quantcli 0.1.14__tar.gz → 0.1.15__tar.gz

{quantcli-0.1.14/quantcli.egg-info → quantcli-0.1.15}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: quantcli
-Version: 0.1.14
+Version: 0.1.15
 Summary: 面向AI的多因子量化选股策略挖掘工具，AI Agent 友好 CLI
 Author-email: QuantCLI Team <quantcli@example.com>
 Project-URL: repository, https://github.com/wumu2013/quantcli

{quantcli-0.1.14 → quantcli-0.1.15}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "quantcli"
-version = "0.1.14"
+version = "0.1.15"
 description = "面向AI的多因子量化选股策略挖掘工具，AI Agent 友好 CLI"
 readme = "README.md"
 requires-python = ">=3.9"

{quantcli-0.1.14 → quantcli-0.1.15}/quantcli/core/data.py

@@ -21,10 +21,7 @@ from pathlib import Path
 import pandas as pd
 
 from ..datasources import create_datasource, DataSource, StockInfo
-from ..utils import (
-    get_logger, project_root, ensure_dir, format_size,
-    parse_date, today
-)
+from ..utils import get_logger, project_root, ensure_dir, format_size, parse_date, today
 
 logger = get_logger(__name__)
 
@@ -32,6 +29,7 @@ logger = get_logger(__name__)
 @dataclass
 class DataConfig:
     """数据管理配置"""
+
     source: str = "akshare"
     cache_dir: str = "./data"
     parallel: int = 4
@@ -54,6 +52,7 @@ class DataConfig:
 @dataclass
 class DataQualityReport:
     """数据质量报告"""
+
     total_rows: int
     null_count: Dict[str, int]
     outlier_count: Dict[str, int]
@@ -113,12 +112,7 @@ class DataManager:
     # =============================================================================
 
     def get_daily(
-        self,
-        symbol: str,
-        start_date,
-        end_date,
-        use_cache: bool = True,
-        **kwargs
+        self, symbol: str, start_date, end_date, use_cache: bool = True, **kwargs
     ) -> pd.DataFrame:
         """获取单只股票日线数据"""
         cache_file = self.cache_dir / "raw" / "stock_daily" / f"{symbol}.parquet"
@@ -151,7 +145,7 @@ class DataManager:
         start_date,
         end_date,
         use_cache: bool = True,
-        parallel: bool = False
+        parallel: bool = False,
     ) -> pd.DataFrame:
         """批量获取多只股票日线数据"""
         if not symbols:
@@ -161,6 +155,7 @@ class DataManager:
 
         if parallel and len(symbols) > 1:
             from concurrent.futures import ThreadPoolExecutor
+
             with ThreadPoolExecutor(max_workers=self.config.parallel) as executor:
                 futures = [
                     executor.submit(self.get_daily, s, start_date, end_date, use_cache)
@@ -192,10 +187,7 @@ class DataManager:
         return result
 
     def get_universe(
-        self,
-        universe: str = "all",
-        market: str = "all",
-        use_cache: bool = True
+        self, universe: str = "all", market: str = "all", use_cache: bool = True
     ) -> List[StockInfo]:
         """获取股票池"""
         cache_file = self.cache_dir / "cache" / f"stock_list_{market}.json"
@@ -221,9 +213,7 @@ class DataManager:
         return stocks
 
     def get_trading_calendar(
-        self,
-        exchange: str = "SSE",
-        use_cache: bool = True
+        self, exchange: str = "SSE", use_cache: bool = True
     ) -> List:
         """获取交易日历"""
         cache_file = self.cache_dir / "cache" / f"trading_calendar_{exchange}.json"
@@ -244,11 +234,7 @@ class DataManager:
         return trading_days
 
     def get_index_daily(
-        self,
-        symbol: str,
-        start_date,
-        end_date,
-        use_cache: bool = True
+        self, symbol: str, start_date, end_date, use_cache: bool = True
     ) -> pd.DataFrame:
         """获取指数日线数据"""
         cache_file = self.cache_dir / "raw" / "index_daily" / f"{symbol}.parquet"
@@ -268,7 +254,9 @@ class DataManager:
 
         return df
 
-    def _filter_universe(self, stocks: List[StockInfo], universe: str) -> List[StockInfo]:
+    def _filter_universe(
+        self, stocks: List[StockInfo], universe: str
+    ) -> List[StockInfo]:
         """根据股票池名称过滤股票"""
         return stocks
 
@@ -303,8 +291,10 @@ class DataManager:
             try:
                 existing = pd.read_parquet(cache_file)
                 df_copy = pd.concat([existing, df_copy]).drop_duplicates(
-                    subset=["date"] if "symbol" not in df_copy.columns else ["date", "symbol"],
-                    keep="last"
+                    subset=["date"]
+                    if "symbol" not in df_copy.columns
+                    else ["date", "symbol"],
+                    keep="last",
                 )
             except Exception:
                 pass
@@ -313,35 +303,62 @@ class DataManager:
 
         self._cache_metadata[cache_key] = {
             "last_updated": datetime.now().isoformat(),
-            "rows": len(df_copy)
+            "rows": len(df_copy),
         }
         self._save_metadata()
 
         logger.debug(f"Cached {len(df_copy)} rows to {cache_file}")
 
