def order_by_rsi(f):
# 按照rsi反转下单
HOLCData = Global_variables.get_value("HOLCData")
while not len(HOLCData.index) or list(
HOLCData["date"])[-1] != datetime.datetime.strftime(
datetime.datetime.today() - datetime.timedelta(days=1),
"%Y%m%d"):
HOLCData = Global_variables.get_value("HOLCData")
for security_code, group in HOLCData.groupby("security_code"):
if list(group["security_type"])[0] == "Stock":
close_price = HOLCData["CLOSE"]
change_list = list(close_price.diff())
rise_list = [max(change, 0) for change in change_list]
pre_ratio_rise_list = rise_list[(-21):(-1)]
pre_ratio_change_list = change_list[(-21):(-1)]
pre_rsi_ratio = numpy.mean(pre_ratio_rise_list) / numpy.mean(
[numpy.abs(change) for change in pre_ratio_change_list])
cur_ratio_rise_list = rise_list[:(-20)]
cur_ratio_change_list = change_list[:(-20)]
cur_rsi_ratio = numpy.mean(cur_ratio_rise_list) / numpy.mean(
[numpy.abs(change) for change in cur_ratio_change_list])
pre_short_rise_list = rise_list[(-7):(-1)]
pre_short_change_list = change_list[(-7):(-1)]
pre_short_rsi_ratio = numpy.mean(pre_short_rise_list) / numpy.mean(
[numpy.abs(change) for change in pre_short_change_list])
pre_long_rise_list = rise_list[(-13):(-1)]
pre_long_change_list = change_list[(-12):]
pre_long_rsi_ratio = numpy.mean(pre_long_rise_list) / numpy.mean(
[numpy.abs(change) for change in pre_long_change_list])
cur_short_rise_list = rise_list[(-6):]
cur_short_change_list = change_list[(-6):]
cur_short_rsi_ratio = numpy.mean(cur_short_rise_list) / numpy.mean(
[numpy.abs(change) for change in cur_short_change_list])
cur_long_rise_list = rise_list[(-12):]
cur_long_change_list = change_list[(-12):]
cur_long_rsi_ratio = numpy.mean(cur_long_rise_list) / numpy.mean(
[numpy.abs(change) for change in cur_long_change_list])
if cur_rsi_ratio < 0.3 < pre_rsi_ratio:
signal = "buy"
elif pre_rsi_ratio < 0.7 < cur_rsi_ratio:
signal = "sell"
else:
if pre_short_rsi_ratio < pre_long_rsi_ratio and cur_short_rsi_ratio > cur_long_rsi_ratio:
signal = "buy"
elif pre_short_rsi_ratio > pre_long_rsi_ratio and cur_short_rsi_ratio < cur_long_rsi_ratio:
signal = "sell"
else:
signal = "no_action"
print("rsi:", signal)
order_by_signal(security_code, f, signal)
def order_by_mas(f):
# 根据moving average signal下单
def mas_calcu(prices, volumes):
total_volume = sum(volumes)
return sum([
prices[i] * volumes[i] / total_volume for i in range(len(prices))
])
def mas(group):
# 计算moving average信号
pre_price_list = list(group["RT_LAST"].iloc[(-3602):(-2)])
pre_volume_list = list(group["RT_LAST_VOL"].iloc[(-3602):(-2)])
pre_average = mas_calcu(pre_price_list, pre_volume_list)
pre_pri = group["RT_LAST"].iloc[-2]
cur_price_list = list(group["RT_LAST"].iloc[(-3601):(-1)])
cur_volume_list = list(group["RT_LAST_VOL"].iloc[(-3601):(-1)])
cur_average = mas_calcu(cur_price_list, cur_volume_list)
cur_pri = group["RT_LAST"].iloc[-1]
f.write(
"strategy: moving_average; time: {}; pre_aver_price: {}; pre_price: {}; cur_aver_price: {}; "
"cr_price: {}; ".format(str(max(group["Time"])), pre_average,
pre_pri, cur_average, cur_pri))
print("mas end: {}".format(str(datetime.datetime.now())))
if cur_pri > cur_average and pre_pri < pre_average:
return "sell"
elif cur_pri < cur_average and pre_pri > pre_average:
return "buy"
else:
return "no_action"
def mas_order(security_code, group):
signal = mas(group)
f.write("signal: {}\n".format(signal))
if signal != "no_action":
order_by_signal(security_code, f, signal)
print("mas start: {}".format(str(datetime.datetime.now())))
snapshotdata = Global_variables.get_value("SnapshotData")
all_securities = Global_variables.get_value("all_securities")
print("length of snapshotdata in strategies:", len(snapshotdata.index))
if len(snapshotdata.index) < 3602 * len(all_securities):
print("waiting")
return 0
f.write("snapshotdata: \n{}\n".format(snapshotdata))
for security_code, group in snapshotdata.groupby("security_code"):
if list(group["security_type"])[0] == "Bond":
f.write("security: {}\n".format(security_code))
thread = threading.Thread(target=mas_order,
args=(
security_code,
group,
))
thread.start()
def get_snapshot_data():
# 获取行情信息
print("snapshotdata start: {}".format(str(datetime.datetime.now())))
quote = Global_variables.get_value("quote")
SnapshotData = Global_variables.get_value("SnapshotData")
all_securities = Global_variables.get_value("all_securities")
