From b2c41c73b81f07152c3b64836ec6e3ee2708519b Mon Sep 17 00:00:00 2001 From: claude_dev Date: Sun, 2 Aug 2026 20:37:50 +0800 Subject: [PATCH] =?UTF-8?q?fix(backtest):=20CTA=20metrics=20benchmark=20?= =?UTF-8?q?=E7=BC=BA=E5=A4=B1=E9=99=8D=E7=BA=A7(=E4=B8=8D=E9=98=BB?= =?UTF-8?q?=E5=A1=9E=E6=95=B4=E7=BB=84=E6=8C=87=E6=A0=87=E5=9B=BE)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit benchmark 数据缺失时原实现 if 跳过整段 compute_metrics → {task_id}_metrics.json 不写 → 策略净值/回撤/波动图也空(不只基准图)。改为:benchmark 空时传空 Series + log WARN,compute_metrics 内部 reindex→fillna(0) 容空(基准类指标 NaN→None,策略指标正常),metrics.json 照写。效果:策略图照常出,只基准/Alpha/Beta 图空(前端 benchmark-curve/risk-series 已容错)。加测试 test_run_cta_backtest_benchmark_missing_degrades(mock read_index_daily 返空 → 验 compute_metrics 仍 called + metrics.json 仍写 + 策略指标有值)。6 测试全绿(Mac pytest)。 --- sanguo_backtest/cta_engine.py | 52 +++++++++++-------- tests/backtest/test_cta_engine.py | 85 ++++++++++++++++++++++++++++++- 2 files changed, 115 insertions(+), 22 deletions(-) diff --git a/sanguo_backtest/cta_engine.py b/sanguo_backtest/cta_engine.py index 0f854b9..c3cf247 100644 --- a/sanguo_backtest/cta_engine.py +++ b/sanguo_backtest/cta_engine.py @@ -317,32 +317,42 @@ def run_cta_backtest( bench_df = bench_df.sort_values("date") benchmark_returns = bench_df["close"].pct_change().dropna() benchmark_returns.index = pd.to_datetime(bench_df["date"].iloc[1:]) + else: + # 基准缺失：降级。compute_metrics 内部 reindex→ffill→fillna(0) 能容空 + # benchmark(基准类指标返 NaN→None,策略指标正常算),故仍执行 + metrics.json + # 照写。避免原实现整段跳过 → 策略净值/回撤/波动图也空(不只基准图)。 + benchmark_returns = pd.Series(dtype=float) + logging.warning( + "基准数据缺失,降级计算(基准/Alpha/Beta 图将空,策略指标不受影响): " + "benchmark=%s %s~%s", + benchmark_code, start_date.date(), end_date.date(), + ) - # H4: compute_metrics 内部从 daily_df["balance"] 自算 simple return， - # 不再依赖 vnpy 的 log return 列（删原三路 fallback） + # H4: compute_metrics 内部从 daily_df["balance"] 自算 simple return， + # 不再依赖 vnpy 的 log return 列（删原三路 fallback） - # Compute relative metrics - metrics_result = compute_metrics(daily_df, benchmark_returns) + # Compute relative metrics（benchmark 有无都执行；空时基准类指标 NaN→None） + metrics_result = compute_metrics(daily_df, benchmark_returns) - # Merge scalars into statistics (for API response) - statistics.update(metrics_result.scalars) + # Merge scalars into statistics (for API response) + statistics.update(metrics_result.scalars) - # Serialize series to JSON (separate file, same as equity_curve/trades) - import json - series_data = {} - for key, series in metrics_result.series.items(): - if isinstance(series, pd.Series): - series_data[key] = { - "dates": series.index.astype(str).tolist(), - "values": [None if (isinstance(x, float) and not math.isfinite(x)) else x - for x in series.tolist()] - } + # Serialize series to JSON (separate file, same as equity_curve/trades) + import json + series_data = {} + for key, series in metrics_result.series.items(): + if isinstance(series, pd.Series): + series_data[key] = { + "dates": series.index.astype(str).tolist(), + "values": [None if (isinstance(x, float) and not math.isfinite(x)) else x + for x in series.tolist()] + } - # Write metrics series to JSON file - file_dir = os.path.dirname(os.path.abspath(db_path)) - metrics_file = os.path.join(file_dir, f"{task_id}_metrics.json") - with open(metrics_file, "w") as f: - json.dump({"series": series_data}, f, indent=2) + # Write metrics series to JSON file + file_dir = os.path.dirname(os.path.abspath(db_path)) + metrics_file = os.path.join(file_dir, f"{task_id}_metrics.json") + with open(metrics_file, "w") as f: + json.dump({"series": series_data}, f, indent=2) except