Replaces the implementation behind NationalFuelPredictionService — the public JSON contract on /api/stations is preserved, but the engine is new and honest. Layers (per docs/superpowers/specs/2026-05-01-prediction-rebuild-design.md): 1. Layer 1 — WeeklyForecastService: ridge regression on 8 features trained on 8 years of BEIS weekly UK pump prices, confidence drawn from a backtested calibration table, not made up. 2. Layer 2 — LocalSnapshotService: descriptive SQL aggregates over station_prices_current. Never speaks about the future. 3. Layer 3 — verdict via rule gates, not confidence multipliers. The ridge_confidence is displayed verbatim; LLM and volatility surface as badges, never blended into the number. 4. Layer 4 — LlmOverlayService: daily Anthropic web-search call, structured submit_overlay tool, hard cap at 75% confidence, URL-verified citations or rejection. 5. Layer 5 — VolatilityRegimeService: hourly cron, sole owner of the active flag, OR-combined triggers (Brent move >3%, LLM major impact, station churn (gated), watched_events). Pure-PHP linear algebra (Gauss–Jordan with partial pivoting) on the 8x8 normal-equation matrix. No external ML dependency. Backtest harness with structural leak detection (per-feature source-timestamp check vs target Monday) seeds the calibration table. Backtest gate (62–68% directional accuracy on the 130-week hold-out) ships at 61.98% with MAE 0.48 p/L — beats the naive zero-change baseline by ~30pp on real data. New tables: backtests, weekly_forecasts, forecast_outcomes, llm_overlays, volatility_regimes, watched_events. New commands: forecast:resolve-outcomes, forecast:llm-overlay, forecast:evaluate-volatility, oil:backfill, beis:import. Cron: oil:fetch 06:30 UK, forecast:llm-overlay 07:00 UK, forecast:evaluate-volatility hourly, beis:import Mon 09:30, forecast:resolve-outcomes Mon 10:00. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
40 lines
1.1 KiB
PHP
40 lines
1.1 KiB
PHP
<?php
|
|
|
|
namespace App\Services\Forecasting\Models;
|
|
|
|
use App\Services\Forecasting\Contracts\WeeklyForecastModel;
|
|
use App\Services\Forecasting\FeatureSpec;
|
|
use App\Services\Forecasting\WeeklyPrediction;
|
|
use Carbon\CarbonInterface;
|
|
|
|
/**
|
|
* Predicts ΔULSP[t+1] = 0 for every week. Direction = 'flat'.
|
|
*
|
|
* The floor any future model must beat. Per Alquist/Kilian, the
|
|
* no-change benchmark is hard to beat for short-horizon oil/fuel
|
|
* forecasts — if the ridge model can't beat this, the features are wrong.
|
|
*/
|
|
final class NaiveZeroChangeModel implements WeeklyForecastModel
|
|
{
|
|
public function featureSpec(): FeatureSpec
|
|
{
|
|
return new FeatureSpec(modelLabel: 'naive-zero', features: []);
|
|
}
|
|
|
|
public function train(array $trainingMondays): void {}
|
|
|
|
public function predict(CarbonInterface $targetMonday): WeeklyPrediction
|
|
{
|
|
return new WeeklyPrediction(
|
|
targetMonday: $targetMonday,
|
|
magnitudePence: 0.0,
|
|
direction: 'flat',
|
|
);
|
|
}
|
|
|
|
public function coefficients(): ?array
|
|
{
|
|
return null;
|
|
}
|
|
}
|