""" 新闻流量分析引擎 基于"流量为王"理念的短线炒股指导系统 整合数据获取、流量模型、情绪分析、AI分析、预警系统 """ import logging import time from datetime import datetime, timedelta from typing import Dict, List, Optional logging.basicConfig(level=logging.INFO) logger = logging.getLogger(__name__) class NewsFlowEngine: """新闻流量分析引擎""" def __init__(self): """初始化引擎""" # 核心模块 self.fetcher = None self.model = None self.sentiment = None self.agents = None self.alerts = None self.db = None self._init_modules() logger.info("✅ 新闻流量引擎初始化完成") def _init_modules(self): """初始化所有模块""" try: from news_flow_data import NewsFlowDataFetcher self.fetcher = NewsFlowDataFetcher() except Exception as e: logger.error(f"数据获取模块初始化失败: {e}") try: from news_flow_model import NewsFlowModel self.model = NewsFlowModel() except Exception as e: logger.error(f"流量模型模块初始化失败: {e}") try: from news_flow_sentiment import SentimentAnalyzer self.sentiment = SentimentAnalyzer() except Exception as e: logger.error(f"情绪分析模块初始化失败: {e}") try: from news_flow_agents import NewsFlowAgents self.agents = NewsFlowAgents() except Exception as e: logger.error(f"AI分析模块初始化失败: {e}") try: from news_flow_alert import NewsFlowAlertSystem self.alerts = NewsFlowAlertSystem() except Exception as e: logger.error(f"预警系统模块初始化失败: {e}") try: from news_flow_db import news_flow_db self.db = news_flow_db except Exception as e: logger.error(f"数据库模块初始化失败: {e}") def run_quick_analysis(self, platforms: List[str] = None, category: str = None) -> Dict: """ 运行快速分析（不含AI） 用于定时同步和快速查看 Returns: { 'success': bool, 'snapshot_id': int, 'flow_data': Dict, 'model_data': Dict, 'sentiment_data': Dict, 'stock_news': List, 'hot_topics': List, 'fetch_time': str, } """ try: logger.info("🚀 开始快速分析...") start_time = time.time() # 1. 获取多平台新闻数据 logger.info("📊 获取新闻数据...") if not self.fetcher: return {'success': False, 'error': '数据获取模块不可用'} multi_result = self.fetcher.get_multi_platform_news( platforms=platforms, category=category ) if not multi_result['success']: return {'success': False, 'error': '获取新闻数据失败'} platforms_data = multi_result['platforms_data'] success_count = multi_result['success_count'] # 2. 提取股票相关新闻 logger.info("🔍 提取股票相关新闻...") stock_news = self.fetcher.extract_stock_related_news(platforms_data) # 3. 获取热门话题 logger.info("🔥 分析热门话题...") hot_topics = self.fetcher.get_hot_topics(platforms_data, top_n=20) # 4. 计算流量得分（基础） logger.info("📈 计算流量得分...") flow_data = self.fetcher.calculate_flow_score(platforms_data) # 5. 运行流量模型 logger.info("🔬 运行流量模型...") history_scores = self._get_history_scores(hours=24) model_data = None if self.model: model_data = self.model.run_full_model( platforms_data, hot_topics, history_scores ) # 6. 情绪分析 logger.info("💭 分析市场情绪...") sentiment_data = None if self.sentiment: history_sentiments = self._get_history_sentiments(limit=10) sentiment_data = self.sentiment.run_full_sentiment_analysis( platforms_data, stock_news, history_scores, flow_data['total_score'], history_sentiments ) # 7. 保存到数据库 logger.info("💾 保存分析结果...") snapshot_id = None if self.db: snapshot_id = self.db.save_flow_snapshot( flow_data, platforms_data, stock_news, hot_topics ) # 保存情绪记录 if sentiment_data and snapshot_id: sentiment_record = { 'sentiment_index': sentiment_data.get('sentiment', {}).get('sentiment_index', 50), 'sentiment_class': sentiment_data.get('sentiment', {}).get('sentiment_class', '中性'), 'flow_stage': sentiment_data.get('flow_stage', {}).get('stage_name', '未知'), 'momentum': sentiment_data.get('momentum', {}).get('momentum', 1.0), 'viral_k': model_data.get('viral_k', {}).get('k_value', 1.0) if model_data else 1.0, 'flow_type': model_data.get('flow_type', {}).get('flow_type', '未知') if model_data else '未知', 'stage_analysis': sentiment_data.get('flow_stage', {}).get('analysis', ''), } self.db.save_sentiment_record(snapshot_id, sentiment_record) duration = time.time() - start_time logger.info(f"✅ 快速分析完成，耗时 {duration:.2f} 秒") return { 