Help Needed to Debug Existing EA with proven strategy | Page 18 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Help to debug my EA 2 replies

Developer Needed To Debug my EA 2 replies

MT4 EA coder required to fine-tune/debug existing EA 0 replies

Debug code in strategy tester 1 reply

debug message MT4 EA in Strategy Test 7 replies

Attachments: Help Needed to Debug Existing EA with proven strategy
Exit Attachments

Help Needed to Debug Existing EA with proven strategy

  • #341
  • Apr 11, 2018 6:35am Apr 11, 2018 6:35am
Quoting Philip Lee
Hi Tictac-nur, I personally believe that developing the EA in Hybrid Mode is the way to move forward simply because it is a VERY SAFE way to trade with hardly (if any at all) any risk. Tictac-nur, I had done extensive testing in Hybrid Mode in the past 48 hours and after in depth analysis, I truly believe I have found a theoretical solution to reduce drastically the incidence of deadlock situation. By analysing the data from my extensive back tests, it is clearly workable at least 80% of the time - so, should make the Hybrid Mode a lot more profitable....

So basically you are suggesting the further sub level trades Lotsize should have input parameter factor. In your example the first Lotsize is 1.0...and the next trade lotsize will be 2.0...then 3.0, 4.0, 5.0...and so on?
  • #342
  • Apr 11, 2018 6:37am Apr 11, 2018 6:37am
Quoting Philip Lee
Hi Tictac-nur, I personally believe that developing the EA in Hybrid Mode is the way to move forward simply because it is a VERY SAFE way to trade with hardly (if any at all) any risk. Tictac-nur, I had done extensive testing in Hybrid Mode in the past 48 hours and after in depth analysis, I truly believe I have found a theoretical solution to reduce drastically the incidence of deadlock situation. By analysing the data from my extensive back tests, it is clearly workable at least 80% of the time - so, should make the Hybrid Mode a lot more profitable....

Hi Phillip

Can you please explain more about the deadlock situation you are referring to in this post?
  • #343
  • Apr 11, 2018 6:55am Apr 11, 2018 6:55am
Yal,

Let me test out the theory first as and when Tictac-nur codes that simple EA for me. If it works, I will share it here. If it doesn't work, then there is no point talking about it now.

Yal, you obviously have not tested the EA in hybrid mode - if you have done that you'd have known what do I mean by deadlock situation. When trading in hybrid mode, sometimes we will get caught in a deadlock situation (especially when price zigzag endlessly without a breakout) whereby numerous orders would be opened and from my back test, the worst case lasted slightly over one full month with 130 opened orders before the breakout occurred and when this happens it eats up valuable time when the EA should be making lots of profits. But the risk is still minimal even in deadlock situation because all the opened orders basically hedge each other out - it will never cause margin call unless you over traded - and the good news is, the EA traded itself out of all deadlock situation every single time without fail so long as enough room is given to it ($1,000 per tarded pair in 0.01 lot size). Its just a question of having valuable time wasted getting caught in a deadlock situation like that, it does not carry much risks with it. Trading in hybrid mode is actually very profitable but a lot of valuable time is taken up getting caught in deadlock situations.

So, Tictac-nur, I am anxiously waiting for that tool - if you can code it for me ASAP, I will try out the theory I have in my mind - which does not increase the risk in any manner whatsoever and if it works, will share it here.
  • #344
  • Apr 11, 2018 7:03am Apr 11, 2018 7:03am
Quoting Philip Lee
Yal, Let me test out the theory first as and when Tictac-nur codes that simple EA for me. If it works, I will share it here. If it doesn't work, then there is no point talking about it now. Yal, you obviously have not tested the EA in hybrid mode - if you have done that you'd have known what do I mean by deadlock situation. When trading in hybrid mode, sometimes we will get caught in a deadlock situation (especially when price zigzag endlessly without a breakout) whereby numerous orders would be opened and from my back test, the worst case lasted...
Thanks for the explanation! Really appreciate it.
  • #345
  • Apr 11, 2018 7:25am Apr 11, 2018 7:25am
Sorry guys,

Only just realised that what I have described above for the simple EA could not do what I wanted to try. I had missed one important step. I had just PM Tictac-nur to tell her NOT to code that EA as it would serve no purposes as it could not do what I have in mind.

Am going outstation for work tomorrow and will only be back late Friday night. Will deal with this when I am back - so I won't be able to chat till late Friday night or saturday morning.

In the mean time I hope you guys can also come out with some ideas as to how to improve on the EA trading in hybrid mode.