-    def clear_cache(self, older_than: Optional[int] = None, pattern: Optional[str] = None) -> int:
+    def clear_cache(
+        self, older_than: Optional[int] = None, pattern: Optional[str] = None
+    ) -> int:
         """清理缓存"""
         count = 0
-        raw_dir = self.cache_dir / "raw"
+
+        # 清理所有缓存子目录
+        cache_subdirs = ["raw", "prices", "stocklist", "calendar"]
 
         if older_than:
             from datetime import timedelta
+
             cutoff = today() - timedelta(days=older_than)
-            for cache_file in raw_dir.rglob("*.parquet"):
-                mtime = parse_date(cache_file.stat().st_mtime)
-                if mtime < cutoff:
-                    cache_file.unlink()
-                    count += 1
+            for subdir in cache_subdirs:
+                dir_path = self.cache_dir / subdir
+                if dir_path.exists():
+                    for cache_file in dir_path.rglob("*.parquet"):
+                        try:
+                            mtime = parse_date(cache_file.stat().st_mtime)
+                            if mtime < cutoff:
+                                cache_file.unlink()
+                                count += 1
+                        except OSError:
+                            continue
         elif pattern:
-            for cache_file in raw_dir.rglob(pattern):
-                if cache_file.is_file():
-                    cache_file.unlink()
-                    count += 1
+            for subdir in cache_subdirs:
+                dir_path = self.cache_dir / subdir
+                if dir_path.exists():
+                    for cache_file in dir_path.rglob(pattern):
+                        if cache_file.is_file():
+                            try:
+                                cache_file.unlink()
+                                count += 1
+                            except OSError:
+                                continue
         else:
-            for cache_file in raw_dir.rglob("*.parquet"):
-                cache_file.unlink()
-                count += 1
-            ensure_dir(self.cache_dir / "features")
+            # 清理所有缓存目录
+            for subdir in cache_subdirs:
+                dir_path = self.cache_dir / subdir
+                if dir_path.exists():
+                    for cache_file in dir_path.rglob("*.parquet"):
+                        try:
+                            cache_file.unlink()
+                            count += 1
+                        except OSError:
+                            continue
+
+            # 确保目录结构存在
+            for subdir in cache_subdirs:
+                ensure_dir(self.cache_dir / subdir)
 
         logger.info(f"Cleared {count} cache files")
         return count
@@ -354,7 +371,13 @@ class DataManager:
                 relative = str(path.relative_to(self.cache_dir))
                 sizes[relative] = format_size(path.stat().st_size)
 
-        sizes["_total"] = format_size(sum(path.stat().st_size for path in self.cache_dir.rglob("*.parquet") if path.is_file()))
+        sizes["_total"] = format_size(
+            sum(
+                path.stat().st_size
+                for path in self.cache_dir.rglob("*.parquet")
+                if path.is_file()
+            )
+        )
         return sizes
 
     # =============================================================================
@@ -369,7 +392,7 @@ class DataManager:
         filter_st: Optional[bool] = None,
         filter_suspended: Optional[bool] = None,
         filter_new: Optional[int] = None,
-        inplace: bool = False
+        inplace: bool = False,
     ) -> pd.DataFrame:
         """数据清洗"""
         if not inplace:
@@ -381,7 +404,7 @@ class DataManager:
             outlier_count={},
             st_filtered=0,
             suspended_filtered=0,
-            new_stock_filtered=0
+            new_stock_filtered=0,
         )
 
         # 空值填充
@@ -404,14 +427,20 @@ class DataManager:
             report.st_filtered = original_len - len(df)
 
         # 过滤停牌
-        filter_suspended_flag = filter_suspended if filter_suspended is not None else self.config.filter_suspended
+        filter_suspended_flag = (
+            filter_suspended
+            if filter_suspended is not None
+            else self.config.filter_suspended
+        )
         if filter_suspended_flag:
             original_len = len(df)
             df = self._filter_suspended(df)
             report.suspended_filtered = original_len - len(df)
 
         # 过滤新股
-        filter_new_days = filter_new if filter_new is not None else self.config.filter_new
+        filter_new_days = (
+            filter_new if filter_new is not None else self.config.filter_new
+        )
         if filter_new_days and filter_new_days > 0:
             original_len = len(df)
             df = self._filter_new_stocks(df, filter_new_days)
@@ -506,7 +535,10 @@ class DataManager:
             return {
                 **status,
                 "cache": self.get_cache_size(),
-                "config": {"source": self.config.source, "cache_dir": str(self.cache_dir)}
+                "config": {
+                    "source": self.config.source,
+                    "cache_dir": str(self.cache_dir),
+                },
             }
         except Exception as e:
             return {"status": "error", "error": str(e)}

{quantcli-0.1.14 → quantcli-0.1.15/quantcli.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: quantcli
-Version: 0.1.14
+Version: 0.1.15
 Summary: 面向AI的多因子量化选股策略挖掘工具，AI Agent 友好 CLI
 Author-email: QuantCLI Team <quantcli@example.com>
 Project-URL: repository, https://github.com/wumu2013/quantcli