mirror of
https://ghfast.top/https://github.com/aeroxw/easy_tdx_max.git
synced 2026-09-12 16:54:20 +08:00
fix: 策略库「重跑到今天」补齐组合分析 — 多策略组合级WF/一条龙/AI解读
策略库（/strategies）多策略组合卡片此前只有主回测：本轮把 v1.31.0 的组合分析链路延伸到多策略组合（N 策略 × 各自原标的）： - walkforward：组合 WF 泛化为槽位模型（_ComboSlot/_ComboWalkForwardBase）， PortfolioWalkForwardEngine 行为不变；新增 MultiStrategyWalkForwardEngine （N 个策略各跑各自原标的，key 形如 label@symbol），复用切窗语义与 WalkForwardResult 结构（前端 WalkForwardPanel 直接渲染） - benchmark：新增 evaluate_multi 一条龙（MultiStrategyEngine 回测 + 多策略 组合 WF + 逐槽位三段体检多数口径聚合 + 综合评分 + 组合评级 + 各槽位标的 等权买入持有基准对比），报告结构与单标的 evaluate_strategy 同构 - Web：新增 POST /backtest/multi-strategy/wf/run/async 与 /backtest/multi-strategy/evaluate/run/async；多策略组合回测响应附带 grade/score（与单标的/多标的组合响应同构） - 前端 StrategiesView：组合结果区新增「WF 样本外验证 / 一条龙评估 / AI 解读」 按钮与同构面板（按需触发，复用最近一次组合回测的 items/cash）； 绩效指标表补齐 v1.28 深度 6 项（SQN/最大连胜连亏/Ulcer/VaR/CVaR）； aiPrompt 新增 multi 模式（策略明细语境 + 槽位表现段） - 测试：多策略 WF 引擎 3 例、evaluate_multi 3 例、新端点 Web 级 2 例、 aiPrompt multi 模式 1 例（pytest 1611 绿、node --test 5/5、E2E 9/9）
This commit is contained in:
GitHub
@@ -40,7 +40,11 @@ from easy_tdx.backtest.grading import grade_performance, grade_portfolio_equity
|
||||
from easy_tdx.backtest.scoring import score_strategy
|
||||
from easy_tdx.backtest.strategy import Strategy
|
||||
from easy_tdx.backtest.types import to_json_native
|
||||
from easy_tdx.backtest.walkforward import PortfolioWalkForwardEngine, WalkForwardEngine
|
||||
from easy_tdx.backtest.walkforward import (
|
||||
MultiStrategyWalkForwardEngine,
|
||||
PortfolioWalkForwardEngine,
|
||||
WalkForwardEngine,
|
||||
)
|
||||
|
||||
if TYPE_CHECKING:
|
||||
import numpy.typing as npt
|
||||
@@ -54,6 +58,7 @@ else:
|
||||
__all__ = [
|
||||
"evaluate_strategy",
|
||||
"evaluate_portfolio",
|
||||
"evaluate_multi",
|
||||
"run_buy_hold_benchmark",
|
||||
"compute_benchmark_comparison",
|
||||
]
|
||||
@@ -453,3 +458,119 @@ def _aggregate_fitness(
|
||||
)
|
||||
aggregated.high_fitness = aggregated.pass_ratio >= 0.75
|
||||
return aggregated
|
||||
|
||||
|
||||
def evaluate_multi(
|
||||
strategies: list[Any],
|
||||
total_cash: float = 1_000_000.0,
|
||||
commission: float = 0.0003,
|
||||
min_commission: float = 5.0,
|
||||
stamp_tax: float = 0.001,
|
||||
slippage: float = 0.0,
|
||||
execution: str = "next_open",
|
||||
n_windows: int = 7,
|
||||
warmup_ratio: float = 0.3,
|
||||
context_bars: int = 60,
|
||||
split: tuple[float, float, float] = (0.6, 0.2, 0.2),
|
||||
) -> dict[str, Any]:
|
||||
"""多策略组合一条龙评估：组合回测 + 组合 WF + 跨槽位适配性体检 + 综合评分
|
||||
+ 组合评级 + 等权买入持有基准对比。
|
||||
|
||||
与 :func:`evaluate_portfolio`（一个策略 × 多标的）同构，报告结构一致；
|
||||
差异点仅在槽位划分——每个槽位是「一个策略 × 它自己的标的」
|
||||
（:class:`~easy_tdx.backtest.multi_strategy_engine.StrategySlot`）：
|
||||
|
||||
- 全样本回测走 :class:`~easy_tdx.backtest.multi_strategy_engine.MultiStrategyEngine`；
|
||||
- Walk-Forward 走
|
||||
:class:`~easy_tdx.backtest.walkforward.MultiStrategyWalkForwardEngine`；
|
||||
- 适配性体检逐槽位（各自策略 × 各自标的）跑三段后按多数口径聚合；
|
||||
- 买入持有基准 = 各槽位标的的等权买入持有组合（策略换成 _BuyAndHold，
|
||||
其余不变），α/β/信息比率/跟踪误差基于两条组合净值曲线。
|
||||
|
||||
Args:
|
||||
strategies: StrategySlot 列表（每个槽位已绑定策略实例与 K 线）。
|
||||
其余参数: 透传给组合回测 / 组合 WF / 适配性（同口径费率与执行）。
|
||||
|
||||
Returns:
|
||||
完整评估报告字典（结构同 evaluate_strategy，config 记录槽位清单）。
|
||||
"""
|
||||
from easy_tdx.backtest.multi_strategy_engine import MultiStrategyEngine, StrategySlot
|
||||
|
||||
engine_kwargs: dict[str, Any] = {
|
||||
"total_cash": total_cash,
|
||||
"commission": commission,
|
||||
"min_commission": min_commission,
|
||||
"stamp_tax": stamp_tax,
|
||||
"slippage": slippage,
|
||||
"execution": execution,
|
||||
}
|
||||
|
||||
# 1. 全样本组合回测（完整 25 项指标 + 合并净值曲线）
|
||||
bt = MultiStrategyEngine(strategies=list(strategies), **engine_kwargs).run()
|
||||
perf = bt.total_performance
|
||||
|
||||
# 2. 组合 Walk-Forward 样本外
|
||||
wf = MultiStrategyWalkForwardEngine(
|
||||
strategies=list(strategies),
|
||||
n_windows=n_windows,
|
||||
warmup_ratio=warmup_ratio,
|
||||
context_bars=context_bars,
|
||||
**engine_kwargs,
|
||||
).run()
|
||||
|
||||
# 3. 适配性体检：逐槽位（各自策略 × 各自标的）跑三段体检，多数口径聚合
|
||||
per_cash = total_cash / max(len(strategies), 1)
|
||||
fitness_kwargs: dict[str, Any] = {
|
||||
"commission": commission,
|
||||
"min_commission": min_commission,
|
||||
"stamp_tax": stamp_tax,
|
||||
"slippage": slippage,
|
||||
"execution": execution,
|
||||
}
|
||||
per_slot_fitness = [
|
||||
FitnessEngine(
|
||||
strategy=slot.strategy,
|
||||
split=split,
|
||||
context_bars=context_bars,
|
||||
cash=per_cash,
|
||||
**fitness_kwargs,
|
||||
).evaluate(slot.df)
|
||||
for slot in strategies
|
||||
]
|
||||
fitness = _aggregate_fitness(per_slot_fitness, split)
|
||||
|
||||
# 4. 综合评分（叠加组合 WF 一致性）+ 组合评级（净值曲线口径）
|
||||
score = score_strategy(dict(perf), wf=wf)
|
||||
grade = grade_portfolio_equity(bt.combined_equity.to_dict(orient="records"))
|
||||
|
||||
# 5. 基准对比：各槽位标的的等权买入持有组合（策略换成 _BuyAndHold，其余不变）
|
||||
bh_slots = [
|
||||
StrategySlot(label=s.label, symbol=s.symbol, strategy=_BuyAndHold(), df=s.df)
|
||||
for s in strategies
|
||||
]
|
||||
bh_bt = MultiStrategyEngine(strategies=bh_slots, **engine_kwargs).run()
|
||||
bh_keys = ("total_return", "annual_return", "max_drawdown", "sharpe", "calmar", "volatility")
|
||||
bh = dict(to_json_native({k: bh_bt.total_performance.get(k, 0.0) for k in bh_keys}))
|
||||
comparison = compute_benchmark_comparison(bt.combined_equity, bh_bt.combined_equity)
|
||||
|
||||
return {
|
||||
"performance": to_json_native(dict(perf)),
|
||||
"score": score.to_dict(),
|
||||
"grade": grade.to_dict(),
|
||||
"walkforward": wf.to_dict(),
|
||||
"fitness": fitness.to_dict(),
|
||||
"benchmark": {
|
||||
"buy_hold": bh,
|
||||
"excess_return": float(perf.get("total_return", 0.0))
|
||||
- float(bh.get("total_return", 0.0)),
|
||||
**comparison,
|
||||
},
|
||||
"config": {
|
||||
"slots": [f"{s.label}@{s.symbol}" for s in strategies],
|
||||
"total_cash": total_cash,
|
||||
"execution": execution,
|
||||
"n_windows": n_windows,
|
||||
"warmup_ratio": warmup_ratio,
|
||||
"split": list(split),
|
||||
},
|
||||
}
|
||||
|
||||
@@ -45,6 +45,7 @@ __all__ = [
|
||||
"WalkForwardResult",
|
||||
"WalkForwardEngine",
|
||||
"PortfolioWalkForwardEngine",
|
||||
"MultiStrategyWalkForwardEngine",
|
||||
]
|
||||
|
||||
|
||||
@@ -273,7 +274,204 @@ class WalkForwardEngine:
|
||||
result.total_trades = int(sum(w.total_trades for w in ws))
|
||||
|
||||
|
||||
class PortfolioWalkForwardEngine:
|
||||
class _ComboSlot:
|
||||
"""组合 WF 的一个回测槽位（内部结构，由各公开引擎组装）。
|
||||
|
||||
Attributes:
|
||||
key: 槽位标识（组合成交表的 symbol 列值）。
|
||||
strategy: 策略类或实例。
|
||||
df: 该槽位的 K 线。
|
||||
cash: 等权分配到的资金。
|
||||
symbol: 品种感知费率标识（auto_fees 用；不感知则 None）。
|
||||
auto_fees: 是否按品种解析费率。
|
||||
"""
|
||||
|
||||
__slots__ = ("key", "strategy", "df", "cash", "symbol", "auto_fees")
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
