hahaju/aiagents-stock
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

增加选股策略

Browse Source
This commit is contained in:
oficcejo
2025-12-12 19:46:38 +08:00
parent 595278100f
commit ad5865d3b4
25 changed files with 4818 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files
low_price_bull_ui.py
+589
View File
@@ -0,0 +1,589 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
低价擒牛UI模块
"""
import streamlit as st
import pandas as pd
from datetime import datetime
from low_price_bull_selector import LowPriceBullSelector
from low_price_bull_strategy import LowPriceBullStrategy
from notification_service import notification_service
from low_price_bull_monitor import low_price_bull_monitor
from low_price_bull_service import low_price_bull_service
def display_low_price_bull():
"""显示低价擒牛选股界面"""
# 检查是否显示监控面板
if st.session_state.get('show_low_price_monitor'):
from low_price_bull_monitor_ui import display_monitor_panel
display_monitor_panel()
# 返回按钮
if st.button("🔙 返回选股", type="secondary"):
del st.session_state.show_low_price_monitor
st.rerun()
return
st.markdown("顶部按钮区")
col_select, col_monitor = st.columns([3, 1])
with col_select:
st.markdown("## 🐂 低价擒牛 - 低价高成长股票筛选")
with col_monitor:
st.write("") # 占位
if st.button("📊 策略监控", type="primary", width='content'):
st.session_state.show_low_price_monitor = True
st.rerun()
st.markdown("---")
st.markdown("""
### 📋 选股策略说明
**筛选条件**：
- ✅ 股价 < 10元
- ✅ 净利润增长率 ≥ 100%（净利润同比增长率）
- ✅ 非ST股票
- ✅ 非科创板
- ✅ 非创业板
- ✅ 深圳A股
- ✅ 按成交额由小至大排名
**量化交易策略**：
- 💰 资金量：100万元
- 📅 持股周期：5天
- 💼 仓位控制：满仓
- 📊 个股最大持仓：4成（40%）
- 🎯 账户最大持股数：4只
- 🛒 单日最大买入数：2只
- 📈 买入时机：开盘买入
- 📉 卖出时机：MA5下穿MA20或持股满5天
""")
st.markdown("---")
# 参数设置
col1, col2 = st.columns([2, 1])
with col1:
top_n = st.slider(
"筛选数量",
min_value=3,
max_value=10,
value=5,
step=1,
help="选择展示的股票数量"
)
with col2:
st.info(f"💡 将筛选成交额最小的前{top_n}只股票")
st.markdown("---")
# 开始选股按钮
if st.button("🚀 开始低价擒牛选股", type="primary", width='content'):
with st.spinner("正在获取数据，请稍候..."):
# 创建选股器
selector = LowPriceBullSelector()
# 获取股票
success, stocks_df, message = selector.get_low_price_stocks(top_n=top_n)
if success and stocks_df is not None:
# 保存结果
st.session_state.low_price_bull_stocks = stocks_df
st.session_state.low_price_bull_selector = selector
st.success(f"✅ {message}")
# 发送钉钉通知
send_dingtalk_notification(stocks_df, top_n)
st.rerun()
else:
st.error(f"❌ {message}")
# 显示选股结果
if 'low_price_bull_stocks' in st.session_state:
display_stock_results(
st.session_state.low_price_bull_stocks,
st.session_state.get('low_price_bull_selector')
)
def display_stock_results(stocks_df: pd.DataFrame, selector):
"""显示选股结果"""
st.markdown("---")
st.markdown("## 📊 选股结果")
# 统计信息
col1, col2, col3 = st.columns(3)
with col1:
st.metric("筛选数量", f"{len(stocks_df)} 只")
with col2:
# 智能计算平均净利增长率（过滤无效值）
growth_col = stocks_df.get('净利润增长率', stocks_df.get('净利润同比增长率', pd.Series([])))
valid_growth = growth_col[growth_col.notna() & (growth_col != '') & (growth_col != 'N/A')]
if len(valid_growth) > 0:
avg_growth = pd.to_numeric(valid_growth, errors='coerce').mean()
if not pd.isna(avg_growth):
st.metric("平均净利增长率", f"{avg_growth:.1f}%")
else:
st.metric("平均净利增长率", "-")
else:
st.metric("平均净利增长率", "-")
with col3:
# 智能计算平均股价（过滤无效值）
price_col = stocks_df.get('股价', stocks_df.get('最新价', pd.Series([])))
valid_price = price_col[price_col.notna() & (price_col != '') & (price_col != 'N/A')]
if len(valid_price) > 0:
