mirror of
https://github.com/ruvnet/RuView.git
synced 2026-10-22 14:42:44 -05:00
rUv
56e5ae1a6d
Retain the stronger ACF model after an 18-configuration temporal ablation regresses on development data. Add a pinned actual-library integrity probe, strict bounded replay inspection, measured synthetic overhead, and CI tests. Preserve source hashes and separate software integrity from sensing accuracy.
131 lines
7.2 KiB
Python
131 lines
7.2 KiB
Python
"""Synthetic feature and experiment contracts, not real sensing accuracy."""
|
|
import copy
|
|
import json
|
|
from pathlib import Path
|
|
import sys
|
|
import tempfile
|
|
import unittest
|
|
|
|
import numpy as np
|
|
from safetensors.numpy import save_file
|
|
|
|
sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
|
|
from csi_training.espfi_context import BASELINE, CONTRACT, GRID, WIDTHS, ContextBundle, context_features, search
|
|
from csi_training.espfi_kernel import fit
|
|
from csi_training.espfi_motion import ROBUST_CONTRACT, motion_features
|
|
from csi_training.espfi_train import digest, write_json
|
|
from test_espfi_generalize import fixture
|
|
|
|
|
|
def make_baseline(parts, provenance, root):
|
|
root.mkdir()
|
|
state = fit(motion_features(parts['train']['x'], 'motion_acf'), parts['train']['y'], BASELINE)
|
|
save_file(state, str(root / 'model.safetensors'))
|
|
write_json(root / 'status.json', {'status': 'complete', 'promotionAuthorized': False})
|
|
write_json(root / 'metrics.json', {
|
|
'status': 'complete', 'promotionAuthorized': False, 'task': 'public_motion_ablation',
|
|
'contract': ROBUST_CONTRACT, 'selected': {'candidate': BASELINE},
|
|
'model_sha256': digest(root / 'model.safetensors'), 'provenance': provenance,
|
|
})
|
|
|
|
|
|
class ContextTests(unittest.TestCase):
|
|
def test_order_reversal_gain_and_exact_control(self):
|
|
rng = np.random.default_rng(43)
|
|
x = np.exp(rng.normal(size=(2, 52, 190)))
|
|
actual = context_features(x, 'ordered_trend')
|
|
reverse = context_features(x[..., ::-1], 'ordered_trend')
|
|
np.testing.assert_array_equal(actual[:, :468], motion_features(x, 'motion_acf'))
|
|
np.testing.assert_allclose(actual[:, :468], reverse[:, :468], atol=1e-12)
|
|
np.testing.assert_allclose(actual[:, 468:624].reshape(2, 52, 3),
|
|
reverse[:, 468:624].reshape(2, 52, 3)[..., ::-1], atol=1e-12)
|
|
np.testing.assert_allclose(actual[:, 624:], -reverse[:, 624:], atol=1e-12)
|
|
self.assertGreater(np.max(np.abs(actual[:, 468:] - reverse[:, 468:])), .01)
|
|
gain = np.exp(rng.normal(size=(2, 52, 1)))
|
|
np.testing.assert_allclose(actual, context_features(x * gain, 'ordered_trend'), atol=1e-12)
|
|
np.testing.assert_array_equal(context_features(x, 'ordered_mean'), actual[:, :624])
