Backtrader 文档学习-Target Orders
第五部分 ipython 代码中，有详细解释持仓价值Value的计算，算是彩蛋。
1. 概述
sizer不能决定操作是买还是卖，意味着需要一个新的概念，通过增加小智能层可以决定买卖，即通过持仓份额可以决定买卖操作。
这就是策略中order_target_xxx方法族的作用。受zipline的方法的启发，提供了简单指定最终target的机会，target实现:
- size 特定资产组合中的股票、合同数量
- value 投资组合中资产的货币单位价值
- percent 投资组合中资产价值的百分比（来自当前投资组合）
在这种情况下，关键是指定最终target，该方法决定操作是买入还是卖出。同样的逻辑适用于3种方法。order_target_size参数设置：
-
如果目标大于头寸，则发出买入指令，差值为目标-头寸大小 ：
- Pos: 0, target: 7 -> buy(size=7 - 0) -> buy(size=7)
- Pos: 3, target: 7 -> buy(size=7 - 3) -> buy(size=4)
- Pos: -3, target: 7 -> buy(size=7 - -3) -> buy(size=10)
- Pos: -3, target: -2 -> buy(size=-2 - -3) -> buy(size=1)
-
如果目标小于头寸，则根据头寸大小-目标的差额发出卖出指令：
- Pos: 0, target: -7 -> sell(size=0 - -7) -> sell(size=7)
- Pos: 3, target: -7 -> sell(size=3 - -7) -> sell(size=10)
- Pos: -3, target: -7 -> sell(size=-3 - -7) -> sell(size=4)
- Pos: 3, target: 2 -> sell(size=3 - 2) -> sell(size=1)
当使用order_target_value确定价值时，投资组合中资产的当前价值和头寸规模都将被考虑在内，以决定最终的基础操作，理由是:
-
如果头寸规模为负（空头），目标价值必须大于当前价值，这意味着卖出更多
按这个逻辑操作：- If target > value and size >=0 -> buy
- If target > value and size < 0 -> sell
- If target < value and size >= 0 -> sell
- If target < value and size < 0 -> buy
order_target_percent 的逻辑与order_target_value的逻辑相同。考虑投资组合的当前总价值来确定资产的目标价值 。
2. 示例
示例中的逻辑相当简单，只是为了测试效果，规则如下：
- 在奇数月份（1月、3月……），order_target_value =day * 1000，模拟了一个不断增加的target 。
- 在偶数月份（2月、4月……），order_target_value = 31 - day ，模拟了一个递减的target 。
（1）order_target_size
0001 - 2005-01-03 - Position Size: 00 - Value 1000000.00
0001 - 2005-01-03 - Order Target Size: 03
0002 - 2005-01-04 - Position Size: 03 - Value 999994.39
0002 - 2005-01-04 - Order Target Size: 04
0003 - 2005-01-05 - Position Size: 04 - Value 999992.48
0003 - 2005-01-05 - Order Target Size: 05
0004 - 2005-01-06 - Position Size: 05 - Value 999988.79
0004 - 2005-01-06 - Order Target Size: 06
0005 - 2005-01-07 - Position Size: 06 - Value 999991.41
0005 - 2005-01-07 - Order Target Size: 07
0006 - 2005-01-10 - Position Size: 07 - Value 999993.89
0006 - 2005-01-10 - Order Target Size: 10
0007 - 2005-01-11 - Position Size: 10 - Value 999987.32
0007 - 2005-01-11 - Order Target Size: 11
0008 - 2005-01-12 - Position Size: 11 - Value 999992.38
0008 - 2005-01-12 - Order Target Size: 12
0009 - 2005-01-13 - Position Size: 12 - Value 999982.68
... ...
