sell a position after 2 days

PatrikMuniak

I want to buy a position and sell it after 2 days and I've tried to return the day I opened the position into init but it gives me

TypeError: 'GapUp' object is not an iterator

I just started with Backtrader an I have no clue on how to solve this issue, thanks in advance to whoever can give me some hint.

import backtrader as bt
from datetime import datetime, timedelta, date
#datetime.datetime(year, month, day, hour=0, minute=0, second=0, microsecond=0, tzinfo=None, *, fold=0)

class GapUp(bt.Strategy):
    
    def __init__(self):
        self.change = bt.ind.PctChange()
        self.dataclose = self.datas[0].close
        self.currdate = self.datas[0].datetime.date(0)
        position_date = next(self)
        

    def next(self):
        if self.change >= 0.05:
            self.buy(size=1)
            print(self.datas[0].datetime.date(0))
            print(self.datas[0].close[0])
            return self.datas[0].datetime.date(0)
        #if self.currdate >= self.position_date + datetime.timedelta(days=2):
        #    close()
        



cerebro = bt.Cerebro()
cerebro.addstrategy(GapUp)

data = bt.feeds.YahooFinanceCSVData(dataname='^GSPC.csv', fromdate=datetime(2018,1,1), todate=datetime(2019,11,1))
cerebro.adddata(data)
cerebro.broker.setcash(100000.0)
cerebro.run()
cerebro.plot()

ab_trader

__init__() is called once at the very beginning of the run. I doubt that you will be able to return something from the following run to the __init()__.

If you want to exit after N bars, than:

in notify_order() get order execution bar

self.bar_executed = len(self)

in the next() calculate position duration based on current strategy length

self.duration = len(self) - self.bar_executed + 1

if it is 2, then issue the order to close position

if self.duration == 2:
    self.close()

backtrader

@PatrikMuniak said in sell a position after 2 days:

I want to buy a position and sell it after 2 days

Notice that using anything which is datetime related won't usually work unless you don't mind weekends and trading holidays adding days to the count.

PatrikMuniak

@ab_trader
Thanks a lot for replying, I've tried to implement what you've said and I managed to make it work until the graph. The issue is that even if I don't get any error the sell is not working, i.e the plot() shows just buys but not sells.

This is how I tried to implement you answer:

import backtrader as bt
from datetime import datetime, timedelta, date

class GapUp(bt.Strategy):

    
    def __init__(self):
        self.change = bt.ind.PctChange()
        self.dataclose = self.datas[0].close
        self.currdate = self.datas[0].datetime.date(0)
        
    def notify_order(self, order):
        self.bar_executed = len(self)

    def next(self):
        if self.change >= 0.05:
            self.buy(size=1)

        self.notify_order(self)
        self.duration = len(self) - self.bar_executed + 1

        if self.duration == 2:
            self.close()


cerebro = bt.Cerebro()
cerebro.addstrategy(GapUp)

data = bt.feeds.YahooFinanceCSVData(dataname='^GSPC.csv', fromdate=datetime(2018,1,1), todate=datetime(2019,11,1))
cerebro.adddata(data)
cerebro.broker.setcash(100000.0)
cerebro.run()
cerebro.plot()

ab_trader

@PatrikMuniak every next() you call notify_order() and set your self.bar_executed to length of the data. So your self.duration vaule is equal to 1 on every next() call. No surprise you have no sells.

Jake Barrett

@ab_trader thanks for sharing this idea. I am trying to implement the code but am having the same issue above. The self.duration value is constantly one, and one has to call the notify_order() in next() otherwise you get an error about bar_executed. How can one track the number of bars in a trade? Can you use something related to the Trade functions? I'll attach my code. Thank you

ab_trader

there is no need to call self.notify_order() in the next(). in order to avoid an error with bar_executed you need to operate with it only if the position exists. otherwsie use of bar_executed is meaningless.

also don't post pictures, but text; please read the note about backticks on top of the page.