问财量化选股策略逻辑
该选股策略选股逻辑为：换手率3%-12%,前25天有涨停,今日控盘＞21。
选股逻辑分析
该选股策略主要通过筛选市场活跃度较高的股票，选择前25天有涨停的股票，并要求今日控盘＞21。其通过控盘来判断股票的活跃度和买卖情况，结合前25天的涨停信号，可能能够选出涨势比较强的股票。
有何风险？
该选股策略风险相对较大。仅通过今日控盘判断买卖情况可能较为单一，同时前25天的涨停信号不一定代表股票未来的涨幅。筛选出的股票需要进行个股分析，对股票的基本面和未来走势进行深入分析。
如何优化？
可以引入其他技术指标，如MACD等，判断股票短期走势，提高选股成功率。同时，加入其他的技术面和基本面指标，如财务指标、行业趋势等，综合考虑股票情况和未来走势。对选股策略进行量化优化，比如引入机器学习算法对股票形态进行判断和分析。
最终的选股逻辑
该选股策略选股逻辑为：筛选换手率3%-12%、前25天有涨停；同时要求今日控盘＞21。
同花顺指标公式代码参考
通达信指标代码：
H:=HIGH;
L:=LOW;
O:=OPEN;
C:=CLOSE;
Up:=IF(O>REF(C,1),O,REF(C,1));
Dn:=IF(O<REF(C,1),O,REF(C,1));
M1:=SUM(MAX(H-L,0),20);
M2:=SUM(MAX(REF(H,1)-REF(L,1),0),20);
M3:=SUM(MAX(H-REF(C,1),REF(C,1)-L),20);
M4:=SUM(MAX(MAX(H-REF(C,1),0),MAX(REF(C,1)-L,0)),20);
M5:=SUM(MAX(H-Ref(L,1),Abs(H-Ref(C,1))),20);
M6:=SUM(MAX(L-Ref(H,1),Abs(L-Ref(C,1))),20);
OSC:(M1/REF(C,1)+M2/REF(C,1)+M3/REF(C,1)+M4/REF(C,1)+M5/REF(C,1)+M6/REF(C,1))*100;
KLINEHIGH:=REF(HHVBARS(20),1)=0 OR (BARSLAST(H>=HHV(H,20))+1=REF(HHVBARS(20),1));
KLINELOW:=REF(LLVBARS(20),1)=0 OR (BARSLAST(L<=LLV(L,20))+1=REF(LLVBARS(20),1));
KLINEUP:=KLINEHIGH;
KLINEUP:=IF(KLINELOW,KLINEUP,0);
KLINEUP:=IF(H==HHV(H,20),1,KLINEUP);
KLINEDOWN:=KLINELOW;
KLINEDOWN:=IF(KLINEHIGH,KLINEDOWN,0);
KLINEDOWN:=IF(L==LLV(L,20),1,KLINEDOWN);
CONTROLPLATE:=VOL/C*NJ/10000>21;
python代码参考
import pandas as pd
import numpy as np
import time
from datetime import datetime, timedelta
from pytdx.hq import TdxHq_API
from pytdx.util.best_ip import select_best_ip
api = TdxHq_API()
ip = select_best_ip('tdx')
api.connect(ip['ip'], ip['port'])
end_date = datetime.today().strftime('%Y-%m-%d')
start_date = (datetime.today() - timedelta(days=25)).strftime('%Y-%m-%d')
all_stocks = api.get_security_list(0, 0)
df_stocks = pd.DataFrame(all_stocks, columns=['code', 'name', 'market_type', 'exchange_type',
'industry', 'list_date', 'delist_date', 'infolevel'])
selected_stocks = []
for code in df_stocks['code']:
if code.startswith('60') and df_stocks[df_stocks['code'] == code]['market_type'].values[0] != 5 and \
df_stocks[df_stocks['code'] == code]['delist_date'].values[0] == '':
if not api.is_ST(code) and api.get_stock_list_by_market(1)['name'].tolist().count(df_stocks[df_stocks['code'] == code]['name'].values[0]) == 0 and \
api.get_stock_list_by_market(2)['name'].tolist().count(df_stocks[df_stocks['code'] == code]['name'].values[0]) == 0 and \
df_stocks[df_stocks['code'] == code]['list_date'].values[0] <= '2000-01-01':
try:
stock_info = api.get_security_quotes([code])[0]
k_data = api.get_security_bars(9, 0, code, 4, datetime.today().strftime('%Y-%m-%d'))
k_data.set_index('date', inplace=True)
rsv = (k_data['close'] - k_data['low'].rolling(9).min()) / (k_data['high'].rolling(9).max() - k_data['low'].rolling(9).min()) * 100