key: str,
|
||||
strategy: type[Strategy] | Strategy,
|
||||
df: pd.DataFrame,
|
||||
cash: float,
|
||||
symbol: str | None = None,
|
||||
auto_fees: bool = False,
|
||||
) -> None:
|
||||
self.key = key
|
||||
self.strategy = strategy
|
||||
self.df = df
|
||||
self.cash = cash
|
||||
self.symbol = symbol
|
||||
self.auto_fees = auto_fees
|
||||
|
||||
|
||||
class _ComboWalkForwardBase:
|
||||
"""组合级 WF 共用实现：按参考时间轴切窗，逐槽位独立回测后合成组合净值。
|
||||
|
||||
切窗语义与 :class:`WalkForwardEngine`（单标的）一致，差异仅在「回测单元」
|
||||
从单只标的换成 N 个槽位（标的或策略×标的）。
|
||||
"""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
slots: list[_ComboSlot],
|
||||
n_windows: int,
|
||||
warmup_ratio: float,
|
||||
context_bars: int,
|
||||
engine_kwargs: dict[str, Any],
|
||||
) -> None:
|
||||
self._slots = list(slots)
|
||||
self._n_windows = max(int(n_windows), 2)
|
||||
self._warmup_ratio = min(max(float(warmup_ratio), 0.0), 0.8)
|
||||
self._context_bars = max(int(context_bars), 0)
|
||||
self._engine_kwargs = dict(engine_kwargs)
|
||||
|
||||
def run(self) -> WalkForwardResult:
|
||||
"""执行组合 Walk-Forward 验证。
|
||||
|
||||
Returns:
|
||||
:class:`WalkForwardResult`。数据不足以切窗时返回空结果
|
||||
（``windows`` 为空，聚合指标为 0）。
|
||||
"""
|
||||
result = WalkForwardResult(n_windows=self._n_windows, warmup_ratio=self._warmup_ratio)
|
||||
if not self._slots:
|
||||
return result
|
||||
|
||||
# 参考时间轴：全部标的 datetime 的并集（升序）
|
||||
timeline = self._reference_timeline()
|
||||
n = len(timeline)
|
||||
# 最少数据：每窗 ≥ 20 根 + 预热区 ≥ 20 根
|
||||
min_bars = 20 * (1 + self._n_windows)
|
||||
if n < min_bars:
|
||||
return result
|
||||
|
||||
eval_start = int(n * self._warmup_ratio)
|
||||
eval_len = n - eval_start
|
||||
window_len = eval_len // self._n_windows
|
||||
|
||||
for i in range(self._n_windows):
|
||||
s = eval_start + i * window_len
|
||||
e = s + window_len if i < self._n_windows - 1 else n # 末窗吃到尾部
|
||||
if e - s < 5:
|
||||
continue
|
||||
win = self._run_window(timeline, s, e, i)
|
||||
if win is not None:
|
||||
result.windows.append(win)
|
||||
|
||||
WalkForwardEngine._aggregate(result)
|
||||
return result
|
||||
|
||||
def _reference_timeline(self) -> pd.DatetimeIndex:
|
||||
"""全部槽位 datetime 的并集（升序，Timestamp 化）。"""
|
||||
all_dt: list[pd.Timestamp] = []
|
||||
for slot in self._slots:
|
||||
s = self._dt_series(slot.df)
|
||||
if len(s) > 0:
|
||||
all_dt.append(s)
|
||||
if not all_dt:
|
||||
return pd.DatetimeIndex([])
|
||||
return pd.DatetimeIndex(sorted(pd.unique(pd.concat(all_dt))))
|
||||
|
||||
@staticmethod
|
||||
def _dt_series(df: pd.DataFrame) -> pd.Series:
|
||||
"""标的 K 线的 datetime 列统一转 Timestamp（int YYYYMMDD 兼容）。"""
|
||||
col = "datetime" if "datetime" in df.columns else "date"
|
||||
dt = df[col]
|
||||
if dt.dtype.kind in "iu":
|
||||
return pd.to_datetime(dt.astype(str), format="%Y%m%d")
|
||||
if not pd.api.types.is_datetime64_any_dtype(dt):
|
||||
return pd.to_datetime(dt)
|
||||
return pd.Series(pd.to_datetime(dt), index=df.index)
|
||||
|
||||
def _run_window(
|
||||
self, timeline: pd.DatetimeIndex, s: int, e: int, index: int
|
||||
) -> WalkForwardWindow | None:
|
||||
"""独立回测单个窗口 [s, e)（参考时间轴下标），合成组合窗内净值。"""
|
||||
window_start = timeline[s]
|
||||
window_end = timeline[e - 1]
|
||||
ctx_start = timeline[max(0, s - self._context_bars)]
|
||||
|
||||
equity_series: list[pd.Series] = []
|
||||
trade_frames: list[pd.DataFrame] = []
|
||||
for slot in self._slots:
|
||||
dt = self._dt_series(slot.df)
|
||||
mask = (dt >= ctx_start) & (dt <= window_end)
|
||||
sub = slot.df.loc[mask].reset_index(drop=True)
|
||||
dt_sub = dt.loc[mask].reset_index(drop=True)
|
||||
# 上下文 bar 数 = 窗口起点之前保留的 bar 数（warmup 压制其信号）
|
||||
lead = int((dt_sub < window_start).sum())
|
||||
if len(sub) < lead + 5:
|
||||
continue # 该槽位数据不足（晚上市/停牌过多），本窗跳过
|
||||
|
||||
engine = BacktestEngine(
|
||||
strategy=slot.strategy,
|
||||
cash=slot.cash,
|
||||
warmup_bars=lead,
|
||||
symbol=slot.symbol,
|
||||
auto_fees=slot.auto_fees,
|
||||
**self._engine_kwargs,
|
||||
)
|
||||
try:
|
||||
bt = engine.run(sub)
|
||||
except Exception: # noqa: BLE001 — 单槽位失败不拖垮整窗
|
||||
continue
|
||||
|
||||
# 只取窗内净值点（上下文区恒为现金，不参与窗指标，避免稀释波动率）
|
||||
ec = bt.equity_curve
|
||||
if len(ec) > lead:
|
||||
eq = ec.iloc[lead:]
|
||||
equity_series.append(
|
||||
pd.Series(eq["total"].to_numpy(), index=self._dt_series(eq), name=slot.key)
|
||||
)
|
||||
if len(bt.trades) > 0:
|
||||
t = bt.trades.copy()
|
||||
t["symbol"] = slot.key
|
||||
trade_frames.append(t)
|
||||
|
||||
if not equity_series:
|
||||
return None # 所有槽位都跑不了，跳过该窗
|
||||
|
||||
# 合成组合窗内净值：日期并集对齐，ffill 持有不动，上市晚于窗口起点的
|
||||
# 标的其前导缺口用首值回填（首值即其初始资金——还没开仓，持有现金）
|
||||
aligned = pd.concat(equity_series, axis=1).sort_index()
|
||||
aligned = aligned.ffill().bfill()
|
||||
total = aligned.sum(axis=1)
|
||||
peak = total.cummax()
|
||||
drawdown = peak - total
|
||||
peak_safe = peak.where(peak != 0, 1.0)
|
||||
window_equity = pd.DataFrame(
|
||||
{
|
||||
"datetime": total.index,
|
||||
"total": total.to_numpy(),
|
||||
"drawdown": drawdown.to_numpy(),
|
||||
"drawdown_pct": (drawdown / peak_safe).to_numpy(),
|
||||
}
|
||||
)
|
||||
|
||||
all_trades = (
|
||||
pd.concat(trade_frames, ignore_index=True)
|
||||
if trade_frames
|
||||
else pd.DataFrame(columns=["symbol", "direction", "pnl", "rejected"])
|
||||
)
|
||||
from easy_tdx.backtest.performance import PerformanceAnalyzer
|
||||
|
||||
perf = PerformanceAnalyzer(equity_curve=window_equity, trades=all_trades).compute()
|
||||
|
||||
return WalkForwardWindow(
|
||||
index=index,
|
||||
start=window_start.strftime("%Y-%m-%d"),
|
||||
end=window_end.strftime("%Y-%m-%d"),
|
||||
bars=int(e - s),
|
||||
total_return=float(perf.get("total_return", 0.0)),
|
||||
sharpe=float(perf.get("sharpe", 0.0)),
|
||||
max_drawdown=float(perf.get("max_drawdown", 0.0)),
|
||||
total_trades=int(perf.get("total_trades", 0)),
|
||||
win_rate=float(perf.get("win_rate", 0.0)),
|
||||
performance={k: v for k, v in perf.items()},
|
||||
)
|
||||
|
||||
|
||||
class PortfolioWalkForwardEngine(_ComboWalkForwardBase):
|
||||
"""组合级 Walk-Forward：一个策略 × 多只标的，逐窗独立回测并合成组合净值。
|
||||
|
||||
与 :class:`WalkForwardEngine`（单标的）共用切窗语义与
|
||||
@@ -323,163 +521,94 @@ class PortfolioWalkForwardEngine:
|
||||
total_cash: 组合总资金（各标的等权分 1/N）。
|
||||
其余参数: 透传给各窗各标的的 :class:`BacktestEngine`。
|
||||
"""
|
||||
self._strategy = strategy
|
||||
self._stocks = list(stocks)
|
||||
self._n_windows = max(int(n_windows), 2)
|
||||
self._warmup_ratio = min(max(float(warmup_ratio), 0.0), 0.8)
|
||||
self._context_bars = max(int(context_bars), 0)
|
||||
self._total_cash = float(total_cash)
|
||||
self._engine_kwargs: dict[str, Any] = {
|
||||
"commission": commission,
|
||||
"min_commission": min_commission,
|
||||
"stamp_tax": stamp_tax,
|
||||
"slippage": slippage,
