""" Qlib 量化投资平台 - Web 应用 数据来源:Yahoo Finance(实时,覆盖 2010 年至今) """ import warnings warnings.filterwarnings('ignore') import json import logging from datetime import datetime, timedelta import numpy as np import pandas as pd import yfinance as yf from flask import Flask, jsonify, render_template_string, request logging.basicConfig(level=logging.INFO) logger = logging.getLogger(__name__) app = Flask(__name__) # ───────────────────────────────────────────── # 股票列表 # ───────────────────────────────────────────── US_STOCK_LIST = [ {"code": "AAPL", "name": "苹果"}, {"code": "MSFT", "name": "微软"}, {"code": "NVDA", "name": "英伟达"}, {"code": "GOOGL", "name": "谷歌"}, {"code": "AMZN", "name": "亚马逊"}, {"code": "META", "name": "Meta"}, {"code": "TSLA", "name": "特斯拉"}, {"code": "BRK-B", "name": "伯克希尔"}, {"code": "JPM", "name": "摩根大通"}, {"code": "V", "name": "Visa"}, {"code": "WMT", "name": "沃尔玛"}, {"code": "XOM", "name": "埃克森美孚"}, {"code": "BABA", "name": "阿里巴巴"}, {"code": "PDD", "name": "拼多多"}, {"code": "JD", "name": "京东"}, {"code": "BIDU", "name": "百度"}, {"code": "NIO", "name": "蔚来"}, {"code": "XPEV", "name": "小鹏"}, {"code": "LI", "name": "理想"}, {"code": "SPY", "name": "标普500ETF"}, {"code": "QQQ", "name": "纳斯达克ETF"}, {"code": "GLD", "name": "黄金ETF"}, {"code": "AMD", "name": "AMD"}, {"code": "INTC", "name": "英特尔"}, {"code": "NFLX", "name": "奈飞"}, {"code": "DIS", "name": "迪士尼"}, {"code": "BA", "name": "波音"}, {"code": "KO", "name": "可口可乐"}, {"code": "MCD", "name": "麦当劳"}, {"code": "COIN", "name": "Coinbase"}, ] STOCK_LIST = [ {"code": "SH600519", "name": "贵州茅台"}, {"code": "SH600036", "name": "招商银行"}, {"code": "SH601318", "name": "中国平安"}, {"code": "SH600276", "name": "恒瑞医药"}, {"code": "SH600000", "name": "浦发银行"}, {"code": "SH601166", "name": "兴业银行"}, {"code": "SH600887", "name": "伊利股份"}, {"code": "SH601012", "name": "隆基绿能"}, {"code": "SH600030", "name": "中信证券"}, {"code": "SH600900", "name": "长江电力"}, {"code": "SZ000001", "name": "平安银行"}, {"code": "SZ000858", "name": "五粮液"}, {"code": "SZ000333", "name": "美的集团"}, {"code": "SZ300750", "name": "宁德时代"}, {"code": "SZ002415", "name": "海康威视"}, {"code": "SZ000002", "name": "万科A"}, {"code": "SZ000651", "name": "格力电器"}, {"code": "SZ002594", "name": "比亚迪"}, {"code": "SZ000568", "name": "泸州老窖"}, {"code": "SZ000725", "name": "京东方A"}, ] def is_us_stock(code: str) -> bool: """判断是否为美股代码(不含 SH/SZ 前缀)""" return not (code.startswith("SH") or code.startswith("SZ")) def to_yahoo_code(qlib_code: str) -> str: """将 Qlib 格式股票代码转为 Yahoo Finance 格式""" if qlib_code.startswith("SH"): return qlib_code[2:] + ".SS" elif qlib_code.startswith("SZ"): return qlib_code[2:] + ".SZ" # 美股直接使用原始代码 return qlib_code def fetch_stock_data(stock_code: str, start_date: str, end_date: str) -> pd.DataFrame: """从 Yahoo Finance 获取股票数据""" yahoo_code = to_yahoo_code(stock_code) try: df = yf.download( yahoo_code, start=start_date, end=end_date, auto_adjust=True, progress=False, timeout=30 ) if df.empty: return pd.DataFrame() # 展平多级列名 if isinstance(df.columns, pd.MultiIndex): df.columns = [col[0].lower() for col in df.columns] else: df.columns = [col.lower() for col in df.columns] df.index = pd.to_datetime(df.index).tz_localize(None) df = df.dropna(subset=['close']) return df except Exception as e: logger.error(f"获取 {stock_code} 