MA、EMA、MACD、BOLL、KDJ指标计算
MA、EMA、MACD、BOLL、KDJ指标计算
# encoding:utf-8
import os
import sys
import pandas as qh_pd
import time
import datetime
import talib
if __name__ == "__main__":
print("本文件引用")
qh_path = os.path.abspath(os.path.join(os.getcwd(), "..")) # 返回当前工作目录 先到达 QH_网络配置 的父文件夹路径
sys.path.append(qh_path) # 添加自己指定的搜索路径
else:
print("外文件引用")
qh_path = os.path.abspath(os.path.join(os.getcwd(), "..")) # 返回当前工作目录 先到达 QH_网络配置 的父文件夹路径
sys.path.append(qh_path) # 添加自己指定的搜索路径
import QH_Stock_Treasure_Box.qh_parameter_chuanru as qh_chcan # 传入参数
qh_pd.set_option('display.max_columns', None)
qh_pd.set_option('display.width',1000)
def Qh_Target_MA(qh_df,qh_close = "收盘价",qh_timeperiod = 5):
"""
计算MA指标， 作者：阙辉
:param qh_df:
:param qh_close:
:param qh_timeperiod:
:return:
"""
qh_cloumn = "MA_{}".format(qh_timeperiod)
qh_df[qh_cloumn] = talib.MA(qh_df[qh_close], timeperiod=qh_timeperiod, matype=0)
return qh_df
def Qh_Target_EMA(qh_df,qh_close = "收盘价",qh_timeperiod = 5):
"""
计算EMA指标， 作者：阙辉
:param qh_df:
:param qh_close:
:param qh_timeperiod:
:return:
"""
qh_cloumn = "EMA_{}".format(qh_timeperiod)
qh_df[qh_cloumn] = talib.EMA(qh_df[qh_close], timeperiod=qh_timeperiod)
return qh_df
def Qh_Target_MACD(qh_df,qh_close = "收盘价",qh_fastperiod=12, qh_slowperiod=26, qh_signalperiod=9):
"""
计算MACD指标， 作者：阙辉
:param qh_df:
:param qh_close:
:param qh_fastperiod:
:param qh_slowperiod:
:param qh_signalperiod:
:return:
"""
qh_diff = "DIFF_{}".format(qh_slowperiod)
qh_dea = "DEA_{}".format(qh_slowperiod)
qh_macd = "MACD_{}".format(qh_slowperiod)
qh_df[qh_diff], qh_df[qh_dea], qh_df[qh_macd] = talib.MACD(qh_df[qh_close], fastperiod=qh_fastperiod,
slowperiod=qh_slowperiod, signalperiod=qh_signalperiod)
qh_df[qh_macd] = qh_df[qh_macd] * 2 # 修正macd
return qh_df
def Qh_Target_BOLL(qh_df,qh_close = "收盘价",qh_timeperiod=20, qh_nbdev=2):
"""
计算BOLL指标， 作者：阙辉
:param qh_df:
:param qh_close:
:param qh_timeperiod:
:param qh_nbdev:
:return:
"""
qh_uuper = "UPPER_{}".format(qh_timeperiod)
qh_mid = "MID_{}".format(qh_nbdev)
qh_lower = "LOWER_{}".format(qh_nbdev)
qh_df[qh_uuper], qh_df[qh_mid], qh_df[qh_lower] = talib.BBANDS(qh_df[qh_close], timeperiod=qh_timeperiod,
nbdevup=qh_nbdev, nbdevdn=qh_nbdev, matype=0)
return qh_df
def Qh_Target_KDJ(qh_df,qh_hig = "最高价",qh_low = "最低价",qh_close = "收盘价",qh_fastk=9, qh_slowk=3,qh_slowd=3):
"""
计算KDJ指标， 作者：阙辉
:param qh_df:
:param qh_hig:
:param qh_low:
:param qh_close:
:param qh_fastk:
:param qh_slowk:
:param qh_slowd:
:return:
"""
qh_K = "K_{}".format(qh_fastk)
qh_D = "D_{}".format(qh_fastk)
qh_J = "J_{}".format(qh_fastk)
qh_df[qh_K], qh_df[qh_D] = talib.STOCH(qh_df[qh_hig],qh_df[qh_low],qh_df[qh_close],
fastk_period=qh_fastk,slowk_period=qh_slowk,
slowk_matype=1,slowd_period=qh_slowd,slowd_matype=1)