k_data['k'] = rsv.rolling(3).mean()
k_data['d'] = k_data['k'].rolling(3).mean()
k_data['j'] = 3 * k_data['k'] - 2 * k_data['d']
k_data['up'] = k_data['open'].shift(1).combine(k_data['close'].shift(1), max)
k_data['dn'] = k_data['open'].shift(1).combine(k_data['close'].shift(1), min)
k_data['pdi'] = (k_data['high'] - k_data['high'].shift(1)).combine(0, max)
k_data['mdi'] = (k_data['low'].shift(1) - k_data['low']).combine(0, max)
true_range = (k_data['high'] - k_data['low']).combine((k_data['high'] - k_data['close'].shift(1)).abs().combine((k_data['low'] - k_data['close'].shift(1)).abs(), max), max)
k_data['tr'] = true_range.rolling(14).sum()
k_data['atr'] = k_data['tr'].rolling(14).mean()
k_data['di1'] = k_data['pdi'] / k_data['atr']
k_data['di2'] = k_data['mdi'] / k_data['atr']
k_data['adx'] = (k_data['di1'] - k_data['di2']).abs().rolling(14).mean() / (k_data['di1'] + k_data['di2']).rolling(14).mean() * 100
k_data['signal'] = np.where((k_data['k'] > k_data['d']) & (k_data['k'].shift(1) <= k_data['d'].shift(1)), 1, 0)
k_data['osc'] = ((k_data['high'] - k_data['low']).rolling(20).max() + (k_data['close'] - k_data['close'].shift(1)).rolling(20).apply(lambda x: x[x > 0].sum())) / k_data['close'].shift(1).rolling(20).sum() * 100
k_data['kline_up'] = ((k_data['high'] == k_data['high'].rolling(20).max()).shift(1) | ((k_data['high'] > k_data['high'].shift(1)) & (k_data['high'] > k_data['high'].shift(-1)))).astype(int)
k_data['kline_down'] = ((k_data['low'] == k_data['low'].rolling(20).min()).shift(1) | ((k_data['low'] < k_data['low'].shift(1)) & (k_data['low'] < k_data['low'].shift(-1)))).astype(int)
k_data['control_plate'] = (k_data['vol'] / 100 / (k_data['ma5'] * 100000000 / k_data['close']) > 0.21).astype(int)
k_data = k_data.loc[(k_data.index >= start_date) & (k_data.index <= end_date)]
if len(k_data) > 0 and \
k_data['turnover'].quantile(0.7) >= 0.03 and \
k_data['turnover'].quantile(0.7) <= 0.12 and \
len(k_data[k_data['pct_chg'] >= 9.90]) >= 1 and \
k_data.iloc[-1]['control_plate'] == 1:
selected_stocks.append({'code': code,
'name': df_stocks[df_stocks['code'] == code]['name'].values[0],
'market_type': df_stocks[df_stocks['code'] == code]['market_type'].values[0],
'industry': df_stocks[df_stocks['code'] == code]['industry'].values[0],
'infolevel': df_stocks[df_stocks['code'] == code]['infolevel'].values[0]})
except Exception as e:
print(e)
continue
df_selected_stocks = pd.DataFrame(selected_stocks)
api.disconnect()
## 如何进行量化策略实盘？
请把您优化好的选股语句放入文章最下面模板的选股语句中即可。
select_sentence = '市值小于100亿' #选股语句。
模板如何使用？
点击图标右上方的复制按钮，复制到自己的账户即可使用模板进行回测。
## 如果有任何问题请添加 下方的二维码进群提问。
