Hi all

I've used Amibroker for some time and liked their trade list output. So written the analyzer for closed trades which provides similar information:

  chng%    pnl%    mae%    pricein  ticker      pnl/bar    size     value    mfe%    priceout       pnl    cumpnl  dateout       ref    nbars  dir    datein
-------  ------  ------  ---------  --------  ---------  ------  --------  ------  ----------  --------  --------  ----------  -----  -------  -----  ----------
  -6.07   -2.82  -10.12    281.397  SPY         -271.68     159   44742.1    0.68      264.31  -2716.78  -2716.78  2018-02-14      1       10  long   2018-01-31
  -1.59   -0.71   -2.2     119.128  TLT          -68.55     363   43243.6    1.01      117.24   -685.47  -3402.25  2018-02-22      2       10  long   2018-02-07
   1.21   -0.54   -4.01    268.163  SPY          -51.3     -158  -42369.7    1.44      271.41   -513.04  -3915.29  2018-03-01      3       10  short  2018-02-14
   1.89   -0.84   -2.16    117.946  TLT          -53.82    -363  -42814.4    0.8       120.17   -807.27  -4722.56  2018-03-15      4       15  short  2018-02-22
  -1.98   -1.26   -6       270.915  SPY          -49.35     158   42804.6    3.5       265.55  -1184.35  -5906.91  2018-04-05      5       24  long   2018-03-01
  -1.57    0.71   -0.41    265.55   SPY          336.49    -161  -42753.6    2.84      261.37    672.98  -5233.93  2018-04-09      7        2  short  2018-04-05
   0.93    0.42   -1.09    119.413  TLT           19.87     359   42869.2    2.47      120.52    397.45  -4836.48  2018-04-13      6       20  long   2018-03-15
  -0.89    0.4    -1.35    119.962  TLT           77.43    -361  -43306.4    1.28      118.89    387.13  -4449.35  2018-04-20      9        5  short  2018-04-13
   2.9     1.76   -2.06    264.486  SPY           71.27     165   43126.1    3.28      272.16   1710.47  -2738.88  2018-05-11      8       24  long   2018-04-09
  -0.16    0.07   -0.61    272.585  SPY            6.96    -160  -43613.6    0.94      272.15     69.6   -2669.28  2018-05-25     11       10  short  2018-05-11
   0.95    0.32   -2.43    118.985  TLT           10.37     365   43429.6    2.97      120.11    310.98  -2358.3   2018-06-04     10       30  long   2018-04-20
  -1.18    0.26   -0.32    120.11   TLT           86.15    -182  -21860      1.41      118.69    258.44  -2099.86  2018-06-07     13        3  short  2018-06-04
   0.2     0.09   -0.01    118.69   TLT           29.76     372   44152.7    1.52      118.93     89.28  -2010.58  2018-06-12     14        3  long   2018-06-07
   2.36   -1.08   -3       118.93   TLT          -74.67    -372  -44242      0.25      121.74  -1045.32  -3055.9   2018-07-02     15       14  short  2018-06-12
  -0.78   -0.33   -2.73    275.275  SPY          -11.28     159   43768.8    1.53      273.14   -315.8   -3371.7   2018-07-06     12       28  long   2018-05-25
   1.88   -0.85   -2.3     273.14   SPY         -204.31    -159  -43429.3    0.16      278.28   -817.26  -4188.96  2018-07-12     17        4  short  2018-07-06
  -2.24   -1.01   -2.35    122.013  TLT          -64.31     353   43070.5    0.74      119.28   -964.63  -5153.59  2018-07-24     16       15  long   2018-07-02
   2.55    0.86   -0.24    278.28   SPY           45.82     156   43411.7    2.78      285.39    824.76  -4328.83  2018-08-07     18       18  long   2018-07-12
  -0.02    0.01   -1.01    119.053  TLT            0.81    -362  -43097      0.83      119.03      8.15  -4320.68  2018-08-07     19       10  short  2018-07-24

Feel free to use/test it, let me know about any bugs found. The script:

