Merge pull request #133 from Federico-PizarroBejarano/improved_linting

Improved linting

.github/workflows/push.yml

@@ -1,9 +1,4 @@
-# This workflow uses actions that are not certified by GitHub.
-# They are provided by a third-party and are governed by
-# separate terms of service, privacy policy, and support
-# documentation.
-
-name: Test package quality
+name: Test Package Quality
 
 on: push
 
@@ -26,6 +21,11 @@ jobs:
     - name: Install package
       run: |
         pip install -e .
+    - name: Run Linting
+      run: |
+        pre-commit install
+        pre-commit autoupdate
+        pre-commit run
     - name: Unit tests
       run: |
         python -m pytest tests/

.pre-commit-config.yaml

@@ -0,0 +1,57 @@
+# Install the pre-commit hooks below with
+# 'pre-commit install'
+
+# Auto-update the version of the hooks with
+# 'pre-commit autoupdate'
+
+# Run the hooks on all files with
+# 'pre-commit run --all'
+
+repos:
+-   repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.5.0
+    hooks:
+    -   id: check-ast
+    -   id: check-yaml
+    -   id: check-toml
+    -   id: check-added-large-files
+    -   id: check-docstring-first
+    -   id: check-executables-have-shebangs
+    -   id: check-shebang-scripts-are-executable
+    -   id: check-merge-conflict
+    -   id: debug-statements
+    -   id: trailing-whitespace
+    -   id: end-of-file-fixer
+    -   id: double-quote-string-fixer
+
+-   repo: https://github.com/pycqa/isort
+    rev: 5.12.0
+    hooks:
+      - id: isort
+        name: isort
+        args: ['--line-length=110']
+
+-   repo: https://github.com/pre-commit/mirrors-autopep8
+    rev: v2.0.4
+    hooks:
+    -   id: autopep8
+        name: autopep8-default
+        args: ['-i', '--ignore=C0301', '--max-line-length=1000']
+        exclude: (^tests/|^safe_control_gym/math_and_models/transformations.py)
+    -   id: autopep8
+        name: autopep8-tests
+        args: ['-i', '--ignore=C0301,E501,E201,E241,E127', '--max-line-length=1000']
+        files: (^tests/|^safe_control_gym/math_and_models/transformations.py)
+
+-   repo: https://github.com/PyCQA/flake8
+    rev: 6.1.0
+    hooks:
+    -   id: flake8
+        name: flake8_default
+        args: ['--ignore=E501']
+        exclude: (^safe_control_gym/__init__.py|^tests/|^safe_control_gym/math_and_models/transformations.py)
+    -   id: flake8
+        name: flake8_tests
+        args: ['--ignore=E501,E201,E241,E127']
+        files: (^tests/|^safe_control_gym/math_and_models/transformations.py)
+        exclude: ^tests/test_build.py

README.md

@@ -190,6 +190,19 @@ Note that the Bullet engine frequency reported for `safe-control-gym` is typical
 [004]: https://github.com/utiasDSL/safe-control-gym/blob/main/safe_control_gym/envs/gym_control/cartpole.py
 [005]: https://github.com/utiasDSL/safe-control-gym/blob/main/safe_control_gym/envs/gym_pybullet_drones/quadrotor.py
 
+## Run Tests and Linting
+Tests can be run locally by executing:
+```bash
+python3 -m pytest ./tests/  # Run all tests
+```
+
+Linting can be run locally with:
+```bash
+pre-commit install  # Install the pre-commit hooks
+pre-commit autoupdate  # Auto-update the version of the hooks
+pre-commit run --all  # Run the hooks on all files
+```
+
 ## References
 
 - Brunke, L., Greeff, M., Hall, A. W., Yuan, Z., Zhou, S., Panerati, J., & Schoellig, A. P. (2022). [Safe learning in robotics: From learning-based control to safe reinforcement learning](https://www.annualreviews.org/doi/abs/10.1146/annurev-control-042920-020211). Annual Review of Control, Robotics, and Autonomous Systems, 5, 411-444.

