clean up

nkcs/ea.py

@@ -92,7 +92,7 @@ def run_initial(self, nkcs: NKCS) -> None:
         for s in range(Cons.S):
             for p in range(Cons.P):
                 team[s] = self.pop[s][p]
-                genomes = np.asarray([ind.genome for ind in team])
+                genomes = [ind.genome for ind in team]
                 team[s].fitness = nkcs.calc_team_fit(genomes)
                 self.__update_archive(s, team[s])
                 self.evals += 1
@@ -103,7 +103,7 @@ def __eval_team(self, nkcs: NKCS, team: list[Ind]) -> None:
 
         Assign fitness to each individual if it's the best seen.
         """
-        genomes = np.asarray([ind.genome for ind in team])
+        genomes = [ind.genome for ind in team]
         team_fit: Final[float] = nkcs.calc_team_fit(genomes)
         for s in range(Cons.S):
             team[s].fitness = max(team[s].fitness, team_fit)
@@ -126,7 +126,7 @@ def __update_archive(self, s: int, p: Ind) -> None:
             self.archive[s].pop(0)
 
     def update_perf(
-        self, evals: list[int], perf_best: np.ndarray, perf_avg: np.ndarray
+        self, evals: np.ndarray, perf_best: np.ndarray, perf_avg: np.ndarray
     ) -> None:
         """Update current performance tracking."""
         if self.evals % (Cons.P * Cons.S) == 0:
@@ -180,7 +180,9 @@ def get_best_fit(self, s: int) -> float:
 
     def get_avg_fit(self, s: int) -> float:
         """Return the average fitness of a given species."""
-        return np.mean([p.fitness for p in self.pop[s]])
+        fitnesses: list[float] = [p.fitness for p in self.pop[s]]
+        mean = np.mean(fitnesses)
+        return float(mean)
 
     def print_archive(self, s: int) -> None:
         """Print the archived individuals."""

nkcs/main.py

@@ -78,7 +78,7 @@
                 ea[r].run_ea(nkcs[f], evals[r], perf_best[r], perf_avg[r])
             else:
                 ea[r].run_sea(nkcs[f], evals[r], perf_best[r], perf_avg[r])
-            best_fit: Final[float] = ea[r].get_best_fit(0)
+            best_fit: float = ea[r].get_best_fit(0)
             status = f"nkcs {f} experiment {e} complete: {best_fit:.5f}"
             r += 1
             bar.set_description(status)

nkcs/surrogate.py

@@ -90,7 +90,7 @@ def model_mlp(seed: Optional[int]) -> MLPRegressor:
 
 
 def fit_model(
-    x: np.ndarray, y: np.ndarray, seed: int = None
+    x: np.ndarray, y: np.ndarray, seed: int | None = None
 ) -> Union[GaussianProcessRegressor, MLPRegressor]:
     """Train a surrogate model."""
     if Cons.MODEL == "gp":