oficcejo
249 lines
9.6 KiB
Python
249 lines
9.6 KiB
Python
from deepseek_client import DeepSeekClient
|
||
from typing import Dict, Any
|
||
import time
|
||
|
||
class StockAnalysisAgents:
|
||
"""股票分析AI智能体集合"""
|
||
|
||
def __init__(self, model="deepseek-chat"):
|
||
self.model = model
|
||
self.deepseek_client = DeepSeekClient(model=model)
|
||
|
||
def technical_analyst_agent(self, stock_info: Dict, stock_data: Any, indicators: Dict) -> Dict[str, Any]:
|
||
"""技术面分析智能体"""
|
||
print("🔍 技术分析师正在分析中...")
|
||
time.sleep(1) # 模拟分析时间
|
||
|
||
analysis = self.deepseek_client.technical_analysis(stock_info, stock_data, indicators)
|
||
|
||
return {
|
||
"agent_name": "技术分析师",
|
||
"agent_role": "负责技术指标分析、图表形态识别、趋势判断",
|
||
"analysis": analysis,
|
||
"focus_areas": ["技术指标", "趋势分析", "支撑阻力", "交易信号"],
|
||
"timestamp": time.strftime("%Y-%m-%d %H:%M:%S")
|
||
}
|
||
|
||
def fundamental_analyst_agent(self, stock_info: Dict, financial_data: Dict = None) -> Dict[str, Any]:
|
||
"""基本面分析智能体"""
|
||
print("📊 基本面分析师正在分析中...")
|
||
time.sleep(1)
|
||
|
||
analysis = self.deepseek_client.fundamental_analysis(stock_info, financial_data)
|
||
|
||
return {
|
||
"agent_name": "基本面分析师",
|
||
"agent_role": "负责公司财务分析、行业研究、估值分析",
|
||
"analysis": analysis,
|
||
"focus_areas": ["财务指标", "行业分析", "公司价值", "成长性"],
|
||
"timestamp": time.strftime("%Y-%m-%d %H:%M:%S")
|
||
}
|
||
|
||
def fund_flow_analyst_agent(self, stock_info: Dict, indicators: Dict, fund_flow_data: Dict = None) -> Dict[str, Any]:
|
||
"""资金面分析智能体"""
|
||
print("💰 资金面分析师正在分析中...")
|
||
|
||
# 如果有资金流向数据，显示数据来源
|
||
if fund_flow_data and fund_flow_data.get('query_success'):
|
||
print(" ✓ 已获取问财资金流向数据")
|
||
else:
|
||
print(" ⚠ 未获取到问财资金流向数据，将基于技术指标分析")
|
||
|
||
time.sleep(1)
|
||
|
||
analysis = self.deepseek_client.fund_flow_analysis(stock_info, indicators, fund_flow_data)
|
||
|
||
return {
|
||
"agent_name": "资金面分析师",
|
||
"agent_role": "负责资金流向分析、主力行为研究、市场情绪判断",
|
||
"analysis": analysis,
|
||
"focus_areas": ["资金流向", "主力动向", "市场情绪", "流动性"],
|
||
"fund_flow_data": fund_flow_data, # 保存资金流向数据以供后续使用
|
||
"timestamp": time.strftime("%Y-%m-%d %H:%M:%S")
|
||
}
|
||
|
||
def risk_management_agent(self, stock_info: Dict, indicators: Dict) -> Dict[str, Any]:
|
||
"""风险管理智能体"""
|
||
print("⚠️ 风险管理师正在评估中...")
