add numpy ruff rules

pyproject.toml

@@ -90,7 +90,9 @@ select = [
     "F",    # pyflakes
     "I",    # isort
     "N",    # pep8-naming
+    "NPY",  # numpy
     "PD",   # pandas-vet
+    "PIE",  # misc lints
     "PT",   # pytest
     "PTH",  # flake8-use-pathlib
     "PGH",  # pygrep

src/otc/models/objective.py

@@ -40,10 +40,6 @@ def set_seed(seed_val: int = 42) -> int:
     # see https://docs.python.org/3/using/cmdline.html#envvar-PYTHONHASHSEED
     os.environ["PYTHONHASHSEED"] = str(seed_val)
 
-    # pandas and numpy
-    #  https://stackoverflow.com/a/52375474/5755604
-    np.random.seed(seed_val)
-
     # python random module
     random.seed(seed_val)
 

tests/__init__.py

@@ -1,4 +1 @@
-"""Support for tests.
-
-
-"""
+"""Support for tests."""

tests/test_metrics.py

@@ -3,7 +3,6 @@
 import numpy as np
 
 from otc.metrics.metrics import effective_spread
-from otc.models.objective import set_seed
 
 
 class TestMetrics:
@@ -19,11 +18,10 @@ def test_effective_spread(self) -> None:
 
         Value may not be NaN.
         """
-        set_seed(7)
-
-        y_pred = np.random.choice([-1, 1], size=(10))
-        trade_price = np.random.rand(10) * 100
-        fundamental_value = np.random.rand(10) * 100
+        rng = np.random.default_rng(seed=7)
+        y_pred = rng.choice([-1, 1], size=(10))
+        trade_price = rng.random(10) * 100
+        fundamental_value = rng.random(10) * 100
 
         e_s = effective_spread(y_pred, trade_price, fundamental_value)
 

tests/test_objective.py

@@ -35,7 +35,8 @@ def setup(self) -> None:
 
         # make 1 const feature and 1 non-const feature, as catboost requires non-const
         self._x_train = pd.DataFrame(data={"feature_1": 1}, index=index)
-        self._x_train["feature_2"] = np.random.randint(1, 6, self._x_train.shape[0])
+        rng = np.random.default_rng()
+        self._x_train["feature_2"] = rng.integers(1, 6, self._x_train.shape[0])
         self._y_train = self._x_train["feature_2"]
         self._x_val = self._x_train.copy()
         self._y_val = self._y_train.copy()
@@ -113,9 +114,9 @@ def test_gradient_boosting_pretraining_objective(self) -> None:
 
         # labelled (-1,1) and unlabelled (0) instances
         # train set with -1, 1, and 0
-        self._y_train = pd.Series(np.random.randint(-1, 2, self._y_train.shape[0]))
-        # val set with 1
         rng = np.random.default_rng()
+        self._y_train = pd.Series(rng.integers(-1, 2, self._y_train.shape[0]))
+        # val set with 1
         self._y_val = rng.integers(low=1, high=2, size=self._y_train.shape[0])
 
         study = optuna.create_study(direction="maximize")