I also add the psar value to the line:

self.lines.PSAR = self.psar

The directional movement indicator (also known as the directional movement index or DMI) is a valuable.

Visit My Website: Jockrs

@godseye said in Implementing Bracket Orders for binance API:

I might have a naked position in the market for 15 mins in a worst case scenario and that's not an outcome I could take a risk on!

Have you solved the naked position problem?

In TradingView, you can use two period:

4a70f608-7b88-40b6-863f-133043bf65a7-image.png

@run-out said in How to extract historical account value from backtrader or its results:

Dictionary dictionary = results[0].analyzers.getbyname("cash_market").get_analysis() df = pd.DataFrame(dictionary).T df.columns = ["Date", "Cash", "Value"]

It works, thank you so much.

I guess the system double_unders can't be reached from outside of the object

@jplaza This is a bit hacky, but you could also load the data in the __init__ of your strategy. It's not highly recommended, and you would need to be diligent about your dates and forward looking bias.

@emr Thank you for your reply. But I resolved the issue.

@harjot-singh class TestStrategyoptions(bt.Strategy): params = ( ('d0', None ), ('pl_interval',None)) def __init__(self): self.all_contract = self.params.d0 self.close_9_30 = defaultdict(list) self.trade_options = list() # self.contract_op = '' self.price = '' self.active_con = '' self.contract_name = '' self.order_dict = dict() self.order_status = dict() self.buyprice = dict() self.sellprice = dict() self.returns = pd.DataFrame(columns = ['ticker','datetime','returns']) self.all_order = list() self.buy_list = defaultdict(list) self.sell_list = defaultdict(list) self.pnl = dict() # self._dd = self._owner._addanalyzer_slave(bt.analyzers.DrawDown) for i, d in enumerate(self.datas): self.order_dict[d._name] = None self.order_status[d._name] = 0 self.buyprice[d._name] = None self.sellprice[d._name] = None def prenext(self): self.next() def next(self): dt=self.datetime.datetime() # print(dt) # return # for i, d in enumerate(self.datas): # tick_name = d._name # self.log(dt, tick_name) # if dt.time() == time(9, 30): # self.log(d.close[0]) # self.price = float(tick_name.split("_")[2]) # self.order_dict[tick_name] = self.buy(data=d, exec_type = bt.Order.Market, size=25) # self.log(d._name, d.open[0], d.high[0], d.low[0], d.close[0]) # return # if dt.time() == time(9, 30): # ts.sleep(1) # return for i, d in enumerate(self.datas): tick_name = d._name if dt.time() == time(9, 30): # breakpoint() if np.isnan(d.close[0]) == False: # breakpoint() self.close_9_30[tick_name[:-3]].append(d.close[0]) if dt.time() == time(9, 30): for tick in list(self.close_9_30.keys()): if len(self.close_9_30[tick]) != 2: self.close_9_30.pop(tick) else: self.close_9_30[tick] = [abs(self.close_9_30[tick][0] - self.close_9_30[tick][1])] min_list = [(value[0], key) for key, value in self.close_9_30.items()] self.close_9_30.clear() self.trade_options = min(min_list)[1] for i, d in enumerate(self.datas): tick_name = d._name if dt.time() == time(9, 30): if tick_name[:-3] in self.trade_options: self.price = float(tick_name.split("_")[2]) self.order_dict[tick_name] = self.sell(data=d) # self.order_dict[tick_name] = self.sell(data=d, exec_type=bt.Order.Market, price=d.close[0], size=50) # breakpoint() if (dt.time() > time(9, 30)) & (dt.time() < time(15, 15)): if tick_name[:-3] in self.trade_options: if self.sellprice[tick_name] != None: stop_loss_price = self.sellprice[tick_name] + (self.sellprice[tick_name]*.15) if (d.close[0] >= stop_loss_price) & (self.order_status[tick_name] < 0): self.price = float(tick_name.split("_")[2]) self.order_dict[tick_name] = self.buy(data=d, exec_type = bt.Order.Market, plimit=stop_loss_price, size=50) if dt.time() == time(15, 15): if self.order_status[tick_name] < 0: self.order_dict[tick_name] = self.buy(data = d, exectype=bt.Order.Market, size=50) def notify_order(self, order): dt=self.datetime.datetime() for i, d in enumerate(self.datas): ticker = d._name if self.order_dict is None: continue if ticker not in self.order_dict.keys(): continue if self.order_dict[ticker] is None: continue order = self.order_dict[ticker] # breakpoint() if order.status in [order.Accepted, order.Submitted]: if order.isbuy(): self.log('BUY ORDER PLACED, %.2f, %s, underlying price, %.2f' % (order.executed.price, ticker, self.price)) if order.issell(): self.log('SELL ORDER PLACED, %.2f, %s, underlying price, %.2f' % (order.executed.price, ticker, self.price)) # Buy/Sell order submitted/accepted to/by broker - Nothing to do continue # Check if an order has been completed # Attention: broker could reject order if not enough cash if order.status in [order.Completed]: self.all_order.append(order) if order.isbuy(): self.order_status[ticker] = self.order_status[ticker]+1 self.buyprice[ticker] = order.executed.price self.buy_list[dt].append(self.buyprice[ticker]) self.log('BUY EXECUTED, %.2f, %s, underlying price, %.2f' % (order.executed.price, ticker, self.price)) elif order.issell(): self.order_status[ticker] = self.order_status[ticker]-1 self.sellprice[ticker] = order.executed.price self.sell_list[dt].append(self.sellprice[ticker]) self.log('SELL EXECUTED, %.2f, %s, underlying price, %.2f' % (order.executed.price, ticker, self.price)) elif order.status in [order.Canceled, order.Margin, order.Rejected]: self.log('Order Canceled/Margin/Rejected %s' %(ticker)) # Write down: no pending order self.order_dict[ticker] = None

https://www.backtrader.com/docu/order/

@run-out said in Using ParabolicSAR inside strategy:

int

Yes I searched and found QuantConnect. this is exactly I said. for live trading you should have orginal PSAR that is float. in pineScript and QuantConnct this is float but in backtrader because of it's own definition. and when its float you can reach more different takeprofit.

@run-out OK, wow, I thought it would be something way simpler...
I might just keep on deleting them from my results.