#!/usr/bin/env python3 """ 专业交易员看板 v5 — 三层架构：宏观指数 → 板块分析 → 个股分析 增强：关注功能、板块详情、K线演示、个股详情优化 """ import json, os, math, datetime, time from collections import Counter, defaultdict try: from pypinyin import lazy_pinyin, Style HAS_PINYIN = True except ImportError: HAS_PINYIN = False def safe_float(v): if isinstance(v, float): if math.isinf(v): return 99.99 if v > 0 else 0.0 if math.isnan(v): return 0.0 return v def sanitize_data(obj): if isinstance(obj, float): return safe_float(obj) elif isinstance(obj, dict): return {k: sanitize_data(v) for k, v in obj.items()} elif isinstance(obj, list): return [sanitize_data(i) for i in obj] return obj def safe_json_dumps(obj, **kw): return json.dumps(sanitize_data(obj), **kw) def safe_str(v): v = safe_float(v) if isinstance(v, float): return str(round(v, 2)) return str(v) def _num(v): try: return float(v) except (ValueError, TypeError): return 0.0 BASE_DIR = os.path.dirname(os.path.abspath(__file__)) WORK_DIR = os.path.join(BASE_DIR, "data") OUTPUT_DIR = BASE_DIR RESULT_FILE = os.path.join(WORK_DIR, "scan_results_filtered.json") INDUSTRY_FILE = os.path.join(WORK_DIR, "stock_industry_concepts_new.json") BACKTEST_FILE = os.path.join(WORK_DIR, "backtest_results.json") KLINE_FILE = os.path.join(WORK_DIR, "kline_data.json") OUTPUT_HTML = os.path.join(OUTPUT_DIR, "ashare-dashboard.html") with open(RESULT_FILE, "r") as f: data = json.load(f) industry_data = {} if os.path.exists(INDUSTRY_FILE): with open(INDUSTRY_FILE, "r") as f: industry_data = json.load(f).get("stock_info", {}) # ===== 产业链上下游映射（幻方量化·板块轮动推演）===== # 将行业归入 上游(资源/原料) / 中游(制造/加工) / 下游(应用/消费) / 服务(金融/交运) 四个环节 SECTOR_STAGE = { "煤炭开采": "上游", "焦炭Ⅱ": "上游", "贵金属": "上游", "小金属": "上游", "工业金属": "上游", "化学原料": "上游", "农化制品": "上游", "普钢": "中游", "特钢Ⅱ": "中游", "化学制品": "中游", "化学制药": "中游", "生物制品": "中游", "中药Ⅱ": "中游", "电力": "中游", "燃气Ⅱ": "中游", "通用设备": "中游", "汽车零部件": "中游", "通信设备": "中游", "航空装备Ⅱ": "中游", "食品加工": "中游", "白酒Ⅱ": "中游", "房地产开发": "下游", "白色家电": "下游", "消费电子": "下游", "医疗器械": "下游", "专业连锁Ⅱ": "下游", "影视娱乐": "下游", "证券Ⅱ": "金融", "多元金融": "金融", "港口航运": "交运", "航运港口": "交运", "铁路公路": "交运", "航空机场": "交运", } # 金融服务链 / 交运服务链细分标签（二级细分业务） SECTOR_SUB = { "证券Ⅱ": "券商·投行·互联网金融", "多元金融": "信托·保险·期货·AMC", "港口航运": "港口码头·装卸仓储", "航运港口": "港口枢纽·航运物流", "铁路公路": "铁路运输·公路货运", "航空机场": "机场运营·航空客运", } # 产业链上下游传导关系（有向边 source → target） CHAIN_EDGES = [ ("煤炭开采", "电力"), ("煤炭开采", "普钢"), ("焦炭Ⅱ", "普钢"), ("焦炭Ⅱ", "特钢Ⅱ"), ("化学原料", "化学制品"), ("化学原料", "化学制药"), ("化学制品", "化学制药"), ("化学制品", "医疗器械"), ("工业金属", "通用设备"), ("工业金属", "白色家电"), ("工业金属", "消费电子"), ("小金属", "航空装备Ⅱ"), ("小金属", "消费电子"), ("普钢", "汽车零部件"), ("普钢", "房地产开发"), ("普钢", "白色家电"), ("特钢Ⅱ", "汽车零部件"), ("特钢Ⅱ", "航空装备Ⅱ"), ("通用设备", "房地产开发"), ("通信设备", "消费电子"), ("食品加工", "专业连锁Ⅱ"), ("白酒Ⅱ", "专业连锁Ⅱ"), # 交运服务链细分（陆路集疏运 → 港口 → 航运） ("铁路公路", "港口航运"), ("铁路公路", "航运港口"), ("港口航运", "航运港口"), # 金融服务链 → 实体传导（融资支持） ("证券Ⅱ", "房地产开发"), ("多元金融", "房地产开发"), ] # 产业链每条边的传导逻辑说明（source→target: 传导含义 + 交易推演） CHAIN_EDGE_NOTES = { ("煤炭开采", "电力"): "煤价是火电最大成本项（约占50-60%）。煤涨电未涨=电力毛利被压缩（利空电力）；煤跌电稳=电力盈利修复。", ("煤炭开采", "普钢"): "焦煤占炼钢成本约30%。煤涨推高钢厂成本，钢价能否顺价决定钢企毛利，顺价成功则共振上涨。", ("焦炭Ⅱ", "普钢"): "焦炭是炼钢第二大原料。焦炭提涨落地=成本推动钢价；钢厂减产则焦炭需求回落（反向压制）。", ("焦炭Ⅱ", "特钢Ⅱ"): "特钢对焦炭价格敏感度低于普钢，成本传导更依赖高端产品提价能力，观察订单价差。", ("化学原料", "化学制品"): "基础化工品（纯碱/烯烃等）是制品的直接原料。原料涨→制品涨价时滞约1-2季度，先压制后传导。", ("化学原料", "化学制药"): "原料药（API）成本受基础化工影响。原料涨压制制剂毛利，具备提价权的创新药企免疫。", ("化学制品", "化学制药"): "中间体→原料药→制剂的供应链传导。中间体涨价利好一体化药企，利空纯制剂厂商。", ("化学制品", "医疗器械"): "医用耗材（高分子材料）成本传导。集采品种无法顺价=成本上涨直接侵蚀利润。", ("工业金属", "通用设备"): "铜铝占机械制造直接材料成本约20-40%。金属涨价时，有订单储备的龙头可顺价，小厂承压。", ("工业金属", "白色家电"): "铜占家电成本约15-20%。铜涨空调涨不了价=家电毛利收缩；铜跌=盈利弹性释放。", ("工业金属", "消费电子"): "铜箔/铝壳用于消费电子。金属涨价对低端机型压力大，高端机型（苹果链）议价能力强。", ("小金属", "航空装备Ⅱ"): "钛材/高温合金是军航核心材料。小金属涨价利好军工材料环节（稀土/钛材），总装环节影响小。", ("小金属", "消费电子"): "稀土磁材用于声学/马达。涨价利好磁材股，成本压力由整机厂消化。", ("普钢", "汽车零部件"): "钢材占零部件成本约40-60%。钢涨时给主机厂供货的 Tier1 年度锁价，毛利被动收缩。", ("普钢", "房地产开发"): "螺纹钢是建筑主体用钢。地产开工强=钢需强（钢→地产是需求验证方向）。", ("普钢", "白色家电"): "板材（冷轧）用于家电外壳。钢涨压制家电毛利，尤其白电价格战时期无法顺价。", ("特钢Ⅱ", "汽车零部件"): "特钢（齿轮钢/轴承钢）决定零部件品质。特钢涨=高端零部件成本上行，技术壁垒高的企业可顺价。", ("特钢Ⅱ", "航空装备Ⅱ"): "高温合金/高强钢是军机结构件核心。军品定价成本加成模式下，特钢涨价可部分传导。", ("通用设备", "房地产开发"): "工程机械/电梯需求滞后地产开工约6-12个月。地产放松先利好设备订单预期。", ("通信设备", "消费电子"): "通信模组/射频前端是手机核心部件。5G换机周期启动=通信设备与消费电子共振。", ("食品加工", "专业连锁Ⅱ"): "上游食材成本决定餐饮/商超毛利。原料降价=连锁餐饮盈利弹性（逆向受益）。", ("白酒Ⅱ", "专业连锁Ⅱ"): "白酒是商超/烟酒店核心利润品。白酒批价坚挺=渠道利润厚，连锁动销活跃。", ("铁路公路", "港口航运"): "铁路疏港是港口集疏运一环。大宗运量上行时铁路与港口吞吐量同向共振。", ("铁路公路", "航运港口"): "公铁联运直接对接港口枢纽。货运量改善同时利好两端（量价共振）。", ("港口航运", "航运港口"): "同类枢纽（不同细分口径），吞吐量与费率同步变动，观察集装箱/散货结构。", ("证券Ⅱ", "房地产开发"): "地产融资（债券/股权）经券商投行落地。地产政策松=券商投行收入+地产融资改善双击。", ("多元金融", "房地产开发"): "信托/AMC参与地产纾困与并购融资。地产风险出清阶段，多元金融业务量放大。", } backtest_data = {} market_trend = None if os.path.exists(BACKTEST_FILE): with open(BACKTEST_FILE, "r") as f: bt_full = json.load(f) backtest_data = bt_full.get("backtest_results", {}) market_trend = bt_full.get("market_trend", None) bt_signals = bt_full.get("trading_signals", {}) bt_buy_signals = bt_full.get("buy_signals", []) bt_sell_signals = bt_full.get("sell_signals", []) backtest_date = bt_full.get("backtest_date", "") else: bt_signals = {}; bt_buy_signals = []; bt_sell_signals = []; backtest_date = "" kline_data = {} if os.path.exists(KLINE_FILE): with open(KLINE_FILE, "r") as f: kline_data = json.load(f).get("stock_data", {}) BULLISH_PATTERNS = { "底部反转": ["出水芙蓉", "三金叉见底", "价托", "DIF上穿零线", "底部芝麻功", "熊牛转换点"], "趋势确认": ["多方炮", "老鸭头", "均线多头排列", "东方红大阳升", "进入2+3区间"], "突破加速": ["放量过头", "轻松过头", "巨量长阳", "会创新高还有新高"], "洗盘结束": ["底部芝麻功", "天眼地量", "压转托"], } BEARISH_PATTERNS = { "见顶信号": ["断头铡刀", "三死叉见顶", "价压", "DIF下穿零线"], "趋势恶化": ["空方炮", "均线空头排列", "五线顺下", "倒挂老鸭头"], "破位下跌": ["一阴下穿三角压", "死蜘蛛", "会创新低还有新低", "牛熊转换点"], "主力出货": ["量压", "阳阴墓碑", "托转压"], } def analyze_trading_signals(signals): signal_names = set(s["indicator"] for s in signals) bull = sum(1 for s in signals if s["type"] == "看涨") bear = sum(1 for s in signals if s["type"] == "看跌") sentiment = round((bull - bear) / max(len(signals), 1), 2) matched_bull = [{"pattern": n, "signals": [s for s in ps if s in signal_names]} for n, ps in BULLISH_PATTERNS.items() if any(s in signal_names for s in ps)] matched_bear = [{"pattern": n, "signals": [s for s in ps if s in signal_names]} for n, ps in BEARISH_PATTERNS.items() if any(s in signal_names for s in ps)] confidence = 50 if sentiment >= 0: confidence += len(matched_bull) * 12 + bull * 2 - len(matched_bear) * 8 else: confidence += len(matched_bear) * 12 + bear * 2 - len(matched_bull) * 8 confidence = max(0, min(100, confidence)) has_conflict = len(matched_bull) > 0 and len(matched_bear) > 0 risk = "高" if (has_conflict or confidence < 35) else ("低" if (confidence >= 70 and abs(sentiment) >= 0.5) else "中") priority = 5 if (sentiment >= 0.6 and confidence >= 75 and len(signals) >= 10) else (4 if (sentiment >= 0.4 and confidence >= 65 and len(signals) >= 8) else (3 if (sentiment >= 0.2 and confidence >= 55) else (2 if confidence >= 45 else 1))) if sentiment >= 0.6 and priority >= 4: action, detail = "强烈关注", "多重看涨信号共振，建议重点关注突破确认后的介入机会" elif sentiment >= 0.3 and priority >= 3: action, detail = "重点关注", "看涨信号占优，建议观察量能配合，择机关注" elif sentiment <= -0.3 and priority >= 3: action, detail = "注意风险", "看跌信号主导，持仓者注意止损，未持仓者建议规避" elif sentiment <= -0.6: action, detail = "建议规避", "多重看跌信号共振，坚决回避" elif abs(sentiment) < 0.2: action, detail = "观望等待", "多空信号交织，建议等待方向明确后再操作" else: action, detail = "适度关注", "信号偏多但不够强烈，建议持续跟踪" return {"action": action, "action_detail": detail, "confidence": confidence, "risk": risk, "priority": priority, "sentiment_score": sentiment, "matched_bull": matched_bull, "matched_bear": matched_bear, "has_conflict": has_conflict, "pattern_descs": [f'{m["pattern"]}: {",".join(m["signals"])}' for m in (matched_bull + matched_bear)]} indicator_counts = Counter() signal_dates = Counter() stocks_data = [] for s in data["top50"]: bull = sum(1 for sig in s["signals"] if sig["type"] == "看涨") bear = sum(1 for sig in s["signals"] if sig["type"] == "看跌") sentiment = round((bull - bear) / max(s["signal_count"], 1), 2) if sentiment >= 0.6: classification, cls_key = "强势看涨", "strong_bull" elif sentiment >= 0.2: classification, cls_key = "偏多", "bull" elif sentiment > -0.2: classification, cls_key = "中性", "neutral" elif sentiment > -0.6: classification, cls_key = "偏空", "bear" else: classification, cls_key = "强势看跌", "strong_bear" for sig in s["signals"]: indicator_counts[sig["indicator"]] += 1 signal_dates[sig["date"]] += 1 trading = analyze_trading_signals(s["signals"]) ind_info = industry_data.get(s["code"], {}) bt = backtest_data.get(s["code"], {}) bt_sig = bt_signals.get(s["code"], {}) stocks_data.append({ "code": s["code"], "name": s["name"], "signal_count": s["signal_count"], "bullish": bull, "bearish": bear, "sentiment_score": sentiment, "classification": classification, "cls_key": cls_key, "latest_price": s["latest_price"], "pct_change": round(s["pct_change"], 2), "mkt_cap_yi": s.get("mkt_cap_yi", 0), "industry": ind_info.get("industry", "未知"), "concepts": ind_info.get("concepts", []), "signals": sorted(s["signals"], key=lambda x: x["date"]), "bt_win_rate": bt.get("overall_win_rate", 0), "bt_avg_return": bt.get("overall_avg_return", 0), "bt_has_data": bool(bt), "bt_wr_filtered": bt.get("overall_wr_filtered", 0), "bt_wr_3d": bt.get("overall_wr_3d", 0), "bt_wr_5d": bt.get("overall_wr_5d", 0), "bt_pf": bt.get("overall_pf", 0), "bt_exp": bt.get("overall_exp", 0), "bt_quant_score": bt.get("quant_score", 50), "bt_rsi": bt.get("current_rsi", 50), "bt_adx": bt.get("current_adx", 0), "bt_vol_ratio": bt.get("current_vol_ratio", 1.0), "bt_bb_pos": bt.get("current_bb_pos", 0.5), "bt_atr_pct": bt.get("current_atr_pct", 2.0), "trade_signal": bt_sig.get("signal", "观望"), "trade_reason": bt_sig.get("reason", ""), "entry_price": bt_sig.get("entry_price", 0), "stop_loss": bt_sig.get("stop_loss", 0), "target_price": bt_sig.get("target_price", 0), "position_size": bt_sig.get("position_size", 0), "wr_adjusted": bt_sig.get("wr_adjusted", 0), "wr_filtered": bt_sig.get("wr_filtered", 0), "holding_period": bt_sig.get("holding_period", "1-3天"), **trading, }) indicator_info = {} for stock in data["top50"]: for sig in stock["signals"]: if sig["indicator"] not in indicator_info: indicator_info[sig["indicator"]] = {"section": sig["section"], "desc": sig["desc"], "type": sig["type"]} sorted_indicators = sorted(indicator_counts.items(), key=lambda x: x[1], reverse=True) bullish_total = sum(s["bullish"] for s in stocks_data) bearish_total = sum(s["bearish"] for s in stocks_data) industry_dist = Counter(s["industry"] for s in stocks_data if s["industry"] != "未知") concept_dist = Counter() for s in stocks_data: for c in s["concepts"]: concept_dist[c] += 1 sector_analysis = defaultdict(lambda: {"total": 0, "bullish": 0, "bearish": 0, "buy_signals": 0, "sell_signals": 0, "avg_sentiment": 0, "avg_confidence": 0, "avg_quant_score": 0, "avg_pct_change": 0, "avg_win_rate": 0, "bt_count": 0, "stocks": [], "max_sentiment": -99, "min_sentiment": 