Composer Tools

A small library of useful tools for Composer

📦 Installation

pip install composer-tools

🚀 Quick Example

from composer_tools.converters.short_codes import generate_symphony_code

# When SPY's RSI drops below 30, switch to GLD, otherwise stay in BIL
code = generate_symphony_code("RSI_14_SPY_LT_30 GLD,BIL")

This generates:

(defsymphony
 "RSI_14_SPY_LT_30"
 {:asset-class "EQUITIES", :rebalance-frequency :daily}
 (weight-equal
  [(if
      (< (rsi "SPY" {:window 14}) 30.0)
      [(asset "GLD")]
      [(asset "BIL")])])
)

⚙️ How It Works

The format is simple: CONDITION ASSET1,ASSET2

• CONDITION: When to switch assets
• ASSET1: What to buy when condition is true
• ASSET2: What to buy when condition is false

📊 Condition Types

Absolute Conditions

Compare a metric to a fixed value:

generate_symphony_code("RSI_14_SPY_LT_30 GLD,BIL")          # RSI below 30
generate_symphony_code("STD_20_XLU_GT_0.02 TLT,SPY")       # Volatility above 2%
generate_symphony_code("CUMRET_10_SPY_LT_0 CASH,SPY")      # Negative returns

Relative Conditions

Compare metrics between assets:

generate_symphony_code("MA_50_SPY_LT_MA_20_SPY CASH,SPY")        # Death cross
generate_symphony_code("CUMRET_30_XLK_GT_CUMRET_30_XLU XLK,XLU") # Sector rotation

Nested Conditions

Chain conditions with __ (all must be true):

generate_symphony_code("RSI_14_SPY_LT_30__STD_20_SPY_GT_0.02__CUMRET_5_SPY_LT_0 GLD,SPY")

📈 Supported Metrics

• RSI - Relative Strength Index
• STD - Standard deviation of returns
• CUMRET - Cumulative return
• MA - Moving average
• there are more from Composer which need to be added!

💡Real Examples

# Momentum strategy
generate_symphony_code("RSI_14_QQQ_GT_70 QQQ,TLT")

# Volatility protection  
generate_symphony_code("STD_20_SPY_GT_0.025 TLT,SPY")

# Trend following
generate_symphony_code("MA_20_SPY_GT_MA_50_SPY SPY,CASH")

# Nested statement
generate_symphony_code("RSI_14_SPY_LT_30__STD_20_VIX_GT_25__CUMRET_10_SPY_LT_-0.05 GLD,SPY")

🔧 API

generate_symphony_code(input_str: str) -> str
# Converts short notation to Composer code.

parse_condition(condition_str: str) -> dict
# Parses condition strings into structured data (useful for debugging).

parse_condition("RSI_14_SPY_LT_30")
# Returns: {'type': 'absolute', 'metric': 'RSI', 'window': '14', 'ticker': 'SPY', 'operator': 'LT', 'value': '30'}

🤝 Contributing

This project is new and contributors are welcome. Please submit a PR!

There's a lot of interesting work to do:

• More metrics and functions: The rest of the Composer functions
• Better conditions: OR logic, more complex nesting
• Performance: Optimization and caching
• Documentation: More examples and tutorials

Development Setup 🛠️

git clone https://github.com/your-repo/composer-tools
cd composer-tools
uv sync
uv run pytest

All contributions welcome, from typo fixes to major features. The codebase is clean and the tests are comprehensive, so it's easy to jump in.

📄 License

MIT License - use this code however you want.