# Trade list similar to Amibroker output
class trade_list(bt.Analyzer):

    def get_analysis(self):

        return self.trades


    def __init__(self):

        self.trades = []
        self.cumprofit = 0.0


    def notify_trade(self, trade):

        if trade.isclosed:

            brokervalue = self.strategy.broker.getvalue()

            dir = 'short'
            if trade.history[0].event.size > 0: dir = 'long'

            pricein = trade.history[len(trade.history)-1].status.price
            priceout = trade.history[len(trade.history)-1].event.price
            datein = bt.num2date(trade.history[0].status.dt)
            dateout = bt.num2date(trade.history[len(trade.history)-1].status.dt)
            if trade.data._timeframe >= bt.TimeFrame.Days:
                datein = datein.date()
                dateout = dateout.date()

            pcntchange = 100 * priceout / pricein - 100
            pnl = trade.history[len(trade.history)-1].status.pnlcomm
            pnlpcnt = 100 * pnl / brokervalue
            barlen = trade.history[len(trade.history)-1].status.barlen
            pbar = pnl / barlen
            self.cumprofit += pnl

            size = value = 0.0
            for record in trade.history:
                if abs(size) < abs(record.status.size):
                    size = record.status.size
                    value = record.status.value

            highest_in_trade = max(trade.data.high.get(ago=0, size=barlen+1))
            lowest_in_trade = min(trade.data.low.get(ago=0, size=barlen+1))
            hp = 100 * (highest_in_trade - pricein) / pricein
            lp = 100 * (lowest_in_trade - pricein) / pricein
            if dir == 'long':
                mfe = hp
                mae = lp
            if dir == 'short':
                mfe = -lp
                mae = -hp

            self.trades.append({'ref': trade.ref, 'ticker': trade.data._name, 'dir': dir,
                 'datein': datein, 'pricein': pricein, 'dateout': dateout, 'priceout': priceout,
                 'chng%': round(pcntchange, 2), 'pnl': pnl, 'pnl%': round(pnlpcnt, 2),
                 'size': size, 'value': value, 'cumpnl': self.cumprofit,
                 'nbars': barlen, 'pnl/bar': round(pbar, 2),
                 'mfe%': round(mfe, 2), 'mae%': round(mae, 2)})

It returns the list of dictionaries. In order to print it as a table above I've used tabulate module:

from tabulate import tabulate
...
    # add analyzers
    cerebro.addanalyzer(trade_list, _name='trade_list')
...
    # run backtest
    strats = cerebro.run(tradehistory=True)
...
    # get analyzers data
    trade_list = strats[0].analyzers.trade_list.get_analysis()
    print (tabulate(trade_list, headers="keys"))

Updated on 09/17/2018

I've tried pyevolve library for evolutionary optimization. This is first shot for the system shown above (2 params only, but lets be consistent), takes longer time than particle swarm optimization.

# example of optimizing SMA crossover strategy parameters using 
# evolutionary Optimization in the pyevolve python library
# http://pyevolve.sourceforge.net/

from datetime import datetime
import backtrader as bt

from pyevolve import G1DList
from pyevolve import GSimpleGA

class SmaCross(bt.SignalStrategy):
    params = (
        ('sma1', 10),
        ('sma2', 30),
    )

    def __init__(self):
        SMA1 = bt.ind.SMA(period=int(self.params.sma1))
        SMA2 = bt.ind.SMA(period=int(self.params.sma2))
        crossover = bt.ind.CrossOver(SMA1, SMA2)
        self.signal_add(bt.SIGNAL_LONG, crossover)

data0 = bt.feeds.YahooFinanceData(dataname='YHOO',
                                  fromdate=datetime(2011, 1, 1),
                                  todate=datetime(2012, 12, 31))

def runstrat(chromosome):
    
    cerebro = bt.Cerebro()
    cerebro.addstrategy(SmaCross, sma1=chromosome[0], sma2=chromosome[1])

    cerebro.adddata(data0)
    cerebro.run()
    return cerebro.broker.getvalue()

genome = G1DList.G1DList(2)
genome.setParams(rangemin=2, rangemax=55)
genome.evaluator.set(runstrat)
ga = GSimpleGA.GSimpleGA(genome)
ga.setGenerations(5)
ga.setMutationRate(0.2)
ga.setCrossoverRate(1.0)
ga.setPopulationSize(10)
ga.evolve(freq_stats=1)
print ga.bestIndividual()

I am not familiar with real trading approach using bt. @RandyT probably can tell more on this.

