WIP: Tests

benchmarks/test_view.py

@@ -0,0 +1,81 @@
+import pandas as pd
+import pytest
+import numpy as np
+import benchmarks.view_utilities as view
+
+
+def get_params(length):
+    operations = ["add", "matmul", "qobjevo", "mcsolve", "mesolve"]
+    times = np.linspace(1, length, length, dtype=int)
+    size = np.logspace(1, length, length, base=2, dtype=int)
+    density = ["sparse"] * int(length / 2) + ["dense"] * int(length / 2)
+    model = ["qubit"] * int(length / 2) + ["cavity"] * int(length / 2)
+
+    return (operations, times, size, density, model)
+
+
+def get_sum(all_ops, times, filter_ops):
+    res = 0
+    for i, op in enumerate(all_ops):
+        for filter in filter_ops:
+            if op == filter:
+                res += np.sum(times + i * 10)
+
+    return res
+
+
+def create_dataframe(params):
+
+    operations, times, size, density, model = params
+
+    result = pd.DataFrame()
+
+    for i, item in enumerate(operations):
+        tmp_dict = {}
+        tmp_dict["size"] = size.tolist()
+        tmp_dict["times"] = (times + i * 10).tolist()
+        tmp_dict["params_operation"] = [item] * len(times)
+        if "solve" in item:
+            tmp_dict["model"] = model
+        else:
+            tmp_dict["density"] = density
+        tmp = pd.DataFrame(tmp_dict)
+        result = pd.concat([result, tmp])
+
+    return result
+
+
+@pytest.fixture(params=np.logspace(1, 8, 8, base=2, dtype=int).tolist())
+def length(request):
+    return request.param
+
+
+@pytest.mark.parametrize(
+    ["filters"],
+    [
+        pytest.param(
+            (["add"], ["matmul"], ["qobjevo"], ["mcsolve"], ["mesolve"]),
+            id="single_filters",
+        ),
+        pytest.param(
+            (["add", "matmul"], ["qobjevo", "mcsolve", "mesolve"]),
+            id="multi_filters",
+        ),
+    ],
+)
+def test_sort(filters, length):
+    params = get_params(length)
+    operations, times, _, _, _ = params
+    data = create_dataframe(params)
+
+    for filter in filters:
+
+        test_data = view.sort_ops(data, filter)
+        ref_time = get_sum(operations, times, filter)
+
+        test_time = 0
+        for _, value in test_data.items():
+            test_time += value["times"].sum()
+
+        assert filter == list(test_data.keys())
+        assert test_time == ref_time