RevEnginePro 20 Pairs讨论

Started an optimization of some parameters on GBPUSD from January 2014 only and then will see results of the optimized parameters from 2009. Unfortunately optimization will need a lot of time, also if for the moment i am only optimizing 4 parameters. Let see what happens

Frank, those issues from '02.12.2015 13:00' ie., auto closing of orders, has been fixed, right?

Also, you haven't started using REP for trading live on any of your real accounts yet, right? Just curious to know how long you perform forward tests before using it on real accounts.

RedRhino posted:
Open an offline chart (M1) and use the attached script ( drag and drop ) on to the chart, this will tell you what data bars are missing from your history. You will get the results of the script in the Experts Tab ( Messages). such as

2015.02.16 05:01:38.467 sqCheckHistoryGapsScript EURUSDi,H1: EURUSDi/H1 Total: 52 missing bars / 2116 total bars = 2.0% missing bars

Frank, while attaching these scripts to offline M1 chart, I got the error saying 'sqCheckHistoryGapsScript' is not a script and cannot be executed. Same thing with sqShowHistoryGapsScrip as well.

Btw, regarding those 130 errors, I am using FXChoice ECN Demo server. That's not a market maker, right?

doing a lot of good work here guys,keep it up

corre71 posted:
Attached my BT for EURUSD with the suggested set file. The BT is made with tick datas and 1.2 pips spreads

Thank you for providing a Tick Data Modeling Quality backtest. The results are very impressive.

fxdad posted:
Frank, those issues from '02.12.2015 13:00' ie., auto closing of orders, has been fixed, right?

Yes, after digging through the MQL Documentation Book I realized that 'TimeCurrent()' relies on the last Tick from the server and is an inaccurate way of calculating time. Problem resolved.

fxdad posted:
Also, you haven't started using REP for trading live on any of your real accounts yet, right? Just curious to know how long you perform forward tests before using it on real accounts.

I trade manually https://www.myfxbook.com/members/RedRhinofx so it is difficult for me to not intervene when using robots on real money. I want to display accounts that only trade robots without any intervention then followers can see what the system does exactly as advertised. It is temping though because these results are impressive.

fxdad posted:
Btw, regarding those 130 errors, I am using FXChoice ECN Demo server. That's not a market maker, right?

It's associated with the StopLevel Limits set by the broker. Even if they are ECN they may not allow a Zero Stoplevel - https://book.mql4.com/appendix/limits - I'll take care of this issue today.

I added the source code for the ShowHistoryGaps. the script goes in Script folder and the other is an Indicator by the way. After reading the MQL book again, the TimeCurrent() function won't create a bar if the server doesn't provide any ticks during the 60 seconds of a Bar creation which makes massive holes in data. Metaquotes is amazingly stubborn.

corre71 posted:
Started an optimization of some parameters on GBPUSD from January 2014 only and then will see results of the optimized parameters from 2009. Unfortunately optimization will need a lot of time, also if for the moment i am only optimizing 4 parameters. Let see what happens

How did the GBPUSD optimization work out?

RedRhino posted:

corre71 posted:
Started an optimization of some parameters on GBPUSD from January 2014 only and then will see results of the optimized parameters from 2009. Unfortunately optimization will need a lot of time, also if for the moment i am only optimizing 4 parameters. Let see what happens

How did the GBPUSD optimization work out?

It is strange, i get errors and so the optimization fail. Maybe the Bt version cannot run optimizations?

I just ran an optimization and it worked fine for me. After recoding the entire header File ( mqh). I 'll recompile the BT version and upload it my site within 24 hours. It will solve the error reporting and should speed up the backtesting with a new feature I added.

Thanks would be great as I also couldn't get back testing to work

Do you know why the backtest didn't work? I can't duplicate the problem on my side using the same BT version. Remember H1 TF only, and Visual testing is disabled. BT only allows running without visual mode and optimization.

Another great day for RevEnginePro 152.00 pips.

What Parameters should be considered most important for optimizing?

MAfast , MAslow, MA Divergance
Minimum Range
K
MaxSecondsToTrade

Start with a small Lot size, and with a large balance, Once you find any combination that yields a positive return, you can now fine tune your optimization including

MinProfit
MaxLoss
MaxProfit
Grid

and then increase the Lotsize to a suitable size.

Lastly, you can optimize the
InitiateTrailAt
SellLevel
BuyLevel

Each currency pair behavior is unique and optimization will find more profitable setting. I'm using the same settings across all pairs once I found 1 profitable setfile from EURUSD data. It has proven that the overall system is robust, but this doesn't mean we can't better setting for individual pairs.

Thank you for backtesting.

Another one from Synergy FX
https://www.myfxbook.com/members/ForexVentureBot/forexventurebot/1161639

Ok thanks i will run some optimizations. Ist definately better than Rev Trader Pro..