h5i-db

Use when working with an h5i-db database — an embedded, versioned time-series database and event-driven backtesting engine for quant research, driven by the `h5i-db` CLI or the `h5i_db` Python package. Covers creating a database and loading Parquet or CSV tick data, orienting in an existing one, querying it (SQL, ASOF joins, OHLCV/VWAP rollups, time travel) under a context budget, look-ahead-free reads, running event-driven backtests with settlement and fee models, scoring them (factor stats, tearsheets, deflated Sharpe), importing vendor market-data archives, and previewing destructive changes before they land.