Joined Jun 2009
|
Status: <-That's how u gonna b, in the END!
|1,509 Posts
Jay320
Thanks for taking such advise in such great sense.
Not going to waste your time further with "redundant" threads which have been made sticky by FF mods N very very sorry for your time already wasted.
You have been welcomed already,to win it!
Cheers n Regards
Hi guys, I was testing out vGrid2.2 on default settings and there is something wrong - it initiates a grid with certain increment & levels. When that grid takes profit, it doesn't initiate a new "set" with new pending orders. It continues to use the same levels of the old grid! Version 2.1 doesn't seem to have this problem...perhaps a bug in "restoreGridstate"?
If you look at the results of backtesting, all the grids have the same SL and TP. As you know, it should open new straddling pending orders upon starting subsequent grids.
No problems with v2.1. What exactly are you having?
Thanks
Ignored
change the min lot size to 0.02 (or any other lot size), sometime it places pending orders correctly, sometimes it places pending order incorrectly. i do my testing on alpari demo, gu, increment 25pips.
Hey Everyone,
I found the bug with version 2.2. Version 2.3 to fix the bug will be out shortly. I have been forward testing for a number of days and seems to be working perfectly (even using min lot size). I have found a couple of minor bugs with mgrid version also so the results will not be identical. But in my more careful I am verifying every order is correctly placed. It has been a little time consuming so I apologize for the delay... No new features except enhanced comments and bugs fixed :-)
Viper
P.S. After this release, I may begin working on ViperMultiGrid... But that will probably be slow coming...
I was thinking 3 or 4 as the max because I think the ideal spacing is to have a new grid start when you reach b2/s2 (3 levels) and that will actually only allow 3 grids (one will terminate before any more will start... hmmm... I thought the new grid spacing should be based on the base grid levels. Is that how you select them ?
I was thinking 3 or 4 as the max because I think the ideal spacing is to have a new grid start when you reach b2/s2 (3 levels) and that will actually only allow 3 grids (one will terminate before any more will start... hmmm... I thought the new grid spacing should be based on the base grid levels. Is that how you select them ?
Viper
Ignored
3 or 4 is probably optimum. I thought 7 to 10 may work (looking for a close every 10-15 pips) but it's causing me problems right now. Pending orders are too close to each other. Too many positions getting opened in a tight range.
Hey Everyone,
I found the bug with version 2.2. Version 2.3 to fix the bug will be out shortly. I have been forward testing for a number of days and seems to be working perfectly (even using min lot size). I have found a couple of minor bugs with mgrid version also so the results will not be identical. But in my more careful I am verifying every order is correctly placed. It has been a little time consuming so I apologize for the delay... No new features except enhanced comments and bugs fixed :-)
I have attached version 2.3. I have done a fair amount of testing this time (2 weeks forward testing and some back testing) so I am pretty confident that it is bug free and works as intended.
It does not work exactly like mgrid, because mgrid had a couple bugs which are fixed in this release. This is MUCH better because it opens positions slower (due to the bug fix). mgrid (almost) always opened three positions (b1,b2,b3) everytime because of an error in the way it calculated the resulting new profit. The correctly behaving code will often only open new positions at the b3/s3 level. Please beware of this and don't assume that it is not working correctly. This actually reduces whipsaws considerably and reduces the Death Trade frequency. It is a little like ForexSaints suggestion to mitigate the Death Trade.
Again, I set the input parameters to work like mgrid (new features off), but it will still work a little differently now - due to the mgrid bug fixes.
Anyway, here you go, enjoy!
Trade well, Live well,
I have attached version 2.3. I have done a fair amount of testing this time (2 weeks forward testing and some back testing) so I am pretty confident that it is bug free and works as intended.
It does not work exactly like mgrid, because mgrid had a couple bugs which are fixed in this release. This is MUCH better because it opens positions slower (due to the bug fix). mgrid (almost) always opened three positions (b1,b2,b3) everytime because of an error in the way it calculated the resulting new profit. The correctly behaving...
Ignored
Thnaks for the update man
But i have an issue !
when i try to add the ea to the chart i cant? not like the other EAs
|
Joined May 2008
|
Status: do i look confused?
|616 Posts
Thanks Viper. I've been looking forward to this.
Can you explain a bit more about the bugs in orignal mgrid & the way it calculates profit?
Is it that it is supposed to give a fixed profit on the close of every set regardless of the number of swings? Original mgrid would give a profit if you had a straight run, and almost double the ammount of profit if you had a swing between the buy and sell levels.
Will start testing as soon as the Market opens later.
Have done some back testing just to see the order process. Works like a charm and doesn't double on the initial opening (which it use to for me occasionaly).
Desi/At the best.
Not getting that problem at all. Working fine for me. Can you explain exactly whats going on. Is it just not attaching to your chart?
Have you tried compiling it before attaching it?
Oh hang on.... do you have the Orderreliable file in the experts/include folder? If you don't, you will definately need it.
If you don't have it, here is the file you need. Just simply paste the file into the experts/include folder. It will still give 3 warnings when you compile the EA but ignore them, doesn't effect anything.
Can you explain a bit more about the bugs in orignal mgrid & the way it calculates profit?
Is it that it is supposed to give a fixed profit on the close of every set regardless of the number of swings? Original mgrid would give a profit if you had a straight run, and almost double the ammount of profit if you had a swing between the buy and sell levels.
Will start testing as soon as the Market opens later.
Ignored
When the program decides whether it needs to buy or sell, it calls a method called checkPendingProfit() (checkProfits() in original). If the profit it will achieve at TP is less than the goal (2*increment) it goes into a buy/sell loop. The loop says start at the max increment down to one. After each buy/sell it uses an equation to see if the new purchase will put the profit at TP over the target. If it does, it ends the loop right there, it does NOt continue through all levels.
In the original mgrid the equation which tries to add the additional profit from the position just opened is incorrect. I noticed even though the increment was 15 pips in my case and the third level uses three positions, the equation was adding only a few pips to the newly computed profit at TP, causing it to almost always loop through all levels.
The fact that it is always adding positions at level 1, the overall position grows more quickly than it needs to or should.
I hope that answers your question without adding to the confusion...
When the program decides whether it needs to buy or sell, it calls a method called checkPendingProfit() (checkProfits() in original). If the profit it will achieve at TP is less than the goal (2*increment) it goes into a buy/sell loop. The loop says start at the max increment down to one. After each buy/sell it uses an equation to see if the new purchase will put the profit at TP over the target. If it does, it ends the loop right there, it does NOt continue through all levels.
In the original mgrid the equation which tries to add the additional...
