Fix ruff errors

pyproject.toml

@@ -57,7 +57,13 @@ profile = "black"
 
 [tool.ruff]
 line-length = 88
+# ignore = ["UP035"]
 select = ["E", "F", "B", "UP"]
 
 [tool.ruff.per-file-ignores]
 "__init__.py" = ["F401"]
+# For pyserde
+"src/emevo/exp_utils.py" = ["UP006", "UP035"]
+# For typer
+"experiments/**/*.py" = ["B008", "UP006", "UP007"]
+"smoke-tests/*.py" = ["B008", "UP006", "UP007"]

smoke-tests/circle_loop.py

@@ -79,7 +79,9 @@ def main(
 
         if replace and i % replace_interval == 0:
             if i < steps // 2:
-                flag = jnp.zeros(n_max_agents, dtype=bool).at[deactivate_index].set(True)
+                flag = (
+                    jnp.zeros(n_max_agents, dtype=bool).at[deactivate_index].set(True)
+                )
                 state = env.deactivate(state, flag)
                 deactivate_index -= 1
             else:

src/emevo/environments/__init__.py

@@ -1,8 +1,8 @@
 """ Implementation of registry and built-in emevo environments.
 """
 
-from emevo.environments.registry import description, make, register
 from emevo.environments.circle_foraging import CircleForaging
+from emevo.environments.registry import description, make, register
 
 register(
     "CircleForaging-v0",

src/emevo/environments/moderngl_vis.py

@@ -282,6 +282,33 @@ def _collect_heads(circle: Circle, state: State) -> NDArray:
     return np.concatenate((p1, p2), axis=1).reshape(-1, 2)
 
 
+# def _collect_policies(
+#     circle: Circle,
+#     state: State,
+#     max_arrow_length: float,
+# ) -> NDArray:
+# max_f = max(map(lambda bp: bp[1].max(), bodies_and_policies))
+# policy_scaling = max_arrow_length / max_f
+# points = []
+# radius = None
+# for body, policy in bodies_and_policies:
+#     a = body.position
+#     if radius is None:
+#         radius = next(
+#             filter(lambda shape: isinstance(shape, pymunk.Circle), body.shapes)
+#         ).radius
+#     f1, f2 = policy
+#     from1 = a + pymunk.Vec2d(0, radius).rotated(body.angle + np.pi * 0.75)
+#     to1 = from1 + pymunk.Vec2d(0, -f1 * policy_scaling).rotated(body.angle)
+#     from2 = a + pymunk.Vec2d(0, radius).rotated(body.angle - np.pi * 0.75)
+#     to2 = from2 + pymunk.Vec2d(0, -f2 * policy_scaling).rotated(body.angle)
+#     points.append(from1)
+#     points.append(to1)
+#     points.append(from2)
+#     points.append(to2)
+# return np.array(points, dtype=np.float32)
+
+
 def _get_clip_ranges(lengthes: list[float]) -> list[tuple[float, float]]:
     """Clip ranges to [-1, 1]"""
     total = sum(lengthes)
@@ -435,7 +462,9 @@ def overlay(self, name: str, value: Any) -> Any:
         """Render additional value as an overlay"""
         key = name.lower()
         if key == "arrow":
-            segments = _collect_policies(value, self._range_min * 0.1)
+            # Not implmented yet
+            # segments = _collect_policies(value, self._range_min * 0.1)
+            segments = np.zeros(1)
             if "arrow" in self._overlays:
                 do_render = self._overlays["arrow"].update(segments)
             else:

src/emevo/exp_utils.py

@@ -18,7 +18,7 @@ class CfConfig:
     n_agent_sensors: int
     sensor_length: float
     food_loc_fn: str
-    food_num_fn: Tuple["str", int]
+    food_num_fn: Tuple[str, int]
     xlim: Tuple[float, float]
     ylim: Tuple[float, float]
     env_radius: float

src/emevo/rl/ppo_normal.py

@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from typing import cast, NamedTuple
+from typing import NamedTuple, cast
 
 import chex
 import distrax

tests/test_ppo.py

@@ -6,7 +6,6 @@
 import pytest
 
 from emevo.rl.ppo_normal import (
-    Batch,
     NormalPPONet,
     Rollout,
     get_minibatches,