File size: 3,370 Bytes
7956103 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 | """CPU environment and finite regression smoke checks, not research results."""
import hashlib
import json
import os
from pathlib import Path
import platform
import subprocess
import sys
for key in ("OMP_NUM_THREADS", "OPENBLAS_NUM_THREADS", "MKL_NUM_THREADS"):
os.environ[key] = "1"
os.environ["MPLBACKEND"] = "Agg"
import numpy as np
import scipy
from scipy.optimize import linprog
from scipy.stats import binom
import matplotlib
matplotlib.use("Agg")
from matplotlib import pyplot as plt
def main():
project = Path(__file__).resolve().parents[1]
pmf = binom.pmf(np.arange(129), 128, 0.25)
np.testing.assert_allclose(pmf.sum(), 1.0, atol=1e-12, rtol=0)
np.testing.assert_allclose(pmf @ np.arange(129), 32.0, atol=1e-11, rtol=0)
# Two allowed actions for one fair pair: serial mismatch mean 1/8, or
# co-sampled mismatch mean 1/2. Hidden mixing can match target mean 1/4.
# Variables: serial weight, co-sample weight, absolute discrepancy.
lp = linprog(
c=[0.0, 0.0, 1.0],
A_ub=[[0.125, 0.5, -1.0], [-0.125, -0.5, -1.0]],
b_ub=[0.25, -0.25],
A_eq=[[1.0, 1.0, 0.0]], b_eq=[1.0],
bounds=[(0, None), (0, None), (0, None)], method="highs",
)
if not lp.success:
raise RuntimeError(lp.message)
np.testing.assert_allclose(lp.fun, 0.0, atol=1e-9, rtol=0)
np.testing.assert_allclose(lp.x[:2], [2 / 3, 1 / 3], atol=1e-9, rtol=0)
fig, ax = plt.subplots()
ax.plot(np.arange(129), pmf)
fig.canvas.draw()
plt.close(fig)
checker = project / "src" / "verify_main_examples.py"
completed = subprocess.run(
[sys.executable, str(checker)], check=True, capture_output=True,
text=True, timeout=180,
)
regression = json.loads(completed.stdout)
if regression.get("status") != "PASS":
raise RuntimeError(regression)
record = {
"status": "PASS",
"scope": "CPU environment and finite regression checks only; no main experiments run",
"python_executable": sys.executable,
"python_version": platform.python_version(),
"platform": platform.platform(),
"hostname": platform.node(),
"packages": {"numpy": np.__version__, "scipy": scipy.__version__,
"matplotlib": matplotlib.__version__},
"environment_mode": "project venv sharing existing Conda base site-packages",
"visible_cpu_count": os.cpu_count(),
"gpu_used": False,
"library_thread_settings": {key: os.environ[key] for key in
("OMP_NUM_THREADS", "OPENBLAS_NUM_THREADS", "MKL_NUM_THREADS")},
"binomial_mass_residual": float(abs(pmf.sum() - 1)),
"highs_randomization_smoke": {
"objective": float(lp.fun), "weights": lp.x[:2].tolist(),
"equality_residual": float(np.max(np.abs(lp.eqlin.residual))),
"scope": "two-action LP only, not a general policy optimizer",
},
"headless_plotting": "PASS (canvas only; no publication figure generated)",
"finite_regression": regression,
"finite_regression_sha256": hashlib.sha256(checker.read_bytes()).hexdigest(),
}
output = project / "logs" / "environment_check.json"
output.write_text(json.dumps(record, indent=2) + "\n", encoding="utf-8")
print(json.dumps(record, indent=2))
if __name__ == "__main__":
main()
|