vvsandClaude Opus 5
1b6c626f67
(2) The proposed method is again the scheduled projection (2806/3220/6026), its budget multiplier DECLARED as the one free constant: the canonical total is the same for every multiplier 1.25-4, the wide-bracket one is not. The earned projection is the constant-free variant (3568/3020/6588). What the old tables hid: problem by problem the proposed rule beats Chandrupatla on the canonical set 87/24/42, exact sign test p = 9e-5, while trailing him by 0.7% in total; on the wide-bracket set he is genuinely stronger (26 to 9, p = 0.006). Gap decomposition recomputed with the library: 244 (was 262), 160 of it on the flat families (two thirds), fun1 58 (a quarter, not a third as the abstract said). Also corrected on the way: the 'subsumed by Chandrupatla' pin claimed he is better on the worst single problem - true only with the looser tolerance (84 against the credit rule's 83 once fixed); §8.3 still called zero slack degenerate, which Table 1a refutes; §5.2's Brent figure is our pure-Python implementation and now says so. verify_article.py 286/286, 802 tests pass twice. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
59 lines
2.2 KiB
Python
59 lines
2.2 KiB
Python
"""Is fun7 a method failure or a SCORING failure?
|
|
|
|
Every method stops on fun7 within 8-22 evaluations and every one is scored unsolved. This
|
|
checks the only thing that matters: is f actually zero at the point each method returned?
|
|
If it is, the methods found a root and the benchmark's |x - declared_root| <= 1e-6 check is
|
|
what fails — 5 of the 45 problems then contribute 200 dead evaluations each to EVERY
|
|
method's total, which is a fact about the scoring, not about the methods.
|
|
|
|
python -u diag_fun7_root.py
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import inspect
|
|
import math
|
|
|
|
from scipy.optimize._tstutils import _CHANDRUPATLA_TESTS_DICTS
|
|
|
|
from simroot.baselines import CHANDRUPATLA_BASELINE
|
|
from simroot.baselines import best_of, chandrupatla, chandrupatla_scipy
|
|
from simroot.pure import pure_root
|
|
from sweep_full import MAXFEV, XTOL
|
|
|
|
CHA = CHANDRUPATLA_BASELINE # the library alone (2026-09-21)
|
|
|
|
|
|
def main():
|
|
for d in _CHANDRUPATLA_TESTS_DICTS:
|
|
if not str(d["ID"]).startswith("fun7"):
|
|
continue
|
|
f, root = d["f"], float(d["root"])
|
|
a, b = (float(v) for v in d["bracket"])
|
|
x_o, n_o, _ok = pure_root(f, a, b, XTOL, MAXFEV, similarity="ratio",
|
|
scale="derived3", guard="safe", prestep="slope",
|
|
direction="dissimilar")
|
|
x_c, n_c, _okc = CHA(f, a, b, XTOL, MAXFEV)
|
|
print(f"{d['ID']:<8} declared root {root:+.3g}")
|
|
for nm, x, n in (("ours", x_o, n_o), ("chandrupatla", float(x_c), n_c)):
|
|
fx = f(x)
|
|
print(f" {nm:<13} returned x={x:+.10g} after {n:3d} evals "
|
|
f"f(x)={fx:+.3e} |x-root|={abs(x - root):.3g} "
|
|
f"{'f IS ZERO THERE' if fx == 0.0 else 'f is not zero'}")
|
|
# how wide is the flat zero region around the declared root?
|
|
w = 0.0
|
|
for k in range(0, 40):
|
|
t = 10.0 ** (-12 + k * 0.5)
|
|
if f(root + t) == 0.0:
|
|
w = t
|
|
else:
|
|
break
|
|
print(f" f is identically zero at least up to root + {w:.3g}")
|
|
src = inspect.getsource(_CHANDRUPATLA_TESTS_DICTS[0]["f"]) if False else None
|
|
if src:
|
|
print(src)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|