# src/backtest_engine.py from datetime import datetime from typing import Type, Dict, Any, List, Optional import numpy as np import pandas as pd from src.indicators.base_indicators import Indicator # 导入所有需要协调的模块 from .core_data import Bar, Order, Trade, PortfolioSnapshot from .data_manager import DataManager from .execution_simulator import ExecutionSimulator from .backtest_context import BacktestContext from .strategies.base_strategy import Strategy class BacktestEngine: """ 回测引擎：协调数据流、策略执行、订单模拟和结果记录。 """ def __init__(self, data_manager: DataManager, strategy_class: Type[Strategy], strategy_params: Dict[str, Any], # current_segment_symbol: str, # 这个参数不再需要，因为 symbol 会动态更新 initial_capital: float = 100000.0, slippage_rate: float = 0.0001, commission_rate: float = 0.0002, roll_over_mode: bool = False, start_time: Optional[datetime] = None, # 新增开始时间 end_time: Optional[datetime] = None, # 新增结束时间 indicators: List[Indicator] = [], ): # 新增换月模式参数 """ 初始化回测引擎。 Args: data_manager (DataManager): 已经初始化好的数据管理器实例。 strategy_class (Type[Strategy]): 策略类（而不是实例），引擎会负责实例化。 strategy_params (Dict[str, Any]): 传递给策略的参数字典。 initial_capital (float): 初始交易资金。 slippage_rate (float): 交易滑点率。 commission_rate (float): 交易佣金率。 roll_over_mode (bool): 是否启用主连合约换月模式。 """ self.data_manager = data_manager self.initial_capital = initial_capital self.simulator = ExecutionSimulator( initial_capital=initial_capital, slippage_rate=slippage_rate, commission_rate=commission_rate ) # 传入引擎自身给 context，以便 context 可以获取引擎的状态（如 is_rollover_bar） self.context = BacktestContext(self.data_manager, self.simulator) self.context.set_engine(self) # 建立 Context 到 Engine 的引用 # self.current_segment_symbol = current_segment_symbol # 此行移除或作为内部变量动态管理 # 实例化策略。初始 symbol 会在 run_backtest 中根据第一根 Bar 动态设置。 self.strategy = strategy_class(self.context, symbol="INITIAL_PLACEHOLDER_SYMBOL", **strategy_params) self.indicators = indicators self.portfolio_snapshots: List[PortfolioSnapshot] = [] self.trade_history: List[Trade] = [] self.all_bars: List[Bar] = [] self.close_list: List[float] = [] self.open_list: List[float] = [] self.high_list: List[float] = [] self.low_list: List[float] = [] self.volume_list: List[float] = [] self._history_bars: List[Bar] = [] # 引擎层面保留的历史 Bar，通常供策略在 on_bar 中使用 self._max_history_bars: int = strategy_params.get('history_bars_limit', 200) # 换月相关状态 self.roll_over_mode = roll_over_mode # 是否启用换月模式 self._last_processed_bar_symbol: Optional[str] = None # 记录上一根 K 线的 symbol self.is_rollover_bar: bool = False # 标记当前 K 线是否为换月 K 线（禁止开仓） # 新增时间过滤属性 self.start_time = start_time self.end_time = end_time print("\n--- 回测引擎初始化完成 ---") print(f" 策略: {strategy_class.__name__}") print(f" 初始资金: {initial_capital:.2f}") print(f" 换月模式: {'启用' if roll_over_mode else '禁用'}") def run_backtest(self): """ 运行整个回测流程，包含换月逻辑。 """ print("\n--- 回测开始 ---") # 调用策略的初始化方法 self.strategy.on_init() self.strategy.trading = True last_processed_bar: Optional[Bar] = None # 用于在换月时引用旧合约的最后一根 K 线 # 主回测循环 while True: current_bar = self.data_manager.get_next_bar() if current_bar is None: break # 没有更多数据，回测结束 if self.start_time and current_bar.datetime < self.start_time: continue # 如果设置了结束时间，且当前K线在结束时间之后，则终止回测 if self.end_time and current_bar.datetime >= self.end_time: print(f"到达结束时间 {self.end_time}，回测终止。") break # --- 换月逻辑判断和处理 (在处理 current_bar 之前进行) --- # 1. 重置 is_rollover_bar 标记 self.is_rollover_bar = False # 4. 更新 Context 和 Simulator 的当前 Bar 和时间 self.context.set_current_bar(current_bar) self.simulator.update_time(current_time=current_bar.datetime) # 2. 如果启用换月模式，并且检测到合约 symbol 变化 if self.roll_over_mode and \ self._last_processed_bar_symbol is not None and \ current_bar.symbol != self._last_processed_bar_symbol: old_symbol = self._last_processed_bar_symbol new_symbol = current_bar.symbol # 确认 last_processed_bar 确实是旧合约的最后一根 K 线 if last_processed_bar and last_processed_bar.symbol == old_symbol: self.strategy.log(f"检测到换月！从 [{old_symbol}] 切换到 [{new_symbol}]。" f"在旧合约最后一根K线 ({last_processed_bar.datetime}) 执行强制平仓和取消操作。") # A. 