njit back on _homogeneous_decimation; is ~40% faster on graham

araghukas · Oct 22, 2021 · 1f157c5 · 1f157c5
1 parent 73c385b
commit 1f157c5
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 1 deletion.
diff --git a/agf/sancho_rubio.py b/agf/sancho_rubio.py
@@ -6,6 +6,7 @@
     Sancho, M. P. L. et al., Journal of Physics F: Metal Physics 15 (4): 851–58 (1985)
 """
 import numpy as np
+from numba import njit
 
 from typing import Tuple
 from dataclasses import dataclass
@@ -65,6 +66,7 @@ def decimate(arr: np.ndarray,
     return DecimationResult(Ws, Wb, a, b)
 
 
+@njit
 def _homogeneous_decimation(omega: np.ndarray,
                             a: np.ndarray,
                             b: np.ndarray,

diff --git a/setup.py b/setup.py
@@ -9,6 +9,6 @@
     author_email="[email protected]",
     author="Ara Ghukasyan",
     license="MIT",
-    install_requires=["numpy"],
+    install_requires=["numpy", "numba"],
     packages=find_packages()
 )