mirror of
https://ghfast.top/https://github.com/aeroxw/easy-tdx.git
synced 2026-09-12 19:14:16 +08:00
GitHub
00b8c6374a
排查发现用户反馈的"回测统计数据缺失/异常"并非服务器连接问题， 而是回测引擎与组合优化器自身的代码缺陷： #23: 首根 bar 访问 close[-1] 崩溃 _SeriesAccessor 负向越界改返回 NaN（不抛 IndexError）； BacktestEngine 新增 warmup_bars 参数跳过指标预热期。 #25-A: FactorWeightedOptimizer 权重坍缩 n_stocks=2 且得分接近时，减最小值把低分标的权重压到 ~6e-8， 等于单股满仓、n_stocks 被无视，进而出现持仓1只/-99.98%回撤。 新增 _apply_weight_floor 权重下限保证入选标的都有实质权重。 #25-B: RebalanceEngine total_trades 统计错误 total_trades = len(equity_curve)（天数）改为 len(trades_df)（真实笔数）。 #22: 绩效别名键 + 数据异常诊断 performance dict 新增 sharpe_ratio/start_cash/end_value 别名键； 资金曲线异常时返回 diagnostic_warning 而非静默全 0，CLI 显示提示。 文档：README 加入回测手册导航；backtest_usage.md 补充 warmup 说明。 测试：新增 6 个回归测试，更新 3 个；932 passed。
542 lines
17 KiB
Python
542 lines
17 KiB
Python
"""测试策略基类和数据代理。"""
|
||
|
||
from __future__ import annotations
|
||
|
||
import numpy as np
|
||
import pandas as pd
|
||
import pytest
|
||
|
||
from easy_tdx.backtest.strategy import (
|
||
Strategy,
|
||
StrategyDataProxy,
|
||
_SeriesAccessor,
|
||
crossover,
|
||
)
|
||
|
||
# ── 辅助函数 ─────────────────────────────────────────────────────────────────────
|
||
|
||
|
||
def _make_df(n: int = 20, seed: int = 42) -> pd.DataFrame:
|
||
"""构造随机 OHLCV DataFrame。
|
||
|
||
Args:
|
||
n: K线数量
|
||
seed: 随机种子
|
||
|
||
Returns:
|
||
DataFrame with columns: datetime, open, close, high, low, vol, amount
|
||
"""
|
||
rng = np.random.default_rng(seed)
|
||
base = 10.0
|
||
prices = base + rng.random(n) * 5
|
||
|
||
df = pd.DataFrame(
|
||
{
|
||
"datetime": pd.date_range("2024-01-01", periods=n, freq="D"),
|
||
"open": prices,
|
||
"close": prices + rng.random(n) - 0.5,
|
||
"high": prices + rng.random(n),
|
||
"low": prices - rng.random(n),
|
||
"vol": rng.integers(1000, 10000, n),
|
||
"amount": rng.integers(100000, 1000000, n),
|
||
}
|
||
)
|
||
# 确保 high >= max(open, close), low <= min(open, close)
|
||
df["high"] = df[["open", "close", "high"]].max(axis=1)
|
||
df["low"] = df[["open", "close", "low"]].min(axis=1)
|
||
return df
|
||
|
||
|
||
def _make_df_with_extras(n: int = 20) -> pd.DataFrame:
|
||
"""构造带额外列的 DataFrame（MACD_DIF, BOLL_UPPER）。
|
||
|
||
Args:
|
||
n: K线数量
|
||
|
||
Returns:
|
||
DataFrame with standard OHLCV columns + MACD_DIF, BOLL_UPPER
|
||
"""
|
||
df = _make_df(n)
|
||
rng = np.random.default_rng(42)
|
||
|
||
# 添加额外列
|
||
df["MACD_DIF"] = rng.random(n) * 2 - 1
|
||
df["BOLL_UPPER"] = df["close"] + rng.random(n) * 2
|
||
|
||
return df
|
||
|
||
|
||
# ── TestSeriesAccessor ────────────────────────────────────────────────────────────
|
||
|
||
|
||
class TestSeriesAccessor:
|
||
"""测试 _SeriesAccessor 数据访问器。"""
|
||
|
||
def test_current_value(self) -> None:
|
||
"""测试获取当前值 [0]。"""
|
||
arr = np.array([1.0, 2.0, 3.0, 4.0, 5.0])
|
||
acc = _SeriesAccessor(arr, bar_index=2)
|
||
assert acc[0] == 3.0
|
||
|
||
def test_previous_value(self) -> None:
|
||
"""测试获取前一根 [-1]。"""
|
||
arr = np.array([1.0, 2.0, 3.0, 4.0, 5.0])
|
||
acc = _SeriesAccessor(arr, bar_index=2)
|
