It seems this LongOnly sizer could do the trick. Right?

class LongOnly(bt.Sizer):
    params = (('stake', 1),)

    def _getsizing(self, comminfo, cash, data, isbuy):
      if isbuy:
          return self.p.stake

      # Sell situation
      position = self.broker.getposition(data)
      if not position.size:
          return 0  # do not sell if nothing is open

      return self.p.stake

Hi community,

Since short selling may be forbidden in markets like CHINA AShares, I was wondering if there were any settings in Backtrader ensuring that.

Many thanks!

I found a similar exception as well. Does anyone have some idea about this problem?

Re: Mutidata + Multicore Optimization AttributeError

Hi admin,

As you said in this topic:

The problem is that you have to store the PnL value in the stop method of the Strategy. Because unless you do that you talk to the last instance of the PnL which cerebro keeps record of. (With maxcpus > 1 it would be the same)

What if I had a self defined analyzers which requires store some values in stop method. How can I modify my codes so that multi core opt could work properly, instead of error: AttributeError: Can't get attribute 'Lines_LineSeries_LineIterator_DataAccessor_ObserverBase_Observer_DataTrades_bd546845b4ba4f7489a48e0b65e9bec2

Many thanks!

Anybody knows the reason?

Hi community,

I found it did not work correctly when using the filler.
For example, I buy with a sizer of 3 every time, and set the filler to be FixedSize 1.
There should be three OrderExecutionBit Objects in order.executed, with index p1, p2.
However, I found 6 exbits were actually executed.

like,
[exbit0, exbit1, exbit2]

1. execute exbit0, (p1 = 0, p2 = 1, len = 1)
2. execute exbit0, exbit1, (p1 = 0, p2 = 2, len = 2)
3. execute exbit0, exbit1, exbit2 (p1 = 0, p2 = 3, len = 3)

It seems like p1 is set to 0 all the time, which might be the problem.

@scottz1 Found it! That's really helpful! Thx!

@scottz1 Thanks! How should I export the trade results?

Thanks! How should I export the trade results?

Hi community,

When I was doing backtest based on 1min kline data, could I also get analysis on daily performance, like daily trades profits, win/loss counts, daily drawdown, etc, say return in a list (one value per day.)

Many thanks!