|
||||
"execution": execution,
|
||||
"chanlun_level": chanlun_level,
|
||||
"auto_fees": auto_fees,
|
||||
}
|
||||
|
||||
def run(self) -> WalkForwardResult:
|
||||
"""执行组合 Walk-Forward 验证。
|
||||
|
||||
Returns:
|
||||
:class:`WalkForwardResult`。数据不足以切窗时返回空结果
|
||||
（``windows`` 为空，聚合指标为 0）。
|
||||
"""
|
||||
result = WalkForwardResult(n_windows=self._n_windows, warmup_ratio=self._warmup_ratio)
|
||||
if not self._stocks:
|
||||
return result
|
||||
|
||||
# 参考时间轴：全部标的 datetime 的并集（升序）
|
||||
timeline = self._reference_timeline()
|
||||
n = len(timeline)
|
||||
# 最少数据：每窗 ≥ 20 根 + 预热区 ≥ 20 根
|
||||
min_bars = 20 * (1 + self._n_windows)
|
||||
if n < min_bars:
|
||||
return result
|
||||
|
||||
eval_start = int(n * self._warmup_ratio)
|
||||
eval_len = n - eval_start
|
||||
window_len = eval_len // self._n_windows
|
||||
|
||||
for i in range(self._n_windows):
|
||||
s = eval_start + i * window_len
|
||||
e = s + window_len if i < self._n_windows - 1 else n # 末窗吃到尾部
|
||||
if e - s < 5:
|
||||
continue
|
||||
win = self._run_window(timeline, s, e, i)
|
||||
if win is not None:
|
||||
result.windows.append(win)
|
||||
|
||||
WalkForwardEngine._aggregate(result)
|
||||
return result
|
||||
|
||||
def _reference_timeline(self) -> pd.DatetimeIndex:
|
||||
"""全部标的 datetime 的并集（升序，Timestamp 化）。"""
|
||||
all_dt: list[pd.Timestamp] = []
|
||||
for stock in self._stocks:
|
||||
s = self._dt_series(stock.df)
|
||||
if len(s) > 0:
|
||||
all_dt.append(s)
|
||||
if not all_dt:
|
||||
return pd.DatetimeIndex([])
|
||||
return pd.DatetimeIndex(sorted(pd.unique(pd.concat(all_dt))))
|
||||
|
||||
@staticmethod
|
||||
def _dt_series(df: pd.DataFrame) -> pd.Series:
|
||||
"""标的 K 线的 datetime 列统一转 Timestamp（int YYYYMMDD 兼容）。"""
|
||||
col = "datetime" if "datetime" in df.columns else "date"
|
||||
dt = df[col]
|
||||
if dt.dtype.kind in "iu":
|
||||
return pd.to_datetime(dt.astype(str), format="%Y%m%d")
|
||||
if not pd.api.types.is_datetime64_any_dtype(dt):
|
||||
return pd.to_datetime(dt)
|
||||
return pd.Series(pd.to_datetime(dt), index=df.index)
|
||||
|
||||
def _run_window(
|
||||
self, timeline: pd.DatetimeIndex, s: int, e: int, index: int
|
||||
) -> WalkForwardWindow | None:
|
||||
"""独立回测单个窗口 [s, e)（参考时间轴下标），合成组合窗内净值。"""
|
||||
window_start = timeline[s]
|
||||
window_end = timeline[e - 1]
|
||||
ctx_start = timeline[max(0, s - self._context_bars)]
|
||||
|
||||
per_cash = self._total_cash / len(self._stocks)
|
||||
equity_series: list[pd.Series] = []
|
||||
trade_frames: list[pd.DataFrame] = []
|
||||
for stock in self._stocks:
|
||||
key = f"{stock.market}{stock.code}"
|
||||
dt = self._dt_series(stock.df)
|
||||
mask = (dt >= ctx_start) & (dt <= window_end)
|
||||
sub = stock.df.loc[mask].reset_index(drop=True)
|
||||
dt_sub = dt.loc[mask].reset_index(drop=True)
|
||||
# 上下文 bar 数 = 窗口起点之前保留的 bar 数（warmup 压制其信号）
|
||||
lead = int((dt_sub < window_start).sum())
|
||||
if len(sub) < lead + 5:
|
||||
continue # 该标的数据不足（晚上市/停牌过多），本窗跳过
|
||||
|
||||
engine = BacktestEngine(
|
||||
strategy=self._strategy,
|
||||
cash=per_cash,
|
||||
warmup_bars=lead,
|
||||
symbol=key,
|
||||
**self._engine_kwargs,
|
||||
n = max(len(list(stocks)), 1)
|
||||
slots = [
|
||||
_ComboSlot(
|
||||
key=f"{s.market}{s.code}",
|
||||
strategy=strategy,
|
||||
df=s.df,
|
||||
cash=total_cash / n,
|
||||
symbol=f"{s.market}{s.code}",
|
||||
auto_fees=auto_fees,
|
||||
)
|
||||
try:
|
||||
bt = engine.run(sub)
|
||||
except Exception: # noqa: BLE001 — 单标的失败不拖垮整窗
|
||||
continue
|
||||
|
||||
# 只取窗内净值点（上下文区恒为现金，不参与窗指标，避免稀释波动率）
|
||||
ec = bt.equity_curve
|
||||
if len(ec) > lead:
|
||||
eq = ec.iloc[lead:]
|
||||
equity_series.append(
|
||||
pd.Series(eq["total"].to_numpy(), index=self._dt_series(eq), name=key)
|
||||
)
|
||||
if len(bt.trades) > 0:
|
||||
t = bt.trades.copy()
|
||||
t["symbol"] = key
|
||||
trade_frames.append(t)
|
||||
|
||||
if not equity_series:
|
||||
return None # 所有标的都跑不了，跳过该窗
|
||||
|
||||
# 合成组合窗内净值：日期并集对齐，ffill 持有不动，上市晚于窗口起点的
|
||||
# 标的其前导缺口用首值回填（首值即其初始资金——还没开仓，持有现金）
|
||||
aligned = pd.concat(equity_series, axis=1).sort_index()
|
||||
aligned = aligned.ffill().bfill()
|
||||
total = aligned.sum(axis=1)
|
||||
peak = total.cummax()
|
||||
drawdown = peak - total
|
||||
peak_safe = peak.where(peak != 0, 1.0)
|
||||
window_equity = pd.DataFrame(
|
||||
{
|
||||
"datetime": total.index,
|
||||
"total": total.to_numpy(),
|
||||
"drawdown": drawdown.to_numpy(),
|
||||
"drawdown_pct": (drawdown / peak_safe).to_numpy(),
|
||||
}
|
||||
for s in stocks
|
||||
]
|
||||
super().__init__(
|
||||
slots,
|
||||
n_windows=n_windows,
|
||||
warmup_ratio=warmup_ratio,
|
||||
context_bars=context_bars,
|
||||
engine_kwargs={
|
||||
"commission": commission,
|
||||
"min_commission": min_commission,
|
||||
"stamp_tax": stamp_tax,
|
||||
"slippage": slippage,
|
||||
"execution": execution,
|
||||
"chanlun_level": chanlun_level,
|
||||
},
|
||||
)
|
||||
|
||||
all_trades = (
|
||||
pd.concat(trade_frames, ignore_index=True)
|
||||
if trade_frames
|
||||
else pd.DataFrame(columns=["symbol", "direction", "pnl", "rejected"])
|
||||
)
|
||||
from easy_tdx.backtest.performance import PerformanceAnalyzer
|
||||
|
||||
perf = PerformanceAnalyzer(equity_curve=window_equity, trades=all_trades).compute()
|
||||
class MultiStrategyWalkForwardEngine(_ComboWalkForwardBase):
|
||||
"""多策略组合级 Walk-Forward：N 个策略各跑各自的原标的，逐窗独立回测。
|
||||
|
||||
return WalkForwardWindow(
|
||||
index=index,
|
||||
start=window_start.strftime("%Y-%m-%d"),
|
||||
end=window_end.strftime("%Y-%m-%d"),
|
||||
bars=int(e - s),
|
||||
total_return=float(perf.get("total_return", 0.0)),
|
||||
sharpe=float(perf.get("sharpe", 0.0)),
|
||||
max_drawdown=float(perf.get("max_drawdown", 0.0)),
|
||||
total_trades=int(perf.get("total_trades", 0)),
|
||||
win_rate=float(perf.get("win_rate", 0.0)),
|
||||
performance={k: v for k, v in perf.items()},
|
||||
与 :class:`PortfolioWalkForwardEngine` 共用切窗语义、组合窗内净值合成与
|
||||
:class:`WalkForwardResult` 输出结构（前端 WalkForwardPanel 直接复用），
|
||||
唯一差异是槽位划分：每个槽位是「一个策略 × 它自己的标的」
|
||||
（key 形如 ``"{label}@{symbol}"``，与
|
||||
:class:`~easy_tdx.backtest.multi_strategy_engine.MultiStrategyEngine` 的
|
||||
``individual_results`` key 一致）。
|
||||
|
||||
Example:
|
||||
>>> wf = MultiStrategyWalkForwardEngine(strategies=slots, n_windows=5)
|
||||
>>> result = wf.run()
|
||||
>>> result.consistency # 组合盈利窗占比
|
||||
0.6
|
||||
"""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
strategies: list[Any],
|
||||
n_windows: int = 7,
|
||||
warmup_ratio: float = 0.3,
|
||||
context_bars: int = 60,
|
||||
total_cash: float = 1_000_000.0,
|
||||
commission: float = 0.0003,
|
||||
min_commission: float = 5.0,
|
||||
stamp_tax: float = 0.001,
|
||||
slippage: float = 0.0,
|
||||
execution: str = "next_open",
|
||||
) -> None:
|
||||
"""Initialize.
