- Introduced `strategy_backtest.py` as a single source of truth for scoring and backtesting.
- Implemented functions to load live model bundles, build feature frames, and execute backtests.
- Ensured consistency between optimized models and their validation/trading counterparts.
- Added functionality for generating trading signals and retrieving the latest trading decision.
This commit introduces a new CSV file containing the results of a lot escalation what-if analysis. The report includes various regimes, timeframes, trade counts, net profits, final balances, final equities, returns, maximum drawdowns, and profit factors for different escalation scenarios.
- Adjusted session consistency test results to reflect empty DataFrame.
- Updated candidate counts and hashes in pipeline manifest explorer.
- Modified Explorer Diagnostics CSV files to reflect new execution counts and removed unnecessary columns.
- Revised signal loss reports for M15 and M5 to remove stage information while retaining lost counts.
- Enhanced explorer consistency report with updated generated and loaded counts, along with TBM, HMM, XGB pass rates, and executed trades.
- Introduced `pipeline_certification.py` for Phase 13 checks on the observability layer, including stage order, monotonicity, unique IDs, and stale reports.
- Implemented `pipeline_profiler.py` for event-based profiling, ensuring integrity across pipeline stages and providing methods for recording events and generating reports.
- Created `prediction.py` to standardize prediction results and ensure compatibility across different components.
- Added a smoke test script to validate module imports and function signatures before running notebooks.
Luke Mosoti