Physics from Finite Validation — code

Reproducibility repository for the preprint:

Physics from Finite Validation: Classical Recovery and the Quantum Boundary United Field Initiative (UFI) · ufi.observer · ORCID 0009-0000-3163-4426 DOI: 10.5281/zenodo.20628369

Every script is standalone, seeded, and carries its status in the docstring. Several of them document arguments the paper itself withdrew; that is deliberate, and the paper reports them as withdrawn.

State as first published is preserved under tag v1.0.

Requirements

• Python 3.10+
• numpy

pip install numpy

Classical sector (§4)

Script	Section	What it does
calibrate_gr.py	4.2	Gravitational time dilation: the model's leading term coincides with weak-field GR; calibration constant A = 1.0000, no fitting.
two_channel_derivation.py	4.3, 5.2	The temporal and spatial contributions to the light deflection are equal, forced by the invariance of c = L/τ. Factor 2 not inserted by hand.
close_two_holes.py	4.3, 5.2	Status audit. Equality derived under a named reading; the multiplicative composition imported, not a lattice theorem. Withdraws the claim that multiplicativity follows from lattice orthogonality.
hypothesis_B_compounding.py	4.3	Perihelion fork: the path-compounded law (1+φ/N)^N → exp(φ) gives β = 1 → 42.99″/century (observed 42.98).
beta_mercury_test.py	4.3	The bare two-channel product (β = 1/2) predicts ~50″/century — excluded by the data as the composition law.
hole_b_second_order.py	4.3	Second-order analysis: the additive law is excluded; the composition is product-like.
emc2_energy_derivation.py	4.3	E = mc² as a two-channel structural analogy — postulated, not derived; dimensional and electron-rest-energy consistency check.
side_two_independent.py	5.7	The classical bound 2 is a theorem of binarity (exhaustive enumeration of ±1 assignments).

Quantum sector (§5)

Script	Section	Status
lock_to_kuramoto.py	5.2	Reading of the model: a lock is an integrate-and-fire oscillator; the coupling threshold. Withdraws the claim that the leak is necessary for synchronisation.
correlation_identity.py	5.4, 5.5	Identity, exact: E = 1 − 2D − s, and its binary limit E = 1 − 2D. No postselection.
chsh_alphabet_bound.py	5.5	Exhaustive enumeration: neither a binary nor a ternary local model exceeds CHSH = 2. Refutes the hypothesis that the quantum excess lives in the silence.
local_readout_arcsin.py	5.7	Derived: a local sign-readout of pre-distributed data lands on exactly 2.
single_act_vector.py	5.7	Derived: Tsirelson's vector model; why the parallelogram law is not the load; the exclusion of the PR-box; pointwise ±1 components cap CHSH at 2 while one zero component reaches 2√2.
angular_resolution.py	5.7	Derived: the residual against the cosine is first order in the act's grid, vanishes on its native directions, and is therefore exactly zero at the CHSH optimum. Withdraws the earlier second-order estimate and the bounds K ≳ 64, K ≳ 256.
born_from_count.py	5.9	Derived: the Born value is the differentiation fraction, D = sin²(θ/2). Why Gleason cannot reach dimension two.
indivisibility_consequences.py	5.10	Derived: complementarity, the phase-space bound, the standard quantum limit, the Zeno effect. Records a refuted attempt to read the uncertainty relation off frequency estimation.
twist_parity.py	5.11	Taken mathematics: the double cover, and which ℤ₂ actually carries spin. Corrects an earlier conflation of Möbius parity with the belt-trick class.
closure_mass_twist.py	5.11	Derived: circuits as rest mass (the de Broglie clock); the claim that this picture supplies no fermionic option for a massless excitation.

Retained, not deleted: single_act_metric_v1_superseded.py — the earlier and incorrect reading, in which the differentiation count was said to obey the parallelogram law. Its error is documented in place.

Where the load sits

One axiom carries the correlational content of the quantum sector. It is named identically in the abstract, in Status and Scope, in 5.7 and in 6.7:

The settings of a single undivided act are unconstrained unit vectors of a real inner-product space, not pointwise ±1 records carried apart.

If it fails, the Tsirelson bound reverts to 2 and PR-boxes are not excluded. The parallelogram law is not the load: a local hidden-variable model satisfies it too, and the gap is closed by a pointwise fact, not by the geometry.

How to run

python3 <script>.py

Each script prints its own verification and states, at the end, what is derived, what is taken, and what is owed.

Withdrawn arguments, in one place

Script	What it withdraws
chsh_alphabet_bound.py	"the quantum excess lives in the silence"
angular_resolution.py	the bounds K ≳ 64 and K ≳ 256 (a second-order estimate valid only at the cosine's flat points)
lock_to_kuramoto.py	"the leak is necessary for synchronisation"
indivisibility_consequences.py	reading the uncertainty relation off frequency estimation
close_two_holes.py	multiplicativity from lattice orthogonality (two_channel_derivation.py's claim)
single_act_metric_v1_superseded.py	"the differentiation count obeys the parallelogram law"