数据失败: {e}") return pd.DataFrame() def compute_indicators(df: pd.DataFrame) -> dict: """计算技术指标""" close = df['close'] volume = df.get('volume', pd.Series(dtype=float)) # MA ma5 = close.rolling(5).mean() ma10 = close.rolling(10).mean() ma20 = close.rolling(20).mean() ma60 = close.rolling(60).mean() # RSI delta = close.diff() gain = delta.clip(lower=0).rolling(14).mean() loss = (-delta.clip(upper=0)).rolling(14).mean() rs = gain / loss.replace(0, np.nan) rsi = 100 - (100 / (1 + rs)) # MACD ema12 = close.ewm(span=12, adjust=False).mean() ema26 = close.ewm(span=26, adjust=False).mean() macd = ema12 - ema26 signal = macd.ewm(span=9, adjust=False).mean() hist = macd - signal def safe_list(s): return [round(v, 4) if pd.notna(v) else None for v in s] dates = [d.strftime('%Y-%m-%d') for d in df.index] return { "dates": dates, "close": safe_list(close), "open": safe_list(df.get('open', pd.Series(dtype=float))), "high": safe_list(df.get('high', pd.Series(dtype=float))), "low": safe_list(df.get('low', pd.Series(dtype=float))), "volume": safe_list(volume), "ma5": safe_list(ma5), "ma10": safe_list(ma10), "ma20": safe_list(ma20), "ma60": safe_list(ma60), "rsi": safe_list(rsi), "macd": safe_list(macd), "macd_signal": safe_list(signal), "macd_hist": safe_list(hist), } def run_backtest(df: pd.DataFrame, fast: int, slow: int, capital: float) -> dict: """运行双均线策略回测""" if df.empty or len(df) < slow + 10: return {"error": "数据不足,无法回测"} close = df['close'].values dates = [d.strftime('%Y-%m-%d') for d in df.index] ma_fast = pd.Series(close).rolling(fast).mean().values ma_slow = pd.Series(close).rolling(slow).mean().values cash = capital shares = 0.0 equity_curve = [] trades = [] position = False for i in range(len(close)): if np.isnan(ma_fast[i]) or np.isnan(ma_slow[i]): equity_curve.append(round(cash, 2)) continue # 金叉买入 if not position and i > 0 and ma_fast[i] > ma_slow[i] and ma_fast[i-1] <= ma_slow[i-1]: shares = cash / close[i] cash = 0.0 position = True trades.append({"date": dates[i], "action": "买入", "price": round(float(close[i]), 2), "shares": round(shares, 2)}) # 死叉卖出 elif position and i > 0 and ma_fast[i] < ma_slow[i] and ma_fast[i-1] >= ma_slow[i-1]: cash = shares * close[i] shares = 0.0 position = False trades.append({"date": dates[i], "action": "卖出", "price": round(float(close[i]), 2), "amount": round(cash, 2)}) total = cash + shares * close[i] equity_curve.append(round(total, 2)) # 最终平仓 final_value = cash + shares * close[-1] # 买入持有 bh_return = (close[-1] - close[0]) / close[0] * 100 # 最大回撤 eq = pd.Series(equity_curve) rolling_max = eq.cummax() drawdown = (eq - rolling_max) / rolling_max * 100 max_drawdown = drawdown.min() strategy_return = (final_value - capital) / capital * 100 return { "dates": dates, "equity_curve": equity_curve, "buy_hold_curve": [round(capital * float(close[i]) / float(close[0]), 2) for i in range(len(close))], "strategy_return": round(strategy_return, 2), "buy_hold_return": round(bh_return, 2), "max_drawdown": round(float(max_drawdown), 2), "trade_count": len(trades), "final_value": round(final_value, 2), "initial_capital": capital, "trades": trades[-20:], "start_price": round(float(close[0]), 2), "end_price": round(float(close[-1]), 2), } # ───────────────────────────────────────────── # API 