PC HANGING LOADS SINCE 10.3
Forums › ProRealTime English forum › ProRealTime platform support › PC HANGING LOADS SINCE 10.3
- This topic has 10 replies, 3 voices, and was last updated 7 years ago by random45689.
-
-
01/12/2017 at 11:07 PM #2095001/13/2017 at 12:34 AM #2096101/15/2017 at 1:24 PM #21164
What machine you running Hug … 64 bit or 32bit and what RAM?
I got my 64 bit 4GB RAM going yesterday (been sitting on my desk for 6 months!) seemed well fast at first but then when I backtested 100,000 Units it slowed and nearly hung / froze but not quite. I’ll let you know if I have a breakthrough.
As an aside … I bought the 64 bit desktop (ex lease off ebay) for doing backtests, but the mither with 2 keyboards / mice put me off! Anyway I had a brainwave yesterday and I am now using the software below, its free and its magic! Works a treat! Easy to set up etc!
Cheers
GraHal01/15/2017 at 2:19 PM #21165I noticed it lagging the first day or so but seems to be running ok now – am on an Intel i7 16GB RAM Win10 64bit laptop with 3 X monitors attached but thought the issue might have been to do with whatever PRT is pulling through the internet rather than load/issues locally on the machine…?
Re: backtesting I had come across CalculateOnLastBars but being on V10.2 hadn’t tried it – might help?
Don’t know your setp GraHal or if you already know about etc. but an alternative would also be to hook up all monitors to main machine and RDC to the other but yeh InputDirector is a great solution!
01/15/2017 at 3:56 PM #2116901/15/2017 at 4:09 PM #21171Yes I am on 10.3 now but was meaning I came across the CalculateOnLastBars when on 10.2 and haven’t yet tried it in 10.3 (only works in 10.3)
Take a look here https://www.prorealcode.com/documentation/calculateonlastbars/
01/16/2017 at 12:06 PM #21278OK just tried out the calculateonlastbars (backtesting has today slowed to a crawl!) and it is for inidicators only 🙁 not strategies/backtesting – sorry about misleading info!
01/16/2017 at 2:39 PM #21317No probs Max, thanks for reporting back.
Maybe all the many IG users now on Demo 10.3 are crashing the PRT Servers? Do you get error messages? I’ve started to click to send the Error Report to PRT … else PRT may be oblivious to our sorry plight … it’s making me depressed 🙁 I’m smiling now having said that 🙂
Cheers
GraHal01/17/2017 at 2:38 PM #21460Haha 🙂 no error messages just very slow
To be honest I had steered clear of backtesting in 10.2 due to the lack of the tick by tick data and just ran strategies in demo ProOrder but now having witnessed the wild spread changes throughout a day (on FX at least) I am realising backtesting really just doesn’t represent ‘reality’ (unless it does pull in spread variations throughout the backtesting period? oh yeh I had asked about in this thread https://www.prorealcode.com/topic/overnight-spread-increase-backtest/ )
So I am going to stick to forward testing ideas in the demo ProOrder and it has taught me a LOT about how orders are being executed etc. that I would have had no idea about from backtesting only and is actually much more productive/better outcomes etc.
Oh and backtestings speeds at the mo seem to be sometimes very very slow and other times reasonable so I don’t ‘think’ it is a ‘local/machine’ problem and hopefully will improve!
Cheers
Max
1 user thanked author for this post.
01/17/2017 at 3:04 PM #21465Hi Max, thanks for the update, seems we think alike 🙂
When you say … unless it does pull in spread variations throughout the backtesting period? … if you are asking the question (?) then No … backtest doesn’t pull in spread variations, I guess that is one of the main reasons for the disparity in performance between backtest and live trading. I usually enter a compromise / worst case spread in the little setting box for backtest. You know about this spread box … yes? Screen shot attached.
Yes I agree … from last night backtest seemed to get ‘less slow’! 🙂
Cheers
GraHal01/17/2017 at 3:32 PM #21472Yup know about the spread box and yes was asking if backtesting did pull in spread variations throughout the period of the backtest but kinda knew the answer was no…
It’s not only the spread but the minstop distance (IG) which changes and is not taken into account in the backtesting and depending on the strategy being used will give very different results in a backtest vs demo let alone live ProOrder
At the moment forward testing demo prevails in my opinion (although no doubt I will soon come to see the merits of backtesting from another angle and change my tune which happens near on a daily basis the more I learn!)
1 user thanked author for this post.
-
AuthorPosts
Find exclusive trading pro-tools on