examples/cbf/cbf_experiment.py

@@ -4,12 +4,12 @@
 import shutil
 from functools import partial
 
-import numpy as np
 import matplotlib.pyplot as plt
+import numpy as np
 
 from safe_control_gym.experiments.base_experiment import BaseExperiment
-from safe_control_gym.utils.registration import make
 from safe_control_gym.utils.configuration import ConfigFactory
+from safe_control_gym.utils.registration import make
 
 
 def run(plot=True, training=True, n_episodes=1, n_steps=None, curr_path='.', save_data=False):

examples/lqr/lqr_experiment.py

@@ -2,15 +2,15 @@
 
 import os
 import pickle
-from functools import partial
 from collections import defaultdict
+from functools import partial
 
-import numpy as np
 import matplotlib.pyplot as plt
+import numpy as np
 from matplotlib.ticker import FormatStrFormatter
 
-from safe_control_gym.experiments.base_experiment import BaseExperiment
 from safe_control_gym.envs.benchmark_env import Task
+from safe_control_gym.experiments.base_experiment import BaseExperiment
 from safe_control_gym.utils.configuration import ConfigFactory
 from safe_control_gym.utils.registration import make
 

examples/mpsc/config_overrides/cartpole/cartpole_stab.yaml

@@ -4,13 +4,13 @@ task_config:
   ctrl_freq: 15
   pyb_freq: 750
   physics: pyb
-  
+
   # state initialization
   init_state:
     init_x: 0.1
     init_x_dot: -1.5
     init_theta: -0.175
-    init_theta_dot: 0.5  
+    init_theta_dot: 0.5
   randomized_init: True
   randomized_inertial_prop: False
   normalized_rl_action_space: True

examples/mpsc/config_overrides/cartpole/cartpole_track.yaml

@@ -4,13 +4,13 @@ task_config:
   ctrl_freq: 15
   pyb_freq: 750
   physics: pyb
-  
+
   # state initialization
   init_state:
     init_x: 0
     init_x_dot: 0
     init_theta: 0
-    init_theta_dot: 0 
+    init_theta_dot: 0
   randomized_init: True
   randomized_inertial_prop: False
   normalized_rl_action_space: True

examples/mpsc/config_overrides/cartpole/sac_cartpole.yaml

@@ -25,5 +25,5 @@ algo_config:
   save_interval: 1000
   num_checkpoints: 100
   eval_interval: 1000
-  eval_save_best: True 
+  eval_save_best: True
   tensorboard: True

examples/mpsc/config_overrides/quadrotor_2D/quadrotor_2D_stab.yaml

@@ -60,7 +60,7 @@ task_config:
   rew_state_weight: [1, 1, 1, 1, 1, 1]
   rew_act_weight: 0.1
   rew_exponential: True
-  
+
   constraints:
     - constraint_form: default_constraint
       constrained_variable: state

examples/mpsc/config_overrides/quadrotor_2D/quadrotor_2D_track.yaml

@@ -52,7 +52,7 @@ task_config:
     trajectory_plane: 'xz'
     trajectory_position_offset: [0, 1]
     trajectory_scale: 1
-  
+
   inertial_prop:
     M: 0.027
     Iyy: 1.4e-05

examples/mpsc/config_overrides/quadrotor_2D/sac_quadrotor_2D.yaml

@@ -25,5 +25,5 @@ algo_config:
   save_interval: 1000
   num_checkpoints: 100
   eval_interval: 1000
-  eval_save_best: True 
+  eval_save_best: True
   tensorboard: True

examples/mpsc/config_overrides/quadrotor_3D/quadrotor_3D_stab.yaml

@@ -92,7 +92,7 @@ task_config:
   rew_state_weight: [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]
   rew_act_weight: 0.1
   rew_exponential: True
-  
+
   constraints:
     - constraint_form: default_constraint
       constrained_variable: state

examples/mpsc/mpsc_experiment.py

@@ -4,13 +4,13 @@
 import shutil
 from functools import partial
 
-import numpy as np
 import matplotlib.pyplot as plt
+import numpy as np
 
+from safe_control_gym.envs.benchmark_env import Cost, Environment, Task
 from safe_control_gym.experiments.base_experiment import BaseExperiment
-from safe_control_gym.utils.registration import make
 from safe_control_gym.utils.configuration import ConfigFactory
-from safe_control_gym.envs.benchmark_env import Task, Cost, Environment
+from safe_control_gym.utils.registration import make
 
 
 def run(plot=True, training=False, n_episodes=1, n_steps=None, curr_path='.'):

examples/no_controller/verbose_api.py

@@ -5,8 +5,9 @@
     $ python3 verbose_api.py --task quadrotor --overrides verbose_api.yaml
 '''
 
