Merge pull request #16 from UoA-CARES/rf/type-hints-linting

type hinting + linting across the board

.github/workflows/black.yml

@@ -0,0 +1,18 @@
+name: Lint
+run-name: ${{ github.actor }} is linting the code
+
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    branches:
+      - main
+
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+      - uses: psf/black@stable

.github/workflows/pylint.yml

@@ -0,0 +1,50 @@
+name: Pylint
+
+on: 
+  push:
+    branches:
+      - main
+  pull_request:
+    branches:
+      - main
+
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: ["3.10"]
+    steps:
+    - uses: actions/checkout@v3
+
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v3
+      with:
+        python-version: ${{ matrix.python-version }}
+
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install pylint
+      
+    - name: Clone cares_reinforcement_learning repository
+      uses: GuillaumeFalourd/clone-github-repo-action@main
+      with:
+        owner: 'UoA-CARES'
+        repository: 'cares_reinforcement_learning'
+
+    - name: Install cares_reinforcement_learning repository content
+      run: |
+        cd cares_reinforcement_learning
+        pip install -r requirements.txt
+        pip install --editable .
+        cd -
+
+    - name: Install deps
+      run: |
+        pip install -r requirements.txt
+
+    - name: Analysing the code with pylint
+      run: |
+        pylint $(git ls-files '*.py') --rcfile .pylintrc --fail-under=9 --fail-on=error

.pylintrc

@@ -0,0 +1,19 @@
+[MESSAGES CONTROL]
+disable=
+    logging-fstring-interpolation,
+    too-few-public-methods,
+    missing-module-docstring,
+    missing-function-docstring,
+    missing-class-docstring,
+    too-many-locals,
+    W0511,
+    too-many-arguments,
+
+[FORMAT]
+max-line-length=130
+
+[MASTER]
+extension-pkg-whitelist=cv2
+
+[TYPECHECK]
+generated-members=cv2.*

scripts/envrionments/dmcs/DMCS.py → scripts/envrionments/dmcs/dmcs_environment.py

@@ -1,19 +1,14 @@
 import logging
+from functools import cached_property
 
 import cv2
-
-from dm_control import suite
-
 import numpy as np
-from collections import deque
-
-# from typing import override
-from functools import cached_property
-
+from dm_control import suite
+from envrionments.gym_environment import GymEnvironment
 from util.configurations import GymEnvironmentConfig
-from envrionments.GymEnvironment import GymEnvironment
 
-class DMCS(GymEnvironment):
+
+class DMCSEnvironment(GymEnvironment):
     def __init__(self, config: GymEnvironmentConfig) -> None:
         super().__init__(config)
         logging.info(f"Training on Domain {config.domain}")
@@ -22,37 +17,46 @@ def __init__(self, config: GymEnvironmentConfig) -> None:
         self.env = suite.load(self.domain, self.task)
 
     @cached_property
-    def min_action_value(self):
+    def min_action_value(self) -> float:
         return self.env.action_spec().minimum[0]
-    
+
     @cached_property
-    def max_action_value(self):
+    def max_action_value(self) -> float:
         return self.env.action_spec().maximum[0]
 
     @cached_property
-    def observation_space(self):
+    def observation_space(self) -> int:
         time_step = self.env.reset()
-        observation = np.hstack(list(time_step.observation.values())) # # e.g. position, orientation, joint_angles
+        # e.g. position, orientation, joint_angles
+        observation = np.hstack(list(time_step.observation.values()))
         return len(observation)
-    
+
     @cached_property
-    def action_num(self):
+    def action_num(self) -> int:
         return self.env.action_spec().shape[0]
 
-    def set_seed(self, seed):
-        self.env = suite.load(self.domain, self.task, task_kwargs={'random': seed})
+    def set_seed(self, seed: int) -> None:
+        self.env = suite.load(self.domain, self.task, task_kwargs={"random": seed})
 
-    def reset(self):
+    def reset(self) -> np.ndarray:
         time_step = self.env.reset()
-        observation = np.hstack(list(time_step.observation.values())) # # e.g. position, orientation, joint_angles
+        observation = np.hstack(
+            list(time_step.observation.values())
+        )  # # e.g. position, orientation, joint_angles
         return observation
 
-    def step(self, action):
+    def step(self, action: int) -> tuple:
         time_step = self.env.step(action)
-        state, reward, done = np.hstack(list(time_step.observation.values())), time_step.reward, time_step.last()
-        return state, reward, done, False # for consistency with open ai gym just add false for truncated
-
-    def grab_frame(self, camera_id=0, height=240, width=300):
+        state, reward, done = (
+            np.hstack(list(time_step.observation.values())),
+            time_step.reward,
+            time_step.last(),
+        )
+        # for consistency with open ai gym just add false for truncated
+        return state, reward, done, False
+
+    def grab_frame(self, height=240, width=300, camera_id=0) -> np.ndarray:
         frame = self.env.physics.render(camera_id=camera_id, height=height, width=width)
-        frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB) # Convert to BGR for use with OpenCV
+        # Convert to BGR for use with OpenCV
+        frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
         return frame