99}) for s in stocks_data: ind = s["industry"] if ind == "未知": continue sa = sector_analysis[ind] sa["total"] += 1; sa["bullish"] += s["bullish"]; sa["bearish"] += s["bearish"] sa["avg_sentiment"] += s["sentiment_score"]; sa["avg_confidence"] += s["confidence"] sa["avg_pct_change"] += s["pct_change"] sa["max_sentiment"] = max(sa["max_sentiment"], s["sentiment_score"]) sa["min_sentiment"] = min(sa["min_sentiment"], s["sentiment_score"]) if s["trade_signal"] == "买入": sa["buy_signals"] += 1 if s["trade_signal"] == "卖出": sa["sell_signals"] += 1 if s["bt_has_data"]: sa["avg_quant_score"] += s["bt_quant_score"]; sa["avg_win_rate"] += s["bt_win_rate"]; sa["bt_count"] += 1 sa["stocks"].append({"code": s["code"], "name": s["name"], "sentiment": s["sentiment_score"], "signal_count": s["signal_count"], "mkt_cap_yi": s["mkt_cap_yi"], "trade_signal": s["trade_signal"], "bt_quant_score": s["bt_quant_score"]}) for ind in sector_analysis: n = sector_analysis[ind]["total"]; bn = max(sector_analysis[ind]["bt_count"], 1) sector_analysis[ind]["avg_sentiment"] = round(sector_analysis[ind]["avg_sentiment"] / n, 2) sector_analysis[ind]["avg_confidence"] = round(sector_analysis[ind]["avg_confidence"] / n, 0) sector_analysis[ind]["avg_pct_change"] = round(sector_analysis[ind]["avg_pct_change"] / n, 2) sector_analysis[ind]["avg_quant_score"] = round(sector_analysis[ind]["avg_quant_score"] / bn, 1) sector_analysis[ind]["avg_win_rate"] = round(sector_analysis[ind]["avg_win_rate"] / bn, 1) sector_ranking = sorted(sector_analysis.items(), key=lambda x: (x[1]["avg_sentiment"], x[1]["buy_signals"]), reverse=True) # ===== 幻方量化·板块轮动推演（多因子轮动模型）===== def _clamp(v, lo, hi): return max(lo, min(hi, v)) rotation_data = [] for name, sa in sector_analysis.items(): total = max(sa["total"], 1) momentum = sa["avg_pct_change"] sentiment = sa["avg_sentiment"] net_buy = sa["buy_signals"] - sa["sell_signals"] breadth = round(sa["bullish"] / total, 3) if total else 0.0 quant = sa["avg_quant_score"] if sa["bt_count"] > 0 else 50.0 winrate = sa["avg_win_rate"] if sa["bt_count"] > 0 else 50.0 m = _clamp(50 + momentum * 5, 0, 100) # 动量因子：当日涨跌幅 s = _clamp(50 + sentiment * 50, 0, 100) # 情绪因子 n = _clamp(50 + net_buy * 22, 0, 100) # 资金/信号净流入因子 q = _clamp(quant, 0, 100) # 量化评分因子 w = _clamp(winrate, 0, 100) # 历史胜率因子 score = round(0.28 * m + 0.22 * s + 0.16 * n + 0.20 * q + 0.14 * w, 1) if score >= 68: tier = "领跑主线" elif score >= 56: tier = "接力补涨" elif score >= 44: tier = "蓄势观察" elif score >= 34: tier = "动能衰减" else: tier = "退潮回避" rotation_data.append({ "name": name, "total": total, "stage": SECTOR_STAGE.get(name, "其他"), "sub": SECTOR_SUB.get(name, ""), "score": score, "momentum": momentum, "sentiment": sentiment, "net_buy": net_buy, "breadth": breadth, "quant": round(q, 1), "winrate": round(w, 1), "tier": tier, "buy": sa["buy_signals"], "sell": sa["sell_signals"], }) rotation_data.sort(key=lambda x: x["score"], reverse=True) lead = [r for r in rotation_data if r["tier"] == "领跑主线"] follow = [r for r in rotation_data if r["tier"] == "接力补涨"] fade = [r for r in rotation_data if r["tier"] in ("动能衰减", "退潮回避")] lead_names = "、".join(r["name"] for r in lead[:4]) or "暂无明确领跑板块" follow_names = "、".join(r["name"] for r in follow[:4]) or "暂无明确接力板块" fade_names = "、".join(r["name"] for r in fade[:4]) or "暂无明确退潮板块" top_sector = rotation_data[0] if rotation_data else None if rotation_data: direction = f"资金由[{fade_names}]向[{lead_names}]迁移，呈{'主线强化' if top_sector['score'] >= 60 else '震荡分化'}轮动格局" else: direction = "板块数据不足，暂无法推演轮动方向" rotation_insight = { "lead": lead_names, "follow": follow_names, "fade": fade_names, "direction": direction, "top_score": round(top_sector["score"], 1) if top_sector else 0, "top_name": top_sector["name"] if top_sector else "--", } sector_by_name = {r["name"]: r for r in rotation_data} chain_nodes = [] for r in rotation_data: chain_nodes.append({ "name": r["name"], "stage": r["stage"], "sub": r.get("sub", ""), "total": r["total"], "score": r["score"], "tier": r["tier"], "momentum": r["momentum"], "sentiment": r["sentiment"], "buy": r["buy"], "sell": r["sell"], }) chain_edges = [{"source": a, "target": b} for a, b in CHAIN_EDGES if a in sector_by_name and b in sector_by_name] chain_data = {"nodes": chain_nodes, "edges": chain_edges} rotation_data_json = safe_json_dumps(rotation_data, ensure_ascii=False) chain_data_json = safe_json_dumps(chain_data, ensure_ascii=False) rotation_insight_json = safe_json_dumps(rotation_insight, ensure_ascii=False) sector_stage_json = safe_json_dumps(SECTOR_STAGE, ensure_ascii=False) sector_sub_json = safe_json_dumps(SECTOR_SUB, ensure_ascii=False) chain_edge_info_json = safe_json_dumps( [{"source": a, "target": b, "note": n} for (a, b), n in CHAIN_EDGE_NOTES.items()], ensure_ascii=False) market_sentiment = round((bullish_total - bearish_total) / max(bullish_total + bearish_total, 1), 2) action_dist = Counter(s["action"] for s in stocks_data) # Build pinyin initials map for fuzzy search pinyin_map = {} if HAS_PINYIN: for s in stocks_data: name = s["name"] initials = ''.join(lazy_pinyin(name, style=Style.FIRST_LETTER)) full_pinyin = ''.join(lazy_pinyin(name)) pinyin_map[s["code"]] = initials + ' ' + full_pinyin pinyin_map_json = safe_json_dumps(pinyin_map, ensure_ascii=False) stocks_json = safe_json_dumps(stocks_data, ensure_ascii=False) indicator_info_json = safe_json_dumps(indicator_info, ensure_ascii=False) indicator_counts_json = safe_json_dumps(dict(sorted_indicators), ensure_ascii=False) industry_dist_json = safe_json_dumps(dict(industry_dist.most_common()), ensure_ascii=False) concept_dist_json = safe_json_dumps(dict(concept_dist.most_common(30)), ensure_ascii=False) sector_data_json = safe_json_dumps([{"name": k, **v} for k, v in sector_ranking], ensure_ascii=False) # ===== 需求4: 获取A股全市场板块轮动数据 ===== full_market_sectors = [] try: import requests as _req import urllib3 urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning) _sess = _req.Session() _sess.verify = False _sess.headers.update({'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36', 'Referer': 'https://quote.eastmoney.com/'}) _api_hosts = [ 'https://push2delay.eastmoney.com/api/qt/clist/get', 'https://push2.eastmoney.com/api/qt/clist/get', 'http://80.push2.eastmoney.com/api/qt/clist/get', ] _seen_bk = set() for _po in (1, 0): _params = {'pn': 1, 'pz': 100, 'po': _po, 'np': 1, 'fltt': 2, 'invt': 2, 'fid': 'f3', 'fs': 'm:90+t:2', 'fields': 'f12,f14,f3,f20,f21,f128,f136,f140'} for _url in _api_hosts: try: _resp = _sess.get(_url, params=_params, timeout=15) _data = _resp.json().get('data', {}).get('diff', []) if not _data: continue for _b in _data: _c = _b.get('f12', '') if _c in _seen_bk: continue _seen_bk.add(_c) full_market_sectors.append({ "code": _c, "name": _b.get('f14', ''), "pct_change": round(_b.get('f3', 0), 2), "amount": str(round(_b.get('f20', 0) / 1e8, 1)) + '亿' if _b.get('f20', 0) > 0 else '--', "mkt_cap_yi": round(_b.get('f21', 0) / 1e8, 1) if _b.get('f21', 0) > 0 else 0, "lead_stock": _b.get('f128', '') or '--', "lead_code": _b.get('f140', '') or '', }) break except Exception: continue full_market_sectors.sort(key=lambda x: x['pct_change'], reverse=True) print(f" 全市场板块数据: {len(full_market_sectors)} 个板块") # 抓取热门题材的成分股作为"机会个股" hot_sector_stocks = {} for _hs in full_market_sectors[:6]: _bk = _hs.get('code', '') _sn = _hs.get('name', '') if not _bk or not _sn: continue try: _mem_params = {'pn': 1, 'pz': 6, 'po': 1, 'np': 1, 'fltt': 2, 'invt': 2, 'fid': 'f3', 'fs': 'b:' + _bk, 'fields': 'f12,f14,f3,f6'} _mresp = _sess.get('https://push2delay.eastmoney.com/api/qt/clist/get', params=_mem_params, timeout=15) _mdiff = _mresp.json().get('data', {}).get('diff', []) or [] _mstocks = [] for _m in _mdiff[:5]: _mstocks.append({ "code": _m.get('f12', ''), "name": _m.get('f14', ''), "pct_change": round(_m.get('f3', 0), 2), "amount": str(round(_m.get('f6', 0) / 1e8, 1)) + '亿' if _m.get('f6', 0) > 0 else '--', }) if _mstocks: hot_sector_stocks[_sn] = _mstocks except Exception: continue print(f" 热门题材机会个股: {len(hot_sector_stocks)} 个热门题材") except Exception as _e: print(f" 全市场板块数据获取失败: {_e}") full_market_sectors = [] hot_sector_stocks = {} full_market_sectors_json = safe_json_dumps(full_market_sectors, ensure_ascii=False) hot_sector_stocks_json = safe_json_dumps(hot_sector_stocks, ensure_ascii=False) # ===== 需求2: 全市场涨跌分布数据（分页抓全市场，精确统计涨/跌/平/停牌） ===== market_breadth = {} try: _b_fs = 'm:0+t:6,m:0+t:80,m:1+t:2,m:1+t:23,m:0+t:81+s:2048' _b_hosts = [ 'https://push2delay.eastmoney.com/api/qt/clist/get', 'https://push2.eastmoney.com/api/qt/clist/get', 'http://80.push2.eastmoney.com/api/qt/clist/get', ] _b_all = [] _b_total = 0 for _b_url in _b_hosts: try: _first = _sess.get(_b_url, params={'pn': 1, 'pz': 100, 'po': 1, 'np': 1, 'fltt': 2, 'invt': 2, 'fs': _b_fs, 'fields': 'f3,f2,f292'}, timeout=20) _first_data = _first.json().get('data', {}) _b_total = _first_data.get('total', 0) _b_all = list(_first_data.get('diff', []) or []) _page = 1 while _b_total > 0 and len(_b_all) < _b_total and _page <= 100: _page += 1 _pg = _sess.get(_b_url, params={'pn': _page, 'pz': 100, 'po': 1, 'np': 1, 'fltt': 2, 'invt': 2, 'fs': _b_fs, 'fields': 'f3,f2,f292'}, timeout=20) _pg_diff = _pg.json().get('data', {}).get('diff', []) or [] if not _pg_diff: break _b_all.extend(_pg_diff) if _b_total > 0 and len(_b_all) >= int(_b_total * 0.9): break except Exception: continue _up_cnt = 0 _down_cnt = 0 _flat_cnt = 0 _suspended_cnt = 0 _delisted_cnt = 0 # 涨跌幅区间直方图（左跌→右涨，13 档） _hist = [0] * 13 def _bucket(fv): if fv == 0: return 6 if fv < 0: if fv <= -7: return 0 if fv <= -5: return 1 if fv <= -3: return 2 if fv <= -2: return 3 if fv <= -1: return 4 return 5 if fv >= 7: return 12 if fv >= 5: return 11 if fv >= 3: return 10 if fv >= 2: return 9 if fv >= 1: return 8 return 7 for _s in _b_all: _st = _s.get('f292') if _st == 7 or _st == '7': _delisted_cnt += 1 continue _v = _s.get('f3', 0) try: _fv = float(_v) except (ValueError, TypeError): _suspended_cnt += 1 continue if _fv > 0: _up_cnt += 1 elif _fv < 0: _down_cnt += 1 else: _flat_cnt += 1 _hist[_bucket(_fv)] += 1 _hist_labels = ["≤-7%", "-7~-5%", "-5~-3%", "-3~-2%", "-2~-1%", "-1~0%", "平盘", "0~1%", "1~2%", "2~3%", "3~5%", "5~7%", "≥7%"] _hist_json = [{"name": _hist_labels[i], "value": _hist[i]} for i in range(13)] _real_total = max(_b_total - _delisted_cnt, _up_cnt + _down_cnt + _flat_cnt + _suspended_cnt) market_breadth = { "total": _real_total, "up": _up_cnt, "down": _down_cnt, "flat": _flat_cnt, "suspended": _suspended_cnt, "limit_up": 0, "limit_down": 0, "hist": _hist_json, } print(f" 全市场涨跌: 总{_real_total} 涨{_up_cnt} 跌{_down_cnt} 平{_flat_cnt} 停牌{_suspended_cnt} 退市忽略{_delisted_cnt}") except Exception as _e: print(f" 全市场涨跌数据获取失败: {_e}") market_breadth = {} market_breadth_json = safe_json_dumps(market_breadth, ensure_ascii=False) # ===== 需求2: 场内T+0 ETF基金数据 ===== t0_etf_list = [] try: # T+0 ETF: 跨境ETF、债券ETF、黄金ETF、商品ETF等支持当日回转交易（货币ETF波动≈0已剔除） # fs=全市场场内基金（沪m:1+t:9 + 深m:0+t:10）按成交额降序；旧板块MK0024~MK0405只含黄金ETF+LOF _e_params = {'pn': 1, 'pz': 100, 'po': 1, 'np': 1, 'fltt': 2, 'invt': 2, 'fid': 'f6', 'fs': 'm:1+t:9,m:0+t:10', 'fields': 'f12,f14,f2,f3,f6,f8,f152'} _e_data = [] for _eurl in ('https://push2delay.eastmoney.com/api/qt/clist/get', 'https://push2.eastmoney.com/api/qt/clist/get'): try: _e_resp = _sess.get(_eurl, params=_e_params, timeout=15) _e_data = _e_resp.json().get('data', {}).get('diff', []) if _e_data: break except Exception: continue for _e in _e_data: _code = str(_e.get('f12', '')) _name = str(_e.get('f14', '')) _fv = _e.get('f3', 0) # 停牌/退市ETF的 f3 为 '-'，直接跳过 if not isinstance(_fv, (int, float)): continue _pct = round(float(_fv), 2) _amount = _num(_e.get('f6', 0)) _turnover = _num(_e.get('f8', 0)) _price = _num(_e.get('f2', 0)) _name_up = _name.upper() # 货币ETF净值化运作、日内振幅≈0，永无做T机会，排除避免霸占候选名额 if any(k in _name for k in ('货币', '添益', '日利', '理财', '保证金', '现金')): continue _etf_type = '' if '黄金' in _name or 'GOLD' in _name_up or _code.startswith('518'): _etf_type = '黄金ETF' elif '债' in _name or '短融' in _name or '国开' in _name or '国债' in _name or '政金' in _name or _code.startswith('511'): _etf_type = '债券ETF' elif '原油' in _name or '豆粕' in _name or '有色' in _name or '能源' in _name or '商品' in _name or '白银' in _name or '化工' in _name: _etf_type = '商品ETF' elif ('跨境' in _name or '纳指' in _name or '恒生' in _name or '日经' in _name or '标普' in _name or '德国' in _name or '法国' in _name or '港股' in _name or '中概' in _name or 'QDII' in _name_up or '海外' in _name or '东南亚' in _name or '沙特' in _name or '印度' in _name or '美国' in _name or _code.startswith('513') or _code.startswith('1596')): _etf_type = '跨境ETF' if _etf_type and _amount > 1e7 and _price > 0: _strategy = '' if _etf_type == '债券ETF': _strategy = '日内波段：利用债券ETF低波动特性，在盘中低点买入、高点卖出，日收益0.1-0.3%；适合大资金低风险套利' elif _etf_type == '跨境ETF': _strategy = '隔日溢价套利：关注夜间海外市场走势，次日开盘顺势操作；若溢价>2%可做空溢价回落；波动大设2%止损' elif _etf_type == '黄金ETF': _strategy = '趋势跟踪：跟随国际金价波动，RSI<30低吸、RSI>70高抛；地缘事件驱动时可加大仓位；止损3%' elif _etf_type == '商品ETF': _strategy = '事件驱动：关注大宗商品价格和库存变化，突破前日高点买入、跌破前日低点止损；趋势行情可留仓过夜' t0_etf_list.append({ "code": _code, "name": _name, "price": _price, "pct_change": _pct, "amount": str(round(_amount / 1e8, 1)) + '亿', "_amount_num": _amount, "turnover": round(_turnover, 1) if _turnover else 0, "type": _etf_type, "strategy": _strategy, }) # 类型内按成交额降序，再按类型配额挑选（跨境10+黄金3+商品3+债券4），与前端JSONP逻辑一致 _rank_t = {'跨境ETF': 0, '黄金ETF': 1, '商品ETF': 1, '债券ETF': 2} t0_etf_list.sort(key=lambda x: (_rank_t.get(x.get('type'), 3), -x.get('_amount_num', 0))) _quota = {'跨境ETF': 10, '黄金ETF': 3, '商品ETF': 3, '债券ETF': 4} _picked = {} _final = [] for _x in t0_etf_list: _t = _x.get('type') if _picked.get(_t, 0) < _quota.get(_t, 0): _picked[_t] = _picked.get(_t, 0) + 1 _final.append(_x) t0_etf_list = _final[:20] print(f" T+0 ETF: {len(t0_etf_list)} 只") except Exception as _e: print(f" T+0 ETF数据获取失败: {_e}") t0_etf_list = [] t0_etf_list_json = safe_json_dumps(t0_etf_list, ensure_ascii=False) # ===== 需求3: 涨停池数据（Python端fallback，全市场分页抓取） ===== emo_pools_data = {"zt": [], "dt": [], "zb": [], "ok": False} try: _today_yyyymmdd = datetime.datetime.now().strftime('%Y%m%d') _pool_cfg = [ ("getTopicZTPool", "fbt%3Aasc", 500, "zt"), ("getTopicDTPool", "fund%3Aasc", 200, "dt"), ("getTopicZBPool", "fbt%3Aasc", 300, "zb"), ] for _kind, _sort, _ps, _key in _pool_cfg: try: _pool, _idx = [], 0 while _idx < 8: # 最多翻8页，覆盖全市场极端行情 _u = f'https://push2ex.eastmoney.com/{_kind}?ut=7eea3edcaed734bea9cbfc24409ed989&dpt=wz.ztzt&Pageindex={_idx}&pagesize={_ps}&sort={_sort}&date={_today_yyyymmdd}' _r = _sess.get(_u, timeout=15) _d = _r.json() _p = _d.get('data', {}).get('pool', []) if not _p: break _pool.extend(_p) _tc = _d.get('data', {}).get('tc', 0) if _tc and len(_pool) >= _tc: break if len(_p) < _ps: break _idx += 1 if _pool: emo_pools_data[_key] = _pool emo_pools_data["ok"] = True except Exception: pass print(f" 涨停池: ZT={len(emo_pools_data['zt'])} DT={len(emo_pools_data['dt'])} ZB={len(emo_pools_data['zb'])}") except Exception as _e: print(f" 涨停池数据获取失败: {_e}") emo_pools_json = safe_json_dumps(emo_pools_data, ensure_ascii=False) # 大V动态情绪数据（微博52198/唐史主任司马迁/狂龙十八段） weibo_dav_data = {"ok": False, "davs": [], "overall_sentiment": "neutral", "overall_score": 0, "total_bull": 0, "total_bear": 0, "total_neutral": 0, "total_posts": 0, "fetch_time": "", "error": "数据未生成"} _dav_file = os.path.join(WORK_DIR, "weibo_dav_data.json") if os.path.exists(_dav_file): try: with open(_dav_file, "r") as f: weibo_dav_data = json.load(f) print(f" 大V情绪: {weibo_dav_data.get('overall_sentiment','N/A')} posts={weibo_dav_data.get('total_posts',0)}") except Exception as _e: print(f" 大V情绪数据读取失败: {_e}") weibo_dav_json = safe_json_dumps(weibo_dav_data, ensure_ascii=False) # ===== 经济事件日历数据 ===== import datetime as _dt _econ_events = [ # 2026年8月 {"date":"2026-08-26","time":"20:30","title":"美国7月PCE/核心PCE","cat":"宏观","importance":"high","country":"美国", "desc":"美联储锚定通胀指标，预计核心PCE环比+0.2%","forecast":"+0.2%","prior":"+0.2%", "impact":"neutral","impact_text":"若低于预期=降息预期升=利好债市+成长股；高于预期=鹰派=利空", "markets":["美股","债券","外汇"],"status":"pending"}, {"date":"2026-08-26","time":"20:30","title":"美国二季度GDP修正值","cat":"宏观","importance":"medium","country":"美国", "desc":"二季度实际GDP修正值，首次预估值+2.4%","forecast":"+2.4%","prior":"+2.4%", "impact":"neutral","impact_text":"数据若下修=经济放缓担忧=利好降息预期；上修=经济强劲=鹰派", "markets":["美股","债券","外汇"],"status":"pending"}, {"date":"2026-08-27","time":"22:00","title":"密歇根大学消费者信心终值","cat":"宏观","importance":"low","country":"美国", "desc":"8月消费者信心指数终值","forecast":"90.2","prior":"90.2", "impact":"neutral","impact_text":"消费者信心走强=消费股利好；走弱=经济衰退担忧", "markets":["美股"],"status":"pending"}, {"date":"2026-08-29","time":"20:30","title":"美国7月个人收入/支出","cat":"宏观","importance":"medium","country":"美国", "desc":"个人收入和支出数据，反映消费动能","forecast":"+0.3%","prior":"+0.3%", "impact":"neutral","impact_text":"收入增长=消费预期利好；支出放缓=经济动能减弱", "markets":["美股","债券"],"status":"pending"}, # 9月 {"date":"2026-09-01","time":"09:00","title":"中国8月官方PMI","cat":"宏观","importance":"high","country":"中国", "desc":"制造业+非制造业PMI，荣枯线50","forecast":"49.5","prior":"49.3", "impact":"bullish","impact_text":"若回升至50以上=经济企稳=利好A股周期股+港股；低于预期=政策加码预期", "markets":["A股","港股","商品"],"status":"pending"}, {"date":"2026-09-02","time":"09:45","title":"中国8月财新制造业PMI","cat":"宏观","importance":"medium","country":"中国", "desc":"财新制造业PMI初值/终值","forecast":"49.8","prior":"49.8", "impact":"neutral","impact_text":"与官方PMI互补，偏中小盘。50以上=制造业扩张", "markets":["A股","港股"],"status":"pending"}, {"date":"2026-09-05","time":"20:30","title":"美国8月非农就业","cat":"重磅","importance":"high","country":"美国", "desc":"非农就业人口变动+失业率，每月最重要的数据之一","forecast":"+18万","prior":"+20.6万", "impact":"bullish","impact_text":"就业走弱=降息预期升=利好美股科技+A股外资流入；就业超预期=鹰派=利空", "markets":["美股","债券","外汇","商品"],"status":"pending"}, {"date":"2026-09-10","time":"09:30","title":"中国8月CPI/PPI","cat":"宏观","importance":"high","country":"中国", "desc":"消费者物价指数+生产者物价指数","forecast":"CPI+0.6%","prior":"CPI+0.5%", "impact":"neutral","impact_text":"CPI回升+PPI转正=通胀预期=利好资源股；CPI走低=宽松预期=利好债市", "markets":["A股","债券","商品"],"status":"pending"}, {"date":"2026-09-11","time":"20:30","title":"美国8月CPI","cat":"重磅","importance":"high","country":"美国", "desc":"消费者物价指数同比/环比，9月议息前最重要通胀数据","forecast":"+3.4%","prior":"+3.4%", "impact":"bullish","impact_text":"CPI低于预期=9月降息预期确认=利好美股+A股+黄金+债券；超预期=利空", "markets":["美股","债券","外汇","商品","A股"],"status":"pending"}, {"date":"2026-09-12","time":"20:30","title":"美国8月PPI","cat":"宏观","importance":"medium","country":"美国", "desc":"生产者物价指数，CPI的前瞻指标","forecast":"+2.3%","prior":"+2.3%", "impact":"neutral","impact_text":"PPI低于预期=通胀降温=利好；高于预期=成本推升=利空消费", "markets":["美股","债券"],"status":"pending"}, {"date":"2026-09-15","time":"09:20","title":"中国MLF续作/逆回购","cat":"宏观","importance":"medium","country":"中国", "desc":"央行MLF操作+公开市场逆回购，关注利率变动","forecast":"待定","prior":"待定", "impact":"neutral","impact_text":"MLF降息=宽松信号=利好A股债市；缩量续作=流动性收紧", "markets":["A股","债券"],"status":"pending"}, {"date":"2026-09-16","time":"09:30","title":"中国8月经济数据","cat":"宏观","importance":"high","country":"中国", "desc":"工业增加值+社零+固定资产投资","forecast":"待定","prior":"待定", "impact":"neutral","impact_text":"数据走强=经济企稳=利好A股；走弱=政策刺激预期", "markets":["A股","港股","商品"],"status":"pending"}, {"date":"2026-09-17","time":"02:00","title":"美联储FOMC议息会议(第一天)","cat":"重磅","importance":"high","country":"美国", "desc":"美联储9月议息会议第一天","forecast":"待定","prior":"5.25-5.50%", "impact":"neutral","impact_text":"9月会议是下半年最关键会议之一。市场预计可能启动首次降息", "markets":["美股","债券","外汇","商品","A股","港股"],"status":"pending"}, {"date":"2026-09-18","time":"02:00","title":"美联储利率决议+点阵图+发布会","cat":"重磅","importance":"high","country":"美国", "desc":"利率决议+经济预测+点阵图+鲍威尔新闻发布会","forecast":"待定","prior":"5.25-5.50%", "impact":"bullish","impact_text":"若降息=利好全球股市+债券+黄金+A股外资流入；维持不变但鸽派=偏利好；鹰派=利空", "markets":["美股","债券","外汇","商品","A股","港股"],"status":"pending"}, {"date":"2026-09-20","time":"11:00","title":"日本央行利率决议","cat":"重磅","importance":"medium","country":"日本", "desc":"日本央行货币政策会议","forecast":"待定","prior":"0.25%", "impact":"neutral","impact_text":"日银加息=日元升值=全球套息交易平仓风险；维持不变=风险偏好稳定", "markets":["外汇","美股","港股"],"status":"pending"}, {"date":"2026-09-26","time":"20:30","title":"美国8月PCE/核心PCE","cat":"重磅","importance":"high","country":"美国", "desc":"美联储9月会议后的首份通胀数据，验证降息决策","forecast":"待定","prior":"待定", "impact":"neutral","impact_text":"若PCE继续回落=降息路径确认=利好；反弹=鹰派重来=利空", "markets":["美股","债券","外汇","商品"],"status":"pending"}, ] _econ_cal_data = { "ok": True, "fetch_time": _dt.datetime.now().strftime("%Y-%m-%d %H:%M"), "events": _econ_events, "summary": { "total": len(_econ_events), "high_impact": sum(1 for e in _econ_events if e["importance"] == "high"), "bullish": sum(1 for e in _econ_events if e["impact"] == "bullish"), "bearish": sum(1 for e in _econ_events if e["impact"] == "bearish"), "neutral": sum(1 for e in _econ_events if e["impact"] == "neutral"), "next_major": "9月5日 非农 / 9月11日 CPI / 9月18日 FOMC", }, } econ_cal_json = safe_json_dumps(_econ_cal_data, ensure_ascii=False) # 用涨停/跌停池数量回填涨跌分布的涨停/跌停家数 if isinstance(market_breadth, dict) and emo_pools_data.get("ok"): market_breadth["limit_up"] = len(emo_pools_data.get("zt", [])) market_breadth["limit_down"] = len(emo_pools_data.get("dt", [])) market_breadth_json = safe_json_dumps(market_breadth, ensure_ascii=False) action_dist_json = safe_json_dumps(dict(action_dist), ensure_ascii=False) buy_signals_json = safe_json_dumps(bt_buy_signals, ensure_ascii=False) sell_signals_json = safe_json_dumps(bt_sell_signals, ensure_ascii=False) market_trend_json = safe_json_dumps(market_trend or {}, ensure_ascii=False) kline_data_json = safe_json_dumps(kline_data, ensure_ascii=False) bt_has_data_count = sum(1 for s in stocks_data if s["bt_has_data"]) avg_win_rate = round(sum(s["bt_win_rate"] for s in stocks_data if s["bt_has_data"]) / max(bt_has_data_count, 1), 