fund-cli 2.0.0__py3-none-any.whl

fund_cli/analysis/risk.py

@@ -0,0 +1,300 @@
+"""
+风险分析引擎
+
+实现专业的风险分析功能。
+"""
+
+from typing import Any
+
+import numpy as np
+import pandas as pd
+
+from fund_cli.core.analyzer import Analyzer
+
+
+class RiskAnalyzer(Analyzer):
+    """
+    风险分析引擎
+
+    计算各类风险指标，包括：
+    - 波动率风险：年化波动率、下行波动率
+    - 回撤风险：最大回撤、平均回撤、回撤持续时间
+    - 尾部风险：VaR、CVaR、偏度、峰度
+    - 相关性风险：相关性矩阵、Beta
+    """
+
+    def __init__(
+        self,
+        confidence_level: float = 0.95,
+        periods_per_year: int = 252,
+    ):
+        """
+        初始化风险分析引擎
+
+        Args:
+            confidence_level: VaR 置信水平
+            periods_per_year: 年交易日数
+        """
+        self.confidence_level = confidence_level
+        self.periods_per_year = periods_per_year
+
+    def analyze(
+        self,
+        returns: pd.Series,
+        benchmark: pd.Series | None = None,
+        **kwargs,
+    ) -> dict[str, Any]:
+        """
+        执行风险分析
+
+        Args:
+            returns: 收益率序列
+            benchmark: 基准收益率序列（可选）
+            **kwargs: 额外参数
+
+        Returns:
+            风险分析结果字典
+        """
+        # 确保输入为 Series
+        if isinstance(returns, pd.DataFrame):
+            returns = returns.iloc[:, 0]
+
+        returns = returns.dropna()
+
+        metrics = {
+            # 波动率风险
+            "volatility_annual": self.annualized_volatility(returns) * 100,
+            "downside_volatility": self.downside_volatility(returns) * 100,
+            # 回撤风险
+            "max_drawdown": self.max_drawdown(returns) * 100,
+            "avg_drawdown": self.avg_drawdown(returns) * 100,
+            "max_drawdown_duration": self.max_drawdown_duration(returns),
+            # 尾部风险
+            "var_95": self.var(returns, 0.95) * 100,
+            "var_99": self.var(returns, 0.99) * 100,
+            "cvar_95": self.cvar(returns, 0.95) * 100,
+            "skewness": self.skewness(returns),
+            "kurtosis": self.kurtosis(returns),
+            # 分布特征
+            "best_day": returns.max() * 100 if not returns.empty else 0,
+            "worst_day": returns.min() * 100 if not returns.empty else 0,
+            "std_dev": returns.std() * 100 if not returns.empty else 0,
+        }
+
+        # 相对风险指标
+        if benchmark is not None:
+            if isinstance(benchmark, pd.DataFrame):
+                benchmark = benchmark.iloc[:, 0]
+            benchmark = benchmark.dropna()
+
+            common_dates = returns.index.intersection(benchmark.index)
+            if len(common_dates) > 0:
+                returns_aligned = returns.loc[common_dates]
+                benchmark_aligned = benchmark.loc[common_dates]
+
+                metrics["beta"] = self.beta(returns_aligned, benchmark_aligned)
+                metrics["correlation"] = self.correlation(returns_aligned, benchmark_aligned)
+                metrics["tracking_error"] = (
+                    self.tracking_error(returns_aligned, benchmark_aligned) * 100
+                )
+
+        return metrics
+
+    def get_metrics(self) -> list[str]:
+        """获取可计算的指标列表"""
+        return [
+            "volatility_annual",
+            "downside_volatility",
+            "max_drawdown",
+            "var_95",
+            "var_99",
+            "cvar_95",
+            "skewness",
+            "kurtosis",
+            "beta",
+            "correlation",
+            "tracking_error",
+        ]
+
+    # ========== 波动率计算 ==========
+
+    def annualized_volatility(self, returns: pd.Series) -> float:
+        """计算年化波动率"""
+        if returns.empty:
+            return 0.0
+        return returns.std() * np.sqrt(self.periods_per_year)
+
+    def downside_volatility(
+        self,
+        returns: pd.Series,
+        mar: float = 0.0,
+    ) -> float:
+        """
+        计算下行波动率
+
+        Args:
+            returns: 收益率序列
+            mar: 最低可接受收益率
+
+        Returns:
+            下行波动率
+        """
+        if returns.empty:
