Indicators/Strategies/Analyzers

• 

  Closed trade list (including MFE/MAE) analyzer
  analyzer trades mfe mae • • ab_trader  

  14
  8
  Votes
  14
  Posts
  1919
  Views

  S

  @scottz1 Answering my own question, I found how analyzers can easily access attributes in the strategy. This would allow me to avoid creating a duplicate ATR indicator in the analyzer, but I've run into a problem with this. If I use a buystop order to open a trade, the strategy doesn't know the fill price until notify_order. I therefore use strategy notify_order order.Completede to calculate a stop loss in ATR units and submit a stop loss order. My analyzer needs to know that stop loss measurement, but... Analyzer notify_trade is called before strategy notify_order. What would be a good design for this?
• C

  Alligator Strategy
  • camizulub  

  2
  0
  Votes
  2
  Posts
  150
  Views

  B

  I would argue your problem is in how you call the indicator. But it is a long shot, given that part isn't shown.
• 

  Money Flow Index (MFI)
  • Rodrigo Brito  

  2
  0
  Votes
  2
  Posts
  35
  Views

  B

  A reference is always a good starting point. It does let people be on common ground. For example https://school.stockcharts.com/doku.php?id=technical_indicators:money_flow_index_mfi You probably want to read the documentation about line delaying and stage-1, stage-2 operations. Docs - Platform Concepts What some people would call a "canonical" implementation can be found here ... (which you probably can better grasp after seeing the Stage-1 explanation about operators) https://www.backtrader.com/blog/2019-07-08-canonical-or-not/canonical-or-not/
• 

  ssl-channel-chart indicator wanted
  • Kas David  

  1
  0
  Votes
  1
  Posts
  19
  Views

  No one has replied

• 

  LaguerreRSI
  • Developing Coder  

  3
  0
  Votes
  3
  Posts
  100
  Views

  B

  @developing-coder said in LaguerreRSI: Am I missing something? Yes, you are trying to use something without even having looked at literature for it. The definition of the LaguerreRSI has no period. This was a contribution taken at the very beginnings and period was simply there.
• 

  Open Order's side as a Factor on the Strategy condition
  • Vasco Laranjo  

  2
  0
  Votes
  2
  Posts
  45
  Views

  B

  @vasco-laranjo said in Open Order's side as a Factor on the Strategy condition: open/current of an open order, Sorry you are obviously not looking for what open orders have. @vasco-laranjo said in Open Order's side as a Factor on the Strategy condition: for instance on the below I would like to apply it only if "and current position is short (sell side)". You are looking for the actual position. Use the size of the position (which is being tested as a boolean but has a sign and is used in several samples) Docs - Position
• 

  Plotting specific data from Multidata
  • Sujay Bawaskar  

  3
  0
  Votes
  3
  Posts
  135
  Views

  

  @backtrader said in Plotting specific data from Multidata: Docs - Plotting Cool, thanks @backtrader , will try and let you know.
• 

  Resample live feed from CCXT problem
  • mr-m0nst3r  

  4
  0
  Votes
  4
  Posts
  141
  Views

  

