Frostyverse Laboratory — Gravitational Redshift Python Test
Expected / preserved result fingerprints
===========================================================

Scientific boundary
-------------------
This package reproduces the provisionally frozen H-117/H-118 STATIC gravitational
redshift / stationary-clock candidate. It is not the separate cosmological
"Uncaught Redshift Tax" idea from the manuscript.

H-117 flat-clock exponent
-------------------------
Derived from H-017 before gravity/GR scoring:
  AXIS alpha = 0.500000000000
  FACE alpha = 0.500000000000
  BODY alpha = 0.500000000000
  mean alpha = 0.500000000000

The numeric 1/2 is not supplied to the fit.

Focused reference used by lab_test.py
-------------------------------------
Resolution R20
inner radius = 1.60
outer radius = 2.70

Preserved H-117 report:
  FV redshift z = 2.493038134e-07
  GR redshift z = 2.546298506e-07
  FV/GR         = 0.979083
  relative error = 2.0917%

The focused wrapper regenerates the FV field/profile and seals the FV result
before calling the historical GR ruler.

Full H-117 preserved summary
----------------------------
  R20 mean FV/GR = 0.987119 ; worst error = 2.0917% ; spread = 1.4421%
  R40 mean FV/GR = 0.993491 ; worst error = 1.0428% ; spread = 0.7026%
  R80 mean FV/GR = 0.996654 ; worst error = 0.5274% ; spread = 0.3457%
  diagnostic 1/R continuum intercept = 0.999839941
  sealed FV fingerprint = 8c928bfd5828c57ee5f14a3f388c3ef5e2c39b3f348de6d79118b843a3292ed1

H-118 high-resolution confirmation
----------------------------------
  R80  worst redshift error = 0.5274%
  R120 worst redshift error = 0.3380%
  R160 worst redshift error = 0.2723%
  sealed FV fingerprint = 6ad5fc453df9fff609d124677373c6ebae46c88131551e991cc97d4ab02d3f15

Resource warning
----------------
The unchanged historical high-resolution programs are intentionally bundled,
but they are not lightweight:
  H-117 R80 historical main array: ~3.810 GiB
  H-118 R120 historical main array: ~12.864 GiB
  H-118 R160 historical main array: ~30.499 GiB
plus Python/process overhead and H-118 disk-backed generation work.

The reviewer-facing focused mode defaults to R20 and permits custom R12..R80.
