cpus))
         test_vec_env(vectorize_aec_env_v0(simple_world_comm_v2.env(), NUM_ENVS, num_cpus=num_cpus))

test/test_vector/test_pettingzoo_to_vec.py

@@ -1,5 +1,5 @@
 from pettingzoo.mpe import simple_spread_v2, simple_world_comm_v2
-from pettingzoo.butterfly import knights_archers_zombies_v6
+from pettingzoo.butterfly import knights_archers_zombies_v7
 from supersuit import pettingzoo_env_to_vec_env_v0, black_death_v1
 import pytest
 
@@ -30,7 +30,7 @@ def test_bad_action_spaces_env():
 
 
 def test_env_black_death_assertion():
-    env = knights_archers_zombies_v6.parallel_env(spawn_rate=50, max_cycles=2000)
+    env = knights_archers_zombies_v7.parallel_env(spawn_rate=50, max_cycles=2000)
     env = pettingzoo_env_to_vec_env_v0(env)
     with pytest.raises(AssertionError):
         for i in range(100):
@@ -41,7 +41,7 @@ def test_env_black_death_assertion():
 
 
 def test_env_black_death_wrapper():
-    env = knights_archers_zombies_v6.parallel_env(spawn_rate=50, max_cycles=300)
+    env = knights_archers_zombies_v7.parallel_env(spawn_rate=50, max_cycles=300)
     env = black_death_v1(env)
     env = pettingzoo_env_to_vec_env_v0(env)
     env.reset()