smptests/smplock01/smplock01fair.py

0001 # SPDX-License-Identifier: BSD-2-Clause
0002
0003 # Copyright (C) 2016, 2024 embedded brains GmbH & Co. KG
0004 #
0005 # Redistribution and use in source and binary forms, with or without
0006 # modification, are permitted provided that the following conditions
0007 # are met:
0008 # 1. Redistributions of source code must retain the above copyright
0009 #    notice, this list of conditions and the following disclaimer.
0010 # 2. Redistributions in binary form must reproduce the above copyright
0011 #    notice, this list of conditions and the following disclaimer in the
0012 #    documentation and/or other materials provided with the distribution.
0013 #
0014 # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
0015 # AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
0016 # IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
0017 # ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
0018 # LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
0019 # CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
0020 # SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
0021 # INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
0022 # CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
0023 # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
0024 # POSSIBILITY OF SUCH DAMAGE.
0025
0026 import json
0027 import math
0028 import re
0029 import statistics
0030 import matplotlib.pyplot as plt  # type: ignore
0031 from matplotlib import ticker  # type: ignore
0032
0033
0034 def _normed_coefficient_of_variation(counter: list[int]) -> float:
0035     return (statistics.stdev(counter) / statistics.mean(counter)) / math.sqrt(
0036         len(counter))
0037
0038
0039 def _plot(data: dict) -> None:
0040     _, axes = plt.subplots()
0041     axes.set_title("SMP Lock Fairness")
0042     axes.set_xlabel("Active Workers")
0043     axes.set_ylabel("Normed Coefficient of Variation")
0044     axes.set_yscale("symlog", linthresh=1e-6)
0045     x = list(range(2, len(data[0]["results"]) + 1))
0046     axes.xaxis.set_major_locator(ticker.FixedLocator(x))
0047     for samples in data:
0048         if samples["lock-object"] != "global":
0049             continue
0050         if samples["section-type"] != "local counter":
0051             continue
0052         y = [
0053             _normed_coefficient_of_variation(results["counter"])
0054             for results in samples["results"][1:]
0055         ]
0056         axes.plot(x, y, label=samples["lock-type"], marker="o")
0057     axes.legend(loc="best")
0058     plt.savefig("smplock01fair.png")
0059     plt.savefig("smplock01fair.pdf")
0060     plt.close()
0061
0062
0063 _JSON_DATA = re.compile(
0064     r"\*\*\* BEGIN OF JSON DATA \*\*\*(.*)"
0065     r"\*\*\* END OF JSON DATA \*\*\*", re.DOTALL)
0066
0067 with open("smplock01.scn", "r", encoding="utf-8") as src:
0068     match = _JSON_DATA.search(src.read())
0069     data = json.loads(match.group(1))
0070
0071 _plot(data)