+            return 0.0
+
+        downside_returns = returns[returns < mar] - mar
+        return np.sqrt((downside_returns**2).mean()) * np.sqrt(self.periods_per_year)
+
+    # ========== 回撤计算 ==========
+
+    def max_drawdown(self, returns: pd.Series) -> float:
+        """计算最大回撤"""
+        if returns.empty:
+            return 0.0
+
+        cumulative = (1 + returns).cumprod()
+        running_max = cumulative.cummax()
+        drawdown = (cumulative - running_max) / running_max
+        return drawdown.min()
+
+    def avg_drawdown(self, returns: pd.Series) -> float:
+        """计算平均回撤"""
+        if returns.empty:
+            return 0.0
+
+        cumulative = (1 + returns).cumprod()
+        running_max = cumulative.cummax()
+        drawdown = (cumulative - running_max) / running_max
+        return drawdown.mean()
+
+    def max_drawdown_duration(self, returns: pd.Series) -> int:
+        """
+        计算最大回撤持续天数
+
+        Returns:
+            最大回撤持续天数
+        """
+        if returns.empty:
+            return 0
+
+        cumulative = (1 + returns).cumprod()
+        running_max = cumulative.cummax()
+
+        # 找到回撤期间
+        in_drawdown = cumulative < running_max
+
+        if not in_drawdown.any():
+            return 0
+
+        # 计算最长回撤期
+        drawdown_periods = []
+        current_period = 0
+
+        for is_dd in in_drawdown:
+            if is_dd:
+                current_period += 1
+            else:
+                if current_period > 0:
+                    drawdown_periods.append(current_period)
+                current_period = 0
+
+        if current_period > 0:
+            drawdown_periods.append(current_period)
+
+        return max(drawdown_periods) if drawdown_periods else 0
+
+    # ========== 尾部风险计算 ==========
+
+    def var(
+        self,
+        returns: pd.Series,
+        confidence: float = 0.95,
+    ) -> float:
+        """
+        计算风险价值 (Value at Risk)
+
+        Args:
+            returns: 收益率序列
+            confidence: 置信水平
+
+        Returns:
+            VaR 值（负数表示损失）
+        """
+        if returns.empty:
+            return 0.0
+
+        return np.percentile(returns, (1 - confidence) * 100)
+
+    def cvar(
+        self,
+        returns: pd.Series,
+        confidence: float = 0.95,
+    ) -> float:
+        """
+        计算条件风险价值 (Conditional VaR / Expected Shortfall)
+
+        Args:
+            returns: 收益率序列
+            confidence: 置信水平
+
+        Returns:
+            CVaR 值
+        """
+        if returns.empty:
+            return 0.0
+
+        var = self.var(returns, confidence)
+        return returns[returns <= var].mean()
+
+    def skewness(self, returns: pd.Series) -> float:
+        """计算偏度"""
+        if returns.empty:
+            return 0.0
+        return returns.skew()
+
+    def kurtosis(self, returns: pd.Series) -> float:
+        """计算峰度"""
+        if returns.empty:
+            return 0.0
+        return returns.kurtosis()
+
+    # ========== 相对风险计算 ==========
+
+    def beta(
+        self,
+        returns: pd.Series,
+        benchmark: pd.Series,
+    ) -> float:
+        """计算 Beta"""
+        if returns.empty or benchmark.empty:
+            return 0.0
+
+        covariance = returns.cov(benchmark)
+        variance = benchmark.var()
+
+        if variance == 0:
+            return 0.0
+
+        return covariance / variance
+
+    def correlation(
+        self,
+        returns: pd.Series,
+        benchmark: pd.Series,
+    ) -> float:
+        """计算相关系数"""
+        if returns.empty or benchmark.empty:
+            return 0.0
+        return returns.corr(benchmark)
+
+    def tracking_error(
+        self,
+        returns: pd.Series,
+        benchmark: pd.Series,
+    ) -> float:
+        """计算跟踪误差"""
+        if returns.empty or benchmark.empty:
+            return 0.0
+
+        excess_returns = returns - benchmark
+        return excess_returns.std() * np.sqrt(self.periods_per_year)