avg_price = pd.to_numeric(valid_price, errors='coerce').mean()
if not pd.isna(avg_price):
st.metric("平均股价", f"{avg_price:.2f} 元")
else:
st.metric("平均股价", "-")
else:
st.metric("平均股价", "-")
st.markdown("---")
# 显示股票列表
st.markdown("### 📋 精选股票列表")
for idx, row in stocks_df.iterrows():
# 获取股票代码和简称
code = row.get('股票代码', 'N/A')
name = row.get('股票简称', 'N/A')
# 获取价格信息作为标题补充
price = row.get('股价', row.get('最新价', None))
price_str = ''
if price is not None and not pd.isna(price):
try:
price_float = float(price)
price_str = f" | 价格: {price_float:.2f}元"
except:
pass
with st.expander(
f"【第{idx+1}名】{code} - {name}{price_str}",
expanded=(idx < 3)
):
display_stock_detail(row)
# 完整数据表格
st.markdown("---")
st.markdown("### 📊 完整数据表格")
# 选择关键列显示
display_cols = ['股票代码', '股票简称']
# 智能匹配列名
for pattern in ['股价', '最新价']:
matching = [col for col in stocks_df.columns if pattern in col]
if matching:
display_cols.append(matching[0])
break
for pattern in ['净利润增长率', '净利润同比增长率']:
matching = [col for col in stocks_df.columns if pattern in col]
if matching:
display_cols.append(matching[0])
break
for pattern in ['成交额']:
matching = [col for col in stocks_df.columns if pattern in col]
if matching:
display_cols.append(matching[0])
break
for col_name in ['总市值', '市盈率', '市净率', '所属行业']:
matching = [col for col in stocks_df.columns if col_name in col]
if matching:
display_cols.append(matching[0])
# 选择存在的列
final_cols = [col for col in display_cols if col in stocks_df.columns]
if final_cols:
st.dataframe(stocks_df[final_cols], width='content', height=400)
# 下载按钮
csv = stocks_df[final_cols].to_csv(index=False, encoding='utf-8-sig')
st.download_button(
label="📥 下载股票列表CSV",
data=csv,
file_name=f"low_price_bull_{datetime.now().strftime('%Y%m%d')}.csv",
mime="text/csv"
)
# 量化交易模拟
st.markdown("---")
display_strategy_simulation(stocks_df, selector)
def display_stock_detail(row: pd.Series):
"""显示单个股票详情"""
def is_valid_value(value):
"""判断值是否有效（非None、非NaN、非空字符串、非'N/A'）"""
if value is None:
return False
if pd.isna(value):
return False
if str(value).strip() in ['', 'N/A', 'nan', 'None']:
return False
return True
def format_value(value, suffix=''):
"""格式化显示值"""
if isinstance(value, float):
if abs(value) >= 100000000: # 亿
return f"{value/100000000:.2f}亿{suffix}"
elif abs(value) >= 10000: # 万
return f"{value/10000:.2f}万{suffix}"
else:
return f"{value:.2f}{suffix}"
return f"{value}{suffix}"
# 先检查是否有任何财务数据
has_any_data = False
financial_fields = [
('所属行业', row.get('所属行业', row.get('所属同花顺行业', None))),
('总市值', row.get('总市值', row.get('总市值[20241211]', None))),
('市盈率', row.get('市盈率', row.get('市盈率pe', None))),
('市净率', row.get('市净率', row.get('市净率pb', None))),
('流通市值', row.get('流通市值', row.get('流通市值[20241211]', None))),
('换手率', row.get('换手率', row.get('换手率[%]', None)))
]
for _, value in financial_fields:
if is_valid_value(value):
has_any_data = True
break
# 只有当存在有效数据时才显示两列布局
if has_any_data:
col1, col2 = st.columns(2)
else:
col1 = st.container()
col2 = None
with col1:
st.markdown("#### 📊 基本信息")
# 股票代码（必显示）
code = row.get('股票代码', '')
if is_valid_value(code):
st.markdown(f"**股票代码**: {code}")
# 股票简称（必显示）
name = row.get('股票简称', '')
if is_valid_value(name):
st.markdown(f"**股票简称**: {name}")
# 当前价格
price = row.get('股价', row.get('最新价', None))
if is_valid_value(price):
st.markdown(f"**当前价格**: {format_value(price, '元')}")
# 净利润增长率
growth = row.get('净利润增长率', row.get('净利润同比增长率', None))
if is_valid_value(growth):
st.markdown(f"**净利润增长率**: {format_value(growth, '%')}")