'success': True, 'snapshot_id': snapshot_id, 'success_count': success_count, 'flow_data': flow_data, 'model_data': model_data, 'sentiment_data': sentiment_data, 'stock_news': stock_news, 'hot_topics': hot_topics, 'platforms_data': platforms_data, 'fetch_time': datetime.now().strftime('%Y-%m-%d %H:%M:%S'), 'duration': round(duration, 2), } except Exception as e: logger.error(f"❌ 快速分析失败: {e}") return {'success': False, 'error': str(e)} def run_full_analysis(self, platforms: List[str] = None, category: str = None, include_ai: bool = True) -> Dict: """ 运行完整分析（含AI） Returns: { 'success': bool, 'snapshot_id': int, 'flow_data': Dict, 'model_data': Dict, 'sentiment_data': Dict, 'ai_analysis': Dict, 'trading_signals': Dict, 'stock_news': List, 'hot_topics': List, } """ try: logger.info("🚀 开始完整分析...") start_time = time.time() # 1. 先运行快速分析 quick_result = self.run_quick_analysis(platforms, category) if not quick_result['success']: return quick_result # 2. AI智能分析 ai_analysis = None if include_ai: if not self.agents: logger.warning("⚠️ AI代理模块未初始化") elif not self.agents.is_available(): logger.warning("⚠️ DeepSeek API不可用，请检查API密钥配置") else: logger.info("🤖 运行AI分析...") model_data = quick_result.get('model_data', {}) sentiment_data = quick_result.get('sentiment_data', {}) # 基础AI分析 ai_analysis = self.agents.run_full_analysis( quick_result['hot_topics'], quick_result['stock_news'], quick_result['flow_data'], sentiment_data, viral_k=model_data.get('viral_k', {}).get('k_value', 1.0) if model_data else 1.0, flow_type=model_data.get('flow_type', {}).get('flow_type', '未知') if model_data else '未知', ) # 多板块深度分析（多次调用DeepSeek） logger.info("🔍 开始多板块深度分析...") multi_sector_analysis = self.agents.run_multi_sector_analysis( quick_result['hot_topics'], quick_result['stock_news'] ) # 合并多板块分析结果 if ai_analysis and multi_sector_analysis.get('success'): ai_analysis['multi_sector'] = multi_sector_analysis # 保存AI分析结果 if ai_analysis and self.db and quick_result.get('snapshot_id'): ai_record = { 'affected_sectors': ai_analysis.get('sector_analysis', {}).get('benefited_sectors', []), 'recommended_stocks': ai_analysis.get('stock_recommend', {}).get('recommended_stocks', []), 'risk_level': ai_analysis.get('risk_assess', {}).get('risk_level', '未知'), 'risk_factors': ai_analysis.get('risk_assess', {}).get('risk_factors', []), 'advice': ai_analysis.get('investment_advice', {}).get('advice', '观望'), 'confidence': ai_analysis.get('investment_advice', {}).get('confidence', 50), 'summary': ai_analysis.get('investment_advice', {}).get('summary', ''), 'model_used': 'deepseek-chat', 'analysis_time': ai_analysis.get('analysis_time', 0), } self.db.save_ai_analysis(quick_result['snapshot_id'], ai_record) # 3. 生成交易信号 trading_signals = self._generate_trading_signals( quick_result.get('flow_data', {}), quick_result.get('model_data', {}), quick_result.get('sentiment_data', {}), ai_analysis ) duration = time.time() - start_time logger.info(f"✅ 完整分析完成，耗时 {duration:.2f} 秒") return { 'success': True, 'snapshot_id': quick_result.get('snapshot_id'), 'flow_data': quick_result.get('flow_data'), 'model_data': quick_result.get('model_data'), 'sentiment_data': quick_result.get('sentiment_data'), 'ai_analysis': ai_analysis, 'trading_signals': trading_signals, 'stock_news': quick_result.get('stock_news'), 'hot_topics': quick_result.get('hot_topics'), 'platforms_data': quick_result.get('platforms_data'), 'fetch_time': quick_result.get('fetch_time'), 'duration': round(duration, 2), } except Exception as e: logger.error(f"❌ 完整分析失败: {e}") return {'success': False, 'error': str(e)} def run_alert_check(self) -> Dict: """ 运行预警检查 Returns: { 'success': bool, 'alerts': List[Dict], } """ try: logger.info("⚠️ 开始预警检查...") if not self.alerts: return {'success': False, 'error': '预警系统不可用'} # 获取当前数据 quick_result = self.run_quick_analysis() if not quick_result['success']: return {'success': False, 