def env_fn(**kwargs): env = raw_env(**kwargs) env = wrappers.AssertOutOfBoundsWrapper(env) backup_policy = "taking zero action (no movement, no attack)" env = wrappers.NanNoOpWrapper(env, 0, backup_policy) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) example_space = list(env.action_spaces.values())[0] env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NanNoOpWrapper(env, np.zeros(example_space.shape, dtype=example_space.dtype), "taking all zeros action") env = wrappers.OrderEnforcingWrapper(env) return env
def env(): env = raw_env() env = wrappers.CaptureStdoutWrapper(env) env = wrappers.TerminateIllegalWrapper(env, illegal_reward=-1) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) env = wrappers.AssertOutOfBoundsWrapper(env) default_val = 1 env = wrappers.NanNoOpWrapper(env, default_val, "setting action to 1") env = wrappers.OrderEnforcingWrapper(env) return env
def env(): env = raw_env() env = wrappers.CaptureStdoutWrapper(env) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NaNRandomWrapper(env) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) env = wrappers.TerminateIllegalWrapper(env, illegal_reward=-1) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NanNoOpWrapper(env, 26**2 * 2, "executing the 'do nothing' action.") env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) env = wrappers.TerminateIllegalWrapper(env, illegal_reward=-1) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NaNRandomWrapper(env) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) pass_move = env._N * env._N env = wrappers.TerminateIllegalWrapper(env, illegal_reward=-1) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NanNoOpWrapper(env, pass_move, "passing turn with action {}".format(pass_move)) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = r_env = raw_env(**kwargs) env = wrappers.CaptureStdoutWrapper(env) env = wrappers.TerminateIllegalWrapper( env, illegal_reward=HanabiScorePenalty(r_env)) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) env = wrappers.TerminateIllegalWrapper(env, illegal_reward=-1) env = wrappers.AssertOutOfBoundsWrapper(env) pass_action = 3 env = wrappers.NanNoOpWrapper(env, pass_action, "'checked' with action {}".format(pass_action)) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) if env.continuous: env = wrappers.ClipOutOfBoundsWrapper(env) else: env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) default_val = np.zeros((1, )) if env.continuous else 1 env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NanNoOpWrapper(env, default_val, "setting action to {}".format(default_val)) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) player_losing_reward = -3 env = wrappers.CaptureStdoutWrapper(env) env = wrappers.TerminateIllegalWrapper(env, illegal_reward=player_losing_reward) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NaNRandomWrapper(env) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) env = wrappers.CaptureStdoutWrapper(env) env = wrappers.TerminateIllegalWrapper(env, illegal_reward=-1) env = wrappers.AssertOutOfBoundsWrapper(env) pass_move = 308 env = wrappers.NanNoOpWrapper(env, pass_move, "passing turn with action number {}".format(pass_move)) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) world = env.world env = wrappers.AssertOutOfBoundsWrapper(env) backup_policy = "taking zero action (no movement, communication 0)" env = wrappers.NanNoOpWrapper(env, 0, backup_policy) env = wrappers.OrderEnforcingWrapper(env) env.world = world return env
def env(**kwargs): env = raw_env(**kwargs) if env.continuous: default_val = np.zeros((1, )) env = wrappers.ClipOutOfBoundsWrapper(env) else: default_val = 0 env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NanNoOpWrapper(env, default_val, "setting action to 0") env = wrappers.OrderEnforcingWrapper(env) return env
def env(): ''' The env function often wraps the environment in wrappers by default. You can find full documentation for these methods elsewhere in the developer documentation. ''' env = raw_env() # This wrapper is only for environments which print results to the terminal env = wrappers.CaptureStdoutWrapper(env) # this wrapper helps error handling for discrete action spaces env = wrappers.AssertOutOfBoundsWrapper(env) # Provides a wide vareity of helpful user errors # Strongly recommended env = wrappers.OrderEnforcingWrapper(env) return env
def test_unwrapped(name, env_module): env = env_module.env() base_env = env.unwrapped env.reset() agents = env.agents if discrete_observation(env, agents): env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.BaseWrapper(env) env = wrappers.CaptureStdoutWrapper(env) if box_observation(env, agents) and box_action(env, agents): env = wrappers.ClipOutOfBoundsWrapper(env) env = wrappers.OrderEnforcingWrapper(env) env = wrappers.TerminateIllegalWrapper(env, 1.0) if env.metadata["is_parallelizable"]: env = conversions.aec_to_parallel(env) env = conversions.parallel_to_aec(env) env = conversions.turn_based_aec_to_parallel(env) assert env.unwrapped == base_env, "Unwrapped Test: unequal envs"
def env(**kwargs): env = SumoEnvironmentPZ(**kwargs) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NanNoOpWrapper(env, 0, "doing nothing") env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.OrderEnforcingWrapper(env) return env
def env(**kwargs): env = raw_env(**kwargs) env = wrappers.AssertOutOfBoundsWrapper(env) env = wrappers.NanNoOpWrapper(env, 0, "setting action to 0") env = wrappers.OrderEnforcingWrapper(env) return env