0021 - 2005-02-01 - Position Size: 31 - Value 999954.68
0021 - 2005-02-01 - Order Target Size: 30
0022 - 2005-02-02 - Position Size: 30 - Value 999979.65
0022 - 2005-02-02 - Order Target Size: 29
0023 - 2005-02-03 - Position Size: 29 - Value 999966.33
0023 - 2005-02-03 - Order Target Size: 28
0024 - 2005-02-04 - Position Size: 28 - Value 999963.99
0024 - 2005-02-04 - Order Target Size: 27
0025 - 2005-02-07 - Position Size: 27 - Value 999949.19
0025 - 2005-02-07 - Order Target Size: 24
0026 - 2005-02-08 - Position Size: 24 - Value 999947.06
1月份，该target 从第一个交易日3日开始，仓位3，并不断增加。位置大小最初从0移动到3，然后以1为增量。 1月结束最后一个订单目标是31。
当进入2月1日时头寸规模，此时新的target被要求为30，并随着头寸的减少而递减1。
（2） order_target_value
0001 - 2005-01-03 - Position Size: 00 - Value 1000000.00
0001 - 2005-01-03 - Order Target Size: 03
0002 - 2005-01-04 - Position Size: 03 - Value 999994.39
0002 - 2005-01-04 - Order Target Size: 04
0003 - 2005-01-05 - Position Size: 04 - Value 999992.48
0003 - 2005-01-05 - Order Target Size: 05
0004 - 2005-01-06 - Position Size: 05 - Value 999988.79
0004 - 2005-01-06 - Order Target Size: 06
0005 - 2005-01-07 - Position Size: 06 - Value 999991.41
0005 - 2005-01-07 - Order Target Size: 07
0006 - 2005-01-10 - Position Size: 07 - Value 999993.89
0006 - 2005-01-10 - Order Target Size: 10
0007 - 2005-01-11 - Position Size: 10 - Value 999987.32
... ...
0020 - 2005-01-31 - Position Size: 28 - Value 999968.70
0020 - 2005-01-31 - Order Target Size: 31
0021 - 2005-02-01 - Position Size: 31 - Value 999954.68
0021 - 2005-02-01 - Order Target Size: 30
0022 - 2005-02-02 - Position Size: 30 - Value 999979.65
0022 - 2005-02-02 - Order Target Size: 29
0023 - 2005-02-03 - Position Size: 29 - Value 999966.33
0023 - 2005-02-03 - Order Target Size: 28
0024 - 2005-02-04 - Position Size: 28 - Value 999963.99
0024 - 2005-02-04 - Order Target Size: 27
0025 - 2005-02-07 - Position Size: 27 - Value 999949.19
0025 - 2005-02-07 - Order Target Size: 24
（3）order_target_percent
0001 - 2005-01-03 - Position Size: 00 - Value 1000000.00
0001 - 2005-01-03 - data percent 0.00
0001 - 2005-01-03 - Order Target Percent: 0.03
0002 - 2005-01-04 - Position Size: 785 - Value 998532.05
0002 - 2005-01-04 - data percent 0.03
0002 - 2005-01-04 - Order Target Percent: 0.04
0003 - 2005-01-05 - Position Size: 1091 - Value 998007.44
0003 - 2005-01-05 - data percent 0.04
0003 - 2005-01-05 - Order Target Percent: 0.05
0004 - 2005-01-06 - Position Size: 1381 - Value 996985.64
0004 - 2005-01-06 - data percent 0.05
0004 - 2005-01-06 - Order Target Percent: 0.06
0005 - 2005-01-07 - Position Size: 1688 - Value 997708.36
0005 - 2005-01-07 - data percent 0.06
0005 - 2005-01-07 - Order Target Percent: 0.07
0006 - 2005-01-10 - Position Size: 1942 - Value 998397.32
0006 - 2005-01-10 - data percent 0.07
0006 - 2005-01-10 - Order Target Percent: 0.10
... ...