Below is full code for sample trading system, implementing fixed stop loss and take profit orders.

from __future__ import (absolute_import, division, print_function, unicode_literals)

import datetime as dt
import argparse

import backtrader as bt

def parse_args(pargs=None):

    parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter,
                                     description=('Strategy auto-generation framework'))

    parser.add_argument('--data', required=False, default='SPY',
                        metavar='TICKER', help='Yahoo ticker to download')

    parser.add_argument('--data_from', required=False, default='2015-01-01',
                        metavar='YYYY-MM-DD[THH:MM:SS]',
                        help='Date[time] in YYYY-MM-DD[THH:MM:SS] format')

    parser.add_argument('--data_to', required=False, default='2016-12-31',
                        metavar='YYYY-MM-DD[THH:MM:SS]',
                        help='Ending date[time]')

    parser.add_argument('--plot', required=False, default='',
                        nargs='?', const='volume=True',
                        help='kwargs in key=value format')

    return parser.parse_args(pargs)

# script arguments parsing
args = parse_args(None)

# quotes feed for one ticker
data = bt.feeds.YahooFinanceCSVData(dataname=args.data+'.csv',
                                    fromdate=dt.datetime.strptime(args.data_from, '%Y-%m-%d'),
                                    todate=dt.datetime.strptime(args.data_to, '%Y-%m-%d'),
                                    reverse=True, adjclose=False, plot=True)

# observer showing stop loss and take profit levels
class SLTPTracking(bt.Observer):

    lines = ('stop', 'take')

    plotinfo = dict(plot=True, subplot=False)

    plotlines = dict(stop=dict(ls=':', linewidth=1.5),
                     take=dict(ls=':', linewidth=1.5))

    def next(self):

        if self._owner.sl_price != 0.0:
            self.lines.stop[0] = self._owner.sl_price

        if self._owner.tp_price != 0.0:
            self.lines.take[0] = self._owner.tp_price

# master strategy
class MasterStrategy(bt.Strategy):

    params = (
    ('ema_period', 200),
    ('sl', 1.0), ('rr_ratio', 3), ('n', 5),
             )

    def __init__(self):

        # init order variables
        self.order, self.sl_order, self.tp_order = None, None, None
        self.sl_price, self.tp_price = 0.0, 0.0

        self.ema = bt.indicators.ExponentialMovingAverage(period=int(self.p.ema_period),
                                                          subplot=False)
        
        # entry signals
        self.long = bt.indicators.CrossUp(self.data.close, self.ema)
        self.short = bt.indicators.CrossDown(self.data.close, self.ema)

    def notify_trade(self, trade):

        if trade.isclosed:

            # clear stop loss and take profit order variables for no position state
            if self.sl_order:
                self.broker.cancel(self.sl_order)
                self.sl_order = None
                self.sl_price = 0.0

            if self.tp_order:
                self.broker.cancel(self.tp_order)
                self.tp_order = None
                self.tp_price = 0.0

    def notify_order(self, order):

        if order.status in [order.Completed, order.Margin]:
            
            # clear order variable
            self.order = None

    def next(self):

        # process entries at bar open if not in position
        if not self.position:
            
           # check entry to long position
           if self.long[0] == True:
                self.order = self.order_target_percent(target=1.0)

           # check entry to short position
           elif self.short[0] == True:
                self.order = self.order_target_percent(target=-1.0)
                
        # process exits and position support orders if in position
        if self.position:

            price_sl_long = self.position.price * (1 - self.p.sl / 100)
            price_sl_short = self.position.price * (1 + self.p.sl / 100)

            price_tp_long = self.position.price * (1 + self.p.sl / 100 * self.p.rr_ratio)
            price_tp_short = self.position.price * (1 - self.p.sl / 100 * self.p.rr_ratio)
            
            # cancel existing stop loss and take profit orders
            if self.sl_order:
                self.broker.cancel(self.sl_order)

            if self.tp_order:
                self.broker.cancel(self.tp_order)

            # check & update stop loss order
            self.sl_price = 0.0
            if self.position.size > 0 and price_sl_long !=0: self.sl_price = price_sl_long
            if self.position.size < 0 and price_sl_short !=0: self.sl_price = price_sl_short

            if self.sl_price != 0.0:
                self.sl_order = self.order_target_value(target=0.0, exectype=bt.Order.Stop,
                                                        price=self.sl_price)
            # check & update take profit order
            self.tp_price = 0.0
            if self.position.size > 0 and price_tp_long !=0: self.tp_price = price_tp_long
            if self.position.size < 0 and price_tp_short !=0: self.tp_price = price_tp_short

            if self.tp_price != 0.0:
                self.tp_order = self.order_target_value(target=0.0, exectype=bt.Order.Limit,
                                                        price=self.tp_price)

