MazParticipant
Veteran
Hi all,
Lately I have been dumping code writtin in PRT 10.2 code, into PRT 10.3. The result is terrible! Systems I have in 10.2 just completely fall over in 10.3.
Does anyone have a list of things to look out for when converting? Can anyone offer advice on how to get 10.3 to accurately replicate the behavior of code written in 10.2?
All the best (And Happy New Year all),
M
HNY Maz,
Is 10.3 on IG now ?
I’ve lost your email address BTW so wondered (Duh) if you could email me when you get a mo?
MazParticipant
Veteran
I have set up a temporary address. I can be reached in the next few hours on prtmaz at mailinator dot com
Since 10.3 backtest are more accurate than the 10.2 version, it would mean a regression into focusing to have the same results.
Check how much you have of “0 bar” problems with the old probacktest engine first, differences may surely come from there.
MazParticipant
Veteran
Well for example: A back test in 10.2 that generates many hundreds of positions does not open a single position in 10.3. Will get to the bottom of it I guess and update the thread when it becomes clear what’s going on.
EricParticipant
Master
CFD on both?
beacuse (as you probably know) the quotes are different (forex) between SB and CFD
EURUSD CFD 1.05317
EURUSD SB 10 531.7
this is probably not the problem but i mention it anyway
Check if it’s the same instrument like Eric mentioned it.
If your code refers to trading schedules, it may also differ.