collapse

Author Topic: EA Controller  (Read 162434 times)

Offline FLechdrop

  • DonbonsPrivateGroup
  • Sr. Member
  • *
  • Posts: 604
Re: EA Controller
« Reply #330 on: November 23, 2018, 02:09:34 PM »
Perfect. I want the Controller in any case, and probably the Magic Wand, but am still doubting about the rest.

Online diyforexskills

  • Verified Vendor
  • Sr. Member
  • ***
  • Posts: 1236
Re: EA Controller
« Reply #331 on: November 25, 2018, 03:09:00 AM »
Perfect. I want the Controller in any case, and probably the Magic Wand, but am still doubting about the rest.

Please note. The Controller can only communicate with FXAE EAs and the FX Blue Trade Copier; not with "third party" EAs which do not have the signal code in their source file. So unless you are prepared to use the FX Blue Trade Copier to do controlled trading, you would also need the DIY Trade Manager to do controlled trading.
Of course you can also use the Controller manually. So it just monitors the trading of your EA on the demo account, and then you manually switch the EA on and off on your live account as required. If the EA only trades a few days a week that should not require too much baby-sitting.
Information and product page - http://www.diyforexskills.com/forex-tools/




Online diyforexskills

  • Verified Vendor
  • Sr. Member
  • ***
  • Posts: 1236
Re: EA Controller
« Reply #332 on: November 26, 2018, 07:13:09 AM »
Analysis Tool

Whilst googling I came across this nifty app http://qs.lubotorok.com/

Basic version is free and Premium version costs around 5 euros so won't break the bank. And Lubo provides good support. The app has some really nice features including the ability to analyse both MT4 accounts and Strategy Tester BT accounts, and then comparing them in a single report. Could be useful in the proposed thread suggested by fulltimetrader in the Pants Down thread https://donnaforex.com/index.php?topic=21095.msg382631#msg382631  (Could provide some useful actionable data instead of just seeing whether an EA has done well or not).

I am raising it here in part because the developer is considering adding an MA feature to the performance graph. If he did that and then also enabled doing a simulation like the one that is available already for adjusting TP and SL, that would be really great for helping to determine from a BT as to what the best settings might be for running that EA with the Controller.

And/or running the simulation on actual trade history to see if we could improve on the Controller MA settings we have used for that EA.

I am excited by the prospect. Just need to have patience.
Information and product page - http://www.diyforexskills.com/forex-tools/




Offline FLechdrop

  • DonbonsPrivateGroup
  • Sr. Member
  • *
  • Posts: 604
Re: EA Controller
« Reply #333 on: November 27, 2018, 10:57:44 AM »
Please note. The Controller can only communicate with FXAE EAs and the FX Blue Trade Copier; not with "third party" EAs which do not have the signal code in their source file. So unless you are prepared to use the FX Blue Trade Copier to do controlled trading, you would also need the DIY Trade Manager to do controlled trading.
Of course you can also use the Controller manually. So it just monitors the trading of your EA on the demo account, and then you manually switch the EA on and off on your live account as required. If the EA only trades a few days a week that should not require too much baby-sitting.

Got it. I already have the DIY Trade Manager.

Online diyforexskills

  • Verified Vendor
  • Sr. Member
  • ***
  • Posts: 1236
Re: EA Controller
« Reply #334 on: December 19, 2018, 11:53:34 AM »
This seems to work

I had a promising strategy for GU which after a good start began to fail badly. So after some 150 trades I decided to use the Controller PF to scale-in a reverse entry trades at twice the lot size whenever the profit factor (PF) measured over 20 trades fell below 0.9 (ie losses exceed wins). Shaded in orange on the middle graph.

The top graph shows how closed P/L has declined and is below its SMA20. The second graph shows the trades when PF was below 0.9, thus triggering the reverse scale-ins made with the DIY TM Plus; and the bottom graph shows the closed P/L from those reverse scale-ins.

The table at the top shows that the strategy itself (magic 123456) has lost $322 while the scale-ins (magic 9123456) have recouped $403.

So despite having a losing EA, I am still marginally in profit by some $80 by using the Controller PF in conjunction with the TM.
Information and product page - http://www.diyforexskills.com/forex-tools/




Online diyforexskills

  • Verified Vendor
  • Sr. Member
  • ***
  • Posts: 1236
Re: EA Controller
« Reply #335 on: December 21, 2018, 05:13:13 AM »
This seems to work

I had a promising strategy for GU which after a good start began to fail badly. So after some 150 trades I decided to use the Controller PF to scale-in a reverse entry trades at twice the lot size whenever the profit factor (PF) measured over 20 trades fell below 0.9 (ie losses exceed wins). Shaded in orange on the middle graph.

The top graph shows how closed P/L has declined and is below its SMA20. The second graph shows the trades when PF was below 0.9, thus triggering the reverse scale-ins made with the DIY TM Plus; and the bottom graph shows the closed P/L from those reverse scale-ins.

The table at the top shows that the strategy itself (magic 123456) has lost $322 while the scale-ins (magic 9123456) have recouped $403.

So despite having a losing EA, I am still marginally in profit by some $80 by using the Controller PF in conjunction with the TM.

The image below probably shows it better - the performance of the EA (pretty bad) and that of the reverse scale-ins that have been added whenever the profit factor of the EA (measured over 20 trades) fell below 0.9. And so I have stayed in profit even though my EA was in DD.