|
||||
|
||||
Args:
|
||||
strategies: :class:`~easy_tdx.backtest.multi_strategy_engine.StrategySlot`
|
||||
列表（每个槽位已绑定策略实例与 K 线）。
|
||||
n_windows / warmup_ratio / context_bars: 切窗参数（同单标的 WF）。
|
||||
total_cash: 组合总资金（各槽位等权分 1/N）。
|
||||
其余参数: 透传给各窗各槽位的 :class:`BacktestEngine`
|
||||
（与 MultiStrategyEngine 同口径，不含 auto_fees/chanlun_level）。
|
||||
"""
|
||||
n = max(len(list(strategies)), 1)
|
||||
slots = [
|
||||
_ComboSlot(
|
||||
key=f"{s.label}@{s.symbol}",
|
||||
strategy=s.strategy,
|
||||
df=s.df,
|
||||
cash=total_cash / n,
|
||||
)
|
||||
for s in strategies
|
||||
]
|
||||
super().__init__(
|
||||
slots,
|
||||
n_windows=n_windows,
|
||||
warmup_ratio=warmup_ratio,
|
||||
context_bars=context_bars,
|
||||
engine_kwargs={
|
||||
"commission": commission,
|
||||
"min_commission": min_commission,
|
||||
"stamp_tax": stamp_tax,
|
||||
"slippage": slippage,
|
||||
"execution": execution,
|
||||
},
|
||||
)
|
||||
|
||||
@@ -289,6 +289,69 @@ async def run_multi_strategy_backtest_async(
|
||||
return TaskSubmitResponse(task_id=task_id, status=status)
|
||||
|
||||
|
||||
@router.post(
|
||||
"/backtest/multi-strategy/wf/run/async", response_model=TaskSubmitResponse, status_code=202
|
||||
)
|
||||
async def run_multi_strategy_walkforward_async(
|
||||
req: MultiStrategyBacktestRequest,
|
||||
n_windows: int = 7,
|
||||
client: Any = Depends(get_client),
|
||||
) -> TaskSubmitResponse:
|
||||
"""提交多策略组合级 Walk-Forward 样本外验证后台任务。
|
||||
|
||||
逐槽位取行情后，按全部槽位日期并集切窗（预热区 + N 个连续测试窗），
|
||||
每窗各槽位独立回测并合成组合窗内净值。结果为 ``{"walkforward": {...}}``
|
||||
（与单标的 WF 同构），通过 GET /backtest/tasks/{task_id} 轮询。
|
||||
"""
|
||||
slots = await _fetch_multi_strategy_bars(client, req.items)
|
||||
if not slots:
|
||||
raise ValueError("所有策略槽位均未取到有效行情数据")
|
||||
|
||||
snapshot = req.model_copy()
|
||||
description = f"多策略组合WF | {len(slots)}个策略 × {n_windows}窗"
|
||||
|
||||
runner = get_runner()
|
||||
task_id = runner.submit(
|
||||
lambda: _run_multi_strategy_walkforward(slots, snapshot, n_windows),
|
||||
description=description,
|
||||
)
|
||||
state = runner.get(task_id)
|
||||
status: Any = state.status if state.status in ("pending", "running") else "running"
|
||||
return TaskSubmitResponse(task_id=task_id, status=status)
|
||||
|
||||
|
||||
@router.post(
|
||||
"/backtest/multi-strategy/evaluate/run/async",
|
||||
response_model=TaskSubmitResponse,
|
||||
status_code=202,
|
||||
)
|
||||
async def run_multi_strategy_evaluate_async(
|
||||
req: MultiStrategyBacktestRequest,
|
||||
client: Any = Depends(get_client),
|
||||
) -> TaskSubmitResponse:
|
||||
"""提交多策略组合级一条龙评估后台任务：组合回测 + 组合 WF + 跨槽位适配性
|
||||
体检 + 综合评分 + 组合评级 + 等权买入持有基准对比。
|
||||
|
||||
结果结构见 ``easy_tdx.backtest.benchmark.evaluate_multi`` 文档（与
|
||||
单标的 evaluate_strategy 同构），通过 GET /backtest/tasks/{task_id} 轮询。
|
||||
"""
|
||||
slots = await _fetch_multi_strategy_bars(client, req.items)
|
||||
if not slots:
|
||||
raise ValueError("所有策略槽位均未取到有效行情数据")
|
||||
|
||||
snapshot = req.model_copy()
|
||||
description = f"多策略组合一条龙 | {len(slots)}个策略"
|
||||
|
||||
runner = get_runner()
|
||||
task_id = runner.submit(
|
||||
lambda: _run_multi_strategy_evaluate(slots, snapshot),
|
||||
description=description,
|
||||
)
|
||||
state = runner.get(task_id)
|
||||
status: Any = state.status if state.status in ("pending", "running") else "running"
|
||||
return TaskSubmitResponse(task_id=task_id, status=status)
|
||||
|
||||
|
||||
@router.post("/backtest/optimize/run/async", response_model=TaskSubmitResponse, status_code=202)
|
||||
async def run_optimize_async(
|
||||
req: OptimizeBacktestRequest,
|
||||
@@ -1054,8 +1117,14 @@ async def _fetch_multi_strategy_bars(
|
||||
def _run_multi_strategy_backtest(
|
||||
slots: list[Any], req: MultiStrategyBacktestRequest
|
||||
) -> dict[str, Any]:
|
||||
"""执行多策略组合回测并返回清洗后的结果字典（后台线程内调用）。"""
|
||||
"""执行多策略组合回测并返回清洗后的结果字典（后台线程内调用）。
|
||||
|
||||
与组合回测 ``_run_portfolio_backtest`` 同构：附带组合评级（净值口径）
|
||||
与综合评分，供前端/REST 直接消费。
|
||||
"""
|
||||
from easy_tdx.backtest.grading import grade_portfolio_equity
|
||||
from easy_tdx.backtest.multi_strategy_engine import MultiStrategyEngine
|
||||
from easy_tdx.backtest.scoring import score_strategy
|
||||
|
||||
engine = MultiStrategyEngine(
|
||||
strategies=slots,
|
||||
@@ -1067,7 +1136,50 @@ def _run_multi_strategy_backtest(
|
||||
execution=req.execution,
|
||||
)
|
||||
result = engine.run()
|
||||
return serialize_result(result)
|
||||
out = serialize_result(result)
|
||||
# 组合评级（净值曲线口径）+ 综合评分——与单标的/多标的组合响应同构
|
||||
if len(result.combined_equity) >= 2:
|
||||
out["grade"] = grade_portfolio_equity(
|
||||
result.combined_equity.to_dict(orient="records")
|
||||
).to_dict()
|
||||
out["score"] = score_strategy(dict(result.total_performance)).to_dict()
|
||||
return out
|
||||
|
||||
|
||||
def _run_multi_strategy_walkforward(
|
||||
slots: list[Any], req: MultiStrategyBacktestRequest, n_windows: int = 7
|
||||
) -> dict[str, Any]:
|
||||
"""执行多策略组合级 Walk-Forward 验证（后台线程内调用）。"""
|
||||
from easy_tdx.backtest.walkforward import MultiStrategyWalkForwardEngine
|
||||
|
||||
wf = MultiStrategyWalkForwardEngine(
|
||||
strategies=slots,
|
||||
n_windows=n_windows,
|
||||
total_cash=req.cash,
|
||||
commission=req.commission,
|
||||
min_commission=req.min_commission,
|
||||
stamp_tax=req.stamp_tax,
|
||||
slippage=req.slippage,
|
||||
execution=req.execution,
|
||||
).run()
|
||||
return {"walkforward": wf.to_dict()}
|
||||
|
||||
|
||||
def _run_multi_strategy_evaluate(
|
||||
slots: list[Any], req: MultiStrategyBacktestRequest
|
||||
) -> dict[str, Any]:
|
||||
"""执行多策略组合级一条龙评估（后台线程内调用）。"""
|
||||
from easy_tdx.backtest.benchmark import evaluate_multi
|
||||
|
||||
return evaluate_multi(
|
||||
strategies=slots,
|
||||
total_cash=req.cash,
|
||||
commission=req.commission,
|
||||
min_commission=req.min_commission,
|
||||
stamp_tax=req.stamp_tax,
|
||||
slippage=req.slippage,
|
||||
execution=req.execution,
|
||||
)
|
||||
|
||||
|
||||
def _run_optimize(df: pd.DataFrame, req: OptimizeBacktestRequest) -> dict[str, Any]:
|
||||
|
||||
@@ -251,3 +251,59 @@ def test_evaluate_portfolio_serializable():
|
||||
)
|
||||
text = json.dumps(report, default=str)
|
||||
assert "excess_return" in text
|
||||
|
||||
|
||||
# ── evaluate_multi（v1.31.1：多策略组合一条龙）────────────────────────────────
|
||||
def _slots_for_multi() -> list[Any]:
|
||||
from easy_tdx.backtest.multi_strategy_engine import StrategySlot
|
||||
|
||||
return [
|
||||
StrategySlot(
|
||||
label="动量", symbol="SZ:000001", strategy=_CycleTrader(), df=_df(400, drift=0.002)
|
||||
),
|
||||
StrategySlot(
|
||||
label="反转", symbol="SH:600000", strategy=_CycleTrader(), df=_df(400, drift=0.003)
|
||||
),
|
||||
]
|
||||
|
||||
|
||||
def test_evaluate_multi_full_report_structure():
|
||||
"""多策略组合一条龙报告与 evaluate_portfolio 同构（前端面板可复用）。"""
|
||||
from easy_tdx.backtest.benchmark import evaluate_multi
|
||||
|
||||
report = evaluate_multi(_slots_for_multi(), total_cash=500_000)
|
||||
for key in ("performance", "score", "grade", "walkforward", "fitness", "benchmark", "config"):
|
||||
assert key in report
|
||||
assert "sqn" in report["performance"]
|
||||
assert "max_consecutive_losses" in report["performance"]
|
||||
assert report["performance"]["total_stocks"] == 2
|
||||
assert 0 <= report["score"]["total"] <= 100
|
||||
assert report["score"]["wf_provided"] is True
|
||||
assert report["grade"]["grade"] in ("S", "A", "B", "C", "D")
|
||||
assert report["grade"]["scenario"] == "portfolio"
|
||||
assert len(report["walkforward"]["windows"]) > 0
|
||||
assert report["fitness"]["total_checks"] == 8
|
||||