||
assert acc[-1] == 2.0
|
||
|
||
def test_previous_two(self) -> None:
|
||
"""测试获取前两根 [-2]。"""
|
||
arr = np.array([1.0, 2.0, 3.0, 4.0, 5.0])
|
||
acc = _SeriesAccessor(arr, bar_index=2)
|
||
assert acc[-2] == 1.0
|
||
|
||
def test_index_out_of_bounds_negative(self) -> None:
|
||
"""测试索引越界（负方向）返回 NaN，而非抛 IndexError。
|
||
|
||
回测早期 bar_index=0 时 close[-1] 等回溯访问不应崩溃（见 issue #23），
|
||
返回 NaN 让策略自然跳过预热期。
|
||
"""
|
||
arr = np.array([1.0, 2.0, 3.0, 4.0, 5.0])
|
||
acc = _SeriesAccessor(arr, bar_index=0)
|
||
val = acc[-1]
|
||
assert np.isnan(val)
|
||
|
||
def test_len(self) -> None:
|
||
"""测试 __len__ 返回数组长度。"""
|
||
arr = np.array([1.0, 2.0, 3.0])
|
||
acc = _SeriesAccessor(arr, bar_index=1)
|
||
assert len(acc) == 3
|
||
|
||
def test_array_conversion(self) -> None:
|
||
"""测试 __array__ 转为 numpy 数组。"""
|
||
arr = np.array([1.0, 2.0, 3.0])
|
||
acc = _SeriesAccessor(arr, bar_index=1)
|
||
result = np.asarray(acc)
|
||
np.testing.assert_array_equal(result, arr)
|
||
|
||
def test_raw_property(self) -> None:
|
||
"""测试 raw 属性返回原始数组。"""
|
||
arr = np.array([1.0, 2.0, 3.0])
|
||
acc = _SeriesAccessor(arr, bar_index=1)
|
||
np.testing.assert_array_equal(acc.raw, arr)
|
||
|
||
|
||
# ── TestStrategyDataProxy ─────────────────────────────────────────────────────────
|
||
|
||
|
||
class TestStrategyDataProxy:
|
||
"""测试 StrategyDataProxy 数据代理。"""
|
||
|
||
def test_basic_columns(self) -> None:
|
||
"""测试标准 OHLCV 列访问。"""
|
||
df = _make_df(n=10)
|
||
proxy = StrategyDataProxy(df)
|
||
proxy._set_index(5)
|
||
|
||
assert isinstance(proxy.open[0], float)
|
||
assert isinstance(proxy.close[0], float)
|
||
assert isinstance(proxy.high[0], float)
|
||
assert isinstance(proxy.low[0], float)
|
||
assert isinstance(proxy.vol[0], float)
|
||
assert isinstance(proxy.amount[0], float)
|
||
|
||
def test_previous_bar(self) -> None:
|
||
"""测试访问前一根 bar 的数据。"""
|
||
df = _make_df(n=10)
|
||
proxy = StrategyDataProxy(df)
|
||
proxy._set_index(5)
|
||
|
||
# close[0] 应该等于 df["close"].iloc[5]
|
||
assert proxy.close[0] == pytest.approx(df["close"].iloc[5])
|
||
# close[-1] 应该等于 df["close"].iloc[4]
|
||
assert proxy.close[-1] == pytest.approx(df["close"].iloc[4])
|
||
|
||
def test_extra_columns_via_getattr(self) -> None:
|
||
"""测试通过 __getattr__ 访问额外列。"""
|
||
df = _make_df_with_extras(n=10)
|
||
proxy = StrategyDataProxy(df)
|
||
proxy._set_index(5)
|
||
|
||
# MACD_DIF 和 BOLL_UPPER 应该可以访问
|
||
assert isinstance(proxy.MACD_DIF[0], float)
|
||
assert isinstance(proxy.BOLL_UPPER[0], float)
|
||
|
||
# 验证值正确
|
||
assert proxy.MACD_DIF[0] == pytest.approx(df["MACD_DIF"].iloc[5])
|
||
assert proxy.BOLL_UPPER[0] == pytest.approx(df["BOLL_UPPER"].iloc[5])
|
||
|
||
def test_missing_column_raises(self) -> None:
|
||
"""测试访问不存在的列抛出 AttributeError。"""
|
||
df = _make_df(n=10)
|
||
proxy = StrategyDataProxy(df)
|
||
proxy._set_index(5)
|
||
|
||
with pytest.raises(AttributeError, match="列 'NONEXISTENT' 不存在"):
|
||
_ = proxy.NONEXISTENT[0]
|
||
|
||
|
||
# ── TestCrossover ───────────────────────────────────────────────────────────────
|
