37 lines (25 loc) · 1.02 KB

window sampler

Language: Python · Sphere: programming · Category: Monitoring

Signature: () → None

What it does

Anomaly Detection Sampler Module

This module provides classes for detecting anomalies in time-series data using statistical methods with adaptive baselines and windowed sampling.

Guarantee

When it runs, window sampler guarantees sampler.is_full() and sampler.size() == 5; sum(sampler.get_values()) == 20; z == 3.0 (proven by run).

Checkable constraints:

sampler.get_values() == [2, 3, 4, 5, 6]
sampler.is_full() and sampler.size() == 5
sum(sampler.get_values()) == 20
z == 3.0
is_anom
mean == 10.0
std == 30.0
not is_anom and z == 0.0

Verification evidence

Green-run: ✓ passes (re-run under the extractor's gate)
Constraint strength: recovery (truth-pinned)
Independent oracle: — none yet (green-run candidate; not an axiom under the frozen ruler)
Peer review: unreviewed

△ AURA Pattern Library — © Reality Optimizer