RSI on an array

• 03/10/2021 at 1:29 PM #163713
  Mattias

  Is it possible to calculate an RSI with values from an array? Something like:

  myRSI = RSI[14]($myArray)

  1	myRSI = RSI[14]($myArray)

   

  

  Mattias

  Participant

  Topics: 9

  Replies: 76

  Been thanked: 19 times
  03/10/2021 at 2:06 PM #163718
  Vonasi

  So you want to store values in an array and calculate the RSI of the latest 14 of them?

  You would need to use a loop to look at lastset down to lastset-13 values and calculate