-import time
 import inspect
+import time
+
 import pybullet as p
 
 from safe_control_gym.utils.configuration import ConfigFactory

examples/pid/pid_experiment.py

@@ -4,9 +4,9 @@
 import pickle
 from functools import partial
 
+import matplotlib.pyplot as plt
 import numpy as np
 import pybullet as p
-import matplotlib.pyplot as plt
 
 from safe_control_gym.experiments.base_experiment import BaseExperiment
 from safe_control_gym.utils.configuration import ConfigFactory

examples/rl/rl_experiment.py

@@ -3,13 +3,13 @@
 import shutil
 from functools import partial
 
-import numpy as np
 import matplotlib.pyplot as plt
+import numpy as np
 
+from safe_control_gym.envs.benchmark_env import Environment, Task
 from safe_control_gym.experiments.base_experiment import BaseExperiment
-from safe_control_gym.utils.registration import make
 from safe_control_gym.utils.configuration import ConfigFactory
-from safe_control_gym.envs.benchmark_env import Task, Environment
+from safe_control_gym.utils.registration import make
 
 
 def run(gui=True, n_episodes=1, n_steps=None, curr_path='.'):

pyproject.toml

@@ -26,6 +26,7 @@ pytope = "^0.0.4"
 Mosek = "^10.0.18"
 termcolor = "^1.1.0"
 pytest = "^7.2.2"
+pre-commit = "^3.3.2"
 
 [tool.poetry.dev-dependencies]
 

safe_control_gym/__init__.py

@@ -1,5 +1,5 @@
 from safe_control_gym.controllers import *
-from safe_control_gym.safety_filters import *
 from safe_control_gym.envs import *
 from safe_control_gym.envs.env_wrappers import *
 from safe_control_gym.experiments import *
+from safe_control_gym.safety_filters import *

safe_control_gym/controllers/ddpg/ddpg.py

@@ -16,15 +16,16 @@
 import numpy as np
 import torch
 
-from safe_control_gym.utils.logging import ExperimentLogger
-from safe_control_gym.utils.utils import get_random_state, set_random_state, is_wrapped
-from safe_control_gym.envs.env_wrappers.vectorized_env import make_vec_envs
-from safe_control_gym.envs.env_wrappers.vectorized_env.vec_env_utils import _flatten_obs, _unflatten_obs
-from safe_control_gym.envs.env_wrappers.record_episode_statistics import RecordEpisodeStatistics, VecRecordEpisodeStatistics
-from safe_control_gym.math_and_models.normalization import BaseNormalizer, MeanStdNormalizer, RewardStdNormalizer
-
 from safe_control_gym.controllers.base_controller import BaseController
 from safe_control_gym.controllers.ddpg.ddpg_utils import DDPGAgent, DDPGBuffer, make_action_noise_process
+from safe_control_gym.envs.env_wrappers.record_episode_statistics import (RecordEpisodeStatistics,
+                                                                          VecRecordEpisodeStatistics)
+from safe_control_gym.envs.env_wrappers.vectorized_env import make_vec_envs
+from safe_control_gym.envs.env_wrappers.vectorized_env.vec_env_utils import _flatten_obs, _unflatten_obs
+from safe_control_gym.math_and_models.normalization import (BaseNormalizer, MeanStdNormalizer,
+                                                            RewardStdNormalizer)
+from safe_control_gym.utils.logging import ExperimentLogger
+from safe_control_gym.utils.utils import get_random_state, is_wrapped, set_random_state
 
 
 class DDPG(BaseController):

safe_control_gym/controllers/ddpg/ddpg_utils.py

@@ -1,18 +1,18 @@
-from copy import deepcopy
 from collections import defaultdict
+from copy import deepcopy
 
 import numpy as np
 import torch
 import torch.nn as nn
 
-from safe_control_gym.math_and_models.neural_networks import MLP
 from safe_control_gym.controllers.sac.sac_utils import SACBuffer, soft_update
-
+from safe_control_gym.math_and_models.neural_networks import MLP
 
 # -----------------------------------------------------------------------------------
 #                   Agent
 # -----------------------------------------------------------------------------------
 
+
 class DDPGAgent:
     '''A DDPG class that encapsulates model, optimizer and update functions.'''
 
