@vladisld Enjoyed the video. Believe the issue in the code was that it was iterating over rows and writing 1 row at a time to the database...

Can now write 13 years of 1 minute data in ~50 seconds. Needed to batch_size=10000 for it to work on a large data-set.

Problem solved. Now the real work begins.

Thank you!

@mudassar031
resolve it by using prenext function to seed recursive value

@pomnl said in Race Condition Bug w/ Orders and Trades:

@run-out also, to put this into perspective, I currently run tens of millions of tests across these datasets. Just have to ask if you're able to do that with this strategy?
I was doing a factor analysis and the factor parr (eg first part precalculated) didn't change much. I have also used it in other cases where it slowed down. So not sure how it will work out for you. Good luck.

@rajesh
i figured out the issue and solved the issue, now i got new error, can some one look into it.

df = pd.read_csv(data_file, header=0) df["date"] = pd.to_datetime(df["date"]).dt.tz_convert(None) df.drop(df.iloc[ :, 0:1,], inplace = True, axis = 1) df.set_index("date", inplace = True) df.to_csv("datafile_clean.csv") # df.iloc[0:3, :] # df.to_csv('tcs4.csv') cdata = btfeeds.GenericCSVData( dataname="datafile_clean.csv", fromdate = dt.datetime(2020, 10,1), dtformat="%Y-%m-%d %H:%M:%S", datatime = 0 , open=1, high=2, low=3, close=4, volume=5, reverse = True ) if __name__ == '__main__': # Create a cerebro entity cerebro = bt.Cerebro() # Add a strategy cerebro.adddata(cdata) cerebro.addstrategy(macdStrategy) print('Starting Portfolio Value : %0.2f' % cerebro.broker.getvalue()) cerebro.run() # cerebro.plot() print('Final Portfolio Value : %0.2f' % cerebro.broker.getvalue()) TypeError: super(type, obj): obj must be an instance or subtype of type

I just checked, and strategy.stop() does obtain drawdown analyzer results. Nevermind.

drawdown = self.analyzers.drawdown.get_analysis() print(drawdown.max.drawdown)

Getting a full body massage gives your body so many benefits, at the level of physically and psychologically. Our team have the most knowledgeable and experienced professional massage therapists who give a pleasant massage and give an immersive presence. Our massage will improve your speeling quality and increase the amount of positive energy in your body. Feel free to get in touch for Massage in Bangalore with us or visit. https://www.bellaspa.in

It's a blog post.

ffa201ce-3fa9-4ce0-8221-7b684fc7a03b-image.png
I think it's possible to have 1 large bar to form multiple bricks but not with the implementation in backtrader's filter, you have to implement your own renko filter. every extra bricks need to have a small time offset, due to the precision, the smallest value is 10 microseconds which is small enough to use I think.

@backtrader said in Implement stop loss:

No, under buy_price, because your are talking about stop-loss (you are in the market), people will understand the price at which you bought your asset and not the expected (and yet unknown) price at which the next order, the stop-loss one will execute.

Got this. I've declared a new variable which is store buy_price and updates it in notify order method with order.executed.price.

@asus-rog Does anyone have an idea why adding a HeikinAshi indicator breaks this script?
Does it have to do with HA setting a seed?
For data produced with HA commented out and HA uncommented, see results data screenshots at the end of this post.

def prenext(self): # seed recursive value self.lines.ha_open[0] = (self.data.open[0] + self.data.close[0]) / 2.0

If you want to reproduce it, here's a minimal example script. Run it once with these lines commented out and once uncommented:

self.ha = bt.ind.HeikinAshi(self.datas[0]) self.ha.csv = True

IMPORTANT: YOU NEED TO CREATE A DATA FEED WITH YOUR DATA FOR IT TO WORK!
Please adapt this accordingly:

