main.py

import json
import sys
import os

# 添加当前目录到Python路径
sys.path.append(os.path.dirname(os.path.abspath(__file__)))

from data_fetcher import get_stock_data, convert_to_backtrader_data
from grid_strategy import GridStrategy
from backtest_engine import BacktestEngine
from visualization import YieldVisualization
import backtrader as bt
import pandas as pd

def load_config(config_file='config.json'):
    """
    加载配置文件

    参数:
    config_file: 配置文件路径

    返回:
    dict: 配置参数字典
    """
    try:
        with open(config_file, 'r', encoding='utf-8') as f:
            config = json.load(f)
        return config
    except Exception as e:
        print(f"加载配置文件时出错: {e}")
        return {}

def main():
    # 加载配置
    config = load_config('config.json')

    if not config:
        print("无法加载配置文件，程序退出")
        return

    print("配置加载成功:")
    for key, value in config.items():
        print(f"  {key}: {value}")

    # 获取股票数据
    print("\n正在获取股票数据...")
    stock_data = get_stock_data(
        symbol=config['symbol'],
        start_date=config['start_date'],
        end_date=config['end_date'],
        period=config['period']
    )

    if stock_data is None or stock_data.empty:
        print("获取股票数据失败，程序退出")
        return

    print(f"成功获取股票数据，共 {len(stock_data)} 条记录")

    # 转换为Backtrader数据格式
    bt_data = convert_to_backtrader_data(stock_data)

    if bt_data is None:
        print("转换数据格式失败，程序退出")
        return

    # 获取基准股票数据
    print("\n正在获取基准股票数据...")
    benchmark_data = get_stock_data(
        symbol=config['benchmark_symbol'],
        start_date=config['start_date'],
        end_date=config['end_date'],
        period=config['period']
    )

    if benchmark_data is None or benchmark_data.empty:
        print("获取基准股票数据失败")
        benchmark_bt_data = None
    else:
        print(f"成功获取基准股票数据，共 {len(benchmark_data)} 条记录")
        benchmark_bt_data = convert_to_backtrader_data(benchmark_data)

    # 初始化回测引擎
    print("\n正在初始化回测引擎...")
    engine = BacktestEngine()

    # 添加策略
    engine.add_strategy(
        GridStrategy,
        grid_num=config['grid_num'],
        grid_spacing=config['grid_spacing'],
        allocation_ratio=config['allocation_ratio']
    )

    # 添加数据
    engine.add_data(bt_data)

    # 设置初始资金
    engine.set_initial_cash(config['initial_cash'])

    # 添加基准数据（如果有）
    if benchmark_bt_data is not None:
        # 注意：这里需要修改BacktestEngine以支持添加基准数据
        print("基准数据已添加（注意：实际实现中需要修改BacktestEngine以完全支持基准对比）")

    # 运行回测
    print("\n正在运行回测...")
    results = engine.run()

    if not results:
        print("回测运行失败，程序退出")
        return

    print("回测运行完成")

    # 获取回测结果
    result_dict = engine.get_results(results)

    print("\n回测结果:")
    total_return = result_dict.get('total_return', 0)
    annual_return = result_dict.get('annual_return', 0)
    sharpe_ratio = result_dict.get('sharpe_ratio', 0)
    max_drawdown = result_dict.get('max_drawdown', 0)

    print(f"  总收益率: {total_return:.4f}" if total_return is not None else "  总收益率: N/A")
    print(f"  年化收益率: {annual_return:.4f}" if annual_return is not None else "  年化收益率: N/A")
    print(f"  夏普比率: {sharpe_ratio:.4f}" if sharpe_ratio is not None else "  夏普比率: N/A")
    print(f"  最大回撤: {max_drawdown:.4f}" if max_drawdown is not None else "  最大回撤: N/A")

    # 如果有基准数据，计算基准收益率
    benchmark_returns = None
    if benchmark_data is not None and hasattr(benchmark_data, 'close'):
        # 计算基准收益率
        benchmark_returns = (benchmark_data['close'] / benchmark_data['close'].iloc[0]) - 1
        print(f"\n基准股票({config['benchmark_symbol']})累计收益率: {benchmark_returns.iloc[-1]:.4f}")

    # 可视化结果
    print("\n正在生成可视化图表...")
    viz = YieldVisualization()

    # 计算策略收益率
    value_history = result_dict.get('value_history', [])
    if value_history and len(value_history) > 0:
        strategy_returns = [(v / config['initial_cash']) - 1 for v in value_history]
    else:
        # 如果没有价值历史，创建一个简单的模拟序列
        strategy_returns = [0.0] * len(stock_data) if len(stock_data) > 0 else [0.0] * 100

    # 创建收益率对比图
    if benchmark_returns is not None and len(benchmark_returns) > 0:
        # 确保两个序列长度一致
        min_len = min(len(strategy_returns), len(benchmark_returns))
        if min_len > 0:
            strategy_returns_trimmed = strategy_returns[:min_len]
            benchmark_returns_trimmed = benchmark_returns[:min_len]
        else:
            strategy_returns_trimmed = [0.0] * 100
            benchmark_returns_trimmed = [0.0] * 100

        fig = viz.create_yield_comparison_chart(
            strategy_returns_trimmed,
            benchmark_returns_trimmed.values,
            "网格策略",
            f"基准({config['benchmark_symbol']})"
        )
    else:
        fig = viz.create_yield_comparison_chart(
            strategy_returns,
            [0] * len(strategy_returns),
            "网格策略",
            "基准(无数据)"
        )

    # 添加交互功能
    viz.add_interactive_features()

    # 保存图表
    viz.save_chart("yield_comparison.html")

    print("\n程序运行完成，结果已保存到 yield_comparison.html")

if __name__ == "__main__":
    main()