Robot Lab Project | Page 38 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Financial Markets Observatory Lab 1,095 replies

Indicators Lab 236 replies

Rel's Lab 103 replies

Nikola's Trading Lab 247 replies

Spud's Lab Accident 22 replies

Attachments: Robot Lab Project
Exit Attachments

Robot Lab Project

  • #741
  • Mar 1, 2015 4:43am Mar 1, 2015 4:43am
Quoting zbyma
hello Mr Abokwaik, thank you for the advice, but I enclose a graph. This will not be around too.Thanks Strategy Tester Report Ladder_System_01 AM-Live (Build 765) SymbolEURUSD (Euro vs US Dollar) Obdob5 minut (M5) 2014.09.04 20:35 - 2014.12.19 22:55 (2014.04.01 - 2014.12.20) ModelVechny cenov pohyby (nejpřesněj metoda zaloena na vech dostupnch nich časovch intervalech) ParametryMagic_Number=40005;Ladder_Step_Pips=700, Unified_Stop_Pips=1400; Fixed_Lot_Size=0.1; Lot_as_EQ_Perc=0; Max_Open_Orders=15;...
The used Period is Different. Although you have chosen the same period in my test, your MT4 doesn't have enough history to go back to 2014.04.01, so it started from 2014.09.04 instead, missing around 5 months. You have two options:

  1. Get more history data on M5 by opening the pair on M5 time frame, zoom out to the maximum, scroll back as much as MT4 allows you to, and Refresh the chart by pressing Mouse Right Click and selecting Refresh from the pop up menu. If you couldn't reach 2014.04.01, try the second option
  2. Choose a higher time frame (H1 for example), although by changing the time frame results are not guaranteed to be similar.

Regards,

No guts, no glory
  • #742
  • Mar 1, 2015 5:07am Mar 1, 2015 5:07am
Abokwaik thank you, I'll try and write as managed strategy to deal with. Regards zbyma
  • #743
  • Mar 1, 2015 9:36am Mar 1, 2015 9:36am
Hello Abokwaik
and other traders with robots. Now I wonder which way to turn. Historical data Admiral Markets gave me only two months back, XTB historical data provided by only 5 months.

1 / Getting a database of historical data. There certainly has more people and Abokwaik gives us a demonstration tests. We work only with older data. We have an opportunity in the strategy tester using a 3 "models" setting, and each has a different result. Which will be the right one ??

2 / Test the demo version and try different settings. This is another way to get the correct settings, TF, currency pairs. But still result much different from the live trading.

3 / Real - LIVE trading microlot, and little capital at various Brooker.


So I think about me and based on only the option number 3. And as you're thinking ??
  • #744
  • Mar 1, 2015 9:53am Mar 1, 2015 9:53am
Quoting zbyma
1 / Getting a database of historical data.
http://www.tickstory.com/

Free tick data ... you can get 99% modeling quality:

Inserted Video


Regards,
EZ
  • #745
  • Mar 1, 2015 10:39am Mar 1, 2015 10:39am
The Robots Lab purpose was to develop/back test/fine tune different robots, and then forward test successful robots, and finally trading them live.

The problem is that I do not have enough time to do full testing of those robots, and find the best settings.

I tried to provide as many inputs as possible for you guys to back test those robots, submit best settings, suggest improvements WHILE I invest my free time in developing new robots.

Since Tuesday Last week I didn't have the chance to open the Basket Manager, which I consider a High priority.
My work is taking most of my time, and the time left I have to spend with family.

So I encourage every one who is interested in the posted robots to test them out and share their findings in the thread.
No guts, no glory
  • #746
  • Mar 1, 2015 10:42am Mar 1, 2015 10:42am
I am also sad about what happened to HA_Rider robot after it left the Lab.
It was unfairly moved to Commercial Content.
And I feel obliged to open a new thread for that Robot to bring it back to life. Still I couldn't spare time for that.
No guts, no glory
  • #747
  • Mar 1, 2015 10:56am Mar 1, 2015 10:56am
Ezcurrency Hello, thank you for the link to the data from the history. I certainly delivers and pull data into MT4.Thanks
  • #748
  • Mar 1, 2015 11:02am Mar 1, 2015 11:02am
Hello Abokwaik , thanks for your opinion. I myself am a newbie to forex, but I learn quickly and I am interested in other new indicators, robots.
My horror was that even if I use the strategy tester, still do not know what will be the effect on real account.
Thank you very much for your work and for the rest of us to make your robots tried on a real account. Results should any obligation to disclose is setting the strategy, the use of Brooker.
So I urge others - go to a real store. Just to verify and ensure good setup and good Brooker.
  • #749
  • Mar 2, 2015 12:03am Mar 2, 2015 12:03am
Quoting zbyma
Hello Abokwaik , thanks for your opinion. I myself am a newbie to forex, but I learn quickly and I am interested in other new indicators, robots. My horror was that even if I use the strategy tester, still do not know what will be the effect on real account. Thank you very much for your work and for the rest of us to make your robots tried on a real account. Results should any obligation to disclose is setting the strategy, the use of Brooker. So I urge others - go to a real store. Just to verify and ensure good setup and good Brooker....
Back testing is not a joke as many might think, but also is not a guarantee for future success. Its a tool to :