# 成交额
turnover = row.get('成交额', None)
if is_valid_value(turnover):
st.markdown(f"**成交额**: {format_value(turnover, '元')}")
# 涨跌幅
change_pct = row.get('涨跌幅', row.get('涨跌幅:前复权[%]', None))
if is_valid_value(change_pct):
st.markdown(f"**涨跌幅**: {format_value(change_pct, '%')}")
# 只有当有财务数据时才显示财务指标栏目
if col2 is not None:
with col2:
st.markdown("#### 💼 财务指标")
# 所属行业
industry = row.get('所属行业', row.get('所属同花顺行业', None))
if is_valid_value(industry):
st.markdown(f"**所属行业**: {industry}")
# 总市值
market_cap = row.get('总市值', row.get('总市值[20241211]', None))
if is_valid_value(market_cap):
st.markdown(f"**总市值**: {format_value(market_cap, '元')}")
# 市盈率
pe = row.get('市盈率', row.get('市盈率pe', None))
if is_valid_value(pe):
st.markdown(f"**市盈率**: {format_value(pe, '')}")
# 市净率
pb = row.get('市净率', row.get('市净率pb', None))
if is_valid_value(pb):
st.markdown(f"**市净率**: {format_value(pb, '')}")
# 流通市值
float_cap = row.get('流通市值', row.get('流通市值[20241211]', None))
if is_valid_value(float_cap):
st.markdown(f"**流通市值**: {format_value(float_cap, '元')}")
# 换手率
turnover_rate = row.get('换手率', row.get('换手率[%]', None))
if is_valid_value(turnover_rate):
st.markdown(f"**换手率**: {format_value(turnover_rate, '%')}")
# 添加监控按钮
st.markdown("---")
st.markdown("#### 📊 策略监控")
from low_price_bull_monitor_ui import add_stock_to_monitor_button
stock_code = row.get('股票代码', '')
stock_name = row.get('股票简称', '')
price = row.get('股价', row.get('最新价', None))
# 去掉代码后缀
if isinstance(stock_code, str) and '.' in stock_code:
stock_code = stock_code.split('.')[0]
# 转换价格
try:
price_float = float(price) if price and not pd.isna(price) else None
except:
price_float = None
if stock_code and stock_name:
add_stock_to_monitor_button(stock_code, stock_name, price_float)
def display_strategy_simulation(stocks_df: pd.DataFrame, selector):
"""显示量化交易策略模拟"""
st.markdown("## 🎯 策略监控与模拟")
st.info("""
**监控说明**：
- 在上方股票列表中点击"➕ 加入策略监控"按钮即可加入
- 监控条件：① 持股满5天第6天开盘提醒卖出 ② MA5下穿MA20提醒卖出
- 扫描频率：每分钟扫描1次（可在监控面板配置）
- 提醒卖出后自动移出监控列表
- 点击右上角"📊 策略监控"按钮查看监控面板
""")
col1, col2 = st.columns(2)
with col1:
if st.button("🎮 开始策略模拟", type="primary", width='content'):
st.session_state.show_strategy_simulation = True
with col2:
if st.button("🔗 连接MiniQMT实盘", type="secondary", width='content'):
st.warning("⚠️ MiniQMT实盘交易功能需要先配置环境变量，详见系统配置")
# 显示模拟结果
if st.session_state.get('show_strategy_simulation'):
run_strategy_simulation(stocks_df)
def run_strategy_simulation(stocks_df: pd.DataFrame):
"""运行策略模拟"""
st.markdown("---")
st.markdown("### 📈 策略模拟执行")
# 创建策略实例
strategy = LowPriceBullStrategy(initial_capital=1000000.0)
# 模拟买入（按成交额排序，优先买入成交额小的）
st.markdown("#### 1️⃣ 模拟买入信号")
buy_results = []
current_date = datetime.now().strftime("%Y-%m-%d")
for idx, row in stocks_df.head(strategy.max_daily_buy).iterrows():
code = str(row.get('股票代码', '')).split('.')[0]
name = row.get('股票简称', 'N/A')
price = float(row.get('股价', row.get('最新价', 0)))
if price > 0:
success, message, trade = strategy.buy(code, name, price, current_date)
buy_results.append({
'success': success,
'message': message,
'trade': trade
})
# 显示买入结果
for result in buy_results:
if result['success']:
st.success(result['message'])
else:
st.warning(f"⚠️ {result['message']}")
# 显示持仓
st.markdown("---")
st.markdown("#### 2️⃣ 当前持仓")
positions = strategy.get_positions()
if positions:
positions_df = pd.DataFrame(positions)
st.dataframe(positions_df, width='content')
else:
st.info("暂无持仓")
