mirror of
https://ghfast.top/https://github.com/aeroxw/easy_tdx_max.git
synced 2026-09-12 15:44:18 +08:00
GitHub
c49ba4c4b5
组合回测（一策略×多标的）此前只能看 4 个数字，本轮把单标的的整条 分析链路在组合端补齐（WebUI/REST 双端）： - portfolio_engine：合并净值+汇总成交喂 PerformanceAnalyzer，输出 完整 25 项指标（SQN/最大连胜连亏/Ulcer/VaR/CVaR 等）+ 组合层 trades（symbol 列）；修复假年化与回撤口径（负值+固定分母 → 逐点峰值，与单标的/多策略一致） - walkforward：新增 PortfolioWalkForwardEngine，按标的日期并集切窗、 每窗独立开仓、合成组合窗内净值，复用 WalkForwardResult 结构 - benchmark：新增 evaluate_portfolio 一条龙（组合回测+组合WF+ 跨标的多数口径适配性体检+综合评分+组合评级+等权买入持有基准对比）， 报告结构与单标的 evaluate_strategy 同构 - performance：FIFO 持仓天数配对支持 symbol 分组 - Web：新增 POST /backtest/portfolio/wf/run/async 与 /backtest/portfolio/evaluate/run/async；组合回测响应附带 grade（组合净值口径）与 score；新增 _normalize_bars_dt 修复 按标的取数路径的字符串日期/遗留 date 列崩溃（E2E 揭露） - 前端：组合页新增附加分析勾选区与组合绩效指标/WF/一条龙/成交明细 区块；buildPortfolioAiPrompt 组合版 Prompt；抽通用 AiInterpretModal（回测页迁移共用，行为不变）；TradeTable 支持 showSymbol；EvaluatePanel 支持 gradeOverride - 测试：后端 +17 例（pytest 1603 绿）、aiPrompt 组合版 2 例、 Playwright 组合页 E2E 2 例（9/9 绿）
321 lines
12 KiB
Python
321 lines
12 KiB
Python
"""多标的组合回测引擎。
|
||
|
||
支持同时回测多只股票，共享资金池，按策略信号分配资金。
|
||
每只标的独立产生信号，引擎统一管理仓位和资金。
|
||
"""
|
||
|
||
from __future__ import annotations
|
||
|
||
from dataclasses import dataclass, field
|
||
from typing import Any
|
||
|
||
import pandas as pd
|
||
|
||
from easy_tdx.backtest.engine import BacktestEngine
|
||
from easy_tdx.backtest.strategy import Strategy
|
||
from easy_tdx.backtest.types import BacktestResult
|
||
|
||
|
||
@dataclass
|
||
class StockData:
|
||
"""单只标的的数据和标识。
|
||
|
||
Attributes:
|
||
code: 股票代码（如 "000001"）
|
||
market: 市场（如 "SZ"）
|
||
df: K线 DataFrame
|
||
"""
|
||
|
||
code: str
|
||
market: str
|
||
df: pd.DataFrame
|
||
|
||
|
||
@dataclass
|
||
class PortfolioResult:
|
||
"""组合回测结果。
|
||
|
||
Attributes:
|
||
total_performance: 组合整体绩效指标——与单标的回测同口径的完整
|
||
25 项（夏普/回撤/胜率/盈亏比/SQN/最大连胜连亏等，由合并净值
|
||
曲线 + 汇总成交喂 :class:`PerformanceAnalyzer` 计算），另附
|
||
``total_stocks`` / ``total_cash`` 两个组合字段。
|
||
individual_results: 每只标的的独立回测结果
|
||
equity_allocation: 每只标的的资金分配比例
|
||
combined_equity: 组合整体净值曲线（按日期对齐各标的求和），
|
||
列: datetime/total/drawdown/drawdown_pct。各标的独立回测日期范围
|
||
可能不同，此处按日期并集 forward-fill 对齐后求和。
|
||
trades: 组合层汇总成交（各标的 concat + ``symbol`` 列标注来源标的），
|
||
供组合级绩效统计（逐标的 FIFO 配对持仓天数）与前端明细表使用。
|
||
"""
|
||
|
||
total_performance: dict[str, float]
|
||
individual_results: dict[str, BacktestResult]
|
||
equity_allocation: dict[str, float]
|
||
combined_equity: pd.DataFrame
|
||
trades: pd.DataFrame = field(default_factory=pd.DataFrame)
|
||
|
||
def to_dict(self) -> dict[str, Any]:
|
||
"""转为可序列化字典。"""
|
||
return {
|
||
"total_performance": self.total_performance,