Cheers guys.
  • #346
  • Apr 11, 2018 7:26am Apr 11, 2018 7:26am
Quoting Philip Lee
Sorry guys, Only just realised that what I have described above for the simple EA could not do what I wanted to try. I had missed one important step. I had just PM Tictac-nur to tell her NOT to code that EA as it would serve no purposes as it could not do what I have in mind. Am going outstation for work tomorrow and will only be back late Friday night. Will deal with this when I am back - so I won't be able to chat till late Friday night or saturday morning. In the mean time I hope you guys can also come out with some ideas as to how to improve...
I'll test out hybrid more.
  • #347
  • Apr 11, 2018 7:52am Apr 11, 2018 7:52am
Hi Philip. First of all I'd like to thank you for your system.
Would it be possible you or Tictac-nur share mq4 files even if EA doesn't work correctly now? Maybe someone could find a way to improve it.

It also this will let us better understand meaning of an input data. For example what does it mean risk set to 1. Is it percentage risk of account ballace? How does it work?
  • #348
  • Apr 11, 2018 9:34am Apr 11, 2018 9:34am
Hi Rycho,

I do not have the source codes for the later versions. I had already posted the source codes of the ones that I have but no coders seems to be interested to debug it for me. But I don't really mind so long as at the end of the day I can get a fully working copy of another EA trading on the same logic and strategy as my original EA.

Honestly, I feel pity for Tictac-nur as all the burden is on her to code the EA. I am sure that there are tens of capable coders who have and are reading this thread now but it seems that they are only interested to steal the idea here and not contributing at all - I am sure there are already many coders out there already coded their own private version of the EA based on my idea here and are just lurking here to learn the solutions without contributing anything constructive in return (I can see many 'Invisible' members here all the time). This is just how things are with some people but I don't really mind and who am I to judge anyone.

As to the various versions posted & being tested here, only Tictac-nur has the source codes as it was coded by her. She had promised me that at the end of the day, she will post here a fully unrestricted functional EA. Whether or not she will also post the source codes here I don't think we have the right to demand that as she has already given so much and so generously thus far. I am sure you can understand why coders are reluctant to share their source codes as it is their 'secret sauce/ingredient' which I fully understand. By right I should own the copyright as it was my idea but I am not particular about it, so long as credit is given where it is due and that at the end of the day I can get a fully functional unrestricted EA trading exactly the way I wanted it to trade, I'd be very happy.

If this strategy and EA really turns out to be a gold mine, I think its only proper and the decent thing to do for everyone here to make some donations to Tictac-nur as I think she really deserves it.

Sorry guys for the lecturing.

Cheers!!!
  • #349
  • Apr 11, 2018 9:36am Apr 11, 2018 9:36am
Rycho, the EAs are working properly - there is nothing wrong with them (as far as I can tell but I still have not tried out all the features yet). We are trying to find ways to improve on trading in hybrid mode. Maybe you can contribute some ideas???
  • #350
  • Apr 11, 2018 9:38am Apr 11, 2018 9:38am
Today i use it on my vps (demo account), i will keep report after 1 week
Attached Image (click to enlarge)
Click to Enlarge Name: 123.jpg Size: 116 KB
  • #351
  • Apr 11, 2018 10:36am Apr 11, 2018 10:36am
Quoting jamco388
Today i use it on my vps (demo account), i will keep report after 1 week {image}
HI Jamco what settings are you useing ? i copied your settings on post 282 over the same time period and i blew my accoount after 2 weeks, which i found weird, could you post the set fille like i have, so i can investigate why this is happening on my back test further, cheers M
  • #352
  • Apr 11, 2018 10:56am Apr 11, 2018 10:56am
Quoting detector
{quote} Thanks V, yeah M5, I done the same test again and adjusted the time filter to align with London open and NY close, 10-23 and the DD halved (no as big a profit), I thing you all have done excellent work here BTW, I'm of the opinion that an EA that can turn a cent account into a $ account with very little risk (DD) is of extreme value, there are some "paid" EA's that need a starting balance of 000's in order to perform, you guys may well have found world peace and gave it away, good on you all, wish you all well, D EDIT Had to do it!!!! £50...
Nice detector , cant belive it survived on £50, is this the same set file you posted above ? cheers
  • #353
  • Apr 11, 2018 11:06am Apr 11, 2018 11:06am
Quoting montego
{quote} HI Jamco what settings are you useing ? i copied your settings on post 282 over the same time period and i blew my accoount after 2 weeks, which i found weird, could you post the set fille like i have, so i can investigate why this is happening on my back test further, cheers M
I use the same setting on my post 282. I will try for 2 week with start balance $500
  • #354
  • Apr 11, 2018 11:23am Apr 11, 2018 11:23am
Philip

thank you for very exhausttively explanations.
I was asking about source code really to learn how the EA works and avoid to ask many questions about input data.
I'm not a professioinal coder but have a basic knowledge about it so it lets me to unerstand how EA works.
I have been trying to check another data to find more acurately result but I have done it blindly wthout success.