# Create a Data Feed data = bt.feeds.GenericCSVData( dataname='Historical data/es-1m.csv', separator=";", fromdate=datetime.datetime(2012, 5, 1), todate=datetime.datetime(2012, 5, 31), dtformat=("%d/%m/%Y"), timeformat=("%H:%M:%S"), time=1, datetime=0, high=3, low=4, open=2, close=5, volume=6, openinterest=-1, timeframe=bt.TimeFrame.Minutes ) import datetime # For datetime objects import backtrader as bt import math import matplotlib matplotlib.use('QT5Agg') import matplotlib.pyplot as plt import pandas as pd import numpy as np class ZigZag(bt.ind.PeriodN): ''' Identifies Peaks/Troughs of a timeseries ''' lines = ( 'trend', 'last_pivot_t', 'last_pivot_x', 'last_pivot_ago', 'zigzag_peak', 'zigzag_valley', 'zigzag', 'last_zigzag', ) # Fancy plotting name # plotlines = dict(logreturn=dict(_name='log_ret')) plotinfo = dict( subplot=False, plotlinelabels=True, plotlinevalues=True, plotvaluetags=True, ) plotlines = dict( trend=dict(marker='', markersize=0.0, ls='', _plotskip=True), last_pivot_t=dict(marker='', markersize=0.0, ls='', _plotskip=True), last_pivot_x=dict(marker='', markersize=0.0, ls='', _plotskip=True), last_pivot_ago=dict(marker='', markersize=0.0, ls='', _plotskip=True), zigzag_peak=dict(marker='v', markersize=4.0, color='red', fillstyle='full', ls=''), zigzag_valley=dict(marker='^', markersize=4.0, color='red', fillstyle='full', ls=''), zigzag=dict(_name='zigzag', color='blue', ls='-', _skipnan=True), last_zigzag=dict(_name='last_zigzag', color='blue', ls='--', _skipnan=True), ) # update value to standard for Moving Averages params = ( ('period', 2), ('up_retrace', 0.1), ('dn_retrace', 0.1), ('bardist', 0.015), # distance to max/min in absolute perc ) def __init__(self): super(ZigZag, self).__init__() if not self.p.up_retrace: raise ValueError('Upward retracement should not be zero.') if not self.p.dn_retrace: raise ValueError('Downward retracement should not be zero.') if self.p.up_retrace < 0: self.p.up_retrace = -self.p.up_retrace if self.p.dn_retrace > 0: self.p.dn_retrace = -self.p.dn_retrace self.p.up_retrace = self.p.up_retrace / 100 self.p.dn_retrace = self.p.dn_retrace / 100 self.missing_val = float('nan') def prenext(self): self.lines.trend[0] = 0 self.lines.last_pivot_t[0] = 0 self.lines.last_pivot_x[0] = self.data[0] self.lines.last_pivot_ago[0] = 0 self.lines.zigzag_peak[0] = self.missing_val self.lines.zigzag_valley[0] = self.missing_val self.lines.zigzag[0] = self.missing_val self.lines.last_zigzag[0] = self.missing_val def next(self): data = self.data trend = self.lines.trend last_pivot_t = self.lines.last_pivot_t last_pivot_x = self.lines.last_pivot_x last_pivot_ago = self.lines.last_pivot_ago zigzag_peak = self.lines.zigzag_peak zigzag_valley = self.lines.zigzag_valley zigzag = self.lines.zigzag last_zigzag = self.lines.last_zigzag x = data[0] r = x / last_pivot_x[-1] - 1 curr_idx = len(data) - 1 trend[0] = trend[-1] last_pivot_x[0] = last_pivot_x[-1] last_pivot_t[0] = last_pivot_t[-1] last_pivot_ago[0] = curr_idx - last_pivot_t[0] zigzag_peak[0] = self.missing_val zigzag_valley[0] = self.missing_val zigzag[0] = self.missing_val last_zigzag[0] = x if trend[-1] == 0: if r >= self.p.up_retrace: piv = last_pivot_x[0] * (1 - self.p.bardist) zigzag_valley[-int(last_pivot_ago[0])] = piv zigzag[-int(last_pivot_ago[0])] = last_pivot_x[0] trend[0] = 1 last_pivot_x[0] = x last_pivot_t[0] = curr_idx elif r <= self.p.dn_retrace: piv = last_pivot_x[0] * (1 + self.p.bardist) zigzag_peak[-int(last_pivot_ago[0])] = piv zigzag[-int(last_pivot_ago[0])] = last_pivot_x[0] trend[0] = -1 last_pivot_x[0] = x last_pivot_t[0] = curr_idx elif trend[-1] == -1: if r >= self.p.up_retrace: piv = last_pivot_x[0] * (1 - self.p.bardist) zigzag_valley[-int(last_pivot_ago[0])] = piv zigzag[-int(last_pivot_ago[0])] = last_pivot_x[0] trend[0] = 1 last_pivot_x[0] = x last_pivot_t[0] = curr_idx elif x < last_pivot_x[-1]: last_pivot_x[0] = x last_pivot_t[0] = curr_idx elif trend[-1] == 1: if r <= self.p.dn_retrace: piv = last_pivot_x[0] * (1 + self.p.bardist) zigzag_peak[-int(last_pivot_ago[0])] = piv zigzag[-int(last_pivot_ago[0])] = last_pivot_x[0] trend[0] = -1 last_pivot_x[0] = x last_pivot_t[0] = curr_idx elif x > last_pivot_x[-1]: last_pivot_t[0] = curr_idx last_pivot_x[0] = x idx = 1 while idx < len(self.zigzag) and math.isnan(zigzag[-idx]): last_zigzag[-idx] = self.missing_val idx += 1 if idx < len(self.data): last_zigzag[-idx] = zigzag[-idx] class TestStrategy(bt.Strategy): def log(self, txt, dt=None): ''' Logging function fot this strategy''' dt = dt or self.datas[0].datetime.datetime(0) print('%s, %s' % (dt.isoformat(), txt)) def __init__(self): # Keep a reference to the "close" line in the data[0] dataseries self.dataclose = self.datas[0].close # Add indicators # In main plot self.zz = ZigZag(self.data, plotname='ZZ', period=2, up_retrace=0.2, dn_retrace=0.2) self.zz.csv = True self.ha = bt.ind.HeikinAshi(self.datas[0]) self.ha.csv = True if __name__ == '__main__': cerebro = bt.Cerebro() cerebro.addobserver(bt.observers.BuySell, barplot=True, bardist=0.0005) # Add a strategy cerebro.addstrategy(TestStrategy) # Create a Data Feed data = bt.feeds.GenericCSVData( dataname='Historical data/es-1m.csv', separator=";", fromdate=datetime.datetime(2012, 5, 1), todate=datetime.datetime(2012, 5, 31), dtformat=("%d/%m/%Y"), timeformat=("%H:%M:%S"), time=1, datetime=0, high=3, low=4, open=2, close=5, volume=6, openinterest=-1, timeframe=bt.TimeFrame.Minutes ) # Add the Data Feed to Cerebro cerebro.adddata(data) # Set our desired cash start cerebro.broker.setcash(100000.0) # Add a writer to put out csv file cerebro.addwriter(bt.WriterFile, csv=True, out='strategy_results') # Run over everything cerebro.run() # Plot the result cerebro.plot(style='candlestick', iplot = False) # Safe result files results = pd.read_csv("strategy_results", sep=",", header=1) split_row = results.index[results['Id'].str.contains('=====', na=False)][0] results_data = results.iloc[:split_row,:] results_data.to_csv("results_data") summary = results.iloc[split_row:,:] summary.to_csv("summary")

