Thank you for the clarification and link, will look into it.

Has anyone successfully integrated machine learning packages e.g. btgym (https://community.backtrader.com/topic/509/implementation-of-openai-gym-environment-for-backtrader) into strategy development?

Thanks for reply. Is there a way to assign one broker to each strategy then? Because without ability to correctly evaluate position within each strategy, multi-strategy cannot be truly implemented.

How to get current position of an instrument within a strategy? It appears that "strategy.getposition(data)" returns the position of the instrument across all strategies.

.... the purpose of this need is to build indicator on user data. Any advice? Thanks !

Is it possible add user data (other than open/high/low/close/volume/openinterest) from pandas dataframe to build bt datafeed? Thanks!

Is it possible to compute indicator in Next section? If not, is there any way to calculate adaptive indicators which parameters (or sometimes data) need to be dynamically determined in the next process? Thanks!

I see. Thank you!

How to get the maximum close price for the past n days? I tried to use Max indicator but it is not supported?
clsmax = bt.ind.Max(data.close, period=self.p.nmax)
TypeError: init() got an unexpected keyword argument 'period'

Thanks !