# 显示账户摘要
st.markdown("---")
st.markdown("#### 3️⃣ 账户摘要")
summary = strategy.get_portfolio_summary()
col1, col2, col3, col4 = st.columns(4)
with col1:
st.metric("初始资金", f"{summary['initial_capital']:,.0f} 元")
with col2:
st.metric("可用资金", f"{summary['available_cash']:,.0f} 元")
with col3:
st.metric("持仓市值", f"{summary['position_value']:,.0f} 元")
with col4:
st.metric("总资产", f"{summary['total_value']:,.0f} 元")
st.markdown("---")
# 策略说明
st.markdown("#### 📝 策略执行说明")
st.markdown("""
**后续操作**：
1. **持有期管理**：系统会自动跟踪每只股票的持有天数
2. **卖出信号监测**：
- 每日收盘后计算MA5和MA20
- 如果MA5下穿MA20，触发卖出信号
- 如果持股满5天，强制卖出
3. **轮动买入**：卖出后释放资金，继续买入新的符合条件的股票
**风险提示**：
- ⚠️ 本策略为模拟演示，实际交易存在滑点、手续费等成本
- ⚠️ 历史业绩不代表未来收益
- ⚠️ 请谨慎评估风险，理性投资
""")
def send_dingtalk_notification(stocks_df: pd.DataFrame, top_n: int):
"""发送钉钉通知"""
try:
# 检查webhook配置
webhook_config = notification_service.get_webhook_config_status()
if not webhook_config['enabled'] or not webhook_config['configured']:
st.info("💡 未配置Webhook通知，如需接收钉钉消息请在环境配置中设置")
return
# 构建消息内容
keyword = notification_service.config.get('webhook_keyword', 'aiagents通知')
message_text = f"### {keyword} - 低价擒牛选股完成\n\n"
message_text += f"**筛选策略**: 股价<10元 + 净利润增长率≥100% + 深圳A股\n\n"
message_text += f"**筛选数量**: {len(stocks_df)} 只\n\n"
message_text += f"**精选股票**:\n\n"
for idx, row in stocks_df.head(top_n).iterrows():
code = row.get('股票代码', '')
name = row.get('股票简称', '')
# 只显示有效的信息
message_text += f"{idx+1}. **{code} {name}**\n"
# 股价
price = row.get('股价', row.get('最新价', None))
if price is not None and not pd.isna(price) and str(price).strip() not in ['', 'N/A']:
try:
price_float = float(price)
message_text += f" - 股价: {price_float:.2f}元\n"
except:
pass
# 净利润增长率
growth = row.get('净利润增长率', row.get('净利润同比增长率', None))
if growth is not None and not pd.isna(growth) and str(growth).strip() not in ['', 'N/A']:
try:
growth_float = float(growth)
message_text += f" - 净利增长: {growth_float:.2f}%\n"
except:
pass
# 成交额
turnover = row.get('成交额', None)
if turnover is not None and not pd.isna(turnover) and str(turnover).strip() not in ['', 'N/A']:
try:
turnover_float = float(turnover)
if turnover_float >= 100000000: # 亿
message_text += f" - 成交额: {turnover_float/100000000:.2f}亿元\n"
elif turnover_float >= 10000: # 万
message_text += f" - 成交额: {turnover_float/10000:.2f}万元\n"
else:
message_text += f" - 成交额: {turnover_float:.2f}元\n"
except:
pass
# 所属行业
industry = row.get('所属行业', row.get('所属同花顺行业', None))
if industry is not None and not pd.isna(industry) and str(industry).strip() not in ['', 'N/A']:
message_text += f" - 所属行业: {industry}\n"
message_text += "\n"
message_text += f"**生成时间**: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\n\n"
message_text += "_此消息由AI股票分析系统自动发送_"
# 直接发送钉钉Webhook（不使用notification_service的默认格式）
if notification_service.config['webhook_type'] == 'dingtalk':
import requests
data = {
"msgtype": "markdown",
"markdown": {
"title": f"{keyword}",
"text": message_text
}
}
try:
response = requests.post(
notification_service.config['webhook_url'],
json=data,
headers={'Content-Type': 'application/json'},
timeout=10
)
if response.status_code == 200:
result = response.json()
if result.get('errcode') == 0:
st.success("✅ 已发送钉钉通知")
else:
st.warning(f"⚠️ 钉钉通知发送失败: {result.get('errmsg')}")
else:
st.warning(f"⚠️ 钉钉通知请求失败: HTTP {response.status_code}")
except Exception as e:
st.warning(f"⚠️ 发送钉钉通知失败: {str(e)}")
except Exception as e:
st.warning(f"⚠️ 发送通知时出错: {str(e)}")