refactor: remove unnecessary private _standard_normal function (#95)

Author: vegard-solum-4ss

src/smsfusion/noise/_noise.py

@@ -1,27 +1,26 @@
 import numpy as np
 from numpy.typing import NDArray
 
-
-def _standard_normal(
-    n: int, seed: int | np.random.Generator | None = None
-) -> NDArray[np.float64]:
-    """
-    Draw i.i.d. samples from a standard Normal distribution (mean=0, stdev=1).
-
-    Parameters
-    ----------
-    n : int
-        Number of samples to generate.
-    seed : int, np.random.Generator or None, optional
-        A seed used to initialize a random number generator. If passed a Generator,
-        it will be used unaltered.
-
-    Returns
-    -------
-    numpy.ndarray, shape (n,)
-        Sequence of i.i.d. samples.
-    """
-    return np.random.default_rng(seed).standard_normal(n)
+# def _standard_normal(
+#     n: int, seed: int | np.random.Generator | None = None
+# ) -> NDArray[np.float64]:
+#     """
+#     Draw i.i.d. samples from a standard Normal distribution (mean=0, stdev=1).
+
+#     Parameters
+#     ----------
+#     n : int
+#         Number of samples to generate.
+#     seed : int, np.random.Generator or None, optional
+#         A seed used to initialize a random number generator. If passed a Generator,
+#         it will be used unaltered.
+
+#     Returns
+#     -------
+#     numpy.ndarray, shape (n,)
+#         Sequence of i.i.d. samples.
+#     """
+#     return np.random.default_rng(seed).standard_normal(n)
 
 
 def white_noise(
@@ -66,8 +65,10 @@ def white_noise(
     --------
     smsfusion.gauss_markov, smsfusion.random_walk
     """
+    rng = np.random.default_rng(seed)
     sigma_wn = N * np.sqrt(fs)
-    return sigma_wn * _standard_normal(n, seed=seed)  # type: ignore[no-any-return]
+    wn = sigma_wn * rng.standard_normal(n)  # type: ignore[no-any-return]
+    return wn
 
 
 def random_walk(
@@ -116,10 +117,12 @@ def random_walk(
     --------
     smsfusion.gauss_markov, smsfusion.white_noise
     """
+    rng = np.random.default_rng(seed)
+
     sigma_wn = K / np.sqrt(fs)
 
     x = np.zeros(n)
-    epsilon = _standard_normal(n - 1, seed=seed)
+    epsilon = rng.standard_normal(n - 1)
     for i in range(1, n):
         x[i] = x[i - 1] + sigma_wn * epsilon[i - 1]
 
@@ -184,14 +187,16 @@ def gauss_markov(
     --------
     smsfusion.random_walk, smsfusion.white_noise
     """
+    rng = np.random.default_rng(seed)
+
     dt = 1.0 / fs
     beta = 1.0 / tau_c
 
     phi = np.exp(-beta * dt)
     sigma_wn = sigma * np.sqrt(1.0 - np.exp(-2 * beta * dt))
 
     x = np.zeros(n)
-    epsilon = _standard_normal(n - 1, seed=seed)
+    epsilon = rng.standard_normal(n - 1)
     for i in range(1, n):
         x[i] = phi * x[i - 1] + sigma_wn * epsilon[i - 1]
 

tests/test_noise.py

@@ -14,28 +14,10 @@
     random_walk,
     white_noise,
 )
-from smsfusion.noise._noise import _standard_normal
 
 TEST_PATH = Path(__file__).parent
 
 
-def test__standard_normal():
-    x = _standard_normal(100)
-
-    assert len(x) == 100
-    assert np.mean(x) == pytest.approx(0.0, abs=0.3)  # mean value is zero
-    assert np.std(x) == pytest.approx(1.0, abs=0.2)  # std is 1
-
-
-def test__standard_normal_seeds():
-    x0 = _standard_normal(100, seed=0)
-    x1 = _standard_normal(100, seed=0)
-    x2 = _standard_normal(100, seed=1)
-
-    np.testing.assert_array_almost_equal(x0, x1)
-    assert not np.array_equal(x0, x2)
-
-
 def test_white_noise():
     N, fs, n = 3, 10.0, 100_000
     wn_out = white_noise(N, fs, n, seed=123)