Backtrader Community

Maxim Korobov

From 101 (csv, pandas) to some ML usage and only for trading + code, not just ML + code or trading theory.
https://www.udacity.com/course/machine-learning-for-trading--ud501

Maxim Korobov

I have to get orders and trades list after strategy stop.

According to http://backtrader.readthedocs.io/en/stable/strategy.html, they could be reached by _orders and _traders fields of strategy.

There is no problem with _orders - it's just a list. But _trades is

collections.defaultdict(AutoDictList)

Standard Google and SO answers didn't help to transfer it into list. How to do that?

Maxim Korobov

Thank you!

With this flag it works like a charm. Code:

import backtrader as bt


class WarmupDetector:

	@staticmethod
	def detect(strategies):
		greatest_warm_up_period, strategy_name = 0, ""

		runner = bt.Cerebro()
		for sta in strategies:
			runner.addstrategy(sta, silent=True)
		stub_data = bt.DataBase()
		runner.adddata(stub_data)
		
		sis = runner.run(runonce=False)

		for si in sis:
			period = si.get_warm_up_period()
			if period > greatest_warm_up_period:
				greatest_warm_up_period, strategy_name = period, si.__class__.__name__

		return greatest_warm_up_period, strategy_name

	@staticmethod
	def detect_period(strategies):
		return WarmupDetector.detect(strategies)[0]

Usage:

warm_up_period, warm_up_strategy_name = WarmupDetector.detect(strategies_to_add)

Maxim Korobov

In my particular case

line._plotvalue = False

helps.

Thanks for flexible solution!

Maxim Korobov

What's reason that f is None?

"version": "1.9.47.116"
backtrader\feed.py

...

def preload(self):
    while self.load():
        pass

    self._last()
    self.home()

    # preloaded - no need to keep the object around - breaks multip in 3.x
    self.f.close()
    self.f = None

...

Traceback (most recent call last):
  File "D:/Projects/trading-bot/main/lab/backtrader/playground.py", line 277, in <module>
	run()
  File "D:/Projects/trading-bot/main/lab/backtrader/playground.py", line 119, in run
	strategies = run_strategy(back_trader, run_mode == RunMode.DevelopIdea, from_date, to_date)
  File "D:/Projects/trading-bot/main/lab/backtrader/playground.py", line 218, in run_strategy
	strategies = back_trader.run()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\cerebro.py", line 1070, in run
	runstrat = self.runstrategies(iterstrat)
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\cerebro.py", line 1146, in runstrategies
	data.preload()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\feed.py", line 689, in preload
	self.f.close()
AttributeError: 'NoneType' object has no attribute 'close'

Maxim Korobov

Possible due to this :)

https://ichart.yahoo.com/table.csv?s=NFLX&a=0&b=1&c=2016&d=4&e=16&f=2017&g=h&ignore=.csv

alt text

Maxim Korobov

A little patch:

    if self.p.apikey is not None:
        urlargs.append('api_key=%s' % self.p.apikey)

There was a typo previously.

Maxim Korobov

What about implementing Quandl data source internally.
One cons is that API key is mandatory.

Maxim Korobov

Possible due to this :)

https://ichart.yahoo.com/table.csv?s=NFLX&a=0&b=1&c=2016&d=4&e=16&f=2017&g=h&ignore=.csv

alt text

Maxim Korobov

Vote to add indicator into main package! It's great for Forex market.

Maxim Korobov

What's reason that f is None?

"version": "1.9.47.116"
backtrader\feed.py

...

def preload(self):
    while self.load():
        pass

    self._last()
    self.home()

    # preloaded - no need to keep the object around - breaks multip in 3.x
    self.f.close()
    self.f = None

...

Traceback (most recent call last):
  File "D:/Projects/trading-bot/main/lab/backtrader/playground.py", line 277, in <module>
	run()
  File "D:/Projects/trading-bot/main/lab/backtrader/playground.py", line 119, in run
	strategies = run_strategy(back_trader, run_mode == RunMode.DevelopIdea, from_date, to_date)
  File "D:/Projects/trading-bot/main/lab/backtrader/playground.py", line 218, in run_strategy
	strategies = back_trader.run()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\cerebro.py", line 1070, in run
	runstrat = self.runstrategies(iterstrat)
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\cerebro.py", line 1146, in runstrategies
	data.preload()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\feed.py", line 689, in preload
	self.f.close()
AttributeError: 'NoneType' object has no attribute 'close'

Maxim Korobov

Thanks for suggestion, I replace it with:

back_trader.addanalyzer(TimeReturn, timeframe=bt.TimeFrame.Years, _name='annualreturn')

Looks loke pretty the same for SharpeRatio:

Traceback (most recent call last):
  File "D:/Projects/trading-bot/main/lab/backtrader/morning_buy.py", line 87, in <module>
	strategies = back_trader.run(timeframe=bt.TimeFrame.Minutes)
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\cerebro.py", line 794, in run
	runstrat = self.runstrategies(iterstrat)
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\cerebro.py", line 927, in runstrategies
	strat._stop()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\strategy.py", line 421, in _stop
	analyzer._stop()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\analyzer.py", line 194, in _stop
	self.stop()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\analyzers\sharpe.py", line 173, in stop
	ret_free_avg = average(ret_free)
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\mathsupport.py", line 38, in average
	return math.fsum(x) / (len(x) - bessel)
ZeroDivisionError: float division by zero

Maxim Korobov

I'm optimizing parameters for strategy.
Sometimes when strategy is full of fear, so perform no action. In suck situations ZeroDivisionError: float division by zero occur.

Traceback (most recent call last):
  File "D:/Projects/trading-bot/main/lab/backtrader/morning_buy.py", line 86, in <module>
	strategies = back_trader.run(timeframe=bt.TimeFrame.Minutes)
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\cerebro.py", line 794, in run
	runstrat = self.runstrategies(iterstrat)
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\cerebro.py", line 927, in runstrategies
	strat._stop()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\strategy.py", line 421, in _stop
	analyzer._stop()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\analyzer.py", line 194, in _stop
	self.stop()
  File "C:\Users\Home\Anaconda3\lib\site-packages\backtrader\analyzers\annualreturn.py", line 84, in stop
	annualret = (value_end / value_start) - 1.0
ZeroDivisionError: float division by zero

I checked that

b_trader.broker.setcash(1000.0)

was called earlier.

Maxim Korobov

Schedule main loop! I prefer APScheduler as it's flexibily for reports running based on schedule.

Maxim Korobov

In my particular case

line._plotvalue = False

helps.

Thanks for flexible solution!

Maxim Korobov

Am I missing something or there is not option to reach (cancel) this code in Plot_OldSync.plotind()?

        if not math.isnan(lplot[-1]):
            label += ' %.2f' % lplot[-1]

Maxim Korobov

@Maxim Korobov

Best posts made by Maxim Korobov

Latest posts made by Maxim Korobov