|
|
# Independent mathematical reference for the added descriptors.
|
|
z = np.log(x)
|
|
z -= z.mean(-1, keepdims=True)
|
|
z /= np.sqrt(np.mean(z**2, axis=(1, 2), keepdims=True))
|
|
expected = np.stack((z[..., :63].mean(-1), z[..., 63:127].mean(-1), z[..., 127:].mean(-1)), -1)
|
|
np.testing.assert_allclose(actual[:, 468:624], expected.reshape(2, -1), atol=1e-12)
|
|
t = np.linspace(-1., 1., 190)
|
|
np.testing.assert_allclose(actual[:, 624:], np.sum(z * t, -1) / np.sum(t*t), atol=1e-12)
|
|
|
|
def test_numerical_edges_and_input_bounds(self):
|
|
for family, width in WIDTHS.items():
|
|
for level in (0., 1e-20, 1., 1e100):
|
|
result = context_features(np.full((1, 52, 190), level), family)
|
|
self.assertEqual(result.shape, (1, width))
|
|
self.assertTrue(np.isfinite(result).all())
|
|
x = np.ones((1, 52, 380))[:, :, ::2]
|
|
x[..., 100] += 1e-12
|
|
self.assertTrue(np.isfinite(context_features(x, family)).all())
|
|
for x in (np.ones((0, 52, 190)), np.ones((1, 52, 189)),
|
|
-np.ones((1, 52, 190)), np.full((1, 52, 190), np.nan)):
|
|
with self.assertRaises(ValueError):
|
|
context_features(x, 'ordered_mean')
|
|
with self.assertRaises(ValueError):
|
|
context_features(np.ones((1, 52, 190)), 'room_history')
|
|
|
|
def test_selection_isolation_control_reload_and_tamper(self):
|
|
parts, provenance = fixture()
|
|
with tempfile.TemporaryDirectory() as tmp:
|
|
root = Path(tmp)
|
|
make_baseline(parts, provenance, root / 'baseline')
|
|
report = search(parts, provenance, root / 'baseline', root / 'a')
|
|
changed = copy.deepcopy(parts)
|
|
changed['validation']['x'] = changed['validation']['x'][..., ::-1] * 1.2
|
|
changed['validation']['y'] = (changed['validation']['y'] + 1) % 7
|
|
changed['validation']['identity'][:, 2] = changed['validation']['y'] + 1
|
|
other = search(changed, provenance, root / 'baseline', root / 'b')
|
|
self.assertEqual(len(GRID), 18)
|
|
self.assertEqual(report['selected'], other['selected'])
|
|
self.assertEqual(report['model_sha256'], other['model_sha256'])
|
|
self.assertTrue(report['baseline_state_exact'])
|
|
self.assertEqual(report['baseline_control']['candidate'], BASELINE)
|
|
winner = max(report['candidates'], key=lambda r: (r['minimum_cv_macro_f1'],
|
|
r['mean_cv_macro_f1'], -r['scalar_count']))
|
|
self.assertEqual(report['selected'], winner)
|
|
self.assertEqual(report['contract'], CONTRACT)
|
|
self.assertEqual(report['evidence'], 'SYNTHETIC')
|
|
self.assertFalse(report['promotionAuthorized'])
|
|
self.assertEqual(report['reload_max_abs_logit_error'], 0)
|
|
self.assertEqual(ContextBundle(root / 'a').predict(parts['validation']['x']).shape, (7, 7))
|
|
# Changing the scored development labels cannot alter the persisted choice.
|
|
selection = json.loads((root / 'a' / 'selection.json').read_text())
|
|
self.assertEqual(selection['selected'], report['selected'])
|
|
self.assertNotIn('development', selection)
|
|
original = (root / 'a' / 'model.safetensors').read_bytes()
|
|
(root / 'a' / 'model.safetensors').write_bytes(original + b'corruption')
|
|
with self.assertRaisesRegex(ValueError, 'hash mismatch'):
|
|
ContextBundle(root / 'a')
|
|
(root / 'a' / 'model.safetensors').write_bytes(original)
|
|
write_json(root / 'a' / 'status.json', {'status': 'failed', 'promotionAuthorized': False})
|
|
with self.assertRaises(ValueError):
|
|
ContextBundle(root / 'a')
|
|
|
|
def test_partition_provenance_and_failed_control(self):
|
|
parts, provenance = fixture()
|
|
with tempfile.TemporaryDirectory() as tmp:
|
|
root = Path(tmp)
|
|
make_baseline(parts, provenance, root / 'baseline')
|
|
invalid = dict(parts, test=parts['validation'])
|
|
with self.assertRaises(ValueError):
|
|
search(invalid, provenance, root / 'baseline', root / 'bad_split')
|
|
with self.assertRaisesRegex(ValueError, 'provenance mismatch'):
|
|
search(parts, dict(provenance, manifest_sha256='different'), root / 'baseline', root / 'bad_hash')
|
|
# Wrong learned control cannot silently support a comparison.
|
|
changed = copy.deepcopy(parts)
|
|
changed['train']['x'][:, :, 10] *= 2.
|
|
with self.assertRaisesRegex(ValueError, 'did not reproduce'):
|
|
search(changed, provenance, root / 'baseline', root / 'bad_control')
|
|
self.assertEqual(json.loads((root / 'bad_control' / 'status.json').read_text())['status'], 'failed')
|
|
with self.assertRaises(FileExistsError):
|
|
search(parts, provenance, root / 'baseline', root / 'bad_control')
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|