amquant 0.1.0

A lightweight library for loading market data and computing engineered features.

activate the virtual env with : source venv/bin/activate

interactive program :python3 -i main.py

DATA ACCESS:

After main() executes in the same Python session:

• RAW_SERIES: dict[str, Series] — populated with fully enriched Series objects (OHLCV + metadata).
• FEATURE_SETS: dict[str, FeatureSet] — contains computed technical features per symbol.

Both dicts are module-level globals, accessible immediately post-run via RAW_SERIES['BIAT'] or FEATURE_SETS['MC'].

Data lives in memory for the duration of the Python process / REPL session. Not persisted to disk unless explicitly saved.

For library use: expose RAW_SERIES and FEATURE_SETS via a DataManager singleton or return them from load_universe() for clean dependency injection.

demo

list(RAW_SERIES.keys()) RAW_SERIES["BNP"].bars[0] FEATURE_SETS["BNP"].ret_1d[:5] FEATURE_SETS["BNP"].sma_20[:5] FEATURE_SETS["BNP"].dict.keys() len(RAW_SERIES["BNP"].bars) len(RAW_SERIES) len(FEATURE_SETS)

AmQuant