||
|
||
|
||
class TestCrossover:
|
||
"""测试 crossover 金叉检测函数。"""
|
||
|
||
def test_crossover_true(self) -> None:
|
||
"""测试金叉检测（a 上穿 b）。"""
|
||
a = np.array([1, 2, 3, 4, 5])
|
||
b = np.array([5, 4, 3, 2, 1])
|
||
mask = crossover(a, b)
|
||
|
||
# 索引 3 处：a[2]=3 <= b[2]=3, a[3]=4 > b[3]=2 → 金叉
|
||
assert mask[3]
|
||
# 其他位置无金叉
|
||
assert not mask[0]
|
||
assert not mask[1]
|
||
assert not mask[2]
|
||
assert not mask[4]
|
||
|
||
def test_crossover_false_no_cross(self) -> None:
|
||
"""测试无金叉情况（a 全在 b 下方）。"""
|
||
a = np.array([1, 2, 3, 4, 5])
|
||
b = np.array([6, 7, 8, 9, 10])
|
||
mask = crossover(a, b)
|
||
|
||
# 全部 False
|
||
assert not np.any(mask)
|
||
|
||
def test_crossover_series(self) -> None:
|
||
"""测试接受 pd.Series 参数。"""
|
||
a = pd.Series([1, 2, 3, 4, 5])
|
||
b = pd.Series([5, 4, 3, 2, 1])
|
||
mask = crossover(a, b)
|
||
|
||
assert mask[3]
|
||
|
||
def test_crossover_with_accessor(self) -> None:
|
||
"""测试接受 _SeriesAccessor 参数。"""
|
||
arr_a = np.array([1, 2, 3, 4, 5])
|
||
arr_b = np.array([5, 4, 3, 2, 1])
|
||
acc_a = _SeriesAccessor(arr_a, bar_index=4)
|
||
acc_b = _SeriesAccessor(arr_b, bar_index=4)
|
||
|
||
mask = crossover(acc_a, acc_b)
|
||
assert mask[3]
|
||
|
||
|
||
# ── TestStrategyBase ──────────────────────────────────────────────────────────────
|
||
|
||
|
||
class TestStrategyBase:
|
||
"""测试 Strategy 策略基类。"""
|
||
|
||
def test_subclass_init_and_next(self) -> None:
|
||
"""测试子类 init() 和 next() 被调用。"""
|
||
|
||
class SimpleStrategy(Strategy):
|
||
init_called = False
|
||
next_called_count = 0
|
||
|
||
def init(self) -> None:
|
||
SimpleStrategy.init_called = True
|
||
|
||
def next(self) -> None:
|
||
SimpleStrategy.next_called_count += 1
|
||
|
||
df = _make_df(n=10)
|
||
strategy = SimpleStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
assert SimpleStrategy.init_called is True
|
||
|
||
# 模拟引擎遍历所有 bar
|
||
for i in range(len(df)):
|
||
strategy._set_bar_index(i)
|
||
strategy._call_next()
|
||
|
||
assert SimpleStrategy.next_called_count == 10
|
||
|
||
def test_buy_sell_recording(self) -> None:
|
||
"""测试 buy() 和 sell() 创建 Signal。"""
|
||
|
||
class SignalStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
if self._bar_index == 5:
|
||
self.buy(size=100, price=10.0, stop_loss=9.0, take_profit=11.0)
|
||
elif self._bar_index == 8:
|
||
self.sell(size=100, price=10.5)
|
||
|
||
df = _make_df(n=10)
|
||
strategy = SignalStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
for i in range(len(df)):
|
||
strategy._set_bar_index(i)
|
||
strategy._call_next()
|
||
|
||
signals = strategy._clear_signals()
|
||
|
||
assert len(signals) == 2
|
||
assert signals[0].direction == "BUY"
|
||
assert signals[0].size == 100
|
||
assert signals[0].price == 10.0
|
||
assert signals[0].stop_loss == 9.0
|
||
assert signals[0].take_profit == 11.0
|
||
|
||
assert signals[1].direction == "SELL"
|
||
assert signals[1].size == 100
|
||
assert signals[1].price == 10.5
|
||
|
||