|
||
time.sleep(1)
|
||
|
||
risk_prompt = f"""
|
||
作为风险管理专家，请基于以下信息进行风险评估：
|
||
|
||
股票信息：
|
||
- 股票代码：{stock_info.get('symbol', 'N/A')}
|
||
- 股票名称：{stock_info.get('name', 'N/A')}
|
||
- 当前价格：{stock_info.get('current_price', 'N/A')}
|
||
- Beta系数：{stock_info.get('beta', 'N/A')}
|
||
- 52周最高：{stock_info.get('52_week_high', 'N/A')}
|
||
- 52周最低：{stock_info.get('52_week_low', 'N/A')}
|
||
|
||
技术指标：
|
||
- RSI：{indicators.get('rsi', 'N/A')}
|
||
- 布林带位置：当前价格相对于上下轨的位置
|
||
- 波动率指标等
|
||
|
||
请从以下角度进行风险评估：
|
||
1. 市场风险（系统性风险）
|
||
2. 个股风险（非系统性风险）
|
||
3. 流动性风险
|
||
4. 波动性风险
|
||
5. 估值风险
|
||
6. 行业风险
|
||
7. 风险等级评定（低/中/高）
|
||
8. 风险控制建议
|
||
|
||
给出专业的风险评估报告。
|
||
"""
|
||
|
||
messages = [
|
||
{"role": "system", "content": "你是一名专业的风险管理专家，具有丰富的风险识别和控制经验。"},
|
||
{"role": "user", "content": risk_prompt}
|
||
]
|
||
|
||
analysis = self.deepseek_client.call_api(messages)
|
||
|
||
return {
|
||
"agent_name": "风险管理师",
|
||
"agent_role": "负责风险识别、风险评估、风险控制策略制定",
|
||
"analysis": analysis,
|
||
"focus_areas": ["风险识别", "风险量化", "风险控制", "资产配置"],
|
||
"timestamp": time.strftime("%Y-%m-%d %H:%M:%S")
|
||
}
|
||
|
||
def market_sentiment_agent(self, stock_info: Dict) -> Dict[str, Any]:
|
||
"""市场情绪分析智能体"""
|
||
print("📈 市场情绪分析师正在分析中...")
|
||
time.sleep(1)
|
||
|
||
sentiment_prompt = f"""
|
||
作为市场情绪分析专家，请基于当前市场环境对以下股票进行情绪分析：
|
||
|
||
股票信息：
|
||
- 股票代码：{stock_info.get('symbol', 'N/A')}
|
||
- 股票名称：{stock_info.get('name', 'N/A')}
|
||
- 行业：{stock_info.get('sector', 'N/A')}
|
||
- 细分行业：{stock_info.get('industry', 'N/A')}
|
||
|
||
请从以下角度分析市场情绪：
|
||
1. 整体市场情绪（牛市/熊市/震荡市）
|
||
2. 行业板块情绪和热度
|
||
3. 个股关注度和讨论热度
|
||
4. 投资者情绪指标
|
||
5. 市场预期和共识
|
||
6. 消息面和事件驱动因素
|
||
7. 情绪对股价的影响评估
|
||
8. 情绪反转的可能性
|
||
|
||
结合当前宏观环境和市场热点，给出专业的市场情绪分析。
|
||
"""
|
||
|
||
messages = [
|
||
{"role": "system", "content": "你是一名专业的市场情绪分析师，擅长解读市场心理和投资者行为。"},
|
||
{"role": "user", "content": sentiment_prompt}
|
||
]
|
||
|
||
analysis = self.deepseek_client.call_api(messages)
|
||
|
||
return {
|
||
"agent_name": "市场情绪分析师",
|
||
"agent_role": "负责市场情绪研究、投资者心理分析、热点追踪",
|
||
"analysis": analysis,
|
||
"focus_areas": ["市场情绪", "投资者心理", "热点板块", "消息面"],
|
||
"timestamp": time.strftime("%Y-%m-%d %H:%M:%S")
|
||
}
|
||
|
||
def run_multi_agent_analysis(self, stock_info: Dict, stock_data: Any, indicators: Dict,
|
||
financial_data: Dict = None, fund_flow_data: Dict = None) -> Dict[str, Any]:
|
||
"""运行多智能体分析"""
|
||
print("🚀 启动多智能体股票分析系统...")
