@tsz-fung-chan You can keep track of the time data via the self.data.datetime.datetime() function. You can save this value when you enter a position; if the latest datapoint's timestamp is bigger than a defined length of time you can then exit the position.

You can have a look at the throttle function in exchange.py (https://github.com/ccxt/ccxt/blob/master/python/ccxt/base/exchange.py) for the details. But basically the rateLimit is multiplied by a cost (which should be 1 most of the time). So with a 5000ms you should see a delay of 5 seconds. I'm not sure why you are seeing 25. Do you have multiple symbols?

Regarding the websocket. I expect that it will be handled by ccxt transparently to backtrader, but will have to wait and see once it becomes available.

The errors most effective takes place after the resample filter truely produces a cost. I'm at lack of what I'm doing wrong. Has all people else got resampling operating with stay facts? Similar questions best use CSV information about Neoprene Bag.

@rapha It is possible that it is about the Python version change.

params = dict( target_n_positions=20 )

You could try by defining the params inside in a dict, as shown above.

When I run both the tuple and dict approaches work, but who knows. I run backtrader in a environment in 3.8, as I have found some troubles with the newer versions.

Hi @Nael-Shichida

I would suggest you replace you self.order parameter with a dictionary perhaps. ALSO, you can add a paramter tradeid in the buy_bracket method.

Consider, for example:

def log(self, txt, dt=None): ''' Logging function fot this strategy''' dt = dt or self.datas[0].datetime.date(0) print('%s, %s' % (dt.isoformat(), txt)) def __init__(self): # Keep a reference to the "close" line in the data[0] dataseries self.order = {} self.id = 0 self.dataclose = self.datas[0].close self.dataopen = self.datas[0].open self.zone = self.datas[0].close - self.datas[0].open def next(self): # Simply log the closing price of the series from the reference self.log('Close, %.2f' % self.dataclose[0]) self.log('Open, %.2f' % self.dataopen[0]) self.log('Zone, %.2f' % self.zone[0]) self.log('shift, %.2f' % self.zone[-1]) print(len(self)) print(self.order) print(self.position) # """Enters Long position if below conditions are met""" if self.zone[0] < 0: print("Bar is red") # current close less than previous close if self.zone[1] > 0: print("Next candle is green") # Next close less than the previous close if 1.5*abs(self.zone[0]) < abs(self.zone[1]): print("Buy Zone found") # BUY, BUY, BUY!!! (with all possible default parameters) self.log('BUY CREATE, %.2f' % self.dataclose[0]) # self.buy() self.order[self.id] = self.buy_bracket(tradeid = self.id, limitprice=self.dataopen[0], price=self.dataopen[0], stopprice=self.dataclose[0]) # print(self.order) print(type(self.order.ref)) self.id += 1 # this is a simple way, but you can therefore have an id to identify each individual trade

This specific method might not solve your question as you might loose track of what self.id fits with which order. I do feel, however, that this is a starting point for you to have a dictionary of all your submitted order.

POSSIBLE SOLUTION: I have quite a different way to do it (might get a bit lost now, but putting it here for incase it helps).

I have a custom order tracking dictionary from which I track all my OPEN ORDERS (orders (buylimits, selllimits, etc.) and trades (long, short, etc.)). At times, my strategy has up to 14 trades open and this works 100%.
This is a snippet of my self.open_orders tracking when making a BUY trade (opening a long position). My strategy logic, therefore ALWAYS looks at what trades are currently open, what orders are pending, and makes logical decisions on that

# backtrader built-in action self.buy_bracket(tradeid = self.id, exectype = bt.Order.Market, price = close, size = 0.1, limitprice = self.tp, stopprice = self.sl) # self custom order tracking self.open_orders[self.id] = {'symbol' : 'EURUSD', 'lots' : 0.1, 'type' : 'buy', 'open_price' : close, 'open_time' : self.datetime.time(ago=0).isoformat(), 'SL' : self.sl, 'TP' : self.tp, 'pnl' : 0.0, 'commission' : 0.01 'swap' : 0.0, 'comment' : ''}

Note, as I can always view self.open_orders, I know exactly which trade I can identify when I need to make any action on it. For example, if I need to close the trade, I can say self.close(tradeid=self.id) and it will close the position and calculate my PnL.

You can use self.data.datetime.datetime() to return a datetime object with the timestamp of the last datapoint avaiable on the data

OK, yet I am currently just backtesting my exchanging procedure with the verifiable OHLC costs of some forex cash matches, since it was just authentic rates in imported csv records, how does backtrader recognize the resource class? Or on the other hand might there be a method for setting up counter places of forex cash paris as on numerous forex exchanging stages like MT4 and Oanda? Gratitude for any counsel!
<a href="https://www.mrfurniture.ae/">office furniture in Dubai</a>

@cdubya

The data keeps the 4 decimals but not the plot by default.

In[4] cerebro.datas[0].close[0] In[5] 1.00898 In[6] cerebro.plot()

Figure_0.png