  @backtrader Here's the output of remove fromdate script and output: # !/usr/bin/env python # -*- coding: utf-8; py-indent-offset:4 -*- from __future__ import (absolute_import, division, print_function, unicode_literals) import sys import time from datetime import datetime, timedelta from datetime import datetime, timedelta import backtrader as bt import ccxt class TestStrategy(bt.Strategy): params = ( ('printlog', True), ) def log(self, txt, dt=None, doprint=False): ''' Logging function fot this strategy''' if self.params.printlog or doprint: dt = dt or bt.num2date(self.data.datetime[0]) print('%s, %s' % (dt, txt)) def start(self): self.counter = 0 print('START') def prenext(self): self.counter += 1 print('prenext len %d - counter %d' % (len(self), self.counter)) self.log("data#0: " + str(self.datas[0].datetime[0])) self.log("data#1: " + str(self.datas[1].datetime[0])) def __init__(self): self.macd = bt.indicators.MACDHisto(self.datas[0]) self.macd2 = bt.indicators.MACDHisto(self.datas[1]) def next(self): self.counter += 1 price_txt = "Counter: " + str(self.counter) + " Open/Close/High/Low/Volume: " + str(self.data0.open[0]) + " - "+ str(self.data0.close[0]) + " - " + str(self.data0.high[0]) + " - " + str(self.data0.low[0]) + " - " + str(self.data0.volume[0]) + " Len: "+ str(len(self.data0))# + " Time Frame:" + bt.TimeFrame.getname(self.data0._timeframe) + " Len: "+ str(len(self.data0)) self.log(price_txt) macd_txt = "MACD: {:.2f}, Histo: {:.2f}".format(self.macd.macd[0],self.macd.histo[0]) self.log("MACD#1: " + macd_txt) macd2_txt = "MACD: {:.2f}, Histo: {:.2f}".format(self.macd2.macd[0],self.macd2.histo[0]) self.log("MACD#2: " + macd2_txt) if __name__ == '__main__': cerebro = bt.Cerebro() #exchange = sys.argv[1] if len(sys.argv) > 1 else 'gdax' exchange = sys.argv[1] if len(sys.argv) > 1 else 'gateio' symbol = sys.argv[2] if len(sys.argv) > 2 else 'ETH/USDT' hist_start_date = datetime.utcnow() - timedelta(minutes=10) print('UTC NOW: ', datetime.utcnow()) print('hist_start_data: ', hist_start_date) print('Using symbol: ', symbol) # Create data feeds data_1m = bt.feeds.CCXT(exchange=exchange, symbol=symbol, name="1m", timeframe=bt.TimeFrame.Minutes, #fromdate=hist_start_date, compression=1) cerebro.adddata(data_1m) cerebro.replaydata(data_1m, timeframe=bt.TimeFrame.Minutes, name="1m", compression = 5) cerebro.addstrategy(TestStrategy) cerebro.run() Output: UTC NOW: 2019-06-24 02:30:45.980828 hist_start_data: 2019-06-24 02:20:45.980597 Using symbol: ETH/USDT START ... prenext len 163 - counter 163 2019-06-23 12:33:00, data#0: 737233.5229166667 2019-06-23 12:33:00, data#1: 737233.5229166667 dts[]: [737233.5236111111, 737233.5236111111] prenext len 164 - counter 164 2019-06-23 12:34:00, data#0: 737233.5236111111 2019-06-23 12:34:00, data#1: 737233.5236111111 dts[]: [737233.5243055555, 737233.5243055555] prenext len 165 - counter 165 2019-06-23 12:35:00, data#0: 737233.5243055555 2019-06-23 12:35:00, data#1: 737233.5243055555 dts[]: [737233.525, 737233.525] 2019-06-23 12:36:00, Counter: 166 Open/Close/High/Low/Volume: 312.57 - 312.45 - 312.71 - 312.45 - 99.16966564 Len: 166 2019-06-23 12:36:00, MACD#1: MACD: 0.07, Histo: -0.12 2019-06-23 12:36:00, MACD#2: MACD: 1.19, Histo: -0.09 dts[]: [737233.5256944444, 737233.5256944444] 2019-06-23 12:37:00, Counter: 167 Open/Close/High/Low/Volume: 312.45 - 312.54 - 312.54 - 312.36 - 21.3511 Len: 167 2019-06-23 12:37:00, MACD#1: MACD: 0.04, Histo: -0.12 2019-06-23 12:37:00, MACD#2: MACD: 1.19, Histo: -0.08 dts[]: [737233.526388889, 737233.526388889] 2019-06-23 12:38:00, Counter: 168 Open/Close/High/Low/Volume: 312.45 - 312.2 - 312.45 - 312.2 - 18.69802458 Len: 168 2019-06-23 12:38:00, MACD#1: MACD: -0.01, Histo: -0.13 2019-06-23 12:38:00, MACD#2: MACD: 1.17, Histo: -0.10 dts[]: [737233.5270833333, 737233.5270833333] 2019-06-23 12:39:00, Counter: 169 Open/Close/High/Low/Volume: 312.34 - 312.14 - 312.34 - 311.43 - 122.3391867563 Len: 169 2019-06-23 12:39:00, MACD#1: MACD: -0.05, Histo: -0.14 2019-06-23 12:39:00, MACD#2: MACD: 1.16, Histo: -0.11 dts[]: [737233.5277777778, 737233.5277777778] 2019-06-23 12:40:00, Counter: 170 Open/Close/High/Low/Volume: 311.44 - 311.79 - 311.79 - 311.43 - 22.0295 Len: 170 2019-06-23 12:40:00, MACD#1: MACD: -0.11, Histo: -0.16 2019-06-23 12:40:00, MACD#2: MACD: 1.13, Histo: -0.13 dts[]: [737233.5284722223, 737233.5284722223] ... You see the UTC NOW? I don't know why the MACDs' date is at around 2019-06-23 12:40:00. And I'm confused that: using fromdate will case error without fromdate, the script is still getting old data, but without error And I don't know why it's getting old data from that time How can these two happen in the same script? I'm just lost.
• 

  Resample data but do not plot it
  • Kristofer Monisit  

  3
  0
  Votes
  3
  Posts
  73
  Views

  B

  Docs - Plotting
• 

  It's here - a beta you can use RIGHT NOW - Essential trade statistics - all in one place
  • Richard O'Regan  

  12
  5
  Votes
  12
  Posts
  3572
  Views

  I

  Hi, very nice tool kit -- thanks for sharing! Is it still available? It seems like the links above are dead, unfortunately.
• J

