Dual Thrust Breakout Entries and Position Exits
Summary
This code describes a Dual Thrust style breakout strategy applied across a list of instruments. It calculates a reference range from prior bars using the highest high and close and the lowest low and close, then scales that range by separate parameters to set upper and lower thresholds around the current bar’s open. A long position opens when the bar’s high reaches the upper threshold; a short opens when its low reaches the lower threshold, except that short trades are disabled for stocks.
Open positions exit when price reaches the opposite threshold. Stock positions use a trading unit of 100, while other instruments use one. The example supplies no performance results, backtest details, or guidance on parameter selection. Its comments describe daily ranges, but the implementation slices the supplied bar arrays directly, so the meaning of that lookback depends on the bar period and data passed in. It also uses bar highs and lows to detect threshold crossings, without specifying execution prices or slippage.
Key ideas
- The strategy builds a range from prior highs, lows, and closes.
- Separate multipliers set breakout thresholds above and below the current bar open.
- A threshold crossing opens a position, while a crossing of the opposite threshold closes it.
- Short entries and exits are disabled when the strategy is configured for stocks.
- The example provides no evidence of profitability or explicit execution and risk controls.
Tags
Full text
# DualThrust_Sel.py
```py
from wtpy import BaseSelStrategy
from wtpy import SelContext
import numpy as np
class StraDualThrustSel(BaseSelStrategy):
def __init__(self, name, codes:list, barCnt:int, period:str, days:int, k1:float, k2:float, isForStk:bool = False):
BaseSelStrategy.__init__(self, name)
self.__days__ = days
self.__k1__ = k1
self.__k2__ = k2
self.__period__ = period
self.__bar_cnt__ = barCnt
self.__codes__ = codes
self.__is_stk__ = isForStk
def on_init(self, context:SelContext):
return
def on_calculate(self, context:SelContext):
curTime = context.stra_get_time()
trdUnit = 1
if self.__is_stk__:
trdUnit = 100
for code in self.__codes__:
sInfo = context.stra_get_sessioninfo(code)
if not sInfo.isInTradingTime(curTime):
continue
#读取最近50条1分钟线(dataframe对象)
theCode = code
if self.__is_stk__:
theCode = theCode + "Q"
df_bars = context.stra_get_bars(theCode, self.__period__, self.__bar_cnt__)
#把策略参数读进来，作为临时变量，方便引用
days = self.__days__
k1 = self.__k1__
k2 = self.__k2__
#平仓价序列、最高价序列、最低价序列
closes = df_bars.closes
highs = df_bars.highs
lows = df_bars.lows
#读取days天之前到上一个交易日位置的数据
hh = np.amax(highs[-days:-1])
hc = np.amax(closes[-days:-1])
ll = np.amin(lows[-days:-1])
lc = np.amin(closes[-days:-1])
#读取今天的开盘价、最高价和最低价
# lastBar = df_bars.get_last_bar()
openpx = df_bars.opens[-1]
highpx = df_bars.highs[-1]
lowpx = df_bars.lows[-1]
'''
!!!!!这里是重点
1、首先根据最后一条K线的时间，计算当前的日期
2、根据当前的日期，对日线进行切片,并截取所需条数
3、最后在最终切片内计算所需数据
'''
#确定上轨和下轨
upper_bound = openpx + k1* max(hh-lc,hc-ll)
lower_bound = openpx - k2* max(hh-lc,hc-ll)
#读取当前仓位
curPos = context.stra_get_position(code)/trdUnit
if curPos == 0:
if highpx >= upper_bound:
context.stra_set_position(code, 1*trdUnit, 'enterlong')
context.stra_log_text("{} 向上突破{}>={}，多仓进场".format(code, highpx, upper_bound))
continue
if lowpx <= lower_bound and not self.__is_stk__:
context.stra_set_position(code, -1*trdUnit, 'entershort')
context.stra_log_text("{} 向下突破{}<={}，空仓进场".format(code, lowpx, lower_bound))
continue
elif curPos > 0:
if lowpx <= lower_bound:
context.stra_set_position(code, 0, 'exitlong')
context.stra_log_text("{} 向下突破{}<={}，多仓出场".format(code, lowpx, lower_bound))
#raise Exception("except on purpose")
continue
else:
if highpx >= upper_bound and not self.__is_stk__:
context.stra_set_position(code, 0, 'exitshort')
context.stra_log_text("{} 向上突破{}>={}，空仓出场".format(code, highpx, upper_bound))
continue
def on_tick(self, context:SelContext, code:str, newTick:dict):
return
def on_bar(self, context:SelContext, code:str, period:str, newBar:dict):
return
```Shown in full with attribution under the source's licence. Licence: MIT
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.