Topics created by blonc

Were you ever able to integrate polygon.io? @blonc

SOLVED

the data feed was fine, it was with in the next() have to split from datetime to date and separate time

class Master(bt.Strategy): def __init__(self): self.dataclose = self.datas[0].close def next(self): # self.log('Line') print(self.datas[0].datetime.date(0), self.datas[0].datetime.time(0)) def runstrat(): cerebro = bt.Cerebro() cerebro.broker.setcash(550000.0) data123 = bt.feeds.MarketStore( symbol='BTC', query_timeframe='1Min', timeframe=bt.TimeFrame.Minutes, ) cerebro.adddata(data123) cerebro.addstrategy(Master) cerebro.run() print('finished') if __name__ == '__main__': runstrat()

output

2018-06-03 12:00:00 7685.13 2018-06-03 12:01:00 7685.13 2018-06-03 12:02:00 7685.13 2018-06-03 12:03:00 7685.13 2018-06-03 12:04:00 7685.13 2018-06-03 12:05:00 7685.12 2018-06-03 12:06:00 7685.12 2018-06-03 12:07:00 7681.01 2018-06-03 12:08:00 7681.01 2018-06-03 12:09:00 7681.01 finished

SOLVED

the fromdate with in the datafeed initialize was a integer and needed to be a datetime object . IE datetime.date(2018, 6, 1)

That would have been the suggestion. Glad you could solve it quickly.

@blonc said in Strategy Selection Pattern:

The concept I am trying to achieve is to have many strategies all being monitored by a master strategy. This way, the master strategy could be using a markov chain to monitor returns to turn on and off individual strategies based on the total returns of all strategies. It seems the best path is to , correct me if wrong, put the strategies logic into indicators and bring them into one master strategy.

@backtrader said in Strategy Selection Pattern:

Sub-strategies is not a pattern which does exist in backtrader. Using Indicator instances could do the trick.

Was also thinking about utilizing the concept of having a "master" strategy turning on and off "micro strategies", so I'm glad I found this post :)
I can think of a lot of use cases for using a master strategy that based on a wide overview of a market/s can control what micro strategies are suited to be used.

the idea of using indicators as sub-strategies seem brilliant (haven't tested it yet).
but from my understanding strategies can be better analyzed and give valuable insights. I hope we can wider the discussion just to explore more solutions (as @backtrader mentioned nested strategies are not supported)

one idea I have that I don't think break backtrader structure is:
declaring a 'master' that run first and based on whatever user logic it creates a 'signal dictionary' that is then passed to all other micro strategies. allowing them to decide if to run or not in next().
this way all strategies are kept in the same level of operation (not nested)

@backtrader

if not I believe I found a work around, by passing up through params the len() of data feeds added (IE how many) and then doing the following in int()

datas_dic = {} for num in range(0,self.p.total_feeds): sym_name = self.datas[num]._name if not sym_name in datas_dic: datas_dic[sym_name] = ('self.datas%s' % num)

returns a list looking similar to :

{'WASH': 'self.datas2', 'ACN': 'self.datas1', 'SWK': 'self.datas0', 'WSBC': 'self.datas3'}

FOR CONTEXT:

this is needed because i am creating a large portfolio of pairs. I have a single master strategy that then creates sub strategies and I need to create a new sub strategy for every pair, and to create that sub strategy I need to know which datas it needs to run its calculations.

@blonc said in notify_trade():

notify_trade() is only called on the complete closing of a position

notify_trade is called:

When a trade is started When a trade is closed

@blonc said in notify_trade():

but closes it over multiple signals,

You know you are closing it, but the platform cannot know it. A Trade may go down from having a size of 50 units to 1. Is that closing? No. Because it may go again up to 100.

This was already answered in another post and your assumption is nowhere to be found that things are separated by strategy.