0020 - 2005-01-31 - Position Size: 7985 - Value 991966.28
0020 - 2005-01-31 - data percent 0.28
0020 - 2005-01-31 - Order Target Percent: 0.31
0021 - 2005-02-01 - Position Size: 8733 - Value 988008.94
0021 - 2005-02-01 - data percent 0.31
0021 - 2005-02-01 - Order Target Percent: 0.30
0022 - 2005-02-02 - Position Size: 8530 - Value 995005.45
0022 - 2005-02-02 - data percent 0.30
0022 - 2005-02-02 - Order Target Percent: 0.29
0023 - 2005-02-03 - Position Size: 8120 - Value 991240.75
0023 - 2005-02-03 - data percent 0.29
0023 - 2005-02-03 - Order Target Percent: 0.28
0024 - 2005-02-04 - Position Size: 7910 - Value 990607.25
0024 - 2005-02-04 - data percent 0.28
3. 代码
from __future__ import (absolute_import, division, print_function,
unicode_literals)
import argparse
from datetime import datetime
import backtrader as bt
class TheStrategy(bt.Strategy):
'''
This strategy is loosely based on some of the examples from the Van
K. Tharp book: *Trade Your Way To Financial Freedom*. The logic:
- Enter the market if:
- The MACD.macd line crosses the MACD.signal line to the upside
- The Simple Moving Average has a negative direction in the last x
periods (actual value below value x periods ago)
- Set a stop price x times the ATR value away from the close
- If in the market:
- Check if the current close has gone below the stop price. If yes,
exit.
- If not, update the stop price if the new stop price would be higher
than the current
'''
params = (
('use_target_size', False),
('use_target_value', False),
('use_target_percent', False),
)
def notify_order(self, order):
if order.status == order.Completed:
pass
if not order.alive():
self.order = None # indicate no order is pending
def start(self):
self.order = None # sentinel to avoid operrations on pending order
def next(self):
dt = self.data.datetime.date()
portfolio_value = self.broker.get_value()
print('%04d - %s - Position Size: %02d - Value %.2f' %
(len(self), dt.isoformat(), self.position.size, portfolio_value))
data_value = self.broker.get_value([self.data])
if self.p.use_target_value:
print('%04d - %s - data value %.2f' %
(len(self), dt.isoformat(), data_value))
elif self.p.use_target_percent:
port_perc = data_value / portfolio_value
print('%04d - %s - data percent %.2f' %
(len(self), dt.isoformat(), port_perc))
if self.order:
return # pending order execution
size = dt.day
if (dt.month % 2) == 0:
size = 31 - size
if self.p.use_target_size:
target = size
print('%04d - %s - Order Target Size: %02d' %
(len(self), dt.isoformat(), size))
self.order = self.order_target_size(target=size)
elif self.p.use_target_value:
value = size * 1000
print('%04d - %s - Order Target Value: %.2f' %
(len(self), dt.isoformat(), value))
self.order = self.order_target_value(target=value)
elif self.p.use_target_percent:
percent = size / 100.0
print('%04d - %s - Order Target Percent: %.2f' %
(len(self), dt.isoformat(), percent))
self.order = self.order_target_percent(target=percent)
def runstrat(args=None):
args = parse_args(args)
cerebro = bt.Cerebro()
cerebro.broker.setcash(args.cash)
dkwargs = dict()
if args.fromdate is not None:
dkwargs['fromdate'] = datetime.strptime(args.fromdate, '%Y-%m-%d')
if args.todate is not None:
dkwargs['todate'] = datetime.strptime(args.todate, '%Y-%m-%d')
# data
data = bt.feeds.YahooFinanceCSVData(dataname=args.data, **dkwargs)
cerebro.adddata(data)
# strategy
cerebro.addstrategy(TheStrategy,
use_target_size=args.target_size,
use_target_value=args.target_value,
use_target_percent=args.target_percent)