|
||
"individual_results": {k: v.to_dict() for k, v in self.individual_results.items()},
|
||
"equity_allocation": self.equity_allocation,
|
||
"combined_equity": self.combined_equity.to_dict(orient="records"),
|
||
"trades": self.trades.to_dict(orient="records"),
|
||
}
|
||
|
||
|
||
class PortfolioBacktestEngine:
|
||
"""多标的组合回测引擎。
|
||
|
||
管理多只股票的共享资金池，独立运行策略，
|
||
按均等或自定义比例分配资金。
|
||
|
||
用法::
|
||
|
||
engine = PortfolioBacktestEngine(
|
||
strategy=MyStrategy,
|
||
stocks=[
|
||
StockData("000001", "SZ", df1),
|
||
StockData("600000", "SH", df2),
|
||
],
|
||
total_cash=200000,
|
||
)
|
||
result = engine.run()
|
||
print(result.total_performance)
|
||
"""
|
||
|
||
def __init__(
|
||
self,
|
||
strategy: Strategy | type[Strategy],
|
||
stocks: list[StockData],
|
||
total_cash: float = 200_000.0,
|
||
allocation: str = "equal",
|
||
commission: float = 0.0003,
|
||
min_commission: float = 5.0,
|
||
stamp_tax: float = 0.001,
|
||
slippage: float = 0.0,
|
||
execution: str = "next_open",
|
||
chanlun_level: str | None = None,
|
||
auto_fees: bool = False,
|
||
) -> None:
|
||
"""初始化组合回测引擎。
|
||
|
||
Args:
|
||
strategy: 策略类或已构造的策略实例。传实例时（如带参数的
|
||
ParametrizedStrategy），参数会被透传到每个标的的回测。
|
||
传类时（CLI 用法）用默认参数。
|
||
stocks: 标的列表（StockData）
|
||
total_cash: 总资金
|
||
allocation: 资金分配方式（目前仅 "equal" 均等分配）
|
||
commission: 佣金率
|
||
min_commission: 最低佣金
|
||
stamp_tax: 印花税
|
||
slippage: 滑点
|
||
execution: 执行模式
|
||
chanlun_level: 缠论级别（可选）
|
||
auto_fees: 为 True 时按各标的代码解析品种费率（ETF/债券免
|
||
印花税等），覆盖默认值；显式非默认费率仍优先。
|
||
|
||
.. versionadded:: 1.24
|
||
``auto_fees`` 品种感知费率（按 StockData.market+code 逐标的解析）。
|
||
"""
|
||
self._strategy = strategy
|
||
self._stocks = stocks
|
||
self._total_cash = total_cash
|
||
self._allocation = allocation
|
||
self._commission = commission
|
||
self._min_commission = min_commission
|
||
self._stamp_tax = stamp_tax
|
||
self._slippage = slippage
|
||
self._execution = execution
|
||
self._chanlun_level = chanlun_level
|
||
self._auto_fees = auto_fees
|
||
|
||
def _compute_allocations(self) -> dict[str, float]:
|
||
"""计算每只标的的资金分配。"""
|
||
n = len(self._stocks)
|
||
if n == 0:
|
||
return {}
|
||
|
||
if self._allocation == "equal":
|
||
per_stock_cash = self._total_cash / n
|
||
return {f"{s.market}{s.code}": per_stock_cash for s in self._stocks}
|
||
|
||
# 默认均等分配
|
||
per_stock_cash = self._total_cash / n
|
||
return {f"{s.market}{s.code}": per_stock_cash for s in self._stocks}
|
||
|
||
def run(self) -> PortfolioResult:
|
||
"""运行组合回测。
|
||
|
||
对每只标的独立运行回测，按分配的资金量计算收益，
|
||
最终汇总为组合整体绩效。
|
||
|
||
Returns:
|
||
PortfolioResult 包含整体绩效和各标的详细结果
|
||
"""
|
||
