Discussing the strategy VECTORIAL DAX (M5)

• 02/27/2019 at 1:38 PM #92464
  Nicolas

  Balmora74 wrote:

  I have attached 2 indicator that correspond to the display of the angle (in degree) in  form of histograms and the slope of the moving average (in radius) and its moving average

  Nice! Would you mind if I had them in the library too? Angle of MA is an indicator very often asked by people.

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
  02/27/2019 at 1:46 PM #92467
  Balmora74

  Nicolas wrote:

  Balmora74 wrote:

  I have attached 2 indicator that correspond to the display of the angle (in degree) in form of histograms and the slope of the moving average (in radius) and its moving average

  Nice! Would you mind if I had them in the library too? Angle of MA is an indicator very often asked by people.

  
  @Nicolas
  Yes you can had it into the library 🙂

   

  

  Balmora74

  Participant

  Topics: 5

  Replies: 97

  Been thanked: 78 times
  02/27/2019 at 2:03 PM #92468
  jebus89

  My computer/PRT froze in the middle of the night, so i started it again at work today, but sadly PRT shutdown cus it had been “on” for so long. I didnt manage to save any pics.. Results where a 24% walk forward, so not supergreat. Each period had about 20K combinations so obviously it was a huge test and much that could go wrong. Overall the equitycurve didnt look too shabby, i ran 5 periods, only 1 was negative, rest was positive and OK looking.

  Was really hoping to analyze it more, might run the test again later.

  At the same time i ran an optimization of the same variables from 2016 june -> 2018 July, and kept rest as out of sample. it looked surprisingly good.  But then again, i ran optimization on already optimized variables so.. why wouldnt it look good.

   

  All in all, if it keeps performing good for the next 6 months i might as well run it live. My main issue is the “very little” data (few years) that we got..

  

  jebus89

  Participant

  Topics: 118

  Replies: 586

  Been thanked: 238 times
  02/28/2019 at 10:14 PM #92567
  Gianluca

  i had the same idea weeks ago and i used this formula, i don’t know if could help if is similar or mine is completely wrong 😀

  ema15=average[15,ave](typicalprice) c2 = (sin(atan((ema15-ema15[1])/ema15[1]*100)))*1000

  1 2	ema15=average[15,ave](typicalprice) c2 = (sin(atan((ema15-ema15[1])/ema15[1]*100)))*1000

  1 user thanked author for this post.

  Ichimoku Reading

  

  Gianluca

  Participant

  Topics: 63

  Replies: 416

  Been thanked: 54 times
  02/28/2019 at 10:50 PM #92573
  Balmora74

  Gianluca wrote:

  1 2 ema15=average[15,ave](typicalprice)

  c2 = (sin(atan((ema15-ema15[1])/ema15[1]*100)))*1000

  
  @GianLuca

  What is the variable “ave” ? I think the best thing it to create a indicator with the instruction “RETURN” and after watching the price action.

  You also can test it manually  and / or make a bactkest.

  

  Balmora74

  Participant

  Topics: 5

  Replies: 97

  Been thanked: 78 times
  02/28/2019 at 11:10 PM #92576
  Vonasi

  ‘ave’ is any value from 0 to 6 for the average type.

  See here:

  https://www.prorealcode.com/documentation/average/

   

   

  2 users thanked author for this post.

  kbot2021, Balmora74

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  02/28/2019 at 11:14 PM #92577
  Vonasi

  Here is the indicator as I was already testing it:

  p = 15 //Average period type = 0 //0 to 6 ma=average[p,type](typicalprice) c = (sin(atan((ma-ma[1])/ma[1]*100)))*1000 return 0, c

  1 2 3 4 5 6 7

  p = 15 //Average period type = 0 //0 to 6   ma=average[p,type](typicalprice) c = (sin(atan((ma-ma[1])/ma[1]*100)))*1000   return 0, c

   

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  02/28/2019 at 11:29 PM #92587
  Balmora74

  @Giancula @Vonasi

  This is my method to calcutate the angle of EM10 on 20 periods. The result is in degree and thick headed between – 90° and + 90° (see the picture below).