cerebro.run()
if args.plot:
pkwargs = dict(style='bar')
if args.plot is not True: # evals to True but is not True
npkwargs = eval('dict(' + args.plot + ')') # args were passed
pkwargs.update(npkwargs)
cerebro.plot(**pkwargs)
def parse_args(pargs=None):
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter,
description='Sample for Order Target')
parser.add_argument('--data', required=False,
default='./datas/yhoo-1996-2015.txt',
help='Specific data to be read in')
parser.add_argument('--fromdate', required=False,
default='2005-01-01',
help='Starting date in YYYY-MM-DD format')
parser.add_argument('--todate', required=False,
default='2006-12-31',
help='Ending date in YYYY-MM-DD format')
parser.add_argument('--cash', required=False, action='store',
type=float, default=1000000,
help='Ending date in YYYY-MM-DD format')
pgroup = parser.add_mutually_exclusive_group(required=True)
pgroup.add_argument('--target-size', required=False, action='store_true',
help=('Use order_target_size'))
pgroup.add_argument('--target-value', required=False, action='store_true',
help=('Use order_target_value'))
pgroup.add_argument('--target-percent', required=False,
action='store_true',
help=('Use order_target_percent'))
# Plot options
parser.add_argument('--plot', '-p', nargs='?', required=False,
metavar='kwargs', const=True,
help=('Plot the read data applying any kwargs passed\n'
'\n'
'For example:\n'
'\n'
' --plot style="candle" (to plot candles)\n'))
if pargs is not None:
return parser.parse_args(pargs)
return parser.parse_args()
if __name__ == '__main__':
runstrat()
4. Help
python ./order_target.py --help
usage: order_target.py [-h] [--data DATA] [--fromdate FROMDATE]
[--todate TODATE] [--cash CASH]
(--target-size | --target-value | --target-percent)
[--plot [kwargs]]
Sample for Order Target
optional arguments:
-h, --help show this help message and exit
--data DATA Specific data to be read in (default:
./datas/yhoo-1996-2015.txt)
--fromdate FROMDATE Starting date in YYYY-MM-DD format (default:
2005-01-01)
--todate TODATE Ending date in YYYY-MM-DD format (default: 2006-12-31)
--cash CASH Ending date in YYYY-MM-DD format (default: 1000000)
--target-size Use order_target_size (default: False)
--target-value Use order_target_value (default: False)
--target-percent Use order_target_percent (default: False)
--plot [kwargs], -p [kwargs]
Plot the read data applying any kwargs passed For
example: --plot style="candle" (to plot candles)
(default: None)
5.ipython代码
只加载3个月的数据，增加了说明：
from __future__ import (absolute_import, division, print_function,
unicode_literals)
from datetime import datetime
import backtrader as bt
%matplotlib inline
class TheStrategy(bt.Strategy):
'''
This strategy is loosely based on some of the examples from the Van
K. Tharp book: *Trade Your Way To Financial Freedom*. The logic:
- Enter the market if:
- The MACD.macd line crosses the MACD.signal line to the upside
- The Simple Moving Average has a negative direction in the last x
periods (actual value below value x periods ago)
- Set a stop price x times the ATR value away from the close
- If in the market:
- Check if the current close has gone below the stop price. If yes,
exit.
- If not, update the stop price if the new stop price would be higher
than the current
'''
params = (
('use_target_size', False),
('use_target_value', False),
('use_target_percent', False),
)
def notify_order(self, order):
if order.status == order.Completed:
pass
# 没有挂起的order ，order置空
if not order.alive():
