记录一下炒股用到的指标，主要是介绍原理和计算公式，包括不同的库可能计算方式不一样，整理了一下网上的资料。
CCI指标
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
| def CCI(stock, timeperiod=14):
'''
计算CCI指标
'''
cci = 0
security_data = attribute_history(stock, timeperiod*2, '1d',['close','high','low'] , df=True)
nan_count = list(np.isnan(security_data['close'])).count(True)
if nan_count == len(security_data['close']):
log.info("股票 %s 输入数据全是 NaN，该股票可能已退市或刚上市，返回 NaN 值数据。" %stock)
cci[stock] = array([np.nan])
else:
close_CCI = security_data['close']
high_CCI = security_data['high']
low_CCI = security_data['low']
cci = talib.CCI(high_CCI, low_CCI, close_CCI, timeperiod)
return cci
|
RSI指标
1
2
3
4
5
6
7
8
9
| def RSI(stock, timeperiod=14):
'''
计算RSI指标
'''
security_data = attribute_history(stock, timeperiod*2, '1d', ['close'] , df=True)
rsi = talib.RSI(security_data['close'], timeperiod)
return rsi
|
KDJ指标
1
2
3
4
5
6
7
8
9
10
11
12
13
14
| def KDJ(stock, timeperiod=14):
'''
计算KDJ指标
'''
security_data = attribute_history(stock, timeperiod*2, '1d', ['close', 'high', 'low'], df=True)
k, d = talib.STOCH(security_data['high'].values,
security_data['low'].values,
security_data['close'].values,
fastk_period=9,
slowk_period=2*3-1,
slowk_matype=1,
slowd_period=2*3-1,
slowd_matype=1)
return k, d
|
KDJ使用talib库计算的时候需要注意一下参数的设置，如果使用默认参数，则算出来的数值与通达信等软件上的不一样，具体可以参考：
https://xueqiu.com/1747761477/198676825