1) avg_wr_filtered = round(sum(s["bt_wr_filtered"] for s in stocks_data if s["bt_has_data"]) / max(bt_has_data_count, 1), 1) avg_quant_score = round(sum(s["bt_quant_score"] for s in stocks_data if s["bt_has_data"]) / max(bt_has_data_count, 1), 1) avg_pf = round(sum(safe_float(s["bt_pf"]) for s in stocks_data if s["bt_has_data"]) / max(bt_has_data_count, 1), 2) avg_exp = round(sum(safe_float(s["bt_exp"]) for s in stocks_data if s["bt_has_data"]) / max(bt_has_data_count, 1), 2) # ===== Self-Learning: aggregate indicator performance across all stocks ===== indicator_global_stats = defaultdict(lambda: {"total_triggers": 0, "total_wins": 0, "total_returns": [], "pf_list": [], "exp_list": [], "wr_filtered_list": [], "type": "", "desc": ""}) for code, bt in backtest_data.items(): ind_stats = bt.get("indicator_stats", {}) for ind_name, st in ind_stats.items(): gs = indicator_global_stats[ind_name] gs["total_triggers"] += st.get("total", 0) gs["total_wins"] += st.get("wins", 0) gs["total_returns"].append(st.get("avg_return", 0)) gs["pf_list"].append(st.get("profit_factor", 0)) gs["exp_list"].append(st.get("expectancy", 0)) gs["wr_filtered_list"].append(max(st.get("wr_rsi_filter", 0), st.get("wr_adx_filter", 0), st.get("wr_vol_filter", 0), st.get("wr_all_filter", 0))) if ind_name in indicator_info: gs["type"] = indicator_info[ind_name].get("type", "") gs["desc"] = indicator_info[ind_name].get("desc", "") indicator_performance = [] for ind_name, gs in indicator_global_stats.items(): if gs["total_triggers"] == 0: continue wr = round(gs["total_wins"] / gs["total_triggers"] * 100, 1) avg_ret = round(sum(gs["total_returns"]) / len(gs["total_returns"]), 2) if gs["total_returns"] else 0 avg_pf = round(sum(gs["pf_list"]) / len(gs["pf_list"]), 2) if gs["pf_list"] else 0 avg_exp = round(sum(gs["exp_list"]) / len(gs["exp_list"]), 2) if gs["exp_list"] else 0 avg_wr_f = round(sum(gs["wr_filtered_list"]) / len(gs["wr_filtered_list"]), 1) if gs["wr_filtered_list"] else 0 indicator_performance.append({ "name": ind_name, "type": gs["type"], "desc": gs["desc"], "triggers": gs["total_triggers"], "win_rate": wr, "avg_return": avg_ret, "profit_factor": min(avg_pf, 99.99), "expectancy": avg_exp, "wr_filtered": avg_wr_f, "quality": "高" if wr >= 55 and avg_pf > 1.3 else ("中" if wr >= 48 else "低"), }) indicator_performance.sort(key=lambda x: x["win_rate"], reverse=True) # Signal Quality Score per stock: weighted by indicator historical performance for s in stocks_data: if not s["bt_has_data"]: s["signal_quality"] = 50 continue bt = backtest_data.get(s["code"], {}) ind_stats = bt.get("indicator_stats", {}) current_inds = set(sig["indicator"] for sig in s["signals"]) scores = [] for ind_name in current_inds: if ind_name in ind_stats: st = ind_stats[ind_name] wr = st.get("win_rate", 50) pf = min(st.get("profit_factor", 1), 99.99) exp = st.get("expectancy", 0) score = wr * 0.5 + min(pf * 20, 30) + max(exp * 2, -20) + 20 scores.append(score) s["signal_quality"] = round(sum(scores) / len(scores), 1) if scores else 50 # Learning insights learning_insights = [] high_wr_inds = [i for i in indicator_performance if i["win_rate"] >= 55 and i["triggers"] >= 10] low_wr_inds = [i for i in indicator_performance if i["win_rate"] < 45 and i["triggers"] >= 10] if high_wr_inds: top3 = high_wr_inds[:3] learning_insights.append(f"高胜率指标TOP3: {'、'.join(i['name']+'('+str(i['win_rate'])+'%)' for i in top3)} — 优先关注含这些指标的个股") if low_wr_inds: learning_insights.append(f"低胜率指标预警: {'、'.join(i['name']+'('+str(i['win_rate'])+'%)' for i in low_wr_inds[:3])} — 历史表现不佳，谨慎参考") best_sectors = [s for s in stocks_data if s["bt_has_data"] and s["signal_quality"] >= 65] if best_sectors: learning_insights.append(f"高信号质量个股({len(best_sectors)}只): {'、'.join(s['name'] for s in sorted(best_sectors, key=lambda x: x['signal_quality'], reverse=True)[:5])}") avg_sq = round(sum(s["signal_quality"] for s in stocks_data if s["bt_has_data"]) / max(bt_has_data_count, 1), 1) learning_insights.append(f"全市场平均信号质量分: {avg_sq} — {'优质区间' if avg_sq >= 60 else ('中等区间' if avg_sq >= 50 else '偏低区间')}") # Risk metrics total_buy = sum(1 for s in stocks_data if s["trade_signal"] in ("买入", "适度买入", "小仓买入", "轻仓买入", "轻仓关注")) total_sell = sum(1 for s in stocks_data if s["trade_signal"] in ("建议清仓", "建议减仓", "注意风险")) total_pos = sum(s["position_size"] for s in stocks_data if s["position_size"] > 0) sector_concentration = {} for s in stocks_data: if s["position_size"] > 0: sector_concentration[s["industry"]] = sector_concentration.get(s["industry"], 0) + s["position_size"] risk_metrics = { "total_buy_signals": total_buy, "total_sell_signals": total_sell, "total_exposure": total_pos, "avg_position": round(total_pos / max(total_buy, 1), 1) if total_buy else 0, "sector_concentration": dict(sorted(sector_concentration.items(), key=lambda x: x[1], reverse=True)[:5]), "market_regime": ("多头" if market_trend and market_trend.get("trend") == "上行" else "空头" if market_trend and market_trend.get("trend") in ("下行", "单边下行") else "震荡"), # 数据校验 "validation": { "buy_signal_pct": round(total_buy / max(len(stocks_data), 1) * 100, 1), "sell_signal_pct": round(total_sell / max(len(stocks_data), 1) * 100, 1), "watch_pct": round((len(stocks_data) - total_buy - total_sell) / max(len(stocks_data), 1) * 100, 1), "max_single_position": max(s["position_size"] for s in stocks_data) if stocks_data else 0, "position_valid": all(0 <= s["position_size"] <= 100 for s in