Using Summation to identify double bottom

• 09/16/2024 at 5:29 PM #237650
  sgrech

  Please see attached chart screenshot

  I am trying to identify where a 10 period ma (broken blue line) does not make a lower low

  I thought I could use summation function.

  something like

  Where 10ma>10ma[1] and 10ma[1]<10ma[2]  (ie 10ma has reversed direction)

  using summation exmaine the last n bars (lets n = 6) ma10<ma10[1] and look for assets where the count=>5 (remember the for latest bar the ma10>ma10[1])

  Then for the prior 10 bars prior to bar 6 I need to see assets where the ma10[]>ma10[]>=7

  Then for the 14 bars prior to that i need return assets where ma10[]<ma10[] >=12

  If all three criteria are met then include in results.

  I hope this making sense.  Basically I wish to examine the 6 bars then the prior 10 bars prior  and then the 14bars prior to that so looking at total 30 bars

   

  Any help would be appreciated.

  Thank you very much

  Simon

  

  sgrech

  Participant

  Topics: 13

  Replies: 24

  Been thanked: 1 time
  09/17/2024 at 6:32 AM #237672
  sgrech

  Also it is possible to check whether the value of 10ma[1] is above the 10ma low of the entire period?

  

  sgrech

  Participant

  Topics: 13

  Replies: 24

  Been thanked: 1 time
  09/17/2024 at 8:29 AM #237680
  Iván

  Hi Simon,
  I believe there’s a more efficient way to approach the problem to reduce computational time while still getting the results you’re looking for.
  Instead of manually iterating over the bars, you can use a combination of the 10-period moving average and a check for direction changes. Additionally, with the summation function, you can quickly verify how many times this change has occurred within a range of N bars. Here’s how the code might look:

  // Calculate the 10-period moving average ma10 = average[10](close) // Check for the direction change change = ma10 > ma10[1] and ma10[1] < ma10[2] // Use summation to check how many times the change occurred in the last 15 bars check = summation[15](change) // Use barssince to find how many bars have passed since the last change bar = barssince(change)

  1 2 3 4 5 6 7 8 9 10 11

  // Calculate the 10-period moving average ma10 = average[10](close)   // Check for the direction change change = ma10 > ma10[1] and ma10[1] < ma10[2]   // Use summation to check how many times the change occurred in the last 15 bars check = summation[15](change)   // Use barssince to find how many bars have passed since the last change bar = barssince(change)

  This approach optimizes the calculations and reduces complexity, while being more efficient in terms of runtime. You can adjust the number of bars to examine or combine it with other criteria as needed.
  I hope this helps.

  

  Iván

  Moderator

  Topics: 0

  Replies: 777

  Been thanked: 262 times
  09/17/2024 at 9:34 AM #237684
  sgrech

  Thanks this does help.

   

  Thinking I will only look for those asset where the check = 2

  // Use summation to check how many times the change occurred in the last 15 bars

  check = summation[15](change)

  Is there a way to somehow check whether for the 2nd (latest) change in direction, that MA10 value is above MA10 value at the time of the 1st change in direction?

  Many thanks

  Simon

  

  sgrech

  Participant

  Topics: 13

  Replies: 24

  Been thanked: 1 time
  09/17/2024 at 11:42 AM #237702
  Iván

  Here you have an example:

  // Calculate the 10-period moving average ma10 = average[10](close) // Check for the direction change if ma10 >= ma10[1] and ma10[1] < ma10[2] then prevMAvalue=MAvalue change=1 MAvalue=ma10[1] else change=0 endif // Use summation to check how many times the change occurred in the last 15 bars check = summation[15](change) // Use barssince to find how many bars have passed since the last change bar = barssince(change) return ma10 coloured("orange"), MAvalue coloured("darkblue"), prevMAvalue coloured("red")style(dottedline)

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

  // Calculate the 10-period moving average ma10 = average[10](close)   // Check for the direction change if ma10 >= ma10[1] and ma10[1] < ma10[2] then prevMAvalue=MAvalue change=1 MAvalue=ma10[1] else change=0 endif   // Use summation to check how many times the change occurred in the last 15 bars check = summation[15](change)   // Use barssince to find how many bars have passed since the last change bar = barssince(change)   return ma10 coloured("orange"), MAvalue coloured("darkblue"), prevMAvalue coloured("red")style(dottedline)

  

  Iván

  Moderator

  Topics: 0

  Replies: 777

  Been thanked: 262 times
  09/17/2024 at 3:37 PM #237718
  sgrech

  Thank for this, although this appears to be an indicator?  Just in case I did not make myself clear .

  I am seeking to screen for assets where the 2nd change of direction the 10ma value is higher than the 10ma at the point where the first change of direction occurred.

  

  sgrech

  Participant

  Topics: 13

  Replies: 24

  Been thanked: 1 time
• Author
  Posts