  How much actual gains are in this backtest?
  • jaredunn  

  7
  0
  Votes
  7
  Posts
  226
  Views

  B

  @jaredunn said in How much actual gains are in this backtest?: Some mundane detail must be off, haha. Nearly $1m cash growth since halfway thru 2018 til now. http://prntscr.com/o2wbf9 https://www.youtube.com/watch?v=IPLnYP9Oo8s You are expecting people to go to prntscr.com, to youtube.com, but posting basic details about what you do is simply too much. Not even simple logging. The mundane detail: you have at least one HUGE bug in your code. @jaredunn said in How much actual gains are in this backtest?: I hesitate to share my Github here Nobody wants to see your Github. Generate logging as to what you buy and what you sell and your current position in the market (as reported and not as wished by you) and the bug will be clear. It will be so obvious that anyone seeing your post will be able to tell you. The final portfolio value is telling you how much money you have made ... you are in negative. You lose half of your initial investment.
• C

  Bullish and Bearish Divergence MACD/RSI
  • CooleRnax  

  14
  0
  Votes
  14
  Posts
  2983
  Views

  

  After testing the code above and plotting the divergences that it found, I think that the issue is with this line: self.hfi = bt.ind.FindFirstIndexHighest(self.data.high, period=self.p.hl_period) It looks for the highest point (same for the lowest) within the hl_period (100 bars) Then in the following line it looks again backwards from that point to find the second high h_iterable = self.data.get(size=self.p.hl_period, ago=-int(self.hfi[0]) - self.p.hl_min) So basically the code returns signals for very old divergences. If you are on daily chart, you will try to trade on a divergence that is 100 days old. So the first day after the high point it signals a sell, then for the next 100 days it still signals a sell ! I have amended the first line to look for a high/low point for the last day only. In this way if the divergence is growing then it's fine you still get a sell signal, but if the divergence is gone then you don't try to keep selling for 100 days.
• A

  Multi Example
  • augustus  

  5
  0
  Votes
  5
  Posts
  301
  Views

  A

  I apologize if what I am going to ask now seems plain but I am trying to familiarize myself with the sketch of creation of the backtest strategy, but this line of code is not obvious to me and I would like someone to clarify it for me : **eval('dict(' + args.strat + ')')) from what I know, this is a function eval () whose parameter is the sequence runs python code but this record absolutely puzzled me because it seems to me that it is from packing of a dictionary items .- **
• A

  Multiple time period on a single stock
  • andrewsky  

  6
  0
  Votes
  6
  Posts
  247
  Views

  B

  @andrewsky said in Multiple time period on a single stock: self.data.get(size=self.p.period) If you ask for size=self.p.period you are always going to get period data points. If you need the array to grow with each iteration use the len of the data which keeps on growing with each iteration Docs - Platform Concepts - Lines len @andrewsky said in Multiple time period on a single stock: instead of self.data.get(), can I use self.data.close.get() and self.data.datetime.get() ? self.data.get is giving you the default for a multi-line object, which in most cases will be self.data.close.get. Working with the individual lines is obviously possible.
• M

  Multi-Data Indicator implementation help.
  • Mango Loco  

  8
  0
  Votes
  8
  Posts
  243
  Views

  B

  @mango-loco said in Multi-Data Indicator implementation help.: Is the above code the most efficient/cleanest/best way to go about handling the operations I'm doing? The fact is that backtrader allows you to do it in a declarative way or in an iterative way (and to mix both). There seems to a fixation out thee in many cases in using numpy (imported as np) for whatever reason which I fail to understand, even if it is meant to only calculate a single value. But if it suits you, there is no reason to do it otherwise. @mango-loco said in Multi-Data Indicator implementation help.: Is there a way/hack to eliminate the "instantiation lag" that occurs with indicators in BT? Is it possible to pass in data from before the start of the backtest so that any indicators are already fully instantiated at t=0? There is NO lag. Your reasoning and understanding is wrong. If you want a 20-period moving average you need 20 data points to calculate it. Full stop. If you want your indicators at YOUR DEFINED t=0 to be available, work with a data set which has an earlier start datetime.
• M