stocks_data), "entry_stop_valid": all(s["entry_price"] > s["stop_loss"] > 0 for s in stocks_data if s["entry_price"] > 0), "entry_target_valid": all(s["target_price"] > s["entry_price"] > 0 for s in stocks_data if s["entry_price"] > 0), "sector_concentration_pct": round(max(sector_concentration.values()) / max(total_pos, 1) * 100, 1) if sector_concentration else 0, }, } indicator_performance_json = safe_json_dumps(indicator_performance[:30], ensure_ascii=False) learning_insights_json = safe_json_dumps(learning_insights, ensure_ascii=False) risk_metrics_json = safe_json_dumps(risk_metrics, ensure_ascii=False) scan_date = data["scan_date"]; total_scanned = data["total_scanned"]; total_with_signals = data["total_with_signals"] unique_indicators = len(indicator_counts) print(f"Stocks: {len(stocks_data)}, Sectors: {len(sector_ranking)}, Buy: {len(bt_buy_signals)}") # ===================== HTML ===================== html = r''' A股交易决策看板 — 专业交易员视角
A股交易决策看板
市场全景 四度空间 板块轮动 交易信号 模拟交易 ★ 我关注的 0
0
交易信号台
买入信号 · 风险警示 · 仓位管理 · 因子雷达 · 信号矩阵 · 个股排行
--
--:--
今日操盘指南（AI引擎综合推演 · 实盘操作纪律）基于情绪周期 × 市场状态 × 信号共振自动生成
信号追踪与预测校准（自我学习引擎 · 实盘验证）记录每次A/B级信号 → 实盘兑现 → 校准共振评分权重
买入信号 0只 入场/止损/目标/仓位/R-R
风险警示 0只
操作指标面板（实时交易参数）
信号矩阵（个股×指标交叉验证）
个股交易信号排行（Top 50）
★#代码名称指标 行业信号胜率 量化胜率量化分 盈亏比仓位 置信度星级 风险情绪 信号质量 市值(亿)最新价 涨跌幅
1
市场全景
模块化看板 · 按交易员动线自由编排
实时更新 --:--:--
大盘指数 · TPO四度空间Market Profile 价值区/POC分析i
多空偏向仪表盘信号(60%)+大盘(40%)i
市场宽度分析A股全市场涨跌家数i
涨停情绪周期（A股全市场 · 实时涨停数据推演）i
涨停梯队 · 连板天梯（A股全市场 · 点击股票查看）按连板高度分层i
全市场涨跌幅分布实时全A股涨跌幅区间直方图i
今天炒什么全市场热门方向i
市场异动监测板块异动 · 量比异动 · 资金异动 · 领涨领跌i
蓄势待发高评分+缩量整理+即将突破i
个股做T参考日内支撑压力 · T0买卖点 · 量价配合i
重磅事件日历未来30天 · 利好利空研判 · 参考华尔街见闻日历i
场内T+0 ETF机会支持当日回转交易i
两市成交额监控沪深分时累计 · 对比昨日同期i
主力资金流向板块主力净流入排行i
波动率监控日内已实现波动i
全球市场 · 隔夜与亚太美股/港股/日经/人民币i
指数相对强度当日/5日/20日对比i
风格因子表现大小盘/成长防御轮动i
信号分布i
指标触发频次 Top10 鼠标悬停查看指标说明i
回测胜率统计i
--
平均胜率(%)
--
量化过滤(%)
--
胜率提升(pp)
--
平均盈亏比
--
期望收益(%)
--
量化评分
大V动态 · 情绪风向标微博/雪球/微信大V发言情绪分析i
MP
四度空间 · Market Profile
短/中/长期三级时间价格机会分析 · 从交易可执行性角度给出买卖点、止损、目标、仓位
实时计算
方法论解读 · 什么是四度空间
分析对象 支持指数 / 板块 / 个股四度空间分析
可执行交易建议（四度空间自动扫描） 筛选失衡向上/突破价值区的板块与个股

⚙ 自定义看板模块

勾选要展示的模块，用 ↑↓ 或直接拖拽调整顺序（本列表与页面网格均可拖拽，配置自动保存在本机浏览器）。默认布局按交易员动线编排：风控状态 → 情绪主线 → 交易机会 → 资金验证 → 统计反馈；页面上按住任意模块拖到目标位置即可完成排序。
2
板块轮动
A股全市场实时轮动推演 · 产业链传导逻辑 · 全市场板块热力图
板块轮动推演（A股全市场）全市场行业板块 · 动量（当日涨跌幅）+ 成交强度双因子 · 实时刷新
全市场板块轮动明细 东方财富实时板块数据 · 展示涨幅前20
产业链上下游推演主链：上游资源 → 中游制造 → 下游消费 ｜ 独立分支：金融 / 交运（点击板块卡片进入详情，悬停看完整传导逻辑）
卡片左侧色条代表板块轮动梯队：红=领跑主线，橙=接力补涨，灰=蓄势观察，蓝=动能衰减，绿=退潮回避。卡片下方的彩色圆点代表该板块参与的活跃传导（红=共振上涨，金=待传导，绿=成本挤压，蓝=成本红利），悬停圆点可查看该条传导的完整逻辑与交易推演。
板块情绪云图（A股全市场）
4
量化学习中心
指标历史表现 · 自我进化 · 信号质量分层 · 学习洞察
幻方量化策略看板 高胜率回测策略 · 可执行交易方案
学习洞察（基于全市场回测数据自动生成）
指标历史胜率排行（点击查看详情）
指标类型历史触发胜率过滤胜率盈亏比期望质量
指标胜率 vs 盈亏比散点图
★
我关注的个股
关注列表 · K线走势 · 多维分析 · 实时行情
实时更新
P
模拟交易 & 模拟组合
观测信号推荐效果 · 买入/卖出/撤单/持仓 · 多组合对比vs基准（本地保存）
挂单自动撮合 · 止盈止损自动执行
模拟账户概览 佣金万2.5(最低5元) · 卖出印花税0.05% · T+1
下单面板（买卖） 市价单立即成交 · 限价单挂单等待自动撮合
持仓（浮动盈亏实时）
挂单（可撤单）
信号效果统计（已实现交易） 按信号来源分组 · 胜率/盈亏比/期望
交易分析 权益曲线 · 回撤 · 盈亏分布
成交记录
模拟组合管理 组合增删改 · 成分股配置 · 多组合对比vs上证
''' replacements = { "__SCAN_DATE__": scan_date, "__UNIQUE_INDICATORS__": str(unique_indicators), "__STOCKS_JSON__": stocks_json, "__PINYIN_MAP_JSON__": pinyin_map_json, "__INDICATOR_INFO_JSON__": indicator_info_json, "__INDICATOR_COUNTS_JSON__": indicator_counts_json, "__INDUSTRY_DIST_JSON__": industry_dist_json, "__CONCEPT_DIST_JSON__": concept_dist_json, "__SECTOR_DATA_JSON__": sector_data_json, "__ROTATION_DATA_JSON__": rotation_data_json, "__CHAIN_DATA_JSON__": chain_data_json, "__SECTOR_STAGE_JSON__": sector_stage_json, "__SECTOR_SUB_JSON__": sector_sub_json, "__CHAIN_EDGE_INFO_JSON__": chain_edge_info_json, "__ROTATION_INSIGHT_JSON__": rotation_insight_json, "__ACTION_DIST_JSON__": action_dist_json, "__BUY_SIGNALS_JSON__": buy_signals_json, "__SELL_SIGNALS_JSON__": sell_signals_json, "__MARKET_TREND_JSON__": market_trend_json, "__AVG_WIN_RATE__": safe_str(avg_win_rate), "__AVG_WR_FILTERED__": safe_str(avg_wr_filtered), "__AVG_QUANT_SCORE__": safe_str(avg_quant_score), "__AVG_PF__": safe_str(avg_pf), "__AVG_EXP__": safe_str(avg_exp), "__KLINE_DATA_JSON__": kline_data_json, "__INDICATOR_PERF_JSON__": indicator_performance_json, "__LEARNING_INSIGHTS_JSON__": learning_insights_json, "__RISK_METRICS_JSON__": risk_metrics_json, "__FULL_MARKET_SECTORS_JSON__": full_market_sectors_json, "__HOT_SECTOR_STOCKS_JSON__": hot_sector_stocks_json, "__MARKET_BREADTH_JSON__": market_breadth_json, "__T0_ETF_LIST_JSON__": t0_etf_list_json, "__EMO_POOLS_JSON__": emo_pools_json, "__DAV_SENTI_JSON__": weibo_dav_json, "__ECON_CAL_JSON__": econ_cal_json, } for key, val in replacements.items(): html = html.replace(key, val) os.makedirs(OUTPUT_DIR, exist_ok=True) with open(OUTPUT_HTML, "w", encoding="utf-8") as f: f.write(html) print(f"Report generated: {OUTPUT_HTML}") print(f"File size: {os.path.getsize(OUTPUT_HTML)/1024:.1f} KB") # 部署到 web 服务目录（Tailscale 公网: /ashare-dashboard.html） try: import shutil _web_html = os.path.join(BASE_DIR, "ashare-dashboard.html") if os.path.abspath(OUTPUT_HTML) != os.path.abspath(_web_html): shutil.copy2(OUTPUT_HTML, _web_html) print(f"Deployed to web dir: {_web_html}") else: print(f"Web dir is output dir, skip copy: {_web_html}") # 带时间戳版本副本：浏览器缓存旧版时，用带版本号的新 URL 可强制拉取最新文件 _ver = time.strftime("%Y%m%d%H%M") _ver_html = os.path.join(BASE_DIR, f"ashare-dashboard-v{_ver}.html") shutil.copy2(OUTPUT_HTML, _ver_html) print(f"Versioned copy (cache-busting URL): {_ver_html}") except Exception as _dwe: print(f"Deploy copy failed: {_dwe}")