numpy: parametrize some benchmarks

Author: rohany

Makefile

@@ -1,6 +1,7 @@
 BENCHES := ""
 BENCHFLAGS :=
 
+# To group benchmark output by benchmark, use BENCHFLAGS=--benchmark-group-by=func.
 python-bench: numpy/*.py
 	pytest --ignore=taco $(BENCHFLAGS) $(BENCHES)
 

numpy/windowing.py

@@ -2,36 +2,42 @@
 from scipy.sparse import random
 import pytest
 
-def test_add_sparse_window(benchmark):
+def test_add_window(benchmark):
     dim = 10000
-    matrix = random(dim, dim, format="csr")
+    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)
 
+@pytest.mark.parametrize("dim", [5000, 10000, 20000])
+def test_add_sparse_window(benchmark, 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)
 
-def test_add_multiple_sparse_windows(benchmark):
-    dim = 10000
+@pytest.mark.parametrize("dim", [5000, 10000, 20000])
+def test_add_multiple_sparse_windows(benchmark, 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)
 
-def test_add_sparse_strided_window(benchmark):
-    dim = 10000
+@pytest.mark.parametrize("dim", [5000, 10000, 20000])
+def test_add_sparse_strided_window(benchmark, 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)
 
-def test_add_sparse_index_set(benchmark):
-    dim = 10000
+@pytest.mark.parametrize("dim", [5000, 10000, 20000])
+def test_add_sparse_index_set(benchmark, dim):
     indexes = [i * 2 for i in range(0, dim//2)]
     matrix = random(dim, dim, format="csr")
     def bench():