temp_df = quote.get_snapshot(all_securities)
SnapshotData = pandas.concat([SnapshotData, temp_df])
if len(SnapshotData.index) > 3602 * len(all_securities):
SnapshotData = SnapshotData.iloc[len(all_securities):]
Global_variables.set_value("SnapshotData", SnapshotData)
print("snapshotdata ends: {}".format(str(datetime.datetime.now())))
def get_holc_data():
# 获取高开低收
all_securities = Global_variables.get_value("all_securities")
HOLCData = Global_variables.get_value("HOLCData")
quote = Global_variables.get_value("quote")
if not len(
mongocol_to_dataframe(
quote.MB_holc_Stock.collection.find()).index):
query_dates = []
i = 2
while len(query_dates) < 21:
print("length of query dates: {}".format(len(query_dates)))
query_date = datetime.datetime.today() - datetime.timedelta(days=i)
if is_workday(query_date):
query_dates.append(query_date)
i = i + 1
for query_date in reversed(query_dates):
query_date = datetime.datetime.strftime(query_date, '%Y%m%d')
wind_data = quote.get_ohlc(all_securities, query_date)
HOLCData = pandas.concat([HOLCData, wind_data])
Global_variables.set_value("HOLCData", HOLCData)
if max(HOLCData["date"]) != datetime.datetime.strftime(
datetime.datetime.today() - datetime.timedelta(days=1), '%Y%m%d'):
wind_data = quote.get_ohlc(
all_securities,
datetime.datetime.strftime(
datetime.datetime.today() - datetime.timedelta(days=1),
'%Y%m%d'))
HOLCData = pandas.concat([HOLCData, wind_data])
if len(HOLCData.index) > 22 * len(all_securities):
HOLCData = HOLCData.iloc[len(all_securities):]
Global_variables.set_value("HOLCData", HOLCData)
def order_by_bollinger(f):
# 按照布林通道策略下单
HOLCData = Global_variables.get_value("HOLCData")
while not len(HOLCData.index) or list(
HOLCData["date"])[-1] != datetime.datetime.strftime(
datetime.datetime.today() - datetime.timedelta(days=1),
"%Y%m%d"):
HOLCData = Global_variables.get_value("HOLCData")
for security_code, group in HOLCData.groupby("security_code"):
print(group)
if list(group["security_type"])[0] == "Bond":
high_price = list(group["HIGH"])
low_price = list(group["LOW"])
close_price = list(group["CLOSE"])
price_factor = [
numpy.mean([high_price[i], low_price[i], close_price[i]])
for i in range(len(high_price))
]
pre_aver = numpy.mean(price_factor[:(-2)])
pre_sd = numpy.std(price_factor[:(-2)])
pre_upper = pre_aver + 2 * pre_sd
pre_lower = pre_aver - 2 * pre_sd
cur_aver = numpy.mean(price_factor[1:(-1)])
cur_sd = numpy.std(price_factor[1:(-1)])
cur_upper = cur_aver + 2 * cur_sd
cur_lower = cur_aver - 2 * cur_sd
pre_price = price_factor[-2]
cur_price = price_factor[-1]
if pre_price < pre_upper and cur_price > cur_upper:
signal = "sell"
elif pre_price > pre_lower and cur_price < cur_lower:
signal = "buy"
else:
signal = "no_action"
print(pre_price, pre_lower, pre_upper, cur_price, cur_lower,
cur_upper)
order_by_signal(security_code, f, signal)
def order_by_time_signal(f):
# 按照时间信号下单
all_securities = Global_variables.get_value("all_securities")
def time_order(security_code):
order_by_signal(security_code, f, signal="buy")
for security in all_securities:
if security.security == "Stock":
thread = threading.Thread(target=time_order,
args=(security.code, ))
thread.start()
def order_by_signal(security_code, f, signal):
etBase = Global_variables.get_value("etBase")
SClient = Global_variables.get_value("SClient")
connection_signal = Global_variables.get_value("connection_signal")
volume_signal = Global_variables.get_value("volume_signal")
StartPosition = Global_variables.get_value("StartPosition")
RtPosition = Global_variables.get_value("RtPosition")
SnapshotData = Global_variables.get_value("SnapshotData")
security_data = SnapshotData[SnapshotData["security_code"] ==
security_code]
price = numpy.mean(security_data["RT_LAST"])
while not list(RtPosition.columns):
RtPosition = Global_variables.get_value("RtPosition")
if connection_signal:
if signal == "buy":
if volume_signal:
# 个股持仓增长超过100%时停止买入
position = list(RtPosition[RtPosition["证券代码"] == ".".split(
security_code)[0]]["参考市值"])[0]
start_position = list(
StartPosition[StartPosition["证券代码"] == ".".split(
security_code)[0]]["参考市值"])[0]
if position < 2 * start_position:
order = Order(security_code,
"buy",
100,
price=price * 0.99)
ord = order.make_order(etBase)
f.write(str(ord))
else:
SClient.chat_postMessage(