allocations = self._compute_allocations()
|
||
individual_results: dict[str, BacktestResult] = {}
|
||
|
||
for stock in self._stocks:
|
||
key = f"{stock.market}{stock.code}"
|
||
cash = allocations.get(key, 0)
|
||
|
||
engine = BacktestEngine(
|
||
strategy=self._strategy,
|
||
cash=cash,
|
||
commission=self._commission,
|
||
min_commission=self._min_commission,
|
||
stamp_tax=self._stamp_tax,
|
||
slippage=self._slippage,
|
||
execution=self._execution,
|
||
chanlun_level=self._chanlun_level,
|
||
symbol=key,
|
||
auto_fees=self._auto_fees,
|
||
)
|
||
result = engine.run(stock.df)
|
||
individual_results[key] = result
|
||
|
||
# 组合整体净值曲线（各标的按日期对齐求和）——绩效指标依赖它，先算
|
||
combined_equity = self._build_combined_equity(individual_results, allocations)
|
||
|
||
# 汇总整体绩效（合并净值 + 汇总成交 → PerformanceAnalyzer 完整指标）
|
||
all_trades = self._merge_trades(individual_results)
|
||
total_perf = self._aggregate_performance(
|
||
individual_results, allocations, combined_equity, all_trades
|
||
)
|
||
|
||
# 计算资金占比
|
||
total_alloc = sum(allocations.values())
|
||
equity_pct = {k: v / total_alloc if total_alloc > 0 else 0 for k, v in allocations.items()}
|
||
|
||
return PortfolioResult(
|
||
total_performance=total_perf,
|
||
individual_results=individual_results,
|
||
equity_allocation=equity_pct,
|
||
combined_equity=combined_equity,
|
||
trades=all_trades,
|
||
)
|
||
|
||
@staticmethod
|
||
def _merge_trades(results: dict[str, BacktestResult]) -> pd.DataFrame:
|
||
"""把各标的成交 concat 成组合层成交表，附 ``symbol`` 列标注来源标的。
|
||
|
||
``symbol`` 列让 PerformanceAnalyzer 的 FIFO 持仓天数配对按标的分组
|
||
（避免 A 股的买入被 B 股的卖出错误配对）；无成交时返回空表。
|
||
"""
|
||
frames: list[pd.DataFrame] = []
|
||
for key, result in results.items():
|
||
if len(result.trades) > 0:
|
||
t = result.trades.copy()
|
||
t["symbol"] = key
|
||
frames.append(t)
|
||
if not frames:
|
||
return pd.DataFrame(columns=["symbol", "direction", "pnl", "rejected"])
|
||
return pd.concat(frames, ignore_index=True)
|
||
|
||
def _aggregate_performance(
|
||
self,
|
||
results: dict[str, BacktestResult],
|
||
allocations: dict[str, float],
|
||
combined_equity: pd.DataFrame,
|
||
all_trades: pd.DataFrame,
|
||
) -> dict[str, float]:
|
||
"""汇总所有标的的绩效为组合整体绩效。
|
||
|
||
与多策略引擎 ``MultiStrategyEngine._aggregate_performance`` 同口径：
|
||
合并净值曲线 + 汇总成交喂 :class:`PerformanceAnalyzer`，得到
|
||
与单标的回测一致的完整指标（夏普/回撤/胜率/盈亏比/SQN/最大连胜连亏
|
||
等 25 项），便于前端复用 MetricTable 展示。合并曲线的首个值即总投入
|
||
资金，因此 ``total_return`` 天然等于资金加权收益率。
|
||
|
||
Args:
|
||
results: 各标的回测结果
|
||
allocations: 各标的资金分配
|
||
combined_equity: 组合整体净值曲线（_build_combined_equity 产物）
|
||
all_trades: 组合层汇总成交（_merge_trades 产物，含 symbol 列）
|
||
|
||
Returns:
|