  1. Validate the implementation of robot idea (Does it work as it is supposed to or not)
  2. Find best input settings. This is however not an easy job, and it does not mean to just select the inputs which generate highest profit or minimum draw down. No, it is much more than that. To put it in a simple form, Input changes should have a logical relationship with Results.

    1. For example, if we take an MA Cross System with 2 inputs for Fast and Slow MA :

      1. 20/100 : trades = 100, profit = 2000, draw down = 20%.
      2. 30/150 : trades = 80, profit = (-1000), draw down 25%.
      3. 40/200 : trades = 50, profit = 4000, draw down 17%
      4. This is not a logical relation between inputs and Outputs, so I can not say that 40/200 is the best choice, because it was preceded with a negative output following a positive output (random distribution of results).

    2. Results of huge variations of input settings should be normally distributed , resulting a graph similar to a Bell Shape
    3. The sample we worked on to select best inputs is called In Sample (IS), which for example cover a certain period(year 2013).
    4. After selecting a set or more of best inputs, they should be applied to a Out of Sample (OoS). For example a period of (year 2010 or 2014) which was not covered in the initial back testing. And see if the results are similar (not exactly the same), but close enough.
    5. Only then we can think of Forward Testing or Live Trading.

The main advantage of Back testing is that you can test different methods, inputs, markets... in a short time, to see the validity of your idea before trying it for several months in forward testing, or risk it in live trading.

No guts, no glory
  • #750
  • Mar 2, 2015 1:12am Mar 2, 2015 1:12am
Hi wise Abokwaik,
thank you for your explanation and your view of the development of robots forex.
Do you have a great truth, more work is needed to back tests with historical data and also try these results to the demo version.
I myself was amazed how large deviations may be back tests

1 / it is a matter of historical data, the number of days
2 / methods used to back test

It is also necessary to test the robot on the demo version.
It is everything to me known, I expect myself to do so.
But it is necessary to take the next step - test on LIVE account microlot.
There are also big differences by brookers company.
That's what I had in mind that traders wrote their real-world experiences, as selected by the robot works.
As I wrote, I've still businessman who is still learning and it is a short time in this area.

Himself trying robot LIVE NBB_XBR_01 and I am satisfied with his work. He made 4 stores, 3 store profit 1 - no profit.
I'm trying to learn the system from the simpler complex
I believe that we have the same idea and this thread will be more realistic examples

Sincerely,
  • #751
  • Mar 2, 2015 3:55am Mar 2, 2015 3:55am
if you want quick and easy backtesting data just download the MT4 from MetaQuotes and add the MetaQuotes Demo server. Go to history centre and download everything on 1min. The data is at least 90% accuracy from the get go.

Test your EAs as much as you can then filter the promising ones.

then download tick data from here: http://www.forextester.com/data/datasources

this should bring it upto somewhere 90%-99%. After that and if the results are still holding, stick the EA on a Demo account going forward live. Leave it for at least 3 months, if not 6.

Then make your mind up if it is worth putting any real money on it.
  • #752
  • Mar 2, 2015 4:21am Mar 2, 2015 4:21am
Thank you Terry for information and I appreciate your contribution to the debate.
My opinion on the data I wrote in ghosts contribution here.
Demo version of the bill differs from the account LIVE. Everyone knows that.
Therefore, I point to the fact use demo account LIVE + s Malyn microlot. Here we verify the feasibility of the ideas and the robot.
That is my position and it is up to each of which selects strategy. Good luck to all the demo accounts

Thanks
  • #753
  • Mar 2, 2015 5:16am Mar 2, 2015 5:16am
Quoting TerryTibbs
if you want quick and easy backtesting data just download the MT4 from MetaQuotes and add the MetaQuotes Demo server. Go to history centre and download everything on 1min. The data is at least 90% accuracy from the get go. Test your EAs as much as you can then filter the promising ones. then download tick data from here: http://www.forextester.com/data/datasources this should bring it upto somewhere 90%-99%. After that and if the results are still holding, stick the EA on a Demo account going forward live. Leave it for at least 3 months,...
No guts, no glory
  • #754
  • Mar 3, 2015 11:38am Mar 3, 2015 11:38am
Quoting abokwaik
@SmoothTrader has reported an error in position calculation for sell orders when Lot_as_EQ_Perc is used in MA_CROSS_WT_03 Robot, as it always uses 0.1 lot size instead of a calculated one. He even suggested the correct fix for the error, I thank him for that. Here is the Corrected Update for MA_CROSS_WT_03 Robot I will also update the original post for this robot. EDIT : I couldn't update the original post, so I will just modify the link of it on the first post. {file} {file}