def test_I_registers_indicator(self) -> None:
|
||
"""测试 self.I() 注册指标。"""
|
||
from easy_tdx.MyTT import MA
|
||
|
||
class IndicatorStrategy(Strategy):
|
||
def init(self) -> None:
|
||
self.ma5 = self.I(MA, self.data.close, 5)
|
||
self.ma20 = self.I(MA, self.data.close, 20)
|
||
|
||
def next(self) -> None:
|
||
pass
|
||
|
||
df = _make_df(n=50)
|
||
strategy = IndicatorStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
# 验证指标长度正确（MA 会产生 nan，但长度不变）
|
||
assert len(strategy.ma5) == 50
|
||
assert len(strategy.ma20) == 50
|
||
|
||
# 验证指标已注册
|
||
assert "MA" in strategy._indicators
|
||
|
||
def test_full_position_buy(self) -> None:
|
||
"""测试全仓买入（size=0）。"""
|
||
|
||
class FullPositionStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
if self._bar_index == 5:
|
||
self.buy(size=0) # 全仓
|
||
|
||
df = _make_df(n=10)
|
||
strategy = FullPositionStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
for i in range(len(df)):
|
||
strategy._set_bar_index(i)
|
||
strategy._call_next()
|
||
|
||
signals = strategy._clear_signals()
|
||
|
||
assert len(signals) == 1
|
||
assert signals[0].direction == "BUY"
|
||
assert signals[0].size == 0 # 0 表示全仓
|
||
|
||
def test_data_property(self) -> None:
|
||
"""测试 self.data 属性返回 StrategyDataProxy。"""
|
||
df = _make_df(n=10)
|
||
|
||
class DataAccessStrategy(Strategy):
|
||
def init(self) -> None:
|
||
# 验证 data 可访问
|
||
assert hasattr(self.data, "close")
|
||
assert hasattr(self.data, "open")
|
||
|
||
def next(self) -> None:
|
||
# 验证 next() 中也可访问
|
||
assert self.data.close[0] > 0
|
||
|
||
strategy = DataAccessStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
for i in range(len(df)):
|
||
strategy._set_bar_index(i)
|
||
strategy._call_next()
|
||
|
||
def test_position_property(self) -> None:
|
||
"""测试 self.position 属性。"""
|
||
df = _make_df(n=10)
|
||
|
||
class PositionStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
# 初始状态无持仓
|
||
assert self.position["size"] == 0.0
|
||
|
||
strategy = PositionStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
for i in range(len(df)):
|
||
strategy._set_bar_index(i)
|
||
strategy._call_next()
|
||
|
||
def test_get_datetime(self) -> None:
|
||
"""测试 _get_datetime() 方法。"""
|
||
df = _make_df(n=10)
|
||
|
||
class DateTimeStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
# 验证 datetime 正确（YYYYMMDD 格式）
|
||
dt = self._get_datetime()
|
||
assert isinstance(dt, int)
|
||
assert 20240101 <= dt <= 20241231
|
||
|
||
strategy = DateTimeStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
for i in range(len(df)):
|
||
strategy._set_bar_index(i)
|
||
strategy._call_next()
|
||
|
||
def test_clear_signals(self) -> None:
|
||
"""测试 _clear_signals() 清空信号列表。"""
|
||
|
||
class ClearSignalStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
self.buy(size=100)
|
||
|
||
df = _make_df(n=10)
|
||
strategy = ClearSignalStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
# 第一轮
|
||
for i in range(len(df)):
|
||
strategy._set_bar_index(i)
|
||
strategy._call_next()
|
||