The broker is a single account broker. Anything you do, you do it on a single account. Positions will be closed on that account and tradeid is just a marker for the separation of an artificial thing which is a Trade (a collection of operations)

@blonc said in tradeid - multiple strategies on same symbol:

do I need to also include the data?

If you don't, the 1st data in the system will be used. This is in the documentation.

@laurent-michelizza said in more data more problems:

That could happen if your data doesn't have the same start/end dates or is not properly aligned

Only the 1st. The system won't enter the next phase of the strategy until all data feeds (and also the associated indicators) have covered the warm-up period (aka as minperiod)

The reason is clear: in next all elements are guaranteed to have a value available at index [0]. Hence the need to await the minimum period of all elements to be consumed.

If you need to access data before, use prenext, but each element will have to be checked for actual data.

See:

Docs - Operating the Platform Docs - Strategy

@blonc said in analyzer with multiple strategies - results[n] no matter n gives identical results.:

it seems to be giving me results for the entire system not independent strategies.

Yes. The broker is a single account broker. Nowhere it is stated that strategies operate on individual accounts. Actually, the examples that show multiple strategies operating, show that they operate on the same account.

If you want strategies that operate on different accounts, run them with different cerebros.

The main point here is: there is a single strategy which is used against different pairs.

The optimization is to optimize the parameters of a strategy. What you want is to optimize for pairs and that's something which is not the target of the optimization. For each pair you can optimize the other parameters.

But if you insist (from your other post) in that data feeds are passed as params you will probably hit a wall with the optimization approach.

@blonc said in passing data to strategy via an *args:

the bellow issue:

And what's exactly the issue?

That's a well defined and documented behavior of the platform. It has also been discussed before by people who wanted to access data before all warm-up periods were fulfilled and worked using prenext and adding multiple checks.

@blonc said in drawdown per trade in def notify_trade()::

I know you can get drawdown in the analyzer as wellbut that seems to just be the max drawdown

Wrong knowledge then. Docs - Analyzers

Quoting from the DrawDown docs:

Methods: - ``get_analysis`` Returns a dictionary (with . notation support and subdctionaries) with drawdown stats as values, the following keys/attributes are available: - ``drawdown`` - drawdown value in 0.xx % - ``moneydown`` - drawdown value in monetary units - ``len`` - drawdown length - ``max.drawdown`` - max drawdown value in 0.xx % - ``max.moneydown`` - max drawdown value in monetary units - ``max.len`` - max drawdown length

for anyone who happens to run into this into the future. this script has to be ran in Python 3.5 or > due to the var typing in the functions. Or they need to be removed, it functions fine with out them.

that solved both problems, i see the mistake now. silly clearly. thank you for the help =)

2016-01-06, Buy Sig nan Sell Sig nan Position 0 2016-01-06, Buy Sig nan Sell Sig nan Position 0 2016-01-06, Buy Sig nan Sell Sig nan Position 0 2016-01-06, Buy Sig 1.0 Sell Sig 0.0 Position 0 2016-01-06, Buy Sig 1.0 Sell Sig 0.0 Position 1 2016-01-06, Buy Sig 1.0 Sell Sig 0.0 Position 1 2016-01-06, Buy Sig 1.0 Sell Sig 0.0 Position 1 2016-01-07, Buy Sig 0.0 Sell Sig 0.0 Position 1 2016-01-07, Buy Sig 0.0 Sell Sig 0.0 Position 1 2016-01-07, Buy Sig 0.0 Sell Sig 1.0 Position 1 2016-01-07, Buy Sig 0.0 Sell Sig 1.0 Position -1 2016-01-07, Buy Sig 0.0 Sell Sig 1.0 Position -1 2016-01-07, Buy Sig 0.0 Sell Sig 1.0 Position -1 2016-01-07, Buy Sig 0.0 Sell Sig 1.0 Position -1 2016-01-07, Buy Sig 0.0 Sell Sig 1.0 Position -1 2016-01-07, Buy Sig 0.0 Sell Sig 1.0 Position -1