trading_shared

Shared trading utilities for the Willow ecosystem.

This package centralises feature engineering and other reusable logic so that both:

• willow-trader-lab
• binance_simulator

can consume exactly the same implementations and stay in sync.

Installation (editable)

From the workspace root (or from this directory):

python -m pip install -e .

Current contents

• trading_shared.features
  • FEATURE_KEYS_DEFAULT: canonical list of feature names for baseline strategies.
  • FEATURE_KEYS_BTC_SYS_1H: curated feature set for the BTCUSDT 1h 3‑class system (btc_sys_1h_v1), including richer volatility, EMA distance, Bollinger Bands and candle‑shape metrics.
  • compute_indicator_frame(df): computes a superset of indicator time series used by both training and simulation. Consumers select FEATURE_KEYS_* slices depending on the strategy.

More shared components (timeframe helpers, models, rate limiters, etc.) can be added here as we evolve Willow and Charlie.