Since it relies on the ZigZag indicator to plot the signals (which i think was already mentioned by other posters on this thread), I don't think it's possible without feeding all of the backtest data in "real-time". Which would take pretty significant cpu resources to do in order to get meaningful results, I'd imagine.Curious to know what you will find. I really like this indicator, even on smaller timeframes like the 3min but the re-painting made me remove the indicator because it`s important for me to backtest properly so if you can possibly find a way to modify the code to not re-paint, that would be great!
I have done already some backtesting using the on demand but it`s just way too time consuming.
Alternatively, I've been experimenting with the Supertrend Reversal and have been comparing it with this one and have been getting similar signals on the 10m & 15m timeframes (the Supertrend lags a bar behind the Trend Reversal Indicator). As far as i can tell, the Supertrend does not repaint either since the signals are triggered based on an ATR crossover, instead of high/low range pivots.
But honestly, the biggest challenge I've been facing is debugging/testing the actual thinkscript code. Working on it directly in TOS is hardly comparable to coding with an actual IDE.