Excise identity L2 dual-mode and pipeline PASSTHROUGH cold-starts; enforce wave-only IdentityCheck with MissingWaveStateError; hard-fail GoldTether transitions via GoldTetherViolationError; dual-competing Shadow Coherence Gate with populated contract_assessment; auto-compile field packets before intent ratify; conformal argmax ratification and content-addressed proof atoms; sandwich multi-modality ingress with GoldTether digests. Amend ADR-0243/0244 (docs/adr + research) for wave-only / SUPERSEDED sketches. Pin binary gates in tests/test_geometric_convergence_checklist.py. [Verification]: Smoke suite passed locally (~136–142s, 176 passed); cognition suite passed (122 passed, 1 skipped); lifecycle suite 51 passed.
136 lines
4.2 KiB
Python
136 lines
4.2 KiB
Python
"""ADR-0238 — GoldTether residual, floor, autonomy, may_relax_hitl, blend."""
|
|
|
|
from __future__ import annotations
|
|
|
|
import numpy as np
|
|
import pytest
|
|
from hypothesis import given, settings
|
|
from hypothesis import strategies as st
|
|
|
|
from algebra.rotor import make_rotor_from_angle
|
|
from algebra.versor import versor_condition
|
|
from core.physics.goldtether import (
|
|
AutonomyBand,
|
|
GoldTetherMonitor,
|
|
GoldTetherViolationError,
|
|
OperatingMode,
|
|
coherence_residual,
|
|
require_unitary,
|
|
)
|
|
|
|
|
|
def _id() -> np.ndarray:
|
|
v = np.zeros(32, dtype=np.float64)
|
|
v[0] = 1.0
|
|
return v
|
|
|
|
|
|
def test_coherence_residual_nonnegative_and_zero_on_identity():
|
|
assert coherence_residual(_id()) == 0.0
|
|
r = coherence_residual(make_rotor_from_angle(0.5))
|
|
assert r >= 0.0
|
|
|
|
|
|
def test_residual_dual_corrected_and_replay():
|
|
m = GoldTetherMonitor()
|
|
F = make_rotor_from_angle(0.4)
|
|
assert m.residual(F) == m.residual(F)
|
|
assert m.residual(F) == coherence_residual(F)
|
|
|
|
|
|
def test_fail_closed_on_drift():
|
|
m = GoldTetherMonitor(epsilon_drift=1e-9)
|
|
# Inject a non-closed multivector by raw coefficients
|
|
dirty = np.zeros(32, dtype=np.float64)
|
|
dirty[0] = 0.5
|
|
dirty[1] = 0.5
|
|
with pytest.raises(GoldTetherViolationError) as excinfo:
|
|
m.update(dirty, epistemic_elevation=True)
|
|
assert excinfo.value.residual > m.epsilon_drift
|
|
assert m.autonomy == 0.0
|
|
assert m.may_relax_hitl() is False
|
|
|
|
|
|
def test_require_unitary_rejects_above_epsilon():
|
|
dirty = np.zeros(32, dtype=np.float64)
|
|
dirty[0] = 0.5
|
|
dirty[1] = 0.5
|
|
with pytest.raises(GoldTetherViolationError):
|
|
require_unitary(dirty, epsilon=1e-6)
|
|
# Identity is admitted.
|
|
assert require_unitary(_id(), epsilon=1e-6) == 0.0
|
|
|
|
|
|
def test_epistemic_elevation_raises_floor_and_autonomy():
|
|
m = GoldTetherMonitor(epsilon_drift=1e-5, floor_step=0.1, autonomy_step=0.1)
|
|
F = _id()
|
|
for _ in range(10):
|
|
m.update(F, epistemic_elevation=True)
|
|
assert m.floor > 0.0
|
|
assert m.autonomy > 0.0
|
|
assert m.autonomy <= m.floor
|
|
assert m.supervised_autonomy_level == m.autonomy
|
|
|
|
|
|
def test_never_autonomy_above_floor():
|
|
m = GoldTetherMonitor(floor_step=0.02, autonomy_step=0.5)
|
|
for _ in range(20):
|
|
m.update(_id(), epistemic_elevation=True)
|
|
assert m.autonomy <= m.floor + 1e-12
|
|
|
|
|
|
def test_may_relax_hitl_thresholds():
|
|
m = GoldTetherMonitor(
|
|
hitl_floor_threshold=0.3,
|
|
hitl_autonomy_threshold=0.2,
|
|
floor_step=0.1,
|
|
autonomy_step=0.1,
|
|
)
|
|
assert m.may_relax_hitl() is False
|
|
for _ in range(20):
|
|
m.update(_id(), epistemic_elevation=True)
|
|
assert m.may_relax_hitl() is True
|
|
|
|
|
|
def test_force_reset():
|
|
m = GoldTetherMonitor()
|
|
m.update(_id(), epistemic_elevation=True)
|
|
m.force_reset()
|
|
assert m.floor == 0.0 and m.autonomy == 0.0 and m.history == []
|
|
|
|
|
|
def test_serve_never_autonomous_band():
|
|
m = GoldTetherMonitor(floor_step=0.2, autonomy_step=0.2, hitl_floor_threshold=0.1, hitl_autonomy_threshold=0.1)
|
|
for _ in range(10):
|
|
m.update(_id(), epistemic_elevation=True)
|
|
d = m.decide(0.0, mode=OperatingMode.SERVE)
|
|
assert d.band is not AutonomyBand.AUTONOMOUS
|
|
|
|
|
|
def test_lifelong_curve_telemetry_replay():
|
|
m1 = GoldTetherMonitor()
|
|
m2 = GoldTetherMonitor()
|
|
for _ in range(5):
|
|
# identity always closed; deterministic elevation path
|
|
m1.update(_id(), epistemic_elevation=True)
|
|
m2.update(_id(), epistemic_elevation=True)
|
|
assert m1.telemetry()["history_tail"] == m2.telemetry()["history_tail"]
|
|
assert m1.telemetry()["schema_version"] == "goldtether_coherence_v2"
|
|
|
|
|
|
def test_supervised_blend_closure_and_endpoints():
|
|
m = GoldTetherMonitor()
|
|
src = _id()
|
|
tgt = make_rotor_from_angle(0.6)
|
|
assert np.allclose(m.supervised_blend(src, tgt, 0.0), src)
|
|
assert np.allclose(m.supervised_blend(src, tgt, 1.0), tgt, atol=1e-6)
|
|
mid = m.supervised_blend(src, tgt, 0.5)
|
|
assert versor_condition(mid) < 1e-6
|
|
|
|
|
|
@given(st.floats(min_value=0.0, max_value=1.0, allow_nan=False, allow_infinity=False))
|
|
@settings(max_examples=30)
|
|
def test_blend_property_closed(alpha: float):
|
|
m = GoldTetherMonitor()
|
|
out = m.supervised_blend(_id(), make_rotor_from_angle(0.8), float(alpha))
|
|
assert versor_condition(out) < 1e-6
|