@@ -160,8 +160,10 @@ def __init__(self, obs_space, act_space, hidden_dims=(64, 64), activation='relu'
         low, high = act_space.low, act_space.high
         low = torch.FloatTensor(low)
         high = torch.FloatTensor(high)
-        # Rescale action from [-1, 1] to [low, high]
-        def unscale_fn(x): return low.to(x.device) + (0.5 * (x + 1.0) * (high.to(x.device) - low.to(x.device)))
+
+        def unscale_fn(x):  # Rescale action from [-1, 1] to [low, high]
+            return low.to(x.device) + (0.5 * (x + 1.0) * (high.to(x.device) - low.to(x.device)))
+
         self.actor = MLPActor(obs_dim, act_dim, hidden_dims, activation, postprocess_fn=unscale_fn)
 
         # Q functions

safe_control_gym/controllers/lqr/ilqr.py

@@ -8,7 +8,8 @@
 from termcolor import colored
 
 from safe_control_gym.controllers.base_controller import BaseController
-from safe_control_gym.controllers.lqr.lqr_utils import get_cost_weight_matrix, compute_lqr_gain, discretize_linear_system
+from safe_control_gym.controllers.lqr.lqr_utils import (compute_lqr_gain, discretize_linear_system,
+                                                        get_cost_weight_matrix)
 from safe_control_gym.envs.benchmark_env import Task
 
 
@@ -108,8 +109,7 @@ def learn(self, env=None, **kwargs):
             print(colored('--------------------------', 'green'))
 
             if self.ite_counter == 0 and env.done_on_out_of_bound and self.final_info['out_of_bounds']:
-                print(colored('[ERROR] The initial policy might be unstable. '
-                              + 'Break from iLQR updates.', 'red'))
+                print(colored('[ERROR] The initial policy might be unstable. Break from iLQR updates.', 'red'))
                 break
 
             # Maximum episode length.
@@ -133,10 +133,9 @@ def learn(self, env=None, **kwargs):
                 # If cost is increased, increase lambda
                 self.lamb *= self.lamb_factor
 
-                print(f'Cost increased by {-delta_cost}. '
-                      + 'Set feedforward term and controller gain to that '
-                      'from the previous iteration. '
-                      f'Increased lambda to {self.lamb}.')
+                print(f'Cost increased by {-delta_cost}.')
+                print('Set feedforward term and controller gain to that from the previous iteration.')
+                print(f'Increased lambda to {self.lamb}.')
                 print(f'Current policy is from iteration {self.best_iteration}.')
 
                 # Reset feedforward term and controller gain to that from
@@ -164,8 +163,7 @@ def learn(self, env=None, **kwargs):
                 # Check consecutive cost increment (cost decrement).
                 if abs(delta_cost) < self.epsilon and self.prev_ite_improved:
                     # Cost converged.
-                    print(colored('iLQR cost converged with a tolerance '
-                                  + f'of {self.epsilon}.', 'yellow'))
+                    print(colored(f'iLQR cost converged with a tolerance of {self.epsilon}.', 'yellow'))
                     break
 
                 # Set improved flag to True.

safe_control_gym/controllers/lqr/lqr.py

@@ -1,7 +1,7 @@
 '''Linear Quadratic Regulator (LQR).'''
 
 from safe_control_gym.controllers.base_controller import BaseController
-from safe_control_gym.controllers.lqr.lqr_utils import get_cost_weight_matrix, compute_lqr_gain
+from safe_control_gym.controllers.lqr.lqr_utils import compute_lqr_gain, get_cost_weight_matrix
 from safe_control_gym.envs.benchmark_env import Task