qh_df[qh_J] = 3 * qh_df[qh_K] - 2 * qh_df[qh_D]
return qh_df
def Qh_Target_KDJ_CN(qh_df,qh_hig = "最高价",qh_low = "最低价",qh_close = "收盘价",qh_fastk_period=9,qh_slowk_period=3,qh_fastd_period=3):
"""
计算KDJ指标 作者：阙辉
:param qh_higt:
:param qh_low:
:param qh_close:
:param qh_fastk_period:
:param qh_slowk_period:
:param qh_fastd_period:
:return:
"""
QH_K = "QH_K_{}".format(qh_fastk_period)
QH_D = "QH_D_{}".format(qh_fastk_period)
QH_J = "QH_J_{}".format(qh_fastk_period)
qh_df[qh_low] = qh_df[qh_low].astype("float64")
qh_df[qh_low] = qh_df[qh_low].fillna(0)
QH_MinLow = qh_df[qh_low].rolling(qh_fastk_period, min_periods=qh_fastk_period).min() # 最低价
QH_MinLow.fillna(value=qh_df[qh_low].expanding().min(), inplace=True) # 填充空值 NaN
qh_df[qh_hig] = qh_df[qh_hig].astype("float64")
qh_df[qh_hig] = qh_df[qh_hig].fillna(0)
QH_MaxHigh = qh_df[qh_hig].rolling(qh_fastk_period, min_periods=qh_fastk_period).max() # 最高价
QH_MaxHigh.fillna(value=qh_df[qh_hig].expanding().max(), inplace=True) # 填充空值 NaN
qh_df[qh_close] = qh_df[qh_close].astype("float64")
qh_df[qh_close] = qh_df[qh_close].fillna(0)
# RSV 公式：RSV = (收盘价 - 最低价）/（最高价 - 最低价） * 100 n日RSV=（Cn－Ln）÷（Hn－Ln）×100
QH_RSV = (qh_df[qh_close] - QH_MinLow) / (QH_MaxHigh - QH_MinLow) * 100
QH_RSV = QH_RSV.fillna(0)
# 当日K值=2/3×前一日K值＋1/3×当日RSV
qh_df[QH_K] = QH_RSV.ewm(adjust=False, alpha=1 / qh_slowk_period).mean()
# 当日D值=2/3×前一日D值＋1/3×当日K值
qh_df[QH_D] = qh_df[QH_K].ewm(adjust=False, alpha=1 / qh_fastd_period).mean()
# J=3D—2K
qh_df[QH_J] = 3 * qh_df[QH_K] - 2 * qh_df[QH_D]
return qh_df
if __name__ == "__main__":
qh_code = "SZ000001"
qh_times = "天"
qh_fuquan = "不复权"
qh_chaxun_id = "QH_ChaXun_00016"
#获取sql
qh_sql0 = qh_chcan.Qh_CX_Sql_DB(qh_path,qh_chaxun_id = qh_chaxun_id,qh_sql0 ="",qh_ret_sql = True)
qh_sql = qh_sql0.format(qh_code,qh_times,qh_fuquan)
#查询数据
qh_list_sql_ = qh_chcan.Qh_CX_Sql_DB(qh_path,qh_chaxun_id = qh_chaxun_id,qh_sql0 = qh_sql,qh_ret_sql = False,qh_filed = True)
qh_list_sql_filed = qh_list_sql_[1][1][0] #股票代码
qh_list_sql = qh_list_sql_[1][1][1:] #股票历史行情数据
qh_df_stock = qh_pd.DataFrame(qh_list_sql,columns=qh_list_sql_filed)
qh_df_stock = Qh_Target_MA(qh_df_stock,qh_close = "收盘价",qh_timeperiod = 5)
qh_df_stock = Qh_Target_EMA(qh_df_stock, qh_close="收盘价", qh_timeperiod=20)
qh_df_stock = Qh_Target_MACD(qh_df_stock, qh_close="收盘价", qh_fastperiod=12, qh_slowperiod=26, qh_signalperiod=9)
qh_df_stock = Qh_Target_BOLL(qh_df_stock, qh_close="收盘价", qh_timeperiod=20, qh_nbdev=2)
# qh_df_stock = Qh_Target_KDJ(qh_df_stock, qh_hig="最高价", qh_low="最低价", qh_close="收盘价", qh_fastk=9, qh_slowk=3, qh_slowd=3)
qh_df_stock = Qh_Target_KDJ_CN(qh_df_stock, qh_hig="最高价", qh_low="最低价", qh_close="收盘价", qh_fastk_period=9, qh_slowk_period=3,qh_fastd_period=3)
print(qh_df_stock)计算结果
KDJ
BOLL
MACD