scripts/envrionments/environment_factory.py

@@ -0,0 +1,37 @@
+import logging
+
+from envrionments.dmcs.dmcs_environment import DMCSEnvironment
+from envrionments.gym_environment import GymEnvironment
+from envrionments.image_wrapper import ImageWrapper
+from envrionments.openai.openai_environment import OpenAIEnvrionment
+from envrionments.pyboy.mario.mario_environment import MarioEnvironment
+from envrionments.pyboy.pokemon.pokemon_environment import PokemonEnvironment
+from util.configurations import GymEnvironmentConfig
+
+
+def create_pyboy_environment(config: GymEnvironmentConfig) -> GymEnvironment:
+    # TODO extend to other pyboy games...maybe another repo?
+    if config.task == "pokemon":
+        env = PokemonEnvironment(config)
+    elif config.task == "mario":
+        env = MarioEnvironment(config)
+    else:
+        raise ValueError(f"Unkown pyboy environment: {config.task}")
+    return env
+
+
+class EnvironmentFactory:
+    def __init__(self) -> None:
+        pass
+
+    def create_environment(self, config: GymEnvironmentConfig) -> GymEnvironment:
+        logging.info(f"Training Environment: {config.gym}")
+        if config.gym == "dmcs":
+            env = DMCSEnvironment(config)
+        elif config.gym == "openai":
+            env = OpenAIEnvrionment(config)
+        elif config.gym == "pyboy":
+            env = create_pyboy_environment(config)
+        else:
+            raise ValueError(f"Unkown environment: {config.gym}")
+        return ImageWrapper(env) if config.image_observation else env

scripts/envrionments/GymEnvironment.py → scripts/envrionments/gym_environment.py

@@ -1,11 +1,10 @@
-import logging
-
 import abc
-# from typing import override
+import logging
 from functools import cached_property
 
 from util.configurations import GymEnvironmentConfig
 
+
 class GymEnvironment(metaclass=abc.ABCMeta):
     def __init__(self, config: GymEnvironmentConfig) -> None:
         logging.info(f"Training with Task {config.task}")
@@ -15,7 +14,7 @@ def __init__(self, config: GymEnvironmentConfig) -> None:
     @abc.abstractmethod
     def min_action_value(self):
         raise NotImplementedError("Override this method")
-    
+
     @cached_property
     @abc.abstractmethod
     def max_action_value(self):
@@ -25,7 +24,7 @@ def max_action_value(self):
     @abc.abstractmethod
     def observation_space(self):
         raise NotImplementedError("Override this method")
-    
+
     @cached_property
     @abc.abstractmethod
     def action_num(self):
@@ -42,7 +41,7 @@ def reset(self):
     @abc.abstractmethod
     def step(self, action):
         raise NotImplementedError("Override this method")
-    
+
     @abc.abstractmethod
-    def grab_frame(self, camera_id=0, height=240, width=300):
+    def grab_frame(self, height=240, width=300):
         raise NotImplementedError("Override this method")

scripts/envrionments/ImageWrapper.py → scripts/envrionments/image_wrapper.py

@@ -1,15 +1,12 @@
 import logging
-
-import cv2
-# from typing import override
+from collections import deque
 from functools import cached_property
 
-from collections import deque
 import numpy as np
+from envrionments.gym_environment import GymEnvironment
 
-from envrionments.GymEnvironment import GymEnvironment
 
-class ImageWrapper(GymEnvironment):
+class ImageWrapper:
     def __init__(self, gym: GymEnvironment, k=3):
         self.gym = gym
 
@@ -18,7 +15,7 @@ def __init__(self, gym: GymEnvironment, k=3):
 
         self.frame_width = 84
         self.frame_height = 84
-        logging.info(f"Image Observation is on")
+        logging.info("Image Observation is on")
 
     @cached_property
     def observation_space(self):
@@ -27,34 +24,34 @@ def observation_space(self):
     @cached_property
     def action_num(self):
         return self.gym.action_num
-    
+
     @cached_property
     def min_action_value(self):
         return self.gym.min_action_value
-    
+
     @cached_property
     def max_action_value(self):
         return self.gym.max_action_value
-    
+
     def set_seed(self, seed):
         self.gym.set_seed(seed)
 
     def grab_frame(self, height=240, width=300):
         return self.gym.grab_frame(height=height, width=width)
-    
+
     def reset(self):
         _ = self.gym.reset()
         frame = self.grab_frame(height=self.frame_height, width=self.frame_width)
-        frame = np.moveaxis(frame, -1, 0)                    
+        frame = np.moveaxis(frame, -1, 0)
         for _ in range(self.k):
             self.frames_stacked.append(frame)
         stacked_frames = np.concatenate(list(self.frames_stacked), axis=0)
         return stacked_frames
 
     def step(self, action):
-        state, reward, done, truncated = self.gym.step(action)
+        _, reward, done, truncated = self.gym.step(action)
         frame = self.grab_frame(height=self.frame_height, width=self.frame_width)
         frame = np.moveaxis(frame, -1, 0)
         self.frames_stacked.append(frame)
         stacked_frames = np.concatenate(list(self.frames_stacked), axis=0)
-        return stacked_frames, reward, done, truncated 
+        return stacked_frames, reward, done, truncated