Files
mtusi2026lab4-fork/experiments/plot.py

118 lines
5.1 KiB
Python

#!/usr/bin/env python3
"""Plot saved summaries without changing raw evidence or the H1 decision."""
import argparse
import datetime
import hashlib
import json
import platform
import sys
from pathlib import Path
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
import pandas as pd
sys.path.insert(0, str(Path(__file__).resolve().parent / "src"))
from common import campaign_path
def sha256(path):
return hashlib.sha256(path.read_bytes()).hexdigest()
def create_plots(campaign):
inputs = [campaign / "metrics.csv", campaign / "metrics.json"]
frame = pd.read_csv(inputs[0])
summary = json.loads(inputs[1].read_text())
protocol = json.loads((campaign / "protocol.json").read_text())
profiles = [p["name"] for p in protocol["profiles"]]
status = summary["decision"]["criterion_status"]
output = campaign / "plots" / datetime.datetime.now(
datetime.timezone.utc
).strftime("%Y%m%dT%H%M%S.%fZ")
output.mkdir(parents=True)
def finish(fig, ax, title, ylabel, name):
ax.set_title(title + "\n" + status)
ax.set_ylabel(ylabel)
ax.set_xticks(range(len(profiles)), profiles, rotation=20, ha="right")
ax.grid(axis="y", alpha=0.25)
ax.legend(loc="best", fontsize=8)
fig.tight_layout()
fig.savefig(output / (name + ".png"), dpi=160)
plt.close(fig)
def points(column, scale, title, ylabel, name, limit=None):
fig, ax = plt.subplots(figsize=(10, 5))
for mode, offset, color in [("proxy", -0.12, "#2563eb"),
("direct", 0.12, "#d97706")]:
subset = frame[frame["mode"] == mode]
x, y = [], []
for index, profile in enumerate(profiles):
values = pd.to_numeric(
subset.loc[subset["profile"] == profile, column], errors="coerce"
).dropna().tolist()
for j, value in enumerate(values):
x.append(index + offset + (j - (len(values) - 1) / 2) * 0.035)
y.append(value / scale)
ax.scatter(x, y, label=mode, color=color, alpha=0.85)
if limit is not None:
ax.axhline(limit, color="red", linestyle="--", label="API limit")
finish(fig, ax, title + " (each point = one run)", ylabel, name)
points("api_bytes_per_ready", 1024 ** 2, "API RX+TX per ready attachment",
"MiB / ready", "api-traffic")
points("api_sampled_peak_window_max", 1024 ** 2,
"API sampled memory peak: maximum of two replicas", "MiB",
"api-memory-window", limit=512)
points("api_peak_lifetime_max", 1024 ** 2,
"API lifetime memory peak (includes startup and warmup)", "MiB",
"api-memory-lifetime", limit=512)
points("p95_seconds", 1, "Successful upload flow p95", "seconds", "latency-p95")
points("errors", 1, "Upload errors (missing runs are absent)", "count", "errors")
fig, ax = plt.subplots(figsize=(10, 5))
pairs = summary["decision"].get("pairs", [])
for valid, marker, label in [(True, "o", "instrumentation valid"),
(False, "x", "instrumentation issues")]:
selected = [p for p in pairs if p["instrumentation_valid"] == valid]
ax.scatter([profiles.index(p["profile"]) +
(p["repetition"] - 2) * 0.06 for p in selected],
[p["direct_proxy_ratio"] for p in selected],
marker=marker, label=label)
ax.axhline(protocol["criteria"]["median_direct_proxy_traffic_ratio_max"],
linestyle="--", color="green", label="median threshold")
ax.axhline(protocol["criteria"]["each_pair_traffic_ratio_max"],
linestyle=":", color="red", label="each pair threshold")
finish(fig, ax, "Paired API traffic ratio (all available pairs)",
"direct / proxy", "traffic-ratio")
provenance = {
"created_utc": datetime.datetime.now(datetime.timezone.utc).isoformat(),
"python": platform.python_version(), "pandas": pd.__version__,
"matplotlib": matplotlib.__version__,
"inputs_sha256": {f.name: sha256(f) for f in inputs + [campaign / "protocol.json"]},
"plot_script_sha256": sha256(Path(__file__)),
"requirements_sha256": sha256(Path(__file__).parent / "requirements.txt"),
"criterion_status": status,
"note": "All available runs shown; missing values omitted, never replaced with zero. "
"Instrumentation validity and system failures must be read in metrics.json.",
}
(output / "provenance.json").write_text(json.dumps(provenance, indent=2) + "\n")
return output
def main():
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("campaign", help="prepared campaign under root evidence/upload-comparison")
args = parser.parse_args()
try:
print(create_plots(campaign_path(args.campaign)))
except (OSError, ValueError, KeyError) as exc:
parser.exit(1, "Cannot plot: check campaign path and run summarize first (" +
type(exc).__name__ + ").\n")
if __name__ == "__main__":
main()