assert "只标的通过" in report["fitness"]["checks"][0]["detail"]
|
||||
assert "buy_hold" in report["benchmark"]
|
||||
assert report["config"]["slots"] == ["动量@SZ:000001", "反转@SH:600000"]
|
||||
|
||||
|
||||
def test_evaluate_multi_buy_hold_excess_near_zero():
|
||||
"""各槽位换成买入持有后，组合收益 ≈ 等权买入持有基准（excess 近 0）。"""
|
||||
from easy_tdx.backtest.benchmark import evaluate_multi
|
||||
from easy_tdx.backtest.multi_strategy_engine import StrategySlot
|
||||
|
||||
bh_slots = [
|
||||
StrategySlot(label=s.label, symbol=s.symbol, strategy=_BuyFirstBar(), df=s.df)
|
||||
for s in _slots_for_multi()
|
||||
]
|
||||
report = evaluate_multi(bh_slots, total_cash=500_000)
|
||||
assert abs(report["benchmark"]["excess_return"]) < 0.05
|
||||
|
||||
|
||||
def test_evaluate_multi_serializable():
|
||||
from easy_tdx.backtest.benchmark import evaluate_multi
|
||||
|
||||
report = evaluate_multi(_slots_for_multi(), total_cash=500_000, n_windows=3)
|
||||
text = json.dumps(report, default=str)
|
||||
assert "excess_return" in text
|
||||
|
||||
@@ -3,6 +3,7 @@
|
||||
from __future__ import annotations
|
||||
|
||||
import json
|
||||
from typing import Any
|
||||
|
||||
import numpy as np
|
||||
import pandas as pd
|
||||
@@ -131,3 +132,53 @@ class TestPortfolioWalkForward:
|
||||
strategy=PeriodicStrategy, stocks=_stocks(), n_windows=0
|
||||
).run()
|
||||
assert wf.n_windows == 2
|
||||
|
||||
|
||||
# ── MultiStrategyWalkForwardEngine（v1.31.1：多策略组合槽位 WF）───────────────
|
||||
def _slots() -> list[Any]:
|
||||
from easy_tdx.backtest.multi_strategy_engine import StrategySlot
|
||||
|
||||
return [
|
||||
StrategySlot(
|
||||
label="双均线交叉",
|
||||
symbol="SH:601088",
|
||||
strategy=PeriodicStrategy(),
|
||||
df=_make_df(400, seed=42),
|
||||
),
|
||||
StrategySlot(
|
||||
label="RSI反转",
|
||||
symbol="SZ:000001",
|
||||
strategy=PeriodicStrategy(),
|
||||
df=_make_df(400, seed=99),
|
||||
),
|
||||
]
|
||||
|
||||
|
||||
def test_multi_strategy_wf_basic_structure() -> None:
|
||||
from easy_tdx.backtest.walkforward import MultiStrategyWalkForwardEngine
|
||||
|
||||
wf = MultiStrategyWalkForwardEngine(strategies=_slots(), n_windows=4, total_cash=200_000).run()
|
||||
assert len(wf.windows) == 4
|
||||
assert wf.total_trades > 0
|
||||
assert wf.total_trades == sum(w.total_trades for w in wf.windows)
|
||||
# 窗口时间升序
|
||||
starts = [pd.Timestamp(w.start) for w in wf.windows]
|
||||
assert starts == sorted(starts)
|
||||
|
||||
|
||||
def test_multi_strategy_wf_matches_portfolio_structure() -> None:
|
||||
"""与 PortfolioWalkForwardEngine 输出同构（前端面板可复用）。"""
|
||||
from easy_tdx.backtest.walkforward import MultiStrategyWalkForwardEngine
|
||||
|
||||
wf = MultiStrategyWalkForwardEngine(strategies=_slots(), n_windows=3).run()
|
||||
d = wf.to_dict()
|
||||
json.dumps(d)
|
||||
assert "sqn" in d["windows"][0]["performance"]
|
||||
assert "max_consecutive_wins" in d["windows"][0]["performance"]
|
||||
|
||||
|
||||
def test_multi_strategy_wf_empty_slots() -> None:
|
||||
from easy_tdx.backtest.walkforward import MultiStrategyWalkForwardEngine
|
||||
|
||||
wf = MultiStrategyWalkForwardEngine(strategies=[], n_windows=3).run()
|
||||
assert wf.windows == []
|
||||
|
||||
@@ -11,6 +11,7 @@
|
||||
from __future__ import annotations
|
||||
|
||||
import time
|
||||
from typing import Any
|
||||
|
||||
import numpy as np
|
||||
import pandas as pd
|
||||
@@ -1196,3 +1197,171 @@ def test_portfolio_evaluate_endpoint(client, monkeypatch):
|
||||
assert report["grade"]["scenario"] == "portfolio"
|
||||
assert report["fitness"]["total_checks"] == 8
|
||||
assert report["config"]["stocks"] == ["SZ000001", "SH600519"]
|
||||
|
||||
|
||||
# ── 多策略组合级 WF / 一条龙评估端点（v1.31.1）────────────────────────────────
|
||||
def _fake_multi_slots(n: int = 400):
|
||||
"""构造 _fetch_multi_strategy_bars 的 mock 替身（两槽位合成行情）。"""
|
||||
import pandas as pd
|
||||
|
||||
from easy_tdx.backtest.multi_strategy_engine import StrategySlot
|
||||
|
||||
async def fake_fetch(client_arg, items): # noqa: ANN001
|
||||
slots = []
|
||||
for item in items:
|
||||
mkt, code = item.symbol.split(":")
|
||||
close = 10 + np.cumsum(np.random.randn(n) * 0.3 + 0.02)
|
||||
df = pd.DataFrame(
|
||||
{
|
||||
"datetime": pd.date_range("2023-01-02", periods=n, freq="B"),
|
||||
"open": close - 0.1,
|
||||
"high": close + 0.2,
|
||||
"low": close - 0.2,
|
||||
"close": close,
|
||||
"vol": np.full(n, 5000.0),
|
||||
"amount": close * 5000,
|
||||
}
|
||||
)
|
||||
slots.append(
|
||||
StrategySlot(label=item.strategy, symbol=item.symbol, strategy=None, df=df)
|
||||
)
|
||||
return slots
|
||||
|
||||
return fake_fetch
|
||||
|
||||
|
||||
def _multi_request() -> dict[str, Any]:
|
||||
from datetime import date as _date
|
||||
|
||||
start = f"{_date.today().year - 3}-01-02"
|
||||
return {
|
||||
"items": [
|
||||
{
|
||||
"strategy": "ma_cross",
|
||||
"strategy_label": "双均线交叉",
|
||||
"params": {"fast": 5, "slow": 20},
|
||||
"symbol": "SH:601088",
|
||||
"category": "DAY",
|
||||
"start_date": start,
|
||||
},
|
||||
{
|
||||
"strategy": "macd",
|
||||
"strategy_label": "MACD 金叉",
|
||||
"params": {},
|
||||
"symbol": "SZ:000001",
|
||||
"category": "DAY",
|
||||
"start_date": start,
|
||||
},
|
||||
],
|
||||
"cash": 200000,
|
||||
}
|
||||
|
||||
|
||||
def test_multi_strategy_wf_endpoint(client, monkeypatch):
|
||||
"""POST /backtest/multi-strategy/wf/run/async 端到端（mock 取数 + 真实引擎）。
|
||||
|
||||
StrategySlot 由取数阶段绑定策略实例（_build 时替换 mock 的 None），
|
||||
这里用 router 内的真实 _fetch_multi_strategy_bars 不可行（需 client），
|
||||
故 fake_fetch 直接构造策略实例。
|
||||
"""
|
||||
import pandas as pd
|
||||
|
||||
import easy_tdx.web.routers.backtest as bt_router
|
||||
from easy_tdx.backtest.multi_strategy_engine import StrategySlot
|
||||
from easy_tdx.backtest.strategies import get_registry
|
||||
|
||||
async def fake_fetch(client_arg, items): # noqa: ANN001
|
||||
registry = get_registry()
|
||||
slots = []
|
||||
for item in items:
|
||||
entry = registry.get(item.strategy)
|
||||
strategy = entry.build(item.params)
|
||||
close = 10 + np.cumsum(np.random.randn(400) * 0.3 + 0.02)
|
||||
df = pd.DataFrame(
|
||||
{
|
||||
"datetime": pd.date_range("2023-01-02", periods=400, freq="B"),
|
||||
"open": close - 0.1,
|
||||
"high": close + 0.2,
|
||||
"low": close - 0.2,
|
||||
"close": close,
|
||||
"vol": np.full(400, 5000.0),
|
||||
"amount": close * 5000,
|
||||
}
|
||||
)
|
||||
slots.append(
|
||||
StrategySlot(
|
||||
label=item.strategy_label or item.strategy,
|
||||
symbol=item.symbol,
|
||||
strategy=strategy,
|
||||
df=df,
|
||||
)
|
||||
)
|
||||
return slots
|
||||
|
||||
monkeypatch.setattr(bt_router, "_fetch_multi_strategy_bars", fake_fetch)
|
||||
|
||||
resp = client.post(
|
||||
"/api/v1/backtest/multi-strategy/wf/run/async?n_windows=3",
|
||||
json=_multi_request(),
|
||||
)
|
||||
assert resp.status_code == 202, resp.text
|
||||
final = _wait_task(client, resp.json()["task_id"])
|
||||
assert final["status"] == "done", final
|
||||
wf = final["result"]["walkforward"]
|
||||
assert wf["n_windows"] == 3
|
||||
assert len(wf["windows"]) == 3
|
||||
assert "consistency" in wf
|
||||
assert "sqn" in wf["windows"][0]["performance"]
|
||||
|
||||
|
||||
def test_multi_strategy_evaluate_endpoint(client, monkeypatch):
|
||||
"""POST /backtest/multi-strategy/evaluate/run/async 端到端。"""
|
||||
import pandas as pd
|
||||
|
||||
import easy_tdx.web.routers.backtest as bt_router
|
||||
from easy_tdx.backtest.multi_strategy_engine import StrategySlot
|
||||
from easy_tdx.backtest.strategies import get_registry
|
||||
|
||||
async def fake_fetch(client_arg, items): # noqa: ANN001
|
||||
registry = get_registry()
|
