maximum angle in lookback period.

• 07/14/2018 at 4:31 AM #75942
  Real Pro

  Hiya,

  I’d like to compare today’s high to each of the highs in a lookback range of 20 bars.  Then divide that number by the bars since that high occured, in order to find the highest angle possible between today’s high and a single high in the lookback period.  Could I have some clues where to start please?

  

  Real Pro

  Participant

  Topics: 26

  Replies: 78

  Been thanked: 2 times
  07/14/2018 at 10:03 AM #75954
  robertogozzi

  Angles/Slopes are impossible to be measured, this has been extensively discussed here and elsewhere. Some useful links:

  https://www.prorealcode.com/topic/ema-45-degres-uptrend/

  https://www.prorealcode.com/topic/indicator-for-inclination-of-moving-average/

  https://www.prorealcode.com/topic/calculating-the-angle-linear-regression-of-a-line/

  Anyway, to retrieve the BarIndex ID of a bar you can only use loops, since there’s no function returning that info, as of this version of PRT:

  FOR i = 0 TO 19 //20 bars lookback IF high[i] > high[i+1] THEN //put here anything you like to check . . ENDIF NEXT

  1 2 3 4 5 6

  FOR i = 0 TO 19                  //20 bars lookback    IF high[i] > high[i+1] THEN   //put here anything you like to check       .       .    ENDIF NEXT

   

   

   

  1 user thanked author for this post.

  Real Pro

  

  robertogozzi

  Moderator

  Topics: 69

  Replies: 16475

  Been thanked: 3643 times
  07/14/2018 at 10:16 AM #75956
  GraHal

  Real Pro wrote:

  in order to find the highest angle possible between today’s high and a single high in the lookback period

  I guess if you had said … to get a measure of the rate of change between today’s high and a single high in the lookback period ... may be less confusing and might get you the code you need? 🙂

  1 user thanked author for this post.

  Real Pro

  

  GraHal

  Participant

  Topics: 204

  Replies: 8418

  Been thanked: 2043 times
  07/19/2018 at 3:52 AM #76274
  Real Pro

  Thanks for the help Robert and Grahal.

  Grahal, that’s an excellent idea.  Am I geting closer with the following?

  angle estimate

  FOR i = 0 TO 20 //20 bars lookback IF high[i] > high[i+1] THEN //put here anything you like to check a = high[i+1] /barindex aa = highest[20](a) ENDIF NEXT

  1 2 3 4 5 6

  FOR i = 0 TO 20                  //20 bars lookback IF high[i] > high[i+1] THEN   //put here anything you like to check a = high[i+1] /barindex aa = highest[20](a) ENDIF NEXT

   

   

   

   

  

  Real Pro

  Participant

  Topics: 26

  Replies: 78

  Been thanked: 2 times
  07/19/2018 at 4:14 AM #76275
  Real Pro

  ok that attempt was no good.

  I’m trying to create a substitute for a downsloping trendline which can be auto-traded by finding (in psuedocode):

  highest angle in lookback = (high[lookback] – today’s high ) / barindex

  then for each bar after today, subtracting that angle to create a trigger price, until the close >trigger.

  Suggestions very welcome!

   

  

  Real Pro

  Participant

  Topics: 26

  Replies: 78

  Been thanked: 2 times
  07/19/2018 at 5:03 AM #76276
  robertogozzi

  Barindex can be 10000 or more, maybe highest angle in lookback = (high[lookback] – today’s high ) / i would be better, but I did not test it.

  1 user thanked author for this post.

  Real Pro

  

  robertogozzi

  Moderator

  Topics: 69

  Replies: 16475

  Been thanked: 3643 times
• Author
  Posts