problem - visual tester skipping bars or parts of them - page 2
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
I'd also read that article. But I cannot say I understand the algorithm. I concluded at that time that mt5 generated more Ticks for m1-timeframe (perhaps even in non-linear fashion); where-as mt4 generated o-h-l-c in linear fashion on m1-timeframe. My conclusion could be wrong but I said F-it :P that's how I'm gonna look at it. This helps me reinforce the fact the mt4 is totally insufficient for tracking small movements within the back-tester. Thats one of the reasons all the advanced contributers on this site recommended not using mt4 back-testing for scalping strategies... I asked this question allot when I was a noob.
To your point of missing a whole bars of 15-minutes timeframe which is what I was trying to clarify on my earlier questions. Thats a real problem. I've only experienced this problem once when I back-tested with a math-intensive price forecaster from the code base. From that experience, I drew the conclusion that it was a matter of either speed (cpu or mt4) or design of the indicator.
From your explanation, one could draw the conclusion that mt4 would skip a whole 15-minute bar with a Blank-Ea and Speed set to 0. Boy I cant wait til brokers start offering mt5.
Yep, this is kinda what happens :) IF I can't get a decent solution for this I'll create some stupid visual forward test generator in VB tonight.
Thing is I'm trying to get the feeling for the market by "reliving" old market, to see how candles get formed, to experience my trading wins and errors in a bit faster time stepping than real time. I'm trading RT but it's somewhat limiting :P
I tried to get it working with VT, but it seems like I'll have to grab the big guns now.
@yinyanclimax168
Create your own topic. Post your code within that topic using the SRC button above.
Don't hijack other people's post with something soooo-far off topic.
Sorry to bump this again, perhaps dayteam knows something more, still waiting if barrowboy says anything :)
I also tried a dummy ea with a default chart, so no indis, just plain bars, default color setting and everything, so lowest possible cpu consumption. The time hops still happen in exact same places.
Ok tried this on another computer, my desktop(intel i7 920, 8core and 6gb ram), which is a beast compared to the laptop (core 2 duo, 2 gigs of ram).
The same thing (so same history, all same settings, same indicators (saved in template) and the thing runs better. Almost no time skipping, the worst of it is that some minutes go by ligthiningly fast while others take loads of time, however it's still much better than on laptop, there I had whole chunks of minutes missing.
Now the funny part is, that if I try to control the speed of visual test, on laptop (older machine) the modelled time passes very fast at say X setting where x < 10 (and max speed as we know = 32). If I do the same on the desktop, I have to crank it up to 31 to get any decent movement at all.
Is it possible that MT4 is reading/writing data to disk during the visual test? the only reason I can figure out such enormous difference is that on the desktop i have my installation on SSD drive, and my laptop's drive is old, and possibly with some bad sectors. I should otherwise not be seeing time hops.
I would like to do the these tests on my laptop ( pretty much because of wider screen, and I can work outside the house), as on my desktop I have a really old (but bloody good) flatscreen of 1280x1024 so it's now only an incovenience, but it would make me feel more knowledgable(:P) if I(we) came to the bottom of this.
Fc