def __init__(self, env_fns, start_method=None):
self.waiting = False
self.closed = False
n_envs = len(env_fns)
if start_method is None:
# Fork is not a thread safe method (see issue #217)
# but is more user friendly (does not require to wrap the code in
# a `if __name__ == "__main__":`)
fork_available = 'fork' in multiprocessing.get_all_start_methods()
start_method = 'fork' if fork_available else 'spawn'
ctx = multiprocessing.get_context(start_method)
self.remotes, self.work_remotes = zip(
*[ctx.Pipe() for _ in range(n_envs)])
self.processes = []
for work_remote, remote, env_fn in zip(self.work_remotes, self.remotes,
env_fns):
args = (work_remote, remote, CloudpickleWrapper(env_fn))
# daemon=True: if the main process crashes, we should not cause things to hang
process = ctx.Process(target=_worker, args=args, daemon=True)
process.start()
self.processes.append(process)
work_remote.close()
self.remotes[0].send(('get_spaces', None))
observation_space, action_space = self.remotes[0].recv()
VecEnv.__init__(self, len(env_fns), observation_space, action_space)
def __init__(self, env_fns, start_method=None):
self.waiting = False
self.closed = False
n_envs = len(env_fns)
if start_method is None:
# Use thread safe method, see issue #217.
# forkserver faster than spawn but not always available.
forkserver_available = 'forkserver' in multiprocessing.get_all_start_methods()
start_method = 'forkserver' if forkserver_available else 'spawn'
ctx = multiprocessing.get_context(start_method)
self.remotes, self.work_remotes = zip(*[ctx.Pipe() for _ in range(n_envs)])
self.processes = []
for work_remote, remote, env_fn in zip(self.work_remotes, self.remotes, env_fns):
args = (work_remote, remote, CloudpickleWrapper(env_fn))
# daemon=True: if the main process crashes, we should not cause things to hang
process = ctx.Process(target=_worker, args=args, daemon=True)
process.start()
self.processes.append(process)
work_remote.close()
self.remotes[0].send(('get_spaces', None))
observation_space, action_space = self.remotes[0].recv()
VecEnv.__init__(self, len(env_fns), observation_space, action_space)
def __init__(self, env_id, n_agents):
env_path = UnityVecEnv.GetFilePath(env_id, n_agents=n_agents)
print("**** ", env_path)
env = UnityEnv(env_path, multiagent=True)
self.env = env
env.num_envs = env.number_agents
VecEnv.__init__(self, env.num_envs, env.observation_space,
env.action_space)
obs_space = env.observation_space
# self.keys, shapes, dtypes = obs_space_info(obs_space)
# self.buf_obs = { k: np.zeros((self.num_envs,) + tuple(shapes[k]), dtype=dtypes[k]) for k in self.keys }
# self.buf_dones = np.zeros((self.num_envs,), dtype=np.bool)
# self.buf_rews = np.zeros((self.num_envs,), dtype=np.float32)
self.buf_infos = [{} for _ in range(self.num_envs)]
# Fake Monitor
self.tstart = time.time()
self.results_writer = ResultsWriter("filename",
header={
"t_start": time.time(),
'env_id': env.spec
and env.spec.id
},
extra_keys=() + ())
self.reset_keywords = ()
self.info_keywords = ()
self.allow_early_resets = True
self.rewards = None
self.needs_reset = True
self.episode_rewards = []
self.episode_lengths = []
self.episode_times = []
self.total_steps = 0
self.current_reset_info = {
} # extra info about the current episode, that was passed in during reset()
def __init__(self, env_fns):
"""
:param env_fns: ([function])
"""
assert len(env_fns) == 1, "This dummy class does not support multiprocessing"
self.envs = [fn() for fn in env_fns]
env = self.envs[0]
VecEnv.__init__(self, len(env_fns), env.observation_space, env.action_space)
self.env = self.envs[0]
self.actions = None
self.obs = None
self.reward, self.done, self.infos = None, None, None
def __init__(self, env_fns, start_method=None):
"""
@brief Constructor
@warning Only 'forkserver' and 'spawn' start methods are thread-safe, which is
important when TensorFlow sessions or other non thread-safe libraries
are used in the parent.