||||
slots = []
|
||||
for item in items:
|
||||
entry = registry.get(item.strategy)
|
||||
strategy = entry.build(item.params)
|
||||
close = 10 + np.cumsum(np.random.randn(400) * 0.3 + 0.02)
|
||||
df = pd.DataFrame(
|
||||
{
|
||||
"datetime": pd.date_range("2023-01-02", periods=400, freq="B"),
|
||||
"open": close - 0.1,
|
||||
"high": close + 0.2,
|
||||
"low": close - 0.2,
|
||||
"close": close,
|
||||
"vol": np.full(400, 5000.0),
|
||||
"amount": close * 5000,
|
||||
}
|
||||
)
|
||||
slots.append(
|
||||
StrategySlot(
|
||||
label=item.strategy_label or item.strategy,
|
||||
symbol=item.symbol,
|
||||
strategy=strategy,
|
||||
df=df,
|
||||
)
|
||||
)
|
||||
return slots
|
||||
|
||||
monkeypatch.setattr(bt_router, "_fetch_multi_strategy_bars", fake_fetch)
|
||||
|
||||
resp = client.post(
|
||||
"/api/v1/backtest/multi-strategy/evaluate/run/async",
|
||||
json=_multi_request(),
|
||||
)
|
||||
assert resp.status_code == 202, resp.text
|
||||
final = _wait_task(client, resp.json()["task_id"])
|
||||
assert final["status"] == "done", final
|
||||
report = final["result"]
|
||||
for key in ("performance", "score", "grade", "walkforward", "fitness", "benchmark", "config"):
|
||||
assert key in report
|
||||
assert report["grade"]["scenario"] == "portfolio"
|
||||
assert report["fitness"]["total_checks"] == 8
|
||||
assert report["config"]["slots"] == ["双均线交叉@SH:601088", "MACD 金叉@SZ:000001"]
|
||||
|
||||
@@ -330,3 +330,35 @@ test('组合版：WF / 一条龙 / 评级按需拼接', () => {
|
||||
assert.match(p, /# 评级（不看收益率，面向「普通人拿不拿得住」）/)
|
||||
assert.match(p, /档位：\*\*D\*\*/)
|
||||
})
|
||||
|
||||
test('组合版 multi 模式：N 个策略各跑原标的的语境与槽位明细', () => {
|
||||
const p = buildPortfolioAiPrompt({
|
||||
stocks: ['双均线交叉@SH:601088', 'MACD 金叉@SZ:000001'],
|
||||
category: 'DAY',
|
||||
startDate: '2023-01-02',
|
||||
endDate: '2026-09-03',
|
||||
strategyLabel: '2 策略组合',
|
||||
params: {},
|
||||
cash: 500000,
|
||||
commission: 0.0003,
|
||||
slippage: 0,
|
||||
execution: 'next_open',
|
||||
mode: 'multi',
|
||||
result: {
|
||||
...PORTFOLIO_RESULT,
|
||||
individual_results: {
|
||||
'双均线交叉@SH:601088': RESULT,
|
||||
'MACD 金叉@SZ:000001': RESULT,
|
||||
},
|
||||
},
|
||||
})
|
||||
|
||||
// 角色设定为多策略语境（N 个策略各跑各自的原标的）
|
||||
assert.match(p, /N 个策略各跑各自的原标的，资金均分/)
|
||||
// 配置段列策略明细而非标的清单，不再出现单一策略/参数行
|
||||
assert.match(p, /2 个策略各跑各自的原标的，资金均分（各拿总额的 50\.0%）/)
|
||||
assert.match(p, /双均线交叉@SH:601088、MACD 金叉@SZ:000001/)
|
||||
assert.ok(!p.includes('- 策略：2 策略组合'))
|
||||
// 槽位表现段标题为「各策略槽位」
|
||||
assert.match(p, /# 各策略槽位表现（按收益降序；全部）/)
|
||||
})
|
||||
|
||||
+63
-32
@@ -49,7 +49,8 @@ export interface AiPromptInput {
|
||||
}
|
||||
|
||||
export interface PortfolioAiPromptInput {
|
||||
/** 完整标的代码列表（带市场前缀，如 ["SZ:000001", "SH:600519"]） */
|
||||
/** 完整标的代码列表（带市场前缀，如 ["SZ:000001", "SH:600519"]）；
|
||||
* multi 模式下传「策略@标的」明细列表（如 ["双均线@SH:601088", …]） */
|
||||
stocks: string[]
|
||||
category: Category
|
||||
startDate: string
|
||||
@@ -61,6 +62,9 @@ export interface PortfolioAiPromptInput {
|
||||
slippage: number
|
||||
execution: ExecutionMode
|
||||
result: PortfolioResult
|
||||
/** 组合形态：portfolio = 一个策略 × 多只标的（默认）；
|
||||
* multi = 多个策略各跑各自的原标的（策略库「重跑到今天」） */
|
||||
mode?: 'portfolio' | 'multi'
|
||||
/** 附加分析（未勾选/未跑完时传 null，对应段落自动省略） */
|
||||
wf?: WalkForwardResult | null
|
||||
evaluate?: EvaluateReport | null
|
||||
@@ -163,15 +167,22 @@ function n(v: number | string | null | undefined): number | undefined {
|
||||
|
||||
// ── 各段落构建 ───────────────────────────────────────────────────────────────
|
||||
|
||||
function sectionRole(kind: 'single' | 'portfolio' = 'single'): string {
|
||||
const intro =
|
||||
kind === 'portfolio'
|
||||
? '下面是我跑出来的组合回测报告（同一个策略分别跑在一篮子标的上，资金均分、各标的独立回测后净值加总），帮我看看这个组合策略到底行不行。内容上要说到这六件事，顺序随意，用你自然的说话方式组织：'
|
||||
: '下面是我跑出来的回测报告，帮我看看这个策略到底行不行。内容上要说到这六件事，顺序随意，用你自然的说话方式组织：'
|
||||
const step5 =
|
||||
kind === 'portfolio'
|
||||
? '5. **给可执行的下一步**：几条我马上能做的事（改什么参数、加什么过滤、换哪些标的、先做什么测试再谈实盘），别空谈；'
|
||||
: '5. **给可执行的下一步**：几条我马上能做的事（改什么参数、加什么过滤、先做什么测试再谈实盘），别空谈；'
|
||||
function sectionRole(kind: 'single' | 'portfolio' | 'multi' = 'single'): string {
|
||||
let intro =
|
||||
'下面是我跑出来的回测报告，帮我看看这个策略到底行不行。内容上要说到这六件事，顺序随意，用你自然的说话方式组织：'
|
||||
let step5 =
|
||||
'5. **给可执行的下一步**：几条我马上能做的事（改什么参数、加什么过滤、先做什么测试再谈实盘），别空谈；'
|
||||
if (kind === 'portfolio') {
|
||||
intro =
|
||||
'下面是我跑出来的组合回测报告（同一个策略分别跑在一篮子标的上，资金均分、各标的独立回测后净值加总），帮我看看这个组合策略到底行不行。内容上要说到这六件事，顺序随意，用你自然的说话方式组织：'
|
||||
step5 =
|
||||
'5. **给可执行的下一步**：几条我马上能做的事（改什么参数、加什么过滤、换哪些标的、先做什么测试再谈实盘），别空谈；'
|
||||
} else if (kind === 'multi') {
|
||||
intro =
|
||||
'下面是我跑出来的组合回测报告（N 个策略各跑各自的原标的，资金均分、各策略独立回测后净值加总），帮我看看这个策略组合到底行不行。内容上要说到这六件事，顺序随意，用你自然的说话方式组织：'
|
||||
step5 =
|
||||
'5. **给可执行的下一步**：几条我马上能做的事（改哪些策略的参数、换掉拖后腿的策略、加什么过滤、先做什么测试再谈实盘），别空谈；'
|
||||
}
|
||||
return [
|
||||
'# 角色设定',
|
||||
'',
|
||||
@@ -370,28 +381,46 @@ function sectionFooter(): string {
|
||||
// ── 组合版段落 ───────────────────────────────────────────────────────────────
|
||||
|
||||
function sectionPortfolioConfig(i: PortfolioAiPromptInput): string {
|
||||
const stockList =
|
||||
i.stocks.length <= 12
|
||||
? i.stocks.join('、')
|
||||
: `${i.stocks.slice(0, 12).join('、')} 等 ${i.stocks.length} 只`
|
||||
const lines = [
|
||||
'# 组合回测配置',
|
||||
'',
|
||||
`- 组合形式：同一个策略分别跑在 ${i.stocks.length} 只标的上，资金均分（各拿总额的 ${(
|
||||
100 / i.stocks.length
|
||||
).toFixed(1)}%），标的间独立回测、净值按日加总`,
|
||||
`- 标的列表：${stockList}`,
|
||||
`- 回测区间：${i.startDate} ~ ${i.endDate}（${CATEGORY_LABELS[i.category] ?? i.category}）`,
|
||||
`- 策略：${i.strategyLabel}`,
|
||||
`- 参数：${fmtParams(i.params)}`,
|
||||
const multi = i.mode === 'multi'
|
||||
const lines = ['# 组合回测配置', '']
|
||||
if (multi) {
|
||||
const detail =
|
||||
i.stocks.length <= 12
|
||||
? i.stocks.join('、')
|
||||
: `${i.stocks.slice(0, 12).join('、')} 等 ${i.stocks.length} 个策略槽位`
|
||||
lines.push(
|
||||
`- 组合形式：${i.stocks.length} 个策略各跑各自的原标的，资金均分（各拿总额的 ${(
|
||||
100 / i.stocks.length
|
||||
).toFixed(1)}%），策略间独立回测、净值按日加总`,
|
||||
)
|
||||
lines.push(`- 策略明细：${detail}`)
|
||||
} else {
|
||||
const stockList =
|
||||
i.stocks.length <= 12
|
||||
? i.stocks.join('、')
|
||||
: `${i.stocks.slice(0, 12).join('、')} 等 ${i.stocks.length} 只`
|
||||
lines.push(
|
||||
`- 组合形式：同一个策略分别跑在 ${i.stocks.length} 只标的上，资金均分（各拿总额的 ${(
|
||||
100 / i.stocks.length
|
||||
).toFixed(1)}%），标的间独立回测、净值按日加总`,
|
||||
)
|
||||
lines.push(`- 标的列表：${stockList}`)
|
||||
lines.push(`- 策略：${i.strategyLabel}`)
|
||||
lines.push(`- 参数：${fmtParams(i.params)}`)
|
||||
}
|
||||
if (multi) {
|
||||
lines.push('- （各策略的参数见各策略在策略库中保存的配置，此处不逐一展开）')
|
||||
}
|
||||
lines.push(`- 回测区间：${i.startDate} ~ ${i.endDate}（${CATEGORY_LABELS[i.category] ?? i.category}）`)
|
||||
lines.push(
|
||||
`- 组合总资金：${fmtMoney(i.cash)} 元；佣金 ${i.commission}；滑点 ${i.slippage}；成交价：${EXECUTION_LABELS[i.execution] ?? i.execution}`,
|
||||
'',
|
||||
]
|
||||
)
|
||||
lines.push('')
|
||||
return lines.join('\n')
|
||||
}
|
||||
|
||||
/** 各标的表现摘要：按收益降序，超过 12 只时只列最好 6 只 + 最差 6 只。 */
|
||||
function sectionStocksSummary(result: PortfolioResult): string {
|
||||
/** 各槽位表现摘要：按收益降序，超过 12 个时只列最好 6 个 + 最差 6 个。 */
|
||||
function sectionStocksSummary(result: PortfolioResult, multi = false): string {
|
||||
const entries = Object.entries(result.individual_results)
|
||||
if (entries.length === 0) return ''
|
||||
const sorted = entries
|
||||
@@ -401,9 +430,10 @@ function sectionStocksSummary(result: PortfolioResult): string {
|
||||
sorted.length <= 12
|
||||
? sorted
|
||||
: [...sorted.slice(0, 6), ...sorted.slice(sorted.length - 6)]
|
||||
const what = multi ? '各策略槽位' : '各标的'
|
||||
const lines = [
|
||||
