backtest-diagnose

To debug a failing trading backtest, you inspect Python artifacts like metrics.csv, equity.csv, and trades.csv to surface failure signals, then examine code and configuration to pinpoint the root cause before applying fixes.

Unreliable backtest results are typically caused by runtime errors, logic bugs in signal_engine.py, or data errors, which can be identified by inspecting output artifacts and configuration files using a structured diagnostic taxonomy.

Fixing logic bugs in a signal engine involves diagnosing the issue in signal_engine.py or config.json, applying a targeted code update via edit_file, and re-running the backtest to verify that the metrics are stable.

Yes, you can diagnose data errors in backtest trades.csv and equity.csv files by reading and inspecting these artifacts to surface failure signals, classifying the issue, and prescribing fixes to the underlying source code.

The best way to troubleshoot a Python backtest with runtime errors is to apply a hard-gate checklist that inspects metrics and code files, classifies the error, and guides you through editing the source and re-running the test.

A trading backtest may fail after updating config.json due to introduced data errors or logic bugs, requiring a diagnostic inspection of artifacts and configuration to classify the issue and apply a corrective edit.