# single backtest run
def single_backtest():

    cerebro = bt.Cerebro()
    cerebro.addstrategy(MasterStrategy)
    cerebro.addobserver(SLTPTracking)

    cerebro.adddata(data)
    cerebro.broker.set_cash(1000000.0)
    cerebro.broker.set_coc(True)
    cerebro.run()

    print ('Broker value %0.2f' % cerebro.broker.getvalue())

    return cerebro

if __name__ == '__main__':

    cerebro = single_backtest()

    if args.plot:
        cerebro.plot(style='bar', numfigs=1,
                        barup = 'black', bardown = 'black',
                        barupfill = True, bardownfill = False,
                        volup = 'green', voldown = 'red', voltrans = 50.0, voloverlay = False)

Outputs -
ref - bt's unique trade identifier
ticker - data feed name
datein - date and time of trade opening
pricein - price of trade entry
dir - long or short
dateout - date and time of trade closing
priceout - price of trade exit
chng% - exit price to entry price ratio
pnl - money profit/loss per trade
pnl% - proft/loss in %s to broker's value at the trade closing
cumpnl - cumulative profit/loss
size - max position size during trade
value - max trade value
nbars - trade duration in bars
pnl/bar - profit/loss per bar
mfe% - max favorable excursion
mae% - max adverse excursion

You can also run optimization having data feed as an optimization parameter. In this case it will be single cerebro run.

In order to have something automatically plotted in bt it should be data feed, indicator or observer. So you may want to develop an indicator or observer.

Docs - Indicator Development
DOcs - Observers and Statistics

On the other side, formula you are using will not give you a picture shown, it will be only a channel moving together with the close price. It will never cross the price. Hope you understand it and this formula is only first step.

Some time ago in order to develop similar picture I developed stop loss/take profit observer. Here is the link, hope it might help

Stop loss/Take profit observer

Data feed class need to be re-written as follows:

class dataFeed(btfeed.GenericCSVData):
    params = (
        ('dtformat', '%Y%m%d %H:%M:%S'),
        ('datetime', 0),
        ('open', 1),
        ('high', 2),
        ('low', 3),
        ('close', 4),
        ('volume', -1),
        ('openinterest', -1)
    )

and then data initiated as follows:

    data = dataFeed(dataname=path_to_file_with_quotes, timeframe=bt.TimeFrame.Minutes, compression=60)

Put self.tradesize = None or self.tradesize = 0 in the __init()__. It is undefined for first next() calls before you get your trade open.

@tracy it was discussed several times, try to search forums with the word multi. Also check the following post in the blog.

Straight forward way is to load all stocks into bt and do your calculations.

bt.If, bt.And, bt.Or

Docs - Platform Concepts

@abhishek-anand-1 i never checked this by myself.

@mihir sure, it is possible, look at Docs - Broker, get_cash and get_value.

I would recommend to read docs till the end before go deep with bt, will save a lot of time.

@mihir said in Bactrader broker portfolio value increases instead of decreasing when only buying the stocks.:

In the strategy I am only buying, thus the portfolio should only decrease

Cash should be decreasing, but portfolio value depends on the stock prices. They go up, portfolio value goes up as well.

@ssbilis said in Never Closes Position:

What logging do you recommend?

Usually I recommend to print all data related to the issue in the script. In your case everything related to position closing algorithm which apparently doesn't work. So the following maybe printed (based on your script): self.status, self.beta.spread[0], self.beta.spread_sma15[0], self.beta.beta[0] as well as prices which are used to execute orders. Cash as well.

Than looking thru this data you can see if your conditions met or not. Until that time I would agree with @run-out.

@ssbilis typically orders are not executed if there is not enough cash on the account. i would recommend you to add more logging in your script to see what is actually going on.

@kumzy highly likely the indicator is applied only once for the first data feed. Hard to say more without seeing the actual code.

@alechter self.log() method is introduced and it should be called to print the data. You can call it in the next() or in the other strategy methods.

Ad to @run-out comment - you may want to look into docs more to understand how bt works and how lines should be processed in the init and in the next calls.

@alechter

Logging is described in the Quickstart
https://www.backtrader.com/docu/quickstart/quickstart/

@davidavr I think it is due to testing or debugging reasons. From one of the posts above:

I added 100 in p1 because I wanted to set an unachievable price so that it won't get executed but it is still getting executed.

@davidavr said in Order Price is different from issued price:

But what you are describing is a very odd strategy: you want to wait until the price is much higher before you buy it. Most strategies boil down to "buy low, sell high" in one form or another (but not necessarily in that order).

this is what typically trend following strategies are doing: wait for the price to make a move in the necessary direction and only after this move the position in that direction is open. Buy high, sell much higher.

TS (probably) tries to simulate price channel breakout strategy.