However, compared to 'fork' they incur a small start-up cost and have
restrictions on global variables. With those methods, users must wrap
the code in an ``if __name__ == "__main__":``
For more information, see the multiprocessing documentation.
@param[in] env_fns List of Gym Environments to run in subprocesses
@param[in] start_method Method used to start the subprocesses. Must be one of the
methods returned by multiprocessing.get_all_start_methods().
Optional: Defaults to 'fork' on available platforms, and 'spawn' otherwise.
@return Instance of SubprocVecEnvLock.
"""
global lock
self.waiting = False
self.closed = False
n_envs = len(env_fns)
if start_method is None:
# Fork is not a thread safe method (see issue #217)
# but is more user friendly (does not require to wrap the code in
# a `if __name__ == "__main__":`)
fork_available = 'fork' in multiprocessing.get_all_start_methods()
start_method = 'fork' if fork_available else 'spawn'
ctx = multiprocessing.get_context(start_method)
self.remotes, self.work_remotes = zip(
*[ctx.Pipe() for _ in range(n_envs)])
self.processes = []
for work_remote, remote, env_fn in zip(self.work_remotes, self.remotes,
env_fns):
args = (work_remote, remote, CloudpickleWrapper(env_fn), lock)
# daemon=True: if the main process crashes, we should not cause things to hang
process = ctx.Process(target=_worker, args=args, daemon=True)
process.start()
self.processes.append(process)
work_remote.close()
self.remotes[0].send(('get_spaces', None))
observation_space, action_space = self.remotes[0].recv()
VecEnv.__init__(self, len(env_fns), observation_space, action_space)
def __init__(self, env_fns, **env_args):
self.envs = [fn(**env_args) for fn in env_fns]
env = self.envs[0]
VecEnv.__init__(self, len(env_fns), env.observation_space,
env.action_space)
obs_space = env.observation_space
self.keys, shapes, dtypes = obs_space_info(obs_space)
self.buf_obs = OrderedDict([(k,
np.zeros(
(self.num_envs, ) + tuple(shapes[k]),
dtype=dtypes[k])) for k in self.keys])
self.buf_dones = np.zeros((self.num_envs, ), dtype=np.bool)
self.buf_rews = np.zeros((self.num_envs, ), dtype=np.float32)
self.buf_infos = [{} for _ in range(self.num_envs)]
self.actions = None
def __init__(self, env_fns):
self.waiting = False
self.closed = False
n_envs = len(env_fns)
self.remotes, self.work_remotes = zip(*[Pipe() for _ in range(n_envs)])
self.processes = [Process(target=_worker, args=(work_remote, remote, CloudpickleWrapper(env_fn)))
for (work_remote, remote, env_fn) in zip(self.work_remotes, self.remotes, env_fns)]
for process in self.processes:
process.daemon = True # if the main process crashes, we should not cause things to hang
process.start()
for remote in self.work_remotes:
remote.close()
self.remotes[0].send(('get_spaces', None))
observation_space, action_space = self.remotes[0].recv()
VecEnv.__init__(self, len(env_fns), observation_space, action_space)
def __init__(self, env_fns, create_method):
self.waiting = False
self.closed = False
n_envs = len(env_fns)
self.remotes, self.work_remotes = zip(
*[create_pipe() for _ in range(n_envs)])
self.processes = []
for work_remote, env_fn in zip(self.work_remotes, env_fns):
args = (work_remote, CloudpickleWrapper(env_fn))
# daemon=True: if the main process crashes, we should not cause
# things to hang
process = create_method(target=_worker, args=args, daemon=True)
process.start()
self.processes.append(process)
self.remotes[0].send(('get_spaces', None))
observation_space, action_space = self.remotes[0].recv()
VecEnv.__init__(self, len(env_fns), observation_space, action_space)
def __init__(self, num_envs, num_agents, observation_space, action_space):
VecEnv.__init__(self, num_envs, observation_space, action_space)
self.num_agents = num_agents