  Periode = 10 //EMA Period nbChandelier= 20 // Number of Candles MM = Exponentialaverage[Periode](close) ADJASUROPPO = (MM-MM[nbchandelier]*pipsize) / nbChandelier ANGLE = (ATAN(ADJASUROPPO))   //ARC TANGENT trignometric function RETURN angle as "Angle"

  1 2 3 4 5 6 7

  Periode = 10 //EMA Period nbChandelier= 20 // Number of Candles MM = Exponentialaverage[Periode](close) ADJASUROPPO = (MM-MM[nbchandelier]*pipsize) / nbChandelier ANGLE = (ATAN(ADJASUROPPO))   //ARC TANGENT trignometric function   RETURN angle as "Angle"

   

  When i try with your code i don’t have the same result ans the same curve…

  1 user thanked author for this post.

  Harengusrex

  

  Balmora74

  Participant

  Topics: 5

  Replies: 97

  Been thanked: 78 times
  02/28/2019 at 11:39 PM #92591
  Vonasi

  Balmora74 – Please use the ‘Insert PRT Code button when posting code. I tidied it up for you!

  We should all be aware that it is impossible to calculate a slope in degrees as we do not know what the x value is. All we can do is give the slope a relative value that is an indication of the amount of gradient up or down.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  02/28/2019 at 11:50 PM #92592
  Balmora74

  Vonasi wrote:

  Please use the ‘Insert PRT Code button when posting code. I tidied it up for you! We should all be aware that it is impossible to calculate a slope in degrees as we do not know what the x value is. All we can do is give the slope a value that is an indication of the amount of gradient up or down.

  So there is a mistake in my approach ?

  

  Balmora74

  Participant

  Topics: 5

  Replies: 97

  Been thanked: 78 times
  03/01/2019 at 12:12 AM #92596
  Vonasi

  Balmora74 wrote:

  So there is a mistake in my approach ?

  Only if you think that you are calculating degrees. The only thing we know about the value x is that it is the distance from one bar to the next bar but how do we give that a quantity in pips? We can however know the value of y which we can at least measure in pips.

  1 user thanked author for this post.

  Balmora74

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  03/01/2019 at 12:33 AM #92597
  Balmora74

  Ok i understand ! It’s not the same units. Y is in pips and X is the difference between 2 bars…. So If we change the scale of the graphic X will change but not Y… But the strange thing it that the strategy seems to work… !

  

  Balmora74

  Participant

  Topics: 5

  Replies: 97

  Been thanked: 78 times
  03/01/2019 at 9:21 AM #92602
  raphaelopilski

  which variables are you guys optimizing when you choose a new index or forex for this system?

  I would like to have this system for eur/usd and for france. Anyone can check if its possible? Thanks a lot!

  

  raphaelopilski

  Participant

  Topics: 16

  Replies: 73

  Been thanked: 4 times
  03/01/2019 at 11:13 AM #92625
  Balmora74

  raphaelopilski wrote:

  which variables are you guys optimizing when you choose a new index or forex for this system?

  I would like to have this system for eur/usd and for france. Anyone can check if its possible? Thanks a lot!

  All answers and ideas are in the answers above of this file of discussion…

  

  Balmora74

  Participant

  Topics: 5

  Replies: 97

  Been thanked: 78 times
  03/03/2019 at 12:51 PM #92751
  Gianluca

  Vonasi wrote:

  Here is the indicator as I was already testing it:

  p = 15 //Average period<br /> type = 0 //0 to 6</p> <p>ma=average[p,type](typicalprice)<br /> c = (sin(atan((ma-ma[1])/ma[1]*100)))*1000</p> <p>return 0, c

  1 2 3 4 5 6 7

  p = 15 //Average period type = 0 //0 to 6 ma=average[p,type](typicalprice) c = (sin(atan((ma–ma[1])/ma[1]*100)))*1000 return 0, c

  i am using it as a filter in many trategies and for me is working very good.
  you can also use the R2 function and results are also good. this is the code

  once PERIODO = 90 i1 = R2[PERIODO](close)

  1 2	once  PERIODO  =  90 i1 = R2[PERIODO](close)

  2 users thanked author for this post.

  dreads, Balmora74

  

  Gianluca

  Participant

  Topics: 63

  Replies: 416

  Been thanked: 54 times
• Author
  Posts