||
组合整体绩效指标（另附 total_stocks / total_cash 组合字段）
|
||
"""
|
||
from easy_tdx.backtest.performance import PerformanceAnalyzer
|
||
|
||
total_cash = sum(allocations.values())
|
||
if not results or len(combined_equity) < 2:
|
||
return {
|
||
"total_return": 0.0,
|
||
"annual_return": 0.0,
|
||
"total_stocks": float(len(results)),
|
||
"total_cash": total_cash,
|
||
}
|
||
|
||
analyzer = PerformanceAnalyzer(equity_curve=combined_equity, trades=all_trades)
|
||
metrics = analyzer.compute()
|
||
metrics["total_stocks"] = float(len(results))
|
||
metrics["total_cash"] = total_cash
|
||
return metrics
|
||
|
||
def _build_combined_equity(
|
||
self,
|
||
results: dict[str, BacktestResult],
|
||
allocations: dict[str, float],
|
||
) -> pd.DataFrame:
|
||
"""把各标的独立净值曲线按日期对齐求和，生成组合整体净值曲线。
|
||
|
||
各标的独立回测的日期范围可能不同（取数差异、停牌等），这里取所有标的
|
||
datetime 的并集，每个标的的 total 列 forward-fill 对齐到并集后求和。
|
||
|
||
Returns:
|
||
DataFrame: datetime / total / drawdown / drawdown_pct。
|
||
空结果时返回带表头的空 DataFrame。
|
||
"""
|
||
empty = pd.DataFrame(columns=["datetime", "total", "drawdown", "drawdown_pct"])
|
||
if not results:
|
||
return empty
|
||
|
||
# 收集各标的的 (datetime, total) 系列，以 datetime 为索引
|
||
series_list: list[pd.Series] = []
|
||
for key, result in results.items():
|
||
ec = result.equity_curve
|
||
if len(ec) == 0:
|
||
continue
|
||
# datetime 列可能是 int(YYYYMMDD) 或 datetime；统一转可比字符串/时间戳
|
||
dt = ec["datetime"]
|
||
if dt.dtype.kind in "iu": # int YYYYMMDD
|
||
dt = pd.to_datetime(dt.astype(str), format="%Y%m%d")
|
||
elif dt.dtype != "datetime64[ns]":
|
||
dt = pd.to_datetime(dt)
|
||
s = pd.Series(ec["total"].to_numpy(), index=dt, name=key)
|
||
series_list.append(s)
|
||
|
||
if not series_list:
|
||
return empty
|
||
|
||
# 外连接对齐（并集日期），forward-fill 各标的在缺失日期的净值（持有不动），
|
||
# 再求和得组合总净值。缺失值填 0 是为应对某标的完全无该日期数据的情况。
|
||
aligned = pd.concat(series_list, axis=1).sort_index()
|
||
aligned = aligned.ffill().fillna(0)
|
||
total = aligned.sum(axis=1)
|
||
|
||
# 回撤：drawdown 为绝对回撤额（峰值-当前，正值），drawdown_pct 为相对
|
||
# 当时峰值的回撤比例（drawdown / peak，0~1）。分母必须用逐点 peak 而非
|
||
# 固定初始值：净值大涨后 peak 是初始值的好几倍，若除以 initial 会把回撤
|
||
# 百分比严重放大。与单标的 PortfolioTracker.equity_curve、
|
||
# MultiStrategyEngine._build_combined_equity 的定义保持一致，
|
||
# PerformanceAnalyzer 直接读 drawdown_pct 列算 max_drawdown。
|
||
peak = total.cummax()
|
||
drawdown = peak - total
|
||
peak_safe = peak.where(peak != 0, 1.0)
|
||
drawdown_pct = drawdown / peak_safe
|
||
|
||
return pd.DataFrame(
|
||
{
|
||
"datetime": total.index,
|
||
"total": total.to_numpy(),
|
||
"drawdown": drawdown.to_numpy(),
|
||
"drawdown_pct": drawdown_pct.to_numpy(),
|
||
}
|
||
).reset_index(drop=True)
|