workflow

.github/workflows/on_push.yml

@@ -0,0 +1,8 @@
+name: On Push Any Documentation Check
+on: push
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v1
+      - uses: cnheider/postdoc@master

neodroidagent/agents/numpy_agents/model_free/non_contextual/ucb1_policy.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 import argparse
 
-from neodroid.environments.unity_environment import SingleUnityEnvironment
+from neodroid.environments.droid_environment import SingleUnityEnvironment
 from neodroid.utilities import Displayable
 from neodroidagent.agents.numpy_agents.numpy_agent import NumpyAgent
 from neodroidagent.utilities.exploration.ucb1 import UCB1

neodroidagent/agents/torch_agents/model_free/on_policy/ppo_agent.py

@@ -63,7 +63,7 @@ def __init__(
         optimiser_spec: GDKC = GDKC(constructor=torch.optim.Adam, lr=3e-4),
         continuous_arch_spec: GDKC = GDKC(ActorCriticMLP),
         discrete_arch_spec: GDKC = GDKC(CategoricalActorCriticMLP),
-        gradient_norm_clipping: ClipFeature = ClipFeature(True, -0.5, 0.5),
+        gradient_norm_clipping: ClipFeature = ClipFeature(True, 0, 0.5),
         **kwargs,
     ) -> None:
         """
@@ -313,7 +313,7 @@ def _update(self, metric_writer: Writer = MockWriter()) -> float:
             if early_stop_inner:
                 break
 
-        mean_loss = accum_loss.send()
+        mean_loss = next(accum_loss)
 
         if metric_writer:
             metric_writer.scalar("Inner Updates", i)

neodroidagent/common/session_factory/horizontal/experiment.py

@@ -14,7 +14,7 @@
 from cloudpickle import cloudpickle
 from garage.experiment import to_local_command
 
-from neodroid.environments.unity_environment import UnityEnvironment
+from neodroid.environments.droid_environment import UnityEnvironment
 from neodroidagent.agents import SACAgent
 from neodroidagent.common import CategoricalMLP
 

neodroidagent/common/session_factory/vertical/linear.py

@@ -6,7 +6,7 @@
 
 from neodroid.environments import Environment
 from neodroid.environments.gym_environment import NeodroidGymEnvironment
-from neodroid.environments.unity_environment import VectorUnityEnvironment
+from neodroid.environments.droid_environment import VectorUnityEnvironment
 from trolls import NormalisedActions, VectorWrap
 from warg import super_init_pass_on_kws
 from .procedures import OnPolicyEpisodic, Procedure

neodroidagent/common/session_factory/vertical/parallel.py

@@ -5,7 +5,7 @@
 
 from neodroid.environments import Environment
 from neodroid.environments.gym_environment import NeodroidVectorGymEnvironment
-from neodroid.environments.unity_environment import VectorUnityEnvironment
+from neodroid.environments.droid_environment import VectorUnityEnvironment
 from warg import super_init_pass_on_kws
 from .procedures import OnPolicyEpisodic, Procedure
 from .single_agent_environment_session import SingleAgentEnvironmentSession

neodroidagent/common/session_factory/vertical/procedures/training/on_policy_episodic.py

@@ -17,7 +17,7 @@
     Procedure,
 )
 from neodroidagent.utilities.misc import is_positive_and_mod_zero
-from warg.kw_passing import drop_unused_kws, passes_kws_to
+from warg.decorators.kw_passing import drop_unused_kws, passes_kws_to
 
 __author__ = "Christian Heider Nielsen"
 __all__ = ["rollout_on_policy", "OnPolicyEpisodic"]

neodroidagent/common/session_factory/vertical/procedures/training/sampling/rollout.py

@@ -2,7 +2,7 @@
 
 from tqdm import tqdm
 
-from neodroid.environments.unity_environment import VectorUnityEnvironment
+from neodroid.environments.droid_environment import VectorUnityEnvironment
 
 
 def run(self, environment: VectorUnityEnvironment, render: bool = True) -> None:

neodroidagent/entry_points/agent_tests/torch_agent_tests/ppo_test.py

@@ -2,23 +2,14 @@
 # -*- coding: utf-8 -*-
 from typing import Union
 
-from torch import nn
-from torch.nn.functional import mse_loss
-
 from neodroidagent.agents import PPOAgent
-from neodroidagent.common import OffPolicyBatched, ParallelSession
-from neodroidagent.common.architectures.mlp_variants.actor_critic import (
-    ActorCriticMLP,
-    CategoricalActorCriticMLP,
-)
+from neodroidagent.common import ParallelSession
 from neodroidagent.configs.test_reference.base_continous_test_config import *
 
 __author__ = "Christian Heider Nielsen"
 
 from neodroidagent.entry_points.session_factory import session_factory
 
-from warg import NOD
-
 """
 Description: Config for training
 Author: Christian Heider Nielsen

neodroidagent/utilities/exceptions/exceptions.py

@@ -8,32 +8,12 @@
 __all__ = [
     "NoData",
     "NoAgent",
-    "NoEnvironment",
     "NoProcedure",
     "NoTrajectoryException",
     "ActionSpaceNotSupported",
-    "HasNoEnvError",
 ]
 
 
-class HasNoEnvError(Exception):
-    """
-Raised when an agent has no environment assigned and some implicit next or step called.
-"""
-
-    def __init__(self, msg="Agent has no env assigned"):
-        Exception.__init__(self, msg)
-
-
-class NoEnvironment(Exception):
-    """
-
-"""
-
-    def __init__(self, msg="No Environment"):
-        Exception.__init__(self, msg)
-
-
 class NoProcedure(Exception):
     """
 

requirements_docs.txt

@@ -1 +1,3 @@
 sphinx
+m2r
+sphinxcontrib-programoutput

samples/continuous/c2d.py

@@ -3,7 +3,7 @@
 
 import draugr
 from draugr.stopping.stopping_key import add_early_stopping_key_combination
-from neodroid.environments.unity_environment import SingleUnityEnvironment
+from neodroid.environments.droid_environment import SingleUnityEnvironment
 
 __author__ = "Christian Heider Nielsen"
 

setup.py

@@ -207,5 +207,5 @@ def version(self):
         tests_require=pkg.dependencies_testing,
         setup_requires=pkg.setup_dependencies,
         include_package_data=True,
-        python_requires=">=3",
+        python_requires=">=3.6",
     )