  Can BT resample the data in the user defined Indicator?
  • mothee  

  6
  0
  Votes
  6
  Posts
  264
  Views

  M

  @backtrader @ab_trader I finally implement the Indicator as below code, it seems work. but two small questions, when the ATR is coupled? and how do I set the minperiod when data with different time frames, I worked around by using len(self.l.art)>0, else it would raise errors. import numpy as np import backtrader as bt from ext.ult.ult import * from ext.Ind.OCHLInd import OCHLInd # https://www.tradingview.com/script/mtWhbCRf-RS-MTF-Volatility-Stop-V0/ class MtfVSInd(bt.Indicator): """ """ lines = ('trend', 'vs', 'max_close', 'min_close', 'atr') params = ( ('period', 5), ('factor', 1.618), ('mode', 0) # debug modes ) plotinfo = dict(subplot=False) plotlines = dict(vs=dict(_plotskip=False), trend=dict(_plotskip=True, _method='bar', alpha=0.50, width=1.0), max_close=dict(_plotskip=True, color='black'), min_close=dict(_plotskip=True, color='red'), atr=dict(_plotskip=True), ) def _plotinit(self): if self.p.mode == 1: self.plotinfo.subplot = True self.plotlines.vs._plotskip = True self.plotlines.trend._plotskip = True self.plotlines.max_close._plotskip = True self.plotlines.min_close._plotskip = True self.plotlines.atr._plotskip = False elif self.p.mode == 2: self.plotinfo.subplot = True self.plotlines.vs._plotskip = True self.plotlines.trend._plotskip = False self.plotlines.max_close._plotskip = True self.plotlines.min_close._plotskip = True self.plotlines.atr._plotskip = True elif self.p.mode == 3: self.plotinfo.subplot = True self.plotlines.vs._plotskip = True self.plotlines.trend._plotskip = True self.plotlines.max_close._plotskip = False self.plotlines.min_close._plotskip = False self.plotlines.atr._plotskip = True def __init__(self): # the ATR is Day time frame self.l.atr = bt.ind.ATR(self.data1, period=self.p.period) # get the avg(high+low) self.hl = OCHLInd(self.data0, type='hl') super(MtfVSInd, self).__init__() def next(self): if len(self.l.atr) > 0: if np.isnan(self.l.trend[-1]): self.l.trend[0] = 1 elif self.data0.l.close[0] > self.l.vs[-1]: self.l.trend[0] = 1 elif self.data0.l.close[0] < self.l.vs[-1]: self.l.trend[0] = -1 else: self.l.trend[0] = self.l.trend[-1] chg_trend = self.l.trend[0] - self.l.trend[-1] if chg_trend != 0: self.l.max_close[0] = self.data0.l.close[0] elif self.data0.l.close[0] >= self.l.max_close[-1]: self.l.max_close[0] = self.data0.l.close[0] else: self.l.max_close[0] = self.l.max_close[-1] if chg_trend != 0: self.l.min_close[0] = self.data0.l.close[0] elif self.data0.l.close[0] <= self.l.min_close[-1]: self.l.min_close[0] = self.data0.l.close[0] else: self.l.min_close[0] = self.l.min_close[-1] if np.isnan(self.l.vs[-1]): self.l.vs[0] = self.hl[0] elif chg_trend > 0: self.l.vs[0] = self.data0.l.close[0] - self.p.factor * self.l.atr[0] elif self.l.trend[0] > 0: self.l.vs[0] = max(self.l.vs[-1], self.l.max_close[0] - self.p.factor * self.l.atr[0]) elif chg_trend < 0: self.l.vs[0] = self.data0.l.close[0] + self.p.factor * self.l.atr[0] elif self.l.trend[0] < 0: self.l.vs[0] = min(self.l.vs[-1], self.l.min_close[0] + self.p.factor * self.l.atr[0]) else: self.l.vs[0] = self.l.vs[-1]
• L

  Accessing analysers in a strategy
  • livecat  

  4
  0
  Votes
  4
  Posts
  225
  Views

  B

  Because Strategy.stop() is a place where the strategy could still (in theory) do something which affects what an analyzer sees. For analyzers which calculate things during next it's a good point, for analyzers which wait until everything is complete, you will have to fetch it afterwards.
• K

  HADelta code correction?
  • kicking_potato88  

  2
  0
  Votes
  2
  Posts
  166
  Views

  B

  Indeed. Will be soon corrected.
• 

  Let strategies communicate with each other
  strategy multiple data • • Vincere  

  4
  0
  Votes
  4
  Posts
  237
  Views

  B

  The broker is a single account broker. All Strategies see the same value for a given asset. If that's not of your liking, your best bet will be to use a common parameter which can be used as a communication channel, like a dict or a list which will be written to by strategy A and read by B.
• A

  backtrader.linebuffer.LineBuffer object at
  • abyaby  

  4
  0
  Votes
  4
  Posts
  184
  Views

  A

  @backtrader Thanks that you mentioned the Limit Order. I thought, it doesnt matter, if I add cerebro.broker.setcommission(commission=0.002) Regarding unkown data I have backtest self.data.close and self.data.close[0] again. PnL with Limit order are different PnL with Market order are the same. What could it mean? I was especting data.close with market order a random number but surprisingly its the same. data.close[0] vs. data.close order.market -478.3 order.market -478.3 order.limit -634.26 order.limit 2370.7 And if some one know why data.close and data.close[0] have defferent limit order PnL ?