@Wayne-Filkins-0 pass whatever you want to optimize to the maximize function. So for your case, you could multiply a binary value (profitable/unprofitable) by the percentage of profitable trades and pass that to the gen-opt function

from what it sounds like, this project would better fit your needs:

https://github.com/mementum/bta-lib

try passing exactbars=True to cerebro

not sure either.
I'd suggest cleaning up previous mistakes and walking through the code with breakpoints and more detailed logging. These kind of issues are always much easier to fix if you have personal access to the code/data.

I think the data feeds might be off by a second or so.

I would suggest printing the self.datas[0].datetime instead of the date. Possibly try add one or both of the datas using resampledata method.

But does it explain why only a share of my shares get sold or the order gets rejected?

self.order_target_percent(d, target=pos_size)

This line sold those 3 shares. You're ordering for the target pct to be hit every candle so as the value of your account fluctuates, this method will add/subtract to your position to match the pct.

Correct way to write this:

def next(self):
    pass

The problem in your code is here:

    # Sell stocks no longer meeting ranking filter.
    for i, d in enumerate(rankings):
        if self.getposition(d).size:
            if i > self.p.num_positions:
                self.close(d)

if i > 2, close the position, but since python uses zero-based indexing, it will only close a position if a stock is in 4th spot

The guy in your link didn't include fees & slippage.

Further more it should be noted that every single EMA combination tested (and most SMAs) outperformed the buy and hold annualized return of 6.32%^ during the test period (before allowing for transaction costs and slippage)Further more it should be noted that every single EMA combination tested (and most SMAs) outperformed the buy and hold annualized return of 6.32%^ during the test period (before allowing for transaction costs and slippage)

@vladisld Had the same issue. problem is that arange returns numpy datatypes which causes this operation to return a numpy array:

stddev = self.p.devfactor * StdDev(self.data, ma, period=self.p.period,
                                           movav=self.p.movav)

Converting self.p.devfactor to a standard float fixes the problem

@ab_trader again, just an indicator. this isn't changing your ohlc data.