路由 # ───────────────────────────────────────────── @app.route('/api/stocks') def api_stocks(): return jsonify({"stocks": STOCK_LIST}) @app.route('/api/us_stocks') def api_us_stocks(): return jsonify({"stocks": US_STOCK_LIST}) @app.route('/api/stock_data') def api_stock_data(): code = request.args.get('stock_code', 'SH600519') start = request.args.get('start_date', '2020-01-01') end = request.args.get('end_date', datetime.today().strftime('%Y-%m-%d')) df = fetch_stock_data(code, start, end) if df.empty: return jsonify({"error": f"无法获取 {code} 的数据,请检查股票代码或网络连接"}) result = compute_indicators(df) result["stock_code"] = code all_stocks = STOCK_LIST + US_STOCK_LIST result["stock_name"] = next((s["name"] for s in all_stocks if s["code"] == code), code) result["market"] = "US" if is_us_stock(code) else "CN" return jsonify(result) @app.route('/api/backtest', methods=['POST']) def api_backtest(): data = request.get_json() or {} code = data.get('stock_code', 'SH600519') start = data.get('start_date', '2015-01-01') end = data.get('end_date', datetime.today().strftime('%Y-%m-%d')) fast = int(data.get('fast_ma', 5)) slow = int(data.get('slow_ma', 20)) capital = float(data.get('capital', 100000)) df = fetch_stock_data(code, start, end) if df.empty: return jsonify({"error": f"无法获取 {code} 的数据"}) result = run_backtest(df, fast, slow, capital) result["stock_code"] = code all_stocks = STOCK_LIST + US_STOCK_LIST result["stock_name"] = next((s["name"] for s in all_stocks if s["code"] == code), code) return jsonify(result) @app.route('/api/portfolio', methods=['POST']) def api_portfolio(): data = request.get_json() or {} codes = data.get('stocks', ['SH600519', 'SH600036']) start = data.get('start_date', '2020-01-01') end = data.get('end_date', datetime.today().strftime('%Y-%m-%d')) results = {} for code in codes[:5]: df = fetch_stock_data(code, start, end) if df.empty: continue close = df['close'] all_stocks = STOCK_LIST + US_STOCK_LIST name = next((s["name"] for s in all_stocks if s["code"] == code), code) norm = (close / close.iloc[0] * 100).round(2).tolist() dates = [d.strftime('%Y-%m-%d') for d in df.index] total_ret = (close.iloc[-1] - close.iloc[0]) / close.iloc[0] * 100 days = (df.index[-1] - df.index[0]).days ann_ret = ((1 + total_ret / 100) ** (365 / max(days, 1)) - 1) * 100 ann_vol = close.pct_change().std() * np.sqrt(252) * 100 sharpe = ann_ret / ann_vol if ann_vol > 0 else 0 results[code] = { "name": name, "dates": dates, "normalized": norm, "total_return": round(float(total_ret), 2), "annual_return": round(float(ann_ret), 2), "annual_vol": round(float(ann_vol), 2), "sharpe": round(float(sharpe), 2), } return jsonify(results) # ───────────────────────────────────────────── # 前端页面 # ───────────────────────────────────────────── HTML_TEMPLATE = """ Qlib 量化投资平台

Qlib 量化投资平台

基于 Microsoft Qlib 框架 · Yahoo Finance 实时数据 · 覆盖 A股 + 美股 2010 年至今

📈

行情分析

K线图、均线、MACD、RSI 等技术指标,支持 A股 + 美股,数据实时更新

🔄

策略回测

双均线策略回测,资金曲线、最大回撤、夏普比率全面评估

💼

投资组合

多股票归一化收益对比,风险收益指标综合分析

📈 行情分析

🔄 均线策略回测

💼 投资组合分析

选择 2-5 只股票进行对比分析

""" @app.route('/') def index(): return render_template_string(HTML_TEMPLATE, stocks=STOCK_LIST, us_stocks=US_STOCK_LIST) if __name__ == '__main__': app.run(host='0.0.0.0', port=7860, debug=False)