Exception as metrics_error: # Log but don't fail backtest if metrics calculation fails. diff --git a/tests/backtest/test_cta_engine.py b/tests/backtest/test_cta_engine.py index 375aef7..c70310f 100644 --- a/tests/backtest/test_cta_engine.py +++ b/tests/backtest/test_cta_engine.py @@ -360,4 +360,87 @@ class TestRunCtaBacktest: # Verify read_index_daily was called with hs300 code (sh000300) mock_read.assert_called_once() call_args = mock_read.call_args - assert call_args[0][0] == "sh000300", "Default benchmark should be hs300 (sh000300)" \ No newline at end of file + assert call_args[0][0] == "sh000300", "Default benchmark should be hs300 (sh000300)" + + def test_run_cta_backtest_benchmark_missing_degrades(self, temp_db_path): + """benchmark 缺失时 compute_metrics 仍执行、metrics.json 照写(降级)。 + + 根因:read_index_daily 返空时原实现 if 跳过整段 compute_metrics → + {task_id}_metrics.json 不写 → 策略净值/回撤/波动图也空(不只基准图)。 + 降级:benchmark 空时传空 Series 给 compute_metrics(内部 reindex→fillna(0), + 基准类指标返 NaN→None,策略指标正常),metrics.json 照写。 + """ + mock_strategy_class = Mock() + mock_strategy_class.__name__ = "DegBmStrategy" + + dates = pd.date_range("2024-01-01", "2024-03-31", freq="D") + daily_df = pd.DataFrame( + {"balance": [1_000_000.0 * (1.001 ** i) for i in range(len(dates))]}, + index=dates, + ) + + mock_engine = MagicMock() + mock_engine.calculate_result.return_value = daily_df + mock_engine.calculate_statistics.return_value = {"total_return": 0.15} + mock_engine.trades = {"t1": MagicMock()} # 非空 trades 避免 degenerate + mock_engine.history_data = [] # 空 history → 定寸跳过 + + mock_cfg = Mock() + mock_cfg.data_paths = {"daily_dir": "/mock/daily_dir"} + + mock_ashare = MagicMock() + mock_ashare.AShareBacktestingEngine = Mock(return_value=mock_engine) + + mock_metrics_result = Mock() + mock_metrics_result.scalars = { + "total_return": 0.15, # 策略指标(不依赖基准)有值 + "max_drawdown": -0.05, + "alpha": None, # 基准类指标 NaN→None(基准缺失) + "benchmark_return": None, + } + mock_metrics_result.series = { + "equity_curve": pd.Series([1.0, 1.05, 1.1]), + "drawdown": pd.Series([0.0, -0.01, -0.02]), + } + + mock_tzlocal = MagicMock() + mock_tzlocal.get_localzone_name = Mock(return_value="UTC") + + with patch.dict("sys.modules", { + "sanguo_backtest.ashare_engine": mock_ashare, + "tzlocal": mock_tzlocal, + "vnpy.trader.setting": MagicMock(), + }): + # read_index_daily 返空 DataFrame = benchmark 缺失 + with patch("sanguo_backtest.cta_engine.read_index_daily", return_value=pd.DataFrame()): + with patch("sanguo_backtest.cta_engine.compute_metrics", return_value=mock_metrics_result) as mock_cm: + result = run_cta_backtest( + strategy_class=mock_strategy_class, + symbol="600000", + params={}, + start="2024-01-01", + end="2024-03-31", + cfg=mock_cfg, + db_path=temp_db_path, + ) + + # 1. compute_metrics 仍被调用(降级,没因 benchmark 空跳过) + mock_cm.assert_called_once() + # 2. 传入的 benchmark_returns 是空 Series(降级标记) + benchmark_arg = mock_cm.call_args[0][1] + assert isinstance(benchmark_arg, pd.Series) + assert benchmark_arg.empty, "benchmark 缺失时应传空 Series 给 compute_metrics" + + # 3. metrics.json 仍写(策略净值/回撤/波动图数据源) + file_dir = Path(temp_db_path).parent + metrics_file = file_dir / f"{result.task_id}_metrics.json" + assert metrics_file.exists(), f"降级时 metrics.json 应照写: {metrics_file}" + with open(metrics_file) as f: + metrics_data = json.load(f) + series_keys = list(metrics_data.get("series", {}).keys()) + assert "equity_curve" in series_keys + assert "drawdown" in series_keys + + # 4. 策略指标有值(基准类 None) + assert result.statistics.get("total_return") == 0.15 + assert result.statistics.get("alpha") is None \ No newline at end of file