fund_cli/cli.py

@@ -0,0 +1,98 @@
+"""
+Fund CLI 主入口
+
+专业基金分析命令行工具的主程序入口。
+"""
+
+import typer
+from rich.console import Console
+
+from fund_cli import __version__
+
+# 创建主应用
+app = typer.Typer(
+    name="fund",
+    help="专业基金分析CLI工具 - 面向机构客户",
+    add_completion=True,
+    rich_markup_mode="rich",
+)
+
+# 全局控制台实例
+console = Console()
+
+
+def version_callback(value: bool) -> None:
+    """显示版本信息"""
+    if value:
+        console.print(f"[bold blue]Fund CLI[/bold blue] 版本: {__version__}")
+        raise typer.Exit()
+
+
+@app.callback()
+def main(
+    version: bool | None = typer.Option(
+        None,
+        "--version",
+        "-v",
+        help="显示版本信息",
+        callback=version_callback,
+        is_eager=True,
+    ),
+) -> None:
+    """
+    Fund CLI - 专业基金分析命令行工具
+
+    面向机构客户的基金分析工具，支持基金筛选、业绩分析、组合优化等功能。
+
+    使用 --help 查看各子命令的详细帮助。
+    """
+    pass
+
+
+# 导入子命令模块
+from fund_cli.commands import (  # noqa: E402
+    ai_cmd,
+    analyze_cmd,
+    compare_cmd,
+    config_cmd,
+    data_cmd,
+    filter_cmd,
+    holding_cmd,
+    interactive_cmd,
+    manager_cmd,
+    monitor_cmd,
+    optimize_cmd,
+)
+
+# 注册子命令
+app.add_typer(filter_cmd.app, name="filter")
+app.add_typer(analyze_cmd.app, name="analyze")
+app.add_typer(compare_cmd.app, name="compare")
+app.add_typer(optimize_cmd.app, name="optimize")
+app.add_typer(monitor_cmd.app, name="monitor")
+app.add_typer(data_cmd.app, name="data")
+app.add_typer(config_cmd.app, name="config")
+app.add_typer(ai_cmd.app, name="ai")
+app.add_typer(holding_cmd.app, name="holding")
+app.add_typer(manager_cmd.app, name="manager")
+app.add_typer(interactive_cmd.app, name="interactive")
+
+
+# 直接调用的命令
+@app.command("info")
+def info(
+    fund_code: str = typer.Argument(..., help="基金代码（6位数字）"),
+) -> None:
+    """
+    查看基金基础信息。
+
+    示例:
+        fund info 000001
+    """
+    from fund_cli.commands.analyze_cmd import info_fund
+
+    info_fund(fund_code)
+
+
+if __name__ == "__main__":
+    app()

fund_cli/commands/__init__.py

@@ -0,0 +1,9 @@
+"""CLI命令模块 - 筛选、分析、对比、优化、监控、数据、配置、AI"""
+
+from fund_cli.commands.analyze_cmd import app as analyze_app
+from fund_cli.commands.compare_cmd import app as compare_app
+from fund_cli.commands.config_cmd import app as config_app
+from fund_cli.commands.data_cmd import app as data_app
+from fund_cli.commands.filter_cmd import app as filter_app
+
+__all__ = ["filter_app", "analyze_app", "compare_app", "data_app", "config_app"]