Solving Parameter Bloat in the Strategy Pattern

Prerequisites

To follow this guide, you should be comfortable with Python classes, abstract base classes, and Dependency Injection. Familiarity with Data Classes is also recommended.

Key Libraries & Tools

• ABC: Python's built-in module for defining Abstract Base Classes.
• Dataclasses: A decorator and functions for automatically adding generated special methods to user-defined classes.

Code Walkthrough

The Wrong Way: Keyword Arguments

One might try adding **kwargs to the strategy methods. This allows passing any parameter, but it destroys type safety and code clarity.

def should_buy(self, prices: list[float], **kwargs: float) -> bool:
    window_size = kwargs.get("window_size", 3)
    # Logic here...

The Messy Way: The Monster Parameter Class

Creating a single StrategyParameters data class to hold every possible option creates tight coupling. Every strategy then depends on a giant object containing parameters it doesn't even use.

The Best Way: Strategy Initializers

The most robust solution leverages the power of objects to combine data with behavior. By passing parameters into the strategy's __init__ method, you keep the execution methods clean.

@dataclass
class MinMaxStrategy(TradingStrategy):
    min_price: float
    max_price: float

    def should_buy(self, prices: list[float]) -> bool:
        return prices[-1] < self.min_price

Syntax Notes

Using the @dataclass decorator significantly reduces boilerplate code by automatically generating the __init__ method. This keeps your strategy definitions concise while maintaining explicit parameter requirements.

Practical Examples

In a trading bot, a Bitcoin strategy might need much wider price thresholds than a stablecoin strategy. By using initializers, the main setup function configures these specific values once. The bot itself remains oblivious to these numbers, simply calling should_buy() on whatever strategy it was given.

Tips & Gotchas

Avoid the trap of "parameter leakage." If your bot's run method has to know that a strategy needs a window_size, you have broken the abstraction. Always set specific configuration at the point of instantiation where the concrete class is already known.