Tester in new build 604 doesn't work correct - page 9
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Build 509 vs Build 606 Strategy Tester speed test (Visual Mode).
Strategy Tester in Build 509 was 35.5% faster in visual mode than Strategy Tester in Build 606
Build 509 vs Build 606 Strategy Tester speed test (Visual Mode).
Strategy Tester in Build 509 was 35.5% faster in visual mode than Strategy Tester in Build 606
The data are the same, only the .hst format was changed. As SDC's test shows, backtesting results are the same with both build 509 and 60x. But it also shows that SDC first attempt was corrupted because he doesn't use the same data.
The broker can change your data though.
Thanks AV.
I'm aware the actual data in the .hst's hasn't changed. But if the pre 60x data that we have in the .hst was ordered OLHC and the post 60x date being added to that .hst is OHLC, what has been done to ensure the entire .hst is now in the right OHLC format for testing with 60x build?
Further, if I import an old format (OLHC) historic data file into the new terminal's history centre, does it have the smarts to import it properly, despite the fact it's not in the new format?
Apologies if I'm going over this too much.
Edit - incidentally, the like-for-like back-tests I've run in 60x to compare to 509 have end dates in Oct 13, there is no way my broker has altered my M1 data via the terminal prior to that point (thus resulting in testing discrepancies). This means (assuming my questions above have positive answers), that differing results in back-testing between 509 and 60x are due to something else (and I really hope that's the case). As Raptor mentions however, will it be coding or outstanding MT4.5 bugs? I guess we'll find out eventually but it will be a blessing to cross off dodgy .hst files/formatting from the list of possibilities.
Thanks AV.
I'm aware the actual data in the .hst's hasn't changed. But if the pre 60x data that we have in the .hst was ordered OLHC and the post 60x date being added to that .hst is OHLC, what has been done to ensure the entire .hst is now in the right OHLC format for testing with 60x build?
Further, if I import an old format (OLHC) historic data file into the new terminal's history centre, does it have the smarts to import it properly, despite the fact it's not in the new format?
Apologies if I'm going over this too much.
Edit - incidentally, the like-for-like back-tests I've run in 60x to compare to 509 have end dates in Oct 13, there is no way my broker has altered my M1 data via the terminal prior to that point (thus resulting in testing discrepancies). This means (assuming my questions above have positive answers), that differing results in back-testing between 509 and 60x are due to something else (and I really hope that's the case). As Raptor mentions however, will it be coding or outstanding MT4.5 bugs? I guess we'll find out eventually but it will be a blessing to cross off dodgy .hst files/formatting from the list of possibilities.
Are you using spread control script or fixed spread, when backtesting and comparing results ?
Thanks AV.
I'm aware the actual data in the .hst's hasn't changed. But if the pre 60x data that we have in the .hst was ordered OLHC and the post 60x date being added to that .hst is OHLC, what has been done to ensure the entire .hst is now in the right OHLC format for testing with 60x build?
I have not kept up with every discussion so I dont know the issue with .hst files. I can tell you if you take a .hst file from 509 and put it in the 60x history folder, the chart displays the same. That has to mean either there is no difference between 509 and 60x .hst files or 60x already knows how to correctly read .hst data created by previous versions.
We cannot get our complex, live, real-money EAs back-testing properly (i.e. achieve the same test results with 60x builds that we were getting with 509 and prior). They appear to trade properly (though it's not so easy to tell the detail), but back-testing shows very different results. There are so many different potential problems causing this, it's been impossible so far to work out which ones are the culprits and very difficult to fully eliminate anything from the list.
My post doesn't add any technical value I'm afraid but it does demonstrate that there are still capable people/teams out there struggling away with the new reality.
i compare 15 real money running EA's backtest results. most of them produce different results for 509 and 604 tests.. there are too many things with the new compiler to discover, all are low level (related with compiler) stuff. i wish i had a detailed ducumentation for changes to data types, and functions of old mql4, compared to new mql4. so, debugging of possible migration errors could be possible.
i wish, new mql4 compiler developers compare test results of 509 and 604, to recognize possible problems. shown simple-equal ea tests are not real ea's.. my guess, the problem is not related with tester, but, data types are not fully compatible, after migration.
i've tried strictied compilation, again the same. results are different. we are in another world. code compiles, but works in different world. so, we must be carefull for real money ea's... (509 compile ex4 again works different under 604 world, compared to 509 world.)
and, unfortunately modarator says that, find the error, and we fix it. if had know, what is changed in the machine, it could be possible. i remember what is compiler design, although 15+ years passed.
509 Graph:
604 Graph:
hi SDC, there are many high quality sold or public ea's. you can test a real (EA built 509 ex4) under 509 and 604, you will see differences. we are real developers here. your test code is not enogh to show risks. we proved our codes, comparing with years of future tests.
but even 509 ex4 runs different under 509 and under 604.
currently, i can not run my code under 604, because, proved codes will work in different way. i can not accept this. i have to UNDERSTAND what is different bw. compilers/executions. these risks forces me looking for alternative development environments. we spend our lives for current provings. this was not a game.
meanwhile, i'm coding during last 20 years..
hi SDC, there are many high quality sold or public ea's. you can test a real (EA built 509 ex4) under 509 and 604, you will see differences. we are real developers here. your test code is not enogh to show risks. we proved our codes, comparing with years of future tests.
but even 509 ex4 runs different under 509 and under 604.
currently, i can not run my code under 604, because, proved codes will work in different way. i can not accept this. i have to UNDERSTAND what is different bw. compilers/executions. these risks forces me looking for alternative development environments. we spend our lives for current provings. this was not a game.
meanwhile, i'm coding during last 20 years..
Maybe you are right, but you have to demonstrate this, as SDC shows that the results are the same for a given EA. Of course it's not the last word.
Can you provide a code .mq4 or an .ex4, and a test protocol, so we can try to reproduce the problem ?