I love this robot, I use the crossroad of moving averages 1, 20 and 300 in 4H and seems to work well, it would be possible to add some options such as:
To configure the SL with the ATR? example:
Use_SL_ATR: true / false
ATR_SL_period:
ATR_SL_TF: 1h, 4h, 1d ... etc.etc ..
ATR_SL_multiplier:
and to be already perfect, would fail a trailing stop that could also work with the ATR. and put option Break even and break even profit.

thank you very much for the great work done
  • #755
  • Mar 3, 2015 7:04pm Mar 3, 2015 7:04pm
Quoting Motiram
Hello abokwaik, I see you are expert in Robots makings under this post also I saw you recap idea very elegantly under this link http://www.forexfactory.com/showthre...97#post8016897 Which make me impressed and finally giving an Open Source free EA to helping people for learning a coding your nature demonstrate you are a helping person i want to thank you for that Here I have my require to have one EA with the following requirements I will be grateful to you if you can help making the EA and can draw my request summarized as...
Hello abokwaik I know you are very busy and doing hard work to complete your great project I want to congratulate you for that. I do not want to distract you just want to request you if you get some time please can you look my request if it is not annoying For you I will be great full to you.

Thanks And Regards,
Moti
  • #756
  • Mar 3, 2015 11:07pm Mar 3, 2015 11:07pm
Quoting Clipi
{quote} I love this robot, I use the crossroad of moving averages 1, 20 and 300 in 4H and seems to work well, it would be possible to add some options such as: To configure the SL with the ATR? example: Use_SL_ATR: true / false ATR_SL_period: ATR_SL_TF: 1h, 4h, 1d ... etc.etc .. ATR_SL_multiplier: and to be already perfect, would fail a trailing stop that could also work with the ATR. and put option Break even and break even profit. thank you very much for the great work done
Yes, Clipi, Its possible, and looks interesting.


Quoting Motiram
{quote} Hello abokwaik I know you are very busy and doing hard work to complete your great project I want to congratulate you for that. I do not want to distract you just want to request you if you get some time please can you look my request if it is not annoying For you I will be great full to you. Thanks And Regards, Moti
I didn't forget you my friend.

----
I was recently promoted as GM for the company I work for, and my time became very tight.
I will act on these requests once I can spare some time.

Green Pips to you all,
No guts, no glory
  • #757
  • Mar 4, 2015 1:39am Mar 4, 2015 1:39am
Good morning, attached example, I use the line as support ema50 4TF, usually with loss of em20 continuous downtrend, which every red arrow indicates a new loss of em20 and a new entry. I have configured a spacing of 50 pips between signals. The ema300 = ema50 1D, when the price exceeds the ema300 (ema50 1D), the trend ends.
to use this system in all pairs, the SL and TS would have to match the ATR 1D.

Would be very nice if I could add these features in EA ... thank you very much for everything.

Sorry for my bad English.
Attached Image (click to enlarge)
Click to Enlarge Name: EURUSDH4.png Size: 36 KB
  • #758
  • Mar 4, 2015 3:34am Mar 4, 2015 3:34am
Quoting abokwaik
I didn't forget you my friend. ---- I was recently promoted as GM for the company I work for, and my time became very tight. I will act on these requests once I can spare some time. Green Pips to you all,
Congratulations buddy abokwaik, for being a General Manager of the company. I wish you a good luck and further success in you life ahead

Below is a link that can make you feel to party all night. Enjoy it!
Inserted Video


Regards,
Moti
  • #759
  • Mar 4, 2015 3:46am Mar 4, 2015 3:46am
Quoting abokwaik
{quote} Yes, Clipi, Its possible, and looks interesting. {quote} I didn't forget you my friend. ---- I was recently promoted as GM for the company I work for, and my time became very tight. I will act on these requests once I can spare some time. Green Pips to you all,
Funny how good things come to those that work so hard. I am sure it is a well deserved promotion and I for one understand if your replys are not as prompt as they used to be.

Congratulations again and may life bring you nothing but GREEN pips.
  • #760
  • Mar 4, 2015 7:28am Mar 4, 2015 7:28am
Quoting abokwaik
{quote} Yes, Clipi, Its possible, and looks interesting. {quote} I didn't forget you my friend. ---- I was recently promoted as GM for the company I work for, and my time became very tight. I will act on these requests once I can spare some time. Green Pips to you all,
Congratulations Khalil, wish you a good luck in your new position
4 traders viewing now