self.order = None # indicate no order is pending
def start(self):
# 避免有挂起order未处理，置空
self.order = None # sentinel to avoid operrations on pending order
def next(self):
# 当前data的日期，去时分秒
dt = self.data.datetime.date()
# 投资组合货币价值 ，如何计算出来的？？
portfolio_value = self.broker.get_value()
# 处理data长度 ，日期 ，仓位 ，组合价值
# self.p.use_target_size 默认打印position size
# 按 target_size 交易
print('%04d - %s - Position Size: %02d - Value %.2f' %
(len(self), dt.isoformat(), self.position.size, portfolio_value))
# get_value 带参数，用当前的收盘价计算持仓价值
# data_value = self.data.close[0] * self.position.size
data_value = self.broker.get_value([self.data])
print('datavalue: %.2f, open:%.2f, close:%.2f,close-open:%.2f' %
(data_value,self.data.open[0],self.data.close[0],self.data.close[0]-self.data.open[0]))
# 按 target_value 交易
if self.p.use_target_value:
print('%04d - %s - data value %.2f' %
(len(self), dt.isoformat(), data_value))
# 按 target_percent 交易
elif self.p.use_target_percent:
port_perc = data_value / portfolio_value
print('%04d - %s - data percent %.2f' %
(len(self), dt.isoformat(), port_perc))
# 如果有挂起订单 ，不做后续处理，一个订单处理完，再开始下一个订单
if self.order:
return # pending order execution
# 仓位设置为日期的天
size = dt.day
# 如果是偶数月
if (dt.month % 2) == 0:
# 仓位递减交易
size = 31 - size
# target_size 模式
if self.p.use_target_size:
target = size
print('%04d - %s - Order Target Size: %02d' %
(len(self), dt.isoformat(), size))
self.order = self.order_target_size(target=size)
# target_value 模式
elif self.p.use_target_value:
value = size * 1000
print('%04d - %s - Order Target Value: %.2f' %
(len(self), dt.isoformat(), value))
self.order = self.order_target_value(target=value)
# target_percent 模式
elif self.p.use_target_percent:
percent = size / 100.0
print('%04d - %s - Order Target Percent: %.2f' %
(len(self), dt.isoformat(), percent))
self.order = self.order_target_percent(target=percent)
def runstrat(args=None):
# define parameter
param_fromdate = '2005-01-01'
param_todate = '2005-03-31'
param_data = './datas/yhoo-1996-2015.txt'
param_cash = 100000
# set target parameter
param_target_size = True
param_target_value = False
param_target_percent = False
# set plot
param_plot = True
cerebro = bt.Cerebro()
cerebro.broker.setcash(param_cash)
# 数据加载参数字典
dkwargs = dict()
if param_fromdate is not None:
dkwargs['fromdate'] = datetime.strptime(param_fromdate, '%Y-%m-%d')
if param_todate is not None:
dkwargs['todate'] = datetime.strptime(param_todate, '%Y-%m-%d')
# data
data = bt.feeds.YahooFinanceCSVData(dataname=param_data, **dkwargs)
cerebro.adddata(data)
# strategy
cerebro.addstrategy(TheStrategy,
use_target_size=param_target_size,
use_target_value=param_target_value,
use_target_percent=param_target_percent)
cerebro.run()
if param_plot:
# 绘图参数字典
pkwargs = dict(style='bar')
if param_plot is not True: # evals to True but is not True
npkwargs = eval('dict(' + param_plot + ')') # args were passed
pkwargs.update(npkwargs)
cerebro.plot(iplot=False,**pkwargs)
if __name__ == '__main__':
runstrat()
（1）order_target_size
输出结果：
0001 - 2005-01-03 - Position Size: 00 - Value 100000.00
datavalue: 0.00, open:38.36, close:38.18,close-open:-0.18
0001 - 2005-01-03 - Order Target Size: 03
0002 - 2005-01-04 - Position Size: 03 - Value 99994.39
datavalue: 109.74, open:38.45, close:36.58,close-open:-1.87
0002 - 2005-01-04 - Order Target Size: 04
0003 - 2005-01-05 - Position Size: 04 - Value 99992.48
datavalue: 144.52, open:36.69, close:36.13,close-open:-0.56
0003 - 2005-01-05 - Order Target Size: 05
