While the performance is indeed unaffected, there are analyzers which I have in place that utilize the trade log as a means of calculating certain values of interest (e.g. drawdown, number of trades, etc...) - If we are unable to get to the bottom of this, I will just rewrite them to use orders, instead.
I would go different way - first, add regular bars data feed into
bt, then apply renko to it as a filter or indicator (
bt has both of them) to this data feed. Use renko as a signal and issue orders against the first, regular bar data feed. First, this approach is inline with way
bt should be used, second - it excludes errors due to use of the renko bars as a data for order execution. But you choose.
Can you explain to me why it would reject an order when there was more than enough volume in a bar to fulfill the size of the order?
As stated above
bt doesn't consider the volume during the order execution, than the order will not be rejected due to low volume. I just noticed that in
bt there is no partial execution during back test.