|
||
print("=" * 50)
|
||
|
||
# 并行运行各个分析师
|
||
agents_results = {}
|
||
|
||
# 技术面分析
|
||
agents_results["technical"] = self.technical_analyst_agent(stock_info, stock_data, indicators)
|
||
|
||
# 基本面分析
|
||
agents_results["fundamental"] = self.fundamental_analyst_agent(stock_info, financial_data)
|
||
|
||
# 资金面分析（传入资金流向数据）
|
||
agents_results["fund_flow"] = self.fund_flow_analyst_agent(stock_info, indicators, fund_flow_data)
|
||
|
||
# 风险管理分析
|
||
agents_results["risk_management"] = self.risk_management_agent(stock_info, indicators)
|
||
|
||
# 市场情绪分析
|
||
agents_results["market_sentiment"] = self.market_sentiment_agent(stock_info)
|
||
|
||
print("✅ 所有分析师完成分析")
|
||
print("=" * 50)
|
||
|
||
return agents_results
|
||
|
||
def conduct_team_discussion(self, agents_results: Dict[str, Any], stock_info: Dict) -> str:
|
||
"""进行团队讨论"""
|
||
print("🤝 分析团队正在进行综合讨论...")
|
||
time.sleep(2)
|
||
|
||
# 提取各分析师的报告
|
||
technical_report = agents_results.get("technical", {}).get("analysis", "")
|
||
fundamental_report = agents_results.get("fundamental", {}).get("analysis", "")
|
||
fund_flow_report = agents_results.get("fund_flow", {}).get("analysis", "")
|
||
risk_report = agents_results.get("risk_management", {}).get("analysis", "")
|
||
sentiment_report = agents_results.get("market_sentiment", {}).get("analysis", "")
|
||
|
||
discussion_prompt = f"""
|
||
现在进行投资决策团队会议，参会人员包括：技术分析师、基本面分析师、资金面分析师、风险管理师、市场情绪分析师。
|
||
|
||
股票：{stock_info.get('name', 'N/A')} ({stock_info.get('symbol', 'N/A')})
|
||
|
||
各分析师报告：
|
||
|
||
【技术分析师报告】
|
||
{technical_report}
|
||
|
||
【基本面分析师报告】
|
||
{fundamental_report}
|
||
|
||
【资金面分析师报告】
|
||
{fund_flow_report}
|
||
|
||
【风险管理师报告】
|
||
{risk_report}
|
||
|
||
【市场情绪分析师报告】
|
||
{sentiment_report}
|
||
|
||
请模拟一场真实的投资决策会议讨论：
|
||
1. 各分析师观点的一致性和分歧
|
||
2. 不同维度分析的权重考量
|
||
3. 风险收益评估
|
||
4. 投资时机判断
|
||
5. 策略制定思路
|
||
6. 达成初步共识
|
||
|
||
请以对话形式展现讨论过程，体现专业团队的思辨过程。
|
||
"""
|
||
|
||
messages = [
|
||
{"role": "system", "content": "你需要模拟一场专业的投资团队讨论会议，体现不同角色的观点碰撞和最终共识形成。"},
|
||
{"role": "user", "content": discussion_prompt}
|
||
]
|
||
|
||
discussion_result = self.deepseek_client.call_api(messages, max_tokens=6000)
|
||
|
||
print("✅ 团队讨论完成")
|
||
return discussion_result
|
||
|
||
def make_final_decision(self, discussion_result: str, stock_info: Dict, indicators: Dict) -> Dict[str, Any]:
|
||
"""制定最终投资决策"""
|
||
print("📋 正在制定最终投资决策...")
|
||
time.sleep(1)
|
||
|
||
decision = self.deepseek_client.final_decision(discussion_result, stock_info, indicators)
|
||
|
||
print("✅ 最终投资决策完成")
|
||
return decision
|