signals1 = strategy._clear_signals()
|
||
assert len(signals1) == 10
|
||
|
||
# 第二轮（无新信号）
|
||
signals2 = strategy._clear_signals()
|
||
assert len(signals2) == 0
|
||
|
||
def test_buy_without_bind_raises(self) -> None:
|
||
"""测试未绑定数据时调用 buy() 抛出错误。"""
|
||
|
||
class BadStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
self.buy(size=100)
|
||
|
||
strategy = BadStrategy()
|
||
# 未调用 _bind_data()
|
||
|
||
with pytest.raises(RuntimeError, match="策略未绑定数据"):
|
||
strategy._call_init()
|
||
strategy._set_bar_index(0)
|
||
strategy._call_next()
|
||
|
||
def test_chanlun_property(self) -> None:
|
||
"""测试 chanlun 属性（预留）。"""
|
||
df = _make_df(n=10)
|
||
|
||
class ChanlunStrategy(Strategy):
|
||
def init(self) -> None:
|
||
assert self.chanlun is None
|
||
|
||
def next(self) -> None:
|
||
pass
|
||
|
||
strategy = ChanlunStrategy()
|
||
strategy._bind_data(df)
|
||
strategy._call_init()
|
||
|
||
for i in range(len(df)):
|
||
strategy._set_bar_index(i)
|
||
strategy._call_next()
|
||
|
||
|
||
class TestWarmupAndLookback:
|
||
"""issue #23: close[-1] 在首根 bar 不应崩溃；warmup 期不产生信号。"""
|
||
|
||
def test_lookback_negative_returns_nan_at_bar_zero(self) -> None:
|
||
"""_SeriesAccessor 负向越界返回 NaN（非 IndexError）。"""
|
||
arr = np.array([1.0, 2.0, 3.0, 4.0, 5.0])
|
||
acc = _SeriesAccessor(arr, bar_index=0)
|
||
assert np.isnan(acc[-1])
|
||
assert np.isnan(acc[-2])
|
||
|
||
def test_engine_no_crash_on_close_minus_one(self) -> None:
|
||
"""文档示例：next() 里访问 close[-1]/close[-2] 不应抛 IndexError。
|
||
|
||
回归 issue #23：DualMAStrategy 在首根 bar 访问 close[-1] 崩溃。
|
||
"""
|
||
from easy_tdx.backtest.engine import BacktestEngine
|
||
|
||
class LookbackStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
# 文档记录的访问方式
|
||
_ = self.data.close[0]
|
||
_ = self.data.close[-1]
|
||
_ = self.data.close[-2]
|
||
|
||
df = _make_df(n=50)
|
||
engine = BacktestEngine(LookbackStrategy, cash=100000)
|
||
# 修复前：抛 IndexError；修复后：正常跑完
|
||
result = engine.run(df)
|
||
assert len(result.equity_curve) == 50
|
||
|
||
def test_warmup_bars_skips_early_next(self) -> None:
|
||
"""warmup_bars=N 时前 N 根不调用 next()、不产生信号。"""
|
||
from easy_tdx.backtest.engine import BacktestEngine
|
||
|
||
next_bars: list[int] = []
|
||
|
||
class TrackingStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
next_bars.append(self._bar_index)
|
||
self.buy(size=100)
|
||
|
||
df = _make_df(n=20)
|
||
engine = BacktestEngine(TrackingStrategy, cash=100000, warmup_bars=5)
|
||
engine.run(df)
|
||
# warmup 期（bar 0~4）不被调用
|
||
assert next_bars == list(range(5, 20))
|
||
|
||
def test_warmup_bars_default_zero_backward_compat(self) -> None:
|
||
"""默认 warmup_bars=0：每根 bar 都调用 next()（向后兼容）。"""
|
||
from easy_tdx.backtest.engine import BacktestEngine
|
||
|
||
next_count = 0
|
||
|
||
class CountStrategy(Strategy):
|
||
def init(self) -> None:
|
||
pass
|
||
|
||
def next(self) -> None:
|
||
nonlocal next_count
|
||
next_count += 1
|
||
|
||
df = _make_df(n=15)
|
||
BacktestEngine(CountStrategy, cash=100000).run(df)
|
||
assert next_count == 15
|