'# 各标的表现（按收益降序；' +
|
||||
(sorted.length <= 12 ? '全部' : `省略中间 ${sorted.length - 12} 只，其余为最好/最差各 6 只`) +
|
||||
`# ${what}表现（按收益降序；` +
|
||||
(sorted.length <= 12 ? '全部' : `省略中间 ${sorted.length - 12} 个，其余为最好/最差各 6 个`) +
|
||||
'）',
|
||||
'',
|
||||
]
|
||||
@@ -451,13 +481,14 @@ export function buildAiPrompt(input: AiPromptInput): string {
|
||||
|
||||
/** 组装组合回测的 AI 解读 Prompt（与单标的同构，段落随附加分析增减）。 */
|
||||
export function buildPortfolioAiPrompt(input: PortfolioAiPromptInput): string {
|
||||
const multi = input.mode === 'multi'
|
||||
const parts: string[] = [
|
||||
sectionRole('portfolio'),
|
||||
sectionRole(multi ? 'multi' : 'portfolio'),
|
||||
sectionPortfolioConfig(input),
|
||||
sectionMetrics(input.result.total_performance),
|
||||
sectionEquityPoints(input.result.combined_equity),
|
||||
]
|
||||
const stocksSummary = sectionStocksSummary(input.result)
|
||||
const stocksSummary = sectionStocksSummary(input.result, multi)
|
||||
if (stocksSummary) parts.push(stocksSummary)
|
||||
if (input.wf) parts.push(sectionWf(input.wf))
|
||||
if (input.evaluate) parts.push(sectionEvaluate(input.evaluate))
|
||||
|
||||
@@ -241,6 +241,33 @@ export async function submitMultiStrategyTask(
|
||||
return (await resp.json()) as TaskSubmitResponse
|
||||
}
|
||||
|
||||
/** 提交多策略组合级 Walk-Forward 样本外验证后台任务（n_windows 默认 7）。 */
|
||||
export async function submitMultiStrategyWalkforwardTask(
|
||||
req: MultiStrategyBacktestRequest,
|
||||
nWindows = 7,
|
||||
): Promise<TaskSubmitResponse> {
|
||||
const resp = await fetch(`${BASE}/backtest/multi-strategy/wf/run/async?n_windows=${nWindows}`, {
|
||||
method: 'POST',
|
||||
headers: { 'Content-Type': 'application/json' },
|
||||
body: JSON.stringify(req),
|
||||
})
|
||||
if (!resp.ok) await throwError(resp)
|
||||
return (await resp.json()) as TaskSubmitResponse
|
||||
}
|
||||
|
||||
/** 提交多策略组合级一条龙评估后台任务（组合回测+WF+适配性+评分+基准对比）。 */
|
||||
export async function submitMultiStrategyEvaluateTask(
|
||||
req: MultiStrategyBacktestRequest,
|
||||
): Promise<TaskSubmitResponse> {
|
||||
const resp = await fetch(`${BASE}/backtest/multi-strategy/evaluate/run/async`, {
|
||||
method: 'POST',
|
||||
headers: { 'Content-Type': 'application/json' },
|
||||
body: JSON.stringify(req),
|
||||
})
|
||||
if (!resp.ok) await throwError(resp)
|
||||
return (await resp.json()) as TaskSubmitResponse
|
||||
}
|
||||
|
||||
/** 提交参数网格寻优后台任务，返回 task_id。 */
|
||||
export async function submitOptimizeTask(
|
||||
req: OptimizeBacktestRequest,
|
||||
|
||||
@@ -14,6 +14,8 @@ import {
|
||||
submitOptimizeAllTask,
|
||||
submitOptimizeTask,
|
||||
submitMultiStrategyTask,
|
||||
submitMultiStrategyEvaluateTask,
|
||||
submitMultiStrategyWalkforwardTask,
|
||||
submitWalkforwardTask,
|
||||
submitEvaluateTask,
|
||||
fetchTask,
|
||||
@@ -255,6 +257,8 @@ export const useBacktestStore = defineStore('backtest', () => {
|
||||
multiStrategyRunning.value = true
|
||||
error.value = ''
|
||||
multiStrategyResult.value = null
|
||||
// 新一次组合回测开始时清掉上一轮的附加分析（WF/一条龙面板随主结果一起刷新）
|
||||
clearMultiStrategyExtraAnalysis()
|
||||
try {
|
||||
const { task_id } = await submitMultiStrategyTask(req)
|
||||
const start = Date.now()
|
||||
@@ -281,9 +285,66 @@ export const useBacktestStore = defineStore('backtest', () => {
|
||||
|
||||
function clearMultiStrategy() {
|
||||
multiStrategyResult.value = null
|
||||
clearMultiStrategyExtraAnalysis()
|
||||
error.value = ''
|
||||
}
|
||||
|
||||
// ── 多策略组合附加分析：组合级 Walk-Forward / 一条龙评估 ──────────────────
|
||||
const multiWfResult = ref<WalkForwardResult | null>(null)
|
||||
const multiWfRunning = ref(false)
|
||||
const multiWfError = ref<string>('')
|
||||
const multiEvaluateResult = ref<EvaluateReport | null>(null)
|
||||
const multiEvaluateRunning = ref(false)
|
||||
const multiEvaluateError = ref<string>('')
|
||||
|
||||
/** 提交多策略组合级 WF 样本外验证后台任务并轮询（N 槽位 × N 窗，较慢）。 */
|
||||
async function runMultiStrategyWalkforward(req: MultiStrategyBacktestRequest, nWindows = 7) {
|
||||
multiWfRunning.value = true
|
||||
multiWfError.value = ''
|
||||
multiWfResult.value = null
|
||||
try {
|
||||
const { task_id } = await submitMultiStrategyWalkforwardTask(req, nWindows)
|
||||
const body = await pollTask<{ walkforward: WalkForwardResult }>(
|
||||
task_id,
|
||||
300_000,
|
||||
'组合 WF 验证',
|
||||
)
|
||||
multiWfResult.value = body.walkforward
|
||||
} catch (e) {
|
||||
multiWfError.value = formatError(e)
|
||||
multiWfResult.value = null
|
||||
} finally {
|
||||
multiWfRunning.value = false
|
||||
}
|
||||
}
|
||||
|
||||
/** 提交多策略组合级一条龙评估后台任务并轮询（组合回测+WF+适配性+评分+基准对比）。 */
|
||||
async function runMultiStrategyEvaluate(req: MultiStrategyBacktestRequest) {
|
||||
multiEvaluateRunning.value = true
|
||||
multiEvaluateError.value = ''
|
||||
multiEvaluateResult.value = null
|
||||
try {
|
||||
const { task_id } = await submitMultiStrategyEvaluateTask(req)
|
||||
multiEvaluateResult.value = await pollTask<EvaluateReport>(
|
||||
task_id,
|
||||
600_000,
|
||||
'组合一条龙评估',
|
||||
)
|
||||
} catch (e) {
|
||||
multiEvaluateError.value = formatError(e)
|
||||
multiEvaluateResult.value = null
|
||||
} finally {
|
||||
multiEvaluateRunning.value = false
|
||||
}
|
||||
}
|
||||
|
||||
function clearMultiStrategyExtraAnalysis() {
|
||||
multiWfResult.value = null
|
||||
multiWfError.value = ''
|
||||
multiEvaluateResult.value = null
|
||||
multiEvaluateError.value = ''
|
||||
}
|
||||
|
||||
// ── 参数网格寻优（Phase 4） ─────────────────────────────────────────────
|
||||
const optimizeResult = ref<OptimizeResult | null>(null)
|
||||
const optimizeRunning = ref(false)
|
||||
@@ -385,6 +446,12 @@ export const useBacktestStore = defineStore('backtest', () => {
|
||||
portfolioEvaluateError,
|
||||
multiStrategyResult,
|
||||
multiStrategyRunning,
|
||||
multiWfResult,
|
||||
multiWfRunning,
|
||||
multiWfError,
|
||||
multiEvaluateResult,
|
||||
multiEvaluateRunning,
|
||||
multiEvaluateError,
|
||||
optimizeResult,
|
||||
optimizeRunning,
|
||||
optimizeContext,
|
||||
@@ -413,6 +480,9 @@ export const useBacktestStore = defineStore('backtest', () => {
|
||||
clearPortfolioExtraAnalysis,
|
||||
runMultiStrategy,
|
||||
clearMultiStrategy,
|
||||
runMultiStrategyWalkforward,
|
||||
runMultiStrategyEvaluate,
|
||||
clearMultiStrategyExtraAnalysis,
|
||||
runOptimize,
|
||||
runOptimizeAll,
|
||||
setOptimizeContext,
|
||||
|
||||
@@ -6,20 +6,24 @@
|
||||
import { computed, nextTick, onMounted, ref } from 'vue'
|
||||
import { useRouter } from 'vue-router'
|
||||
|
||||
import AiInterpretModal from '../components/AiInterpretModal.vue'
|
||||
import EquityChart from '../components/EquityChart.vue'
|
||||
import EvaluatePanel from '../components/EvaluatePanel.vue'
|
||||
import GradeDetails from '../components/GradeDetails.vue'
|
||||
import MetricTable from '../components/MetricTable.vue'
|
||||
import PortfolioCompareChart from '../components/PortfolioCompareChart.vue'
|
||||
import PortfolioSummaryTable from '../components/PortfolioSummaryTable.vue'
|
||||
import WalkForwardPanel from '../components/WalkForwardPanel.vue'
|
||||
import {
|
||||
deleteSavedStrategy,
|
||||
fetchSavedStrategies,
|
||||
formatError,
|
||||
saveStrategy,
|
||||
} from '../api'
|
||||
import { gradePortfolio } from '../grading'
|
||||
import { buildPortfolioAiPrompt } from '../aiPrompt'
|
||||
import { GRADE_META, gradePortfolio } from '../grading'
|
||||
import { detectMarket } from '../market'
|
||||
import type { MultiStrategyItem, Performance, SavedStrategy } from '../types'
|
||||
import type { Category, MultiStrategyItem, Performance, SavedStrategy } from '../types'
|
||||