The PF is now above 0.9 and so there have been no reverse scale-ins for the last 10 trades or so by the EA - mostly wins.
Information and product page - http://www.diyforexskills.com/forex-tools/




Offline namrfigk

  • DonbonsPrivateGroup
  • Sr. Member
  • *
  • Posts: 705
Re: EA Controller
« Reply #336 on: January 27, 2019, 12:05:33 AM »
Hi DIY, if I have multiple EAs on my mt4, can your controller filter profitable trades for all of them? What is the maximum number of EAs that I can filter?

Sent from my Redmi 4A using Tapatalk


Online diyforexskills

  • Verified Vendor
  • Sr. Member
  • ***
  • Posts: 1236
Re: EA Controller
« Reply #337 on: January 27, 2019, 01:30:02 AM »
Hi DIY, if I have multiple EAs on my mt4, can your controller filter profitable trades for all of them? What is the maximum number of EAs that I can filter?

Sent from my Redmi 4A using Tapatalk

You would place a copy of the Controller on each of the EA charts so that each EA would be controlled individually using the unique magic number for each EA. If an EA has multiple magics, each Controller can monitor up to 10 magics collectively.
Information and product page - http://www.diyforexskills.com/forex-tools/




Offline namrfigk

  • DonbonsPrivateGroup
  • Sr. Member
  • *
  • Posts: 705
Re: EA Controller
« Reply #338 on: January 27, 2019, 03:39:28 AM »
Is there a possible to increase collective management to 20 to 30 pairs?

Sent from my Redmi 4A using Tapatalk


Online diyforexskills

  • Verified Vendor
  • Sr. Member
  • ***
  • Posts: 1236
Re: EA Controller
« Reply #339 on: January 27, 2019, 03:45:44 AM »
Yes but that would require a small coding change to add the extra magic fields.
Information and product page - http://www.diyforexskills.com/forex-tools/




Offline namrfigk

  • DonbonsPrivateGroup
  • Sr. Member
  • *
  • Posts: 705
Re: EA Controller
« Reply #340 on: January 27, 2019, 03:50:14 AM »
Yes if it can manage 30 magic fields per EA then I will buy it thank you

Sent from my Redmi 4A using Tapatalk


Online diyforexskills

  • Verified Vendor
  • Sr. Member
  • ***
  • Posts: 1236
Re: EA Controller
« Reply #341 on: January 27, 2019, 05:02:37 AM »
Just need to be sure about how you intend to use this.

The Controller does the monitoring (closed P/L vs its MA or by profit factor) and then sends a signal, yes or no, depending whether the EA is trading profitably or not.

But the signal needs a receiver - to date only the FX Autotrader Elite, the DIY Trade Manager Plus and the FX Blue Trade Copier can receive and act on those signals.

So in your case you would have to be running those EAs on a demo platform with the Controllers on each of the EA charts. Let's assume you have three EAs each with 20 or 30 magics. So you would need to have three Trade Copier Senders on that demo, one for each EA, and three Trade Copier Receivers on your live account. With all the appropriate settings so as to distinguish between the EAs.

So when EA1 is trading profitably on demo, those trades will be copied to your live account. At that same time both EA2 and EA3 may be in DD, and so those trades will not be copied to your live account. etc

You would need some good grunt on your VPS to handle the three trade copiers.

Is that something you would want to do? If so I will check with my coder about adding the extra magic fields.


Information and product page - http://www.diyforexskills.com/forex-tools/




Offline namrfigk

  • DonbonsPrivateGroup
  • Sr. Member
  • *
  • Posts: 705
Re: EA Controller
« Reply #342 on: January 27, 2019, 08:00:11 AM »
Yes three receivers, three senders and three controllers with 30 magics each. Does this mean I need 3 controller licences per live account?
In your experience what is the difference between controlling via PL, MA or PF?

Sent from my Redmi 4A using Tapatalk


Offline namrfigk

  • DonbonsPrivateGroup
  • Sr. Member
  • *
  • Posts: 705
Re: EA Controller
« Reply #343 on: January 27, 2019, 08:27:44 AM »
Also how many charts per live and per demo mt4 should I open

Sent from my Redmi 4A using Tapatalk


Online diyforexskills

  • Verified Vendor
  • Sr. Member
  • ***
  • Posts: 1236
Re: EA Controller
« Reply #344 on: January 27, 2019, 08:49:00 AM »
The EA Controller comes in two versions, MA and PF, for $79 for the two, and can be used on however many accounts you want to run.

Not sure what you mean by PL. The MA version measures the closed P/L and then adds a MA (user-defined) to that performance curve. When closed P/L is above its MA, the signal is YES; below, NO. PF is calculated over a user-defined number of trades. And you then set a level above which you want a YES; below a NO.  Obviously a PF of 1 is break even, so needs to be > 1 depending on how deep a DD you are prepared to accept.  I have not had enough experience with PF to advise one or the other so currently have loaded both, using one or the other and just monitoring to see which one will work better over time. I can always switch from one to the other.

The good thing is that if you have a good lot of trading history, you can add both Controllers to the chart and see which one would have given the better YES/NO signals for each of the EAs. The important thing is to try and avoid whipsawing the EA on and off.

On your live account all you need is three charts, one for each of the Copier Receivers.

On the demo you need to have as many charts as you need for just running the EAs. The Controller is an Indicator so can be added to a chart that is running an EA.
Information and product page - http://www.diyforexskills.com/forex-tools/




 

browse forum