numpy: start using n rounds to get aggregate stats about benches

Author: rohany

numpy/conftest.py

@@ -0,0 +1,7 @@
+import pytest
+@pytest.fixture
+def tacoBench(benchmark):
+    def f(func):
+        # Take statistics based on 10 rounds.
+        benchmark.pedantic(func, rounds=10)
+    return f

numpy/windowing.py

@@ -2,46 +2,45 @@
 from scipy.sparse import random
 import pytest
 
-def test_add_window(benchmark):
+def test_add_window(tacoBench):
     dim = 10000
     matrix = random(dim, dim, format="csr").todense()
     def bench():
         x = matrix[1:(dim-1), 1:(dim-1)] 
         res = x + x
-    benchmark.pedantic(bench, iterations=10)
+    tacoBench(bench)
 
 @pytest.mark.parametrize("dim", [5000, 10000, 20000])
-def test_add_sparse_window(benchmark, dim):
+def test_add_sparse_window(tacoBench, dim):
     matrix = random(dim, dim, format="csr")
     def bench():
         x = matrix[1:(dim-1), 1:(dim-1)] 
         res = x + x
         # Sanity check that this has a similar runtime as taco.
         # res = matrix + matrix
-    benchmark.pedantic(bench, iterations=10)
+    tacoBench(bench)
 
 @pytest.mark.parametrize("dim", [5000, 10000, 20000])
-def test_add_multiple_sparse_windows(benchmark, dim):
+def test_add_multiple_sparse_windows(tacoBench, dim):
     matrix1 = random(dim, dim, format="csr")
     matrix2 = random(dim, dim, format="csr")
     def bench():
         res = matrix1[1:(dim-1), 1:(dim-1)] + matrix2[1:(dim-1), 1:(dim-1)] + matrix1[0:(dim-2), 0:(dim-2)]
-    benchmark.pedantic(bench, iterations=10)
+    tacoBench(bench)
 
 @pytest.mark.parametrize("dim", [5000, 10000, 20000])
-def test_add_sparse_strided_window(benchmark, dim):
+def test_add_sparse_strided_window(tacoBench, dim):
     matrix = random(dim, dim, format="csr")
     def bench():
         x = matrix[1:(dim-1):2, 1:(dim-1):2] 
         res = x + x
-    benchmark.pedantic(bench, iterations=10)
+    tacoBench(bench)
 
 @pytest.mark.parametrize("dim", [5000, 10000, 20000])
-def test_add_sparse_index_set(benchmark, dim):
+def test_add_sparse_index_set(tacoBench, dim):
     indexes = [i * 2 for i in range(0, dim//2)]
     matrix = random(dim, dim, format="csr")
     def bench():
         x = matrix[:, indexes] 
         res = x + x
-    benchmark.pedantic(bench, iterations=10)
-
+    tacoBench(bench)