Some question I have ( examples)
1. What is difference between seting BUY,SELL to true and Hybrid_Buy_Sell to true. What will happen if I set all them to true?
2. What meaning have Compounding ?
3. What does it mean when seting Risk =1.0 ?
...
If I have a promising results or fine sets I will share them on this thread.

Good Luck and many thanks for this what you are doing.
  • #355
  • Apr 11, 2018 12:38pm Apr 11, 2018 12:38pm
Hi Phillip

I have been thinking about the deadlock situation. I have the following 2 suggestions to handle it:

1) When in a deadlock situation we would have several open trades and so the EA will not take new cycle of trades. I say we start a brand new cycle of trades each single day irrespective of if the target has met on the last cycle or not. This will allow the EA to keep generating the income without time being wasted like you mentioned. The deadlock situation can take what it needs to take and close on its own when it closes. We do not worry about it since it is not taking a hit on the Equity.

2) The EA can have another smart monitoring of the cycle...meaning if it has certain number of Open Orders (we can setup this as a parameter) and the P/L on this at a certain number (again a parameter) then close all these trades. We might take a very small loss but this is avoid days of running the trade and also avoid unnecessary margin use.

Let me know your thoughts?
  • #356
  • Apr 11, 2018 12:43pm Apr 11, 2018 12:43pm
Quoting montego
{quote} Nice detector , cant belive it survived on £50, is this the same set file you posted above ? cheers
Yeah Montego £50 and it done a sterling job, the set file had an alteration in that I aligned with London open and NY close, also changed the days of trading, it's all posted to copy, good luck, D
EDIT
Waiting for world turbulence to calm down then looking to run live test, D
I don't need to be good at the game, only good enough to beat my opponent,
  • #357
  • Apr 11, 2018 12:54pm Apr 11, 2018 12:54pm
Detector, Montego and Verocity

You guys have done an amazing job back testing this. Thank you very much for the effort. I wish I could also contribute in the same way like you guys in back testing but I very limited knowledge of back testing on MT4 since I never took interest in it (always believed its waste of time) Now I think there is sure some value to it :-)

I will attempt to do it myself today and post the results here. Anything I need to be aware of while I do it? How much time does it take for the test to complete once you start the test? I am thinking if it does not take much of time....we should all come up with a list of all that we would like to test and do a divide and conquer. This way we have tested all possibilities. We can agree upon every possibility we want to test and divide the the work within ourselves.

Let me know your thoughts?
  • #358
  • Apr 11, 2018 1:04pm Apr 11, 2018 1:04pm
Quoting yalgaar
Detector, Montego and Verocity You guys have done an amazing job back testing this. Thank you very much for the effort. I wish I could also contribute in the same way like you guys in back testing but I very limited knowledge of back testing on MT4 since I never took interest in it (always believed its waste of time) Now I think there is sure some value to it :-) I will attempt to do it myself today and post the results here. Anything I need to be aware of while I do it? How much time does it take for the test to complete once you start the test?...
Thanks for the mention, Phillip Lee, top man, all the programmers, respect to you all, they all made it possible.
RE back testing, every tick, correct spread and not too much history to start with, also a good idea to run a visual to see how things actually perform on the chart, other than that trial and error, I do understand mt data is not that reliable so use with caution if live testing, good luck, D
I don't need to be good at the game, only good enough to beat my opponent,
  • #359
  • Apr 11, 2018 2:01pm Apr 11, 2018 2:01pm
Backtesting is by no means perfect AT ALL, we must take it lightly. It will however give us a good idea if we have things functioning properly before forward testing.
  • #360
  • Apr 11, 2018 3:11pm Apr 11, 2018 3:11pm
Quoting yalgaar
Detector, Montego and Verocity You guys have done an amazing job back testing this. Thank you very much for the effort. I wish I could also contribute in the same way like you guys in back testing but I very limited knowledge of back testing on MT4 since I never took interest in it (always believed its waste of time) Now I think there is sure some value to it :-) I will attempt to do it myself today and post the results here. Anything I need to be aware of while I do it? How much time does it take for the test to complete once you start the test?...
Hi Yal as Detector says, also how i learned i watched videos there are some great videos on youtube, just search for how to do a back test
0 traders viewing now