0004 - 2005-01-06 - Position Size: 05 - Value 99988.79
datavalue: 177.15, open:36.32, close:35.43,close-open:-0.89
0004 - 2005-01-06 - Order Target Size: 06
0005 - 2005-01-07 - Position Size: 06 - Value 99991.41
datavalue: 215.76, open:35.99, close:35.96,close-open:-0.03
0005 - 2005-01-07 - Order Target Size: 07
0006 - 2005-01-10 - Position Size: 07 - Value 99993.89
datavalue: 254.24, open:36.00, close:36.32,close-open:0.32
0006 - 2005-01-10 - Order Target Size: 10
0007 - 2005-01-11 - Position Size: 10 - Value 99987.32
datavalue: 356.60, open:36.31, close:35.66,close-open:-0.65
0007 - 2005-01-11 - Order Target Size: 11
0008 - 2005-01-12 - Position Size: 11 - Value 99992.38
datavalue: 397.54, open:35.88, close:36.14,close-open:0.26
0008 - 2005-01-12 - Order Target Size: 12
0009 - 2005-01-13 - Position Size: 12 - Value 99982.68
datavalue: 423.96, open:36.12, close:35.33,close-open:-0.79
0009 - 2005-01-13 - Order Target Size: 13
0010 - 2005-01-14 - Position Size: 13 - Value 99999.96
datavalue: 477.10, open:35.86, close:36.70,close-open:0.84
0010 - 2005-01-14 - Order Target Size: 14
0011 - 2005-01-18 - Position Size: 14 - Value 100006.28
datavalue: 520.52, open:37.10, close:37.18,close-open:0.08
0011 - 2005-01-18 - Order Target Size: 18
0012 - 2005-01-19 - Position Size: 18 - Value 99989.54
datavalue: 656.10, open:38.08, close:36.45,close-open:-1.63
0012 - 2005-01-19 - Order Target Size: 19
0013 - 2005-01-20 - Position Size: 19 - Value 99977.87
datavalue: 679.82, open:35.39, close:35.78,close-open:0.39
0013 - 2005-01-20 - Order Target Size: 20
0014 - 2005-01-21 - Position Size: 20 - Value 99967.98
datavalue: 706.00, open:36.07, close:35.30,close-open:-0.77
0014 - 2005-01-21 - Order Target Size: 21
0015 - 2005-01-24 - Position Size: 21 - Value 99939.03
datavalue: 712.53, open:35.48, close:33.93,close-open:-1.55
0015 - 2005-01-24 - Order Target Size: 24
0016 - 2005-01-25 - Position Size: 24 - Value 99939.81
datavalue: 816.96, open:34.55, close:34.04,close-open:-0.51
0016 - 2005-01-25 - Order Target Size: 25
0017 - 2005-01-26 - Position Size: 25 - Value 99974.89
datavalue: 886.75, open:34.71, close:35.47,close-open:0.76
0017 - 2005-01-26 - Order Target Size: 26
0018 - 2005-01-27 - Position Size: 26 - Value 99955.74
datavalue: 902.98, open:35.38, close:34.73,close-open:-0.65
0018 - 2005-01-27 - Order Target Size: 27
0019 - 2005-01-28 - Position Size: 27 - Value 99952.60
datavalue: 934.74, open:34.90, close:34.62,close-open:-0.28
0019 - 2005-01-28 - Order Target Size: 28
0020 - 2005-01-31 - Position Size: 28 - Value 99968.70
datavalue: 985.88, open:35.04, close:35.21,close-open:0.17
0020 - 2005-01-31 - Order Target Size: 31
0021 - 2005-02-01 - Position Size: 31 - Value 99954.68
datavalue: 1077.25, open:35.13, close:34.75,close-open:-0.38
0021 - 2005-02-01 - Order Target Size: 30
0022 - 2005-02-02 - Position Size: 30 - Value 99979.65
datavalue: 1066.20, open:36.02, close:35.54,close-open:-0.48
0022 - 2005-02-02 - Order Target Size: 29
0023 - 2005-02-03 - Position Size: 29 - Value 99966.33
datavalue: 1017.61, open:35.27, close:35.09,close-open:-0.18
0023 - 2005-02-03 - Order Target Size: 28
0024 - 2005-02-04 - Position Size: 28 - Value 99963.99
datavalue: 980.56, open:34.71, close:35.02,close-open:0.31
0024 - 2005-02-04 - Order Target Size: 27
分析Value：
看value，如何计算出来的 ？每个人都有疑问 ？
增加了一行过程计算数据 ：
datavalue: 109.74, open:38.45, close:36.58,close-open:-1.87
经过一下午研究，在Excel中拼凑计算，终于和BT数据一致了：
Excel模拟计算下载
1528
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