强制平仓旧合约的所有持仓 self.simulator.force_close_all_positions_for_symbol(old_symbol, last_processed_bar) # B. 取消旧合约的所有挂单 self.simulator.cancel_all_pending_orders_for_symbol(old_symbol) # C. 标记【当前这根 Bar (即新合约的第一根 K 线)】为换月 K 线 # 此时 self.is_rollover_bar 变为 True，将通过 Context 传递给策略， # 策略在该 K 线周期内不能开仓。 self.is_rollover_bar = True # D. 通知策略换月事件，让策略有机会重置内部状态 self.strategy.on_rollover(old_symbol, new_symbol) else: self.strategy.log(f"警告: 检测到换月从 {old_symbol} 到 {new_symbol}，但 last_processed_bar 为空或与旧合约不符。" "强制平仓/取消操作可能未正确执行。") # 3. 更新策略关注的当前合约 symbol self.strategy.symbol = current_bar.symbol self.strategy.on_open_bar(current_bar.open, current_bar.symbol) current_indicator_dict = {} close_array = np.array(self.close_list) open_array = np.array(self.open_list) high_array = np.array(self.high_list) low_array = np.array(self.low_list) volume_array = np.array(self.volume_list) for indicator in self.indicators: current_indicator_dict[indicator.get_name()] = indicator.get_latest_value( close_array, open_array, high_array, low_array, volume_array ) self.simulator.process_pending_orders(current_bar, current_indicator_dict) self.all_bars.append(current_bar) self.close_list.append(current_bar.close) self.open_list.append(current_bar.open) self.high_list.append(current_bar.high) self.low_list.append(current_bar.low) self.volume_list.append(current_bar.volume) # 7. 调用策略的 on_bar 方法 # self.strategy.on_bar(current_bar) self.strategy.on_close_bar(current_bar) self.simulator.process_pending_orders(current_bar, current_indicator_dict) # 8. 记录投资组合快照 current_portfolio_value = self.simulator.get_portfolio_value(current_bar) current_positions = self.simulator.get_current_positions() price_at_snapshot = {current_bar.symbol: current_bar.close} # 使用当前 Bar 的收盘价记录快照 snapshot = PortfolioSnapshot( datetime=current_bar.datetime, total_value=current_portfolio_value, cash=self.simulator.cash, positions=current_positions, price_at_snapshot=price_at_snapshot ) self.portfolio_snapshots.append(snapshot) # 9. 更新 `_last_processed_bar_symbol` 和 `last_processed_bar` 为当前 Bar，为下一轮循环做准备 self._last_processed_bar_symbol = current_bar.symbol last_processed_bar = current_bar # --- 回测结束后的清理工作 --- print("\n--- 回测结束，检查并平仓所有剩余持仓 ---") if last_processed_bar: # 确保至少有一根 Bar 被处理过 # 在回测结束时，强制平仓所有可能存在的剩余持仓 # 遍历所有持仓，确保全部清算 remaining_positions_symbols = list(self.simulator.get_current_positions().keys()) for symbol_held in remaining_positions_symbols: if self.simulator.get_current_positions().get(symbol_held, 0) != 0: self.strategy.log(f"回测结束清理: 强制平仓合约 {symbol_held} 的剩余持仓。") # 使用 simulator 的 force_close_all_positions_for_symbol 方法进行清理 self.simulator.force_close_all_positions_for_symbol(symbol_held, last_processed_bar) self.simulator.cancel_all_pending_orders_for_symbol(symbol_held) else: print("没有处理任何 Bar，无需平仓。") # 回测结束后，获取所有交易记录 self.trade_history = self.simulator.get_trade_history() print("--- 回测结束 ---") print(f"总计处理了 {len(self.all_bars)} 根K线。") print(f"总计发生了 {len(self.trade_history)} 笔交易。") final_portfolio_value = 0.0 if last_processed_bar: final_portfolio_value = self.simulator.get_portfolio_value(last_processed_bar) else: final_portfolio_value = self.initial_capital total_return_percentage = ((final_portfolio_value - self.initial_capital) / self.initial_capital) * 100 print(f"最终总净值: {final_portfolio_value:.2f}") print(f"总收益率: {total_return_percentage:.2f}%") def get_backtest_results(self) -> Dict[str, Any]: """ 返回回测结果数据，供结果分析模块使用。 """ return { "portfolio_snapshots": self.portfolio_snapshots, "trade_history": self.trade_history, "initial_capital": self.simulator.initial_capital, "all_bars": self.all_bars } def get_simulator(self) -> ExecutionSimulator: return self.simulator def get_bar_history(self): return self.all_bars def get_price_history(self, key: str): if key == 'close': return self.close_list elif key == 'open': return self.open_list elif key == 'high': return self.high_list elif key == 'low': return self.low_list elif key == 'volume': return self.volume_list return None