Data Classes

Overview The Strategy Pattern allows developers to swap algorithms at runtime without altering the client code. However, a common friction point arises when different strategies require unique parameters. If a `ZipCompression` strategy needs a `compression_level` while `Zlib` needs `chunk_size`, where do these settings live? Failing to handle this correctly leads to leaking implementation details into the high-level classes that use these strategies. 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. ```python 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. ```python @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.