import { useBacktestStore } from '../stores/backtest'
|
||||
|
||||
const router = useRouter()
|
||||
@@ -380,8 +384,9 @@ const holdingViews = computed<HoldingView[]>(() =>
|
||||
}),
|
||||
)
|
||||
|
||||
// 组合整体绩效（19 项指标）。后端 total_performance 现含完整指标，转成
|
||||
// MetricTable 需要的 Performance 类型（缺失字段补 0 兜底，保证渲染不崩）。
|
||||
// 组合整体绩效（25 项指标，与单标的 MetricTable 同口径）。后端
|
||||
// total_performance 含完整指标，转成 MetricTable 需要的 Performance 类型
|
||||
// （缺失字段补 0 兜底，保证老结果渲染不崩）。
|
||||
const comboPerf = computed<Performance | null>(() => {
|
||||
const tp = store.multiStrategyResult?.total_performance
|
||||
if (!tp) return null
|
||||
@@ -409,6 +414,12 @@ const comboPerf = computed<Performance | null>(() => {
|
||||
max_loss: get('max_loss'),
|
||||
avg_holding_days: get('avg_holding_days'),
|
||||
volatility: get('volatility'),
|
||||
ulcer_index: get('ulcer_index'),
|
||||
var_95: get('var_95'),
|
||||
cvar_95: get('cvar_95'),
|
||||
sqn: get('sqn'),
|
||||
max_consecutive_wins: get('max_consecutive_wins'),
|
||||
max_consecutive_losses: get('max_consecutive_losses'),
|
||||
}
|
||||
})
|
||||
|
||||
@@ -417,6 +428,69 @@ const comboPerf = computed<Performance | null>(() => {
|
||||
const comboGrade = computed(() =>
|
||||
store.multiStrategyResult ? gradePortfolio(store.multiStrategyResult) : null,
|
||||
)
|
||||
|
||||
// ── 组合附加分析：组合级 Walk-Forward / 一条龙 / AI 解读 ─────────────────────
|
||||
// 与单标的/组合回测页同构；按钮在组合结果区按需触发（复用最近一次
|
||||
// 组合回测的 items/cash，区间保持不变——检验的是"这组配置"的稳定性）。
|
||||
|
||||
function multiComboRequest() {
|
||||
return { items: lastComboItems.value, cash: lastComboCash.value }
|
||||
}
|
||||
|
||||
async function onComboWf() {
|
||||
if (lastComboItems.value.length === 0) return
|
||||
await store.runMultiStrategyWalkforward(multiComboRequest())
|
||||
}
|
||||
|
||||
async function onComboEvaluate() {
|
||||
if (lastComboItems.value.length === 0) return
|
||||
await store.runMultiStrategyEvaluate(multiComboRequest())
|
||||
}
|
||||
|
||||
// AI 解读：组合版 Prompt（multi 模式：N 个策略各跑各自的原标的）
|
||||
const showAiModal = ref(false)
|
||||
|
||||
const comboStrategyLabel = computed(() =>
|
||||
lastComboItems.value.length > 0
|
||||
? `${lastComboItems.value.length} 策略组合`
|
||||
: '策略组合',
|
||||
)
|
||||
|
||||
const aiPromptText = computed(() => {
|
||||
if (!store.multiStrategyResult) return ''
|
||||
return buildPortfolioAiPrompt({
|
||||
stocks: lastComboItems.value.map(
|
||||
(it) => `${it.strategy_label || it.strategy}@${it.symbol}`,
|
||||
),
|
||||
category: (lastComboItems.value[0]?.category as Category) ?? 'DAY',
|
||||
startDate: (lastComboItems.value[0]?.start_date as string) || '',
|
||||
endDate: (lastComboItems.value[0]?.end_date as string) || isoToday(),
|
||||
strategyLabel: comboStrategyLabel.value,
|
||||
params: {},
|
||||
cash: lastComboCash.value,
|
||||
commission: 0.0003,
|
||||
slippage: 0,
|
||||
execution: 'next_open',
|
||||
result: store.multiStrategyResult,
|
||||
mode: 'multi',
|
||||
wf: store.multiWfResult,
|
||||
evaluate: store.multiEvaluateResult,
|
||||
grade: comboGrade.value,
|
||||
gradeHint: comboGrade.value ? GRADE_META[comboGrade.value.grade].hint : undefined,
|
||||
})
|
||||
})
|
||||
|
||||
/** 随解读落历史库的策略上下文（历史页「去回测」引导用） */
|
||||
const aiContext = computed(() => ({
|
||||
strategy: 'multi',
|
||||
strategy_label: comboStrategyLabel.value,
|
||||
kind: 'multi',
|
||||
symbol: lastComboItems.value.map((it) => it.symbol).join(','),
|
||||
category: (lastComboItems.value[0]?.category as string) || 'DAY',
|
||||
params: {},
|
||||
start_date: (lastComboItems.value[0]?.start_date as string) || '',
|
||||
end_date: isoToday(),
|
||||
}))
|
||||
</script>
|
||||
|
||||
<template>
|
||||
@@ -631,10 +705,30 @@ const comboGrade = computed(() =>
|
||||
· {{ store.multiStrategyResult.total_performance.total_stocks }} 个策略 ·
|
||||
总资金 {{ store.multiStrategyResult.total_performance.total_cash.toFixed(0) }}
|
||||
</span>
|
||||
<span v-if="store.multiStrategyResult && !store.multiStrategyRunning" class="combo-extra-actions">
|
||||
<button
|
||||
class="save-combo-btn"
|
||||
:disabled="store.multiWfRunning"
|
||||
title="按全部槽位日期并集切窗，每窗各策略独立回测后合成组合净值，检验跨时段稳定性"
|
||||
@click="onComboWf"
|
||||
>
|
||||
{{ store.multiWfRunning ? 'WF 验证中…' : '🔬 WF 样本外验证' }}
|
||||
</button>
|
||||
<button
|
||||
class="save-combo-btn"
|
||||
:disabled="store.multiEvaluateRunning"
|
||||
title="组合回测+组合WF+跨策略适配性体检+综合评分+等权买入持有基准对比，一份报告"
|
||||
@click="onComboEvaluate"
|
||||
>
|
||||
{{ store.multiEvaluateRunning ? '评估中…' : '📋 一条龙评估' }}
|
||||
</button>
|
||||
<button class="save-combo-btn" @click="showAiModal = true">🤖 AI 解读</button>
|
||||
<button class="save-combo-btn" @click="openSaveCombo">💾 保存为组合</button>
|
||||
</span>
|
||||
<button
|
||||
v-if="store.multiStrategyResult && !store.multiStrategyRunning"
|
||||
v-else-if="store.multiStrategyResult && store.multiStrategyRunning"
|
||||
class="save-combo-btn"
|
||||
@click="openSaveCombo"
|
||||
disabled
|
||||
>
|
||||
💾 保存为组合
|
||||
</button>
|
||||
@@ -674,6 +768,42 @@ const comboGrade = computed(() =>
|
||||
<EquityChart :equity="store.multiStrategyResult.combined_equity" />
|
||||
</div>
|
||||
|
||||
<!-- 附加分析：组合级 Walk-Forward（与单标的/组合回测页同构面板） -->
|
||||
<div
|
||||
v-if="store.multiWfRunning || store.multiWfResult || store.multiWfError"
|
||||
class="combo-chart-block"
|
||||
>
|
||||
<h4>Walk-Forward 样本外验证</h4>
|
||||
<p v-if="store.multiWfRunning" class="empty-text">
|
||||
验证中…（策略数 × 窗口数 次回测，约需十几秒）
|
||||
</p>
|
||||
<div v-else-if="store.multiWfError" class="warn-box">
|
||||
⚠ {{ store.multiWfError }}
|
||||
</div>
|
||||
<WalkForwardPanel v-else-if="store.multiWfResult" :wf="store.multiWfResult" />
|
||||
</div>
|
||||
|
||||
<!-- 附加分析：组合级一条龙评估（与单标的/组合回测页同构面板） -->
|
||||
<div
|
||||
v-if="
|
||||
store.multiEvaluateRunning || store.multiEvaluateResult || store.multiEvaluateError
|
||||
"
|
||||
class="combo-chart-block"
|
||||
>
|
||||
<h4>一条龙评估</h4>
|
||||
<p v-if="store.multiEvaluateRunning" class="empty-text">
|
||||
评估中…（组合回测 + 组合WF + 跨策略适配性 + 基准对比，可能需要一两分钟）
|
||||
</p>
|
||||
<div v-else-if="store.multiEvaluateError" class="warn-box">
|
||||
⚠ {{ store.multiEvaluateError }}
|
||||
</div>
|
||||
<EvaluatePanel
|
||||
v-else-if="store.multiEvaluateResult"
|
||||
:report="store.multiEvaluateResult"
|
||||
:grade-override="comboGrade"
|
||||
/>
|
||||
</div>
|
||||
|
||||
<div v-if="comboPerf" class="combo-chart-block">
|
||||
<h4>绩效指标</h4>
|
||||
<MetricTable :perf="comboPerf" />
|
||||
@@ -747,6 +877,16 @@ const comboGrade = computed(() =>
|
||||
</div>
|
||||
</div>
|
||||
</section>
|
||||
|
||||
<!-- AI 解读 Prompt 对话框（单标的/组合/策略组合通用组件） -->
|
||||
<AiInterpretModal
|
||||
v-if="showAiModal && store.multiStrategyResult"
|
||||
:prompt="aiPromptText"
|
||||
:filename="`AI解读_${comboStrategyLabel}.md`"
|
||||
:context="aiContext"
|
||||
tip="建议先点「WF 样本外验证」「一条龙评估」，跑完再打开此弹窗，数据会一并打包。"
|
||||
@close="showAiModal = false"
|
||||
/>
|
||||
</div>
|
||||
</template>
|
||||
|
||||
@@ -1184,6 +1324,22 @@ const comboGrade = computed(() =>
|
||||
.save-combo-btn:hover {
|
||||
background: linear-gradient(135deg, #fbbf24, #f59e0b);
|
||||
}
|
||||
.save-combo-btn:disabled {
|
||||
opacity: 0.5;
|
||||
cursor: default;
|
||||
}
|
||||
/* 组合结果区附加分析按钮组（WF/一条龙/AI 解读/保存）：内联排列，窄屏可换行 */
|
||||
.combo-extra-actions {
|
||||
display: inline-flex;
|
||||
align-items: center;
|
||||
gap: 8px;
|
||||
margin-left: 10px;
|
||||
flex-wrap: wrap;
|
||||
vertical-align: middle;
|
||||
}
|
||||
.combo-extra-actions .save-combo-btn {
|
||||
margin-left: 0;
|
||||
}
|
||||
|
||||
/* 警示条基础类（过拟合 / 免责共享） */
|
||||
.warn-box {
|
||||
|
||||
Reference in New Issue
Block a user