Add comparison between cythonized and pure Python

Author: gjbex

source-code/cython/Numpy/run_cythonized_sum.py

@@ -0,0 +1,33 @@
+#!/usr/bin/env python
+
+from argparse import ArgumentParser
+import numpy as np
+import pyximport
+import timeit
+
+pyximport.install(pyimport=True)
+import array_sum_pure
+
+def py_sum(a):
+    total = 0.0
+    for i in range(a.shape[0]):
+        for j in range(a.shape[1]):
+            total += a[i, j]
+    return total
+
+if __name__ == '__main__':
+    arg_parser = ArgumentParser(description='compute array sum using '
+                                            'pure, non-cythonized and pure cythonized implementations')
+    arg_parser.add_argument('--n', type=int, default=10_000,
+                            help='size (nxn array)')
+    arg_parser.add_argument('--iter', type=int, default=10,
+                            help='number of iterations')
+    options = arg_parser.parse_args()
+    a = np.ones((options.n, options.n))
+    for func in [array_sum_pure.array_sum, py_sum]:
+        total = 0.0
+        start_time = timeit.default_timer()
+        for iter_nr in range(options.iter):
+            total += func(a)
+        total_time = timeit.default_timer() - start_time
+        print(f'{func.__qualname__:s}: {total_time:.6f} s ({total})')

source-code/cython/Numpy/run_pure_sum.py

@@ -0,0 +1,30 @@
+#!/usr/bin/env python
+
+from argparse import ArgumentParser
+import array_sum_pure
+import numpy as np
+import timeit
+
+def py_sum(a):
+    total = 0.0
+    for i in range(a.shape[0]):
+        for j in range(a.shape[1]):
+            total += a[i, j]
+    return total
+
+if __name__ == '__main__':
+    arg_parser = ArgumentParser(description='compute array sum using '
+                                            'pure, non-cythonized and pure cythonized implementations')
+    arg_parser.add_argument('--n', type=int, default=10_000,
+                            help='size (nxn array)')
+    arg_parser.add_argument('--iter', type=int, default=10,
+                            help='number of iterations')
+    options = arg_parser.parse_args()
+    a = np.ones((options.n, options.n))
+    for func in [array_sum_pure.array_sum, py_sum]:
+        total = 0.0
+        start_time = timeit.default_timer()
+        for iter_nr in range(options.iter):
+            total += func(a)
+        total_time = timeit.default_timer() - start_time
+        print(f'{func.__qualname__:s}: {total_time:.6f} s ({total})')