channel="#quantitative-investment",
text="{}持仓增长过快".format(security_code))
elif signal == "sell":
if ".".split(security_code)[0] in list(RtPosition["证券代码"]):
order = Order(security_code, "sell", 100, price=price * 1.01)
ord = order.make_order(etBase)
f.write(str(ord))
def check_wind_connection():
# wind连接检测
SClient = Global_variables.get_value("SClient")
SnapshotData = Global_variables.get_value("SnapshotData")
while (not len(SnapshotData.index)) or (
"000300.SH" not in SnapshotData["security_code"]):
pass
while datetime.datetime.now().hour < 15:
connection_signal = Global_variables.get_value("connection_signal")
SnapshotData = Global_variables.get_value(["SnapshotData"])
hs_index_df = SnapshotData.groupby("security_code")["000300.SH"]
hs_index_price = list(hs_index_df["rt_last"])
if hs_index_price[-1] == hs_index_price[-2] and hs_index_price[
-2] == hs_index_price[-3]:
Global_variables.set_value("connection_signal", 0)
SClient.chat_postMessage(channel="#quantitative-investment",
text="wind断线")
else:
if not connection_signal:
SClient.chat_postMessage(channel="#quantitative-investment",
text="wind恢复连接")
Global_variables.set_value("connection_signal", 1)
from model.Monitor import check_easytrader_connection, check_wind_connection
from model.Strategies import get_snapshot_data, get_holc_data, order_by_mas, order_by_time_signal, order_by_bollinger, order_by_rsi
from model import Global_variables
if __name__ == '__main__':
scheduler = Global_variables.get_value("scheduler")
with open(".//log.txt", "w") as f:
scheduler.add_job(check_easytrader_connection,
'cron',
day_of_week='0-4',
hour='10',
minute='23',
second='00',
id='check_easytrader_connection',
args=[])
scheduler.add_job(check_wind_connection,
'cron',
day_of_week='0-4',
hour='10',
minute='23',
second='00',
id='check_wind_connection',
args=[])
scheduler.add_job(get_snapshot_data,
'cron',
day_of_week='0-4',
hour='9-16',
minute='00-59',
second='00-59',
def check_easytrader_connection():
# easytrader断线重连兼仓位监测
etBase = Global_variables.get_value("etBase")
SClient = Global_variables.get_value("SClient")
mg_position = MongoBase("amy_inside", "Arthur", "Position")
try:
start_balance = etBase.balance[0]
except:
try:
try:
etBase.exit()
etBase = easytrader.use('yh_client')
etBase.prepare(user='******',
password='******',
exe_path='C:\\中国银河证券双子星3.2\\xiadan.exe')
except Exception as e:
print(str(e))
# 客户端无法自动登录时模拟enter按键登录
win32api.keybd_event(13, 0, 0, 0)
win32api.keybd_event(13, 0, win32con.KEYEVENTF_KEYUP, 0)
Global_variables.set_value("etBase", etBase)
start_balance = etBase.balance[0]
except Exception as e:
print("cannot reconnect to easyTrader: ", str(e))
Global_variables.set_value("connection_signal", 0)
return 1
StartPosition = pandas.DataFrame(etBase.position)
StartPosition["start_or_end"] = "start"
StartPosition["Date"] = str(datetime.datetime.today())
Global_variables.set_value("StartPosition", StartPosition)
mg_position.insertDB(StartPosition)
print("while reached")
while datetime.datetime.now().hour < 15:
try:
volume_signal = Global_variables.get_value("volume_signal")
balance = etBase.balance[0]
RtPosition = pandas.DataFrame(etBase.position)
print("RtPosition:\n", RtPosition)
Global_variables.set_value("RtPosition", RtPosition)
if volume_signal:
if balance["总市值"] / balance["总资产"] >= 0.8:
Global_variables.set_value("volume_signal", 0)
SClient.chat_postMessage(
channel="#quantitative-investment", text="持仓过多")
if balance["总市值"] / start_balance["总市值"] > 1.3:
Global_variables.set_value("volume_signal", 0)
SClient.chat_postMessage(
channel="#quantitative-investment", text="持仓增加过快")
if not volume_signal:
if balance["总市值"] / balance["总资产"] < 0.8 and balance[
"总市值"] / StartPosition["总市值"] < 1.3:
Global_variables.set_value("volume_signal", 1)
SClient.chat_postMessage(
channel="#quantitative-investment", text="解除持仓过多")
except:
etBase.exit()
try:
etBase = easytrader.use('yh_client')
etBase.prepare(user='******',
password='******',
exe_path='C:\\中国银河证券双子星3.2\\xiadan.exe')
except:
# 客户端无法自动登录时模拟enter按键登录
win32api.keybd_event(13, 0, 0, 0)
win32api.keybd_event(13, 0, win32con.KEYEVENTF_KEYUP, 0)
Global_variables.set_value("etBase", etBase)
end_position = pandas.DataFrame(etBase.position)
end_position["start_or_end"] = "end"
end_position["Date"] = str(datetime.datetime.now().date())
mg_position.insertDB(end_position)