Note how it seems to break the columns zigzag_peak, zigzag_valley, zigzag and last_zigzag.
Due to this zigzag_peak, zigzag_valley, zigzag become all nan, see below.
I presume that it breaks the first row output for the ZZ script, breaking all subsequent lines.

WITHOUT HeikinAshi, it works:
Screenshot 2022-05-20 174413.png

WITH HeikinAshi, it's broken:
Screenshot 2022-05-20 174621_2.png

Any help is greatly appreciated :)

Never mind, I did a little digging. We can achieve what I said in the previous post by using "stop orders". So, case is closed for now .. for me at least. :)

@dowdtw Would be great if you could share your solution...

@bb2 It makes sense if you realize that resampledata() performs the resampling in-place; something not specified very clearly in the docs

@irving-birger said in Closed trade list (including MFE/MAE) analyzer:

Looks amazing! Thx for the effort!

I run into

if trade.history[0].event.size > 0: dr = 'long' IndexError: list index out of range

did I miss sth?

Got it! Forgot to enable history with this:

cerebro = bt.Cerebro(tradehistory=True)

@scottz1 said in My Orders dont get Closed (wrong implementation of SL/TP?):

Looks ok basically. However,

remember to add logic for partial fills so the tp/sl orders will have the correct size. remember to check existing position so as not to overwrite self.order and end up doubling your position.

https://www.backtrader.com/docu/order-creation-execution/oco/oco/

No, it cannot work for me, because that doesn't read from Yahoo. It tries

@Mario-Pellegrini said in float division by zero:

It is not a random code since you also say that the ticker works

It is random, because you apparently want to read from Yahoo and apply Quandl. The code has also nothing to do with the code from the homepage (3 snippets) that directly reads data from Yahoo.

Why don't you start from one of the scripts on the home page, change the ticker to FCA.MI and then slowly modify it? The link again for your reference

https://www.backtrader.com/home/helloalgotrading/

@tianjixuetu Hi, can we have a little discussion about how to speed the backtest? Recently I am trying to find a new way to solve it. My Wechat Id is 13247198760

@run-out said in Cancelling Bracket order following closing of the main side:

Normally when using a list or dictionary to track orders, if you remove the orders from the list at the right spots, there should be no issues with the list.

However, I tend to get a bit cautious (lazy?) and insert a clause at the beginning of next() that checks if:

Are there any open positions, and...

Are there any live orders.

If there are no open positions and no live orders, really the list should be empty. The code looks something like:

def next(self): if self.position.size == 0 and len([o for o in self.o_li if o.status < 4]) == 0: self.o_li = list()

Order status of 4 is completed, and higher are all dead orders.

Then another thing I notice in your code is that you have bracket orders with purchase price which is fine. But if the stock trends permanantly away from your purchase price, you will be left with a set of orders that will never fill, blocking future intended orders, since there are open orders.

Since in your bracket orders you entering the market at last close, probably best to leave exectype as Market and no plimit, or alternatively insert a valid period so the bracket expires after so many bars.

@jcgeraldi His email is in his profile.

@ZHUOER-ZHANG try it. For bt it doesn't matter, but may be a big deal for your trading algo.