'error': quick_result.get('error')} # 获取历史数据 history_data = self._get_previous_snapshot() # 构建检查数据 current_data = { 'flow_data': quick_result.get('flow_data', {}), 'hot_topics': quick_result.get('hot_topics', []), 'viral_k': quick_result.get('model_data', {}).get('viral_k', {}), 'flow_stage': quick_result.get('sentiment_data', {}).get('flow_stage', {}), } # 检查预警 alerts = self.alerts.check_alerts( current_data, history_data, quick_result.get('sentiment_data'), quick_result.get('snapshot_id') ) logger.info(f"✅ 预警检查完成，触发 {len(alerts)} 个预警") return { 'success': True, 'alerts': alerts, 'snapshot_id': quick_result.get('snapshot_id'), } except Exception as e: logger.error(f"❌ 预警检查失败: {e}") return {'success': False, 'error': str(e)} def get_dashboard_data(self) -> Dict: """ 获取仪表盘数据 Returns: { 'latest_snapshot': Dict, 'latest_sentiment': Dict, 'latest_ai_analysis': Dict, 'recent_alerts': List, 'flow_trend': Dict, 'scheduler_status': Dict, } """ try: data = {} if self.db: # 最新快照 data['latest_snapshot'] = self.db.get_latest_snapshot() # 最新情绪 data['latest_sentiment'] = self.db.get_latest_sentiment() # 最新AI分析 data['latest_ai_analysis'] = self.db.get_latest_ai_analysis() # 最近预警 data['recent_alerts'] = self.db.get_alerts(days=1) # 流量趋势（7天） data['flow_trend'] = self.get_flow_trend(days=7) # 调度器状态 try: from news_flow_scheduler import news_flow_scheduler data['scheduler_status'] = news_flow_scheduler.get_status() except: data['scheduler_status'] = None return data except Exception as e: logger.error(f"获取仪表盘数据失败: {e}") return {} def get_flow_trend(self, days: int = 7) -> Dict: """获取流量趋势""" if not self.db: return {'dates': [], 'scores': [], 'trend': '无数据'} stats = self.db.get_daily_statistics(days) if not stats: return {'dates': [], 'scores': [], 'trend': '无数据', 'analysis': '暂无历史数据'} # 反转（从旧到新） stats.reverse() dates = [s['date'] for s in stats] avg_scores = [s['avg_score'] for s in stats] max_scores = [s['max_score'] for s in stats] min_scores = [s['min_score'] for s in stats] # 判断趋势 if len(avg_scores) >= 3: recent_avg = sum(avg_scores[-3:]) / 3 earlier_avg = sum(avg_scores[:3]) / 3 if recent_avg > earlier_avg * 1.2: trend = '上升' analysis = f"近期流量持续上升（近3日均值{recent_avg:.0f} > 前3日均值{earlier_avg:.0f}），市场热度升温。" elif recent_avg < earlier_avg * 0.8: trend = '下降' analysis = f"近期流量持续下降（近3日均值{recent_avg:.0f} < 前3日均值{earlier_avg:.0f}），市场热度降温。" else: trend = '平稳' analysis = f"近期流量波动不大（近3日均值{recent_avg:.0f} ≈ 前3日均值{earlier_avg:.0f}），市场处于平衡状态。" else: trend = '数据不足' analysis = '历史数据不足，无法判断趋势' return { 'dates': dates, 'avg_scores': avg_scores, 'max_scores': max_scores, 'min_scores': min_scores, 'trend': trend, 'analysis': analysis, } def _generate_trading_signals(self, flow_data: Dict, model_data: Dict, sentiment_data: Dict, ai_analysis: Dict = None) -> Dict: """生成交易信号""" signals = { 'overall_signal': '观望', 'confidence': 50, 'risk_level': '中等', 'hot_sectors': [], 'operation_advice': '', 'key_message': '', } # 获取各项指标 total_score = flow_data.get('total_score', 0) flow_level = flow_data.get('level', '中') sentiment_index = 50 flow_stage = '未知' if sentiment_data: sentiment_index = sentiment_data.get('sentiment', {}).get('sentiment_index', 50) flow_stage = sentiment_data.get('flow_stage', {}).get('stage_name', '未知') viral_k = 1.0 if model_data: viral_k = model_data.get('viral_k', {}).get('k_value', 1.0) # 核心判断逻辑 if flow_stage in ['一致', 'consensus']: # 流量高潮 = 逃命时刻 signals['overall_signal'] = '卖出' signals['confidence'] = 90 signals['risk_level'] = '极高' signals['key_message'] = '⚠️ 流量高潮 = 价格高潮 = 逃命时刻！立即减仓或清仓！' signals['operation_advice'] = '立即减仓或清仓，锁定利润。不要贪婪，不要犹豫。' elif flow_stage in ['退潮', 'decline']: signals['overall_signal'] = '观望' signals['confidence'] = 80 signals['risk_level'] = '高' signals['key_message'] = '流量退潮，及时止盈止损' signals['operation_advice'] = '持仓者及时止盈止损，空仓者继续观望。' elif flow_stage in ['加速', 'acceleration'] and viral_k > 1.2: signals['overall_signal'] = '买入' signals['confidence'] = 75 signals['risk_level'] = '中等' signals['key_message'] = '流量加速期，可参与龙头' signals['operation_advice'] = '关注龙头股，轻仓试探。设置止损位（-5%），止盈位（+15%）。' elif flow_stage in ['启动', 'startup']: signals['overall_signal'] = '关注' signals['confidence'] = 65 signals['risk_level'] = '低' signals['key_message'] = '流量启动期，可以关注' signals['operation_advice'] = '密切关注，等待确认后介入。' elif flow_level == "极高" and sentiment_index > 85: signals['overall_signal'] = '观望' signals['confidence'] = 70 signals['risk_level'] = '高' signals['key_message'] = '流量极高+情绪过热，追高风险大' signals['operation_advice'] = '不建议追高，等待回调机会。' else: signals['overall_signal'] = '观望' signals['confidence'] = 50 signals['risk_level'] = '中等' signals['key_message'] = '市场无明确方向，保持观望' signals['operation_advice'] = '保持观望，等待流量信号明确。' # 整合AI分析结果 if ai_analysis: advice = ai_analysis.get('investment_advice', {}) if advice.get('advice'): signals['ai_advice'] = advice.get('advice') signals['ai_confidence'] = advice.get('confidence', 50) signals['ai_summary'] = advice.get('summary', '') sectors = ai_analysis.get('sector_analysis', {}).get('benefited_sectors', []) signals['hot_sectors'] = sectors[:3] return signals def _get_history_scores(self, hours: int = 24) -> List[int]: """获取历史流量得分""" if not self.db: return [] scores = self.db.get_recent_scores(hours) return [s['total_score'] for s in scores] def _get_history_sentiments(self, limit: int = 10) -> List[Dict]: """获取历史情绪记录""" if not self.db: return [] return self.db.get_sentiment_history(limit) def _get_previous_snapshot(self) -> Optional[Dict]: """获取上一次快照""" if not self.db: return None snapshots = self.db.get_history_snapshots(limit=2) if len(snapshots) >= 2: detail = self.db.get_snapshot_detail(snapshots[1]['id']) return { 'hot_topics': detail.get('hot_topics', []), 'snapshot': detail.get('snapshot', {}), } return None def compare_with_history(self, current_score: int) -> Dict: """与历史数据对比""" if not self.db: return { 'percentile': 50, 'level_description': '无历史对比', 'comparison': '暂无足够的历史数据进行对比' } stats = self.db.get_daily_statistics(30) if not stats: return { 'percentile': 50, 'level_description': '无历史对比', 'comparison': '暂无足够的历史数据进行对比' } all_scores = [] for stat in stats: all_scores.extend([stat['avg_score'], stat['max_score'], stat['min_score']]) all_scores.sort() lower_count = sum(1 for s in all_scores if s < current_score) percentile = int(lower_count / len(all_scores) * 100) if all_scores else 50 if percentile >= 90: level_description = "极高水平" comparison = f"当前流量得分{current_score}处于历史极高水平（超过{percentile}%的历史记录），流量极度爆发！" elif percentile >= 70: level_description = "较高水平" comparison = f"当前流量得分{current_score}处于历史较高水平（超过{percentile}%的历史记录），流量活跃。" elif percentile >= 30: level_description = "正常水平" comparison = f"当前流量得分{current_score}处于历史正常水平（超过{percentile}%的历史记录）。" else: level_description = "较低水平" comparison = f"当前流量得分{current_score}处于历史较低水平（仅超过{percentile}%的历史记录），流量低迷。" return { 'percentile': percentile, 'level_description': level_description, 'comparison': comparison } # 全局引擎实例 news_flow_engine = NewsFlowEngine() # 测试代码 if __name__ == "__main__": print("=== 测试新闻流量分析引擎 ===") # 运行快速分析 print("\n--- 快速分析 ---") result = news_flow_engine.run_quick_analysis(category='finance') if result['success']: print(f"✅ 分析成功！快照ID: {result.get('snapshot_id')}") print(f"\n流量得分: {result['flow_data']['total_score']}") print(f"流量等级: {result['flow_data']['level']}") print(f"股票相关新闻: {len(result['stock_news'])} 条") print(f"热门话题: {len(result['hot_topics'])} 个") if result.get('sentiment_data'): sentiment = result['sentiment_data'].get('sentiment', {}) print(f"\n情绪指数: {sentiment.get('sentiment_index', 'N/A')}") print(f"情绪分类: {sentiment.get('sentiment_class', 'N/A')}") flow_stage = result['sentiment_data'].get('flow_stage', {}) print(f"流量阶段: {flow_stage.get('stage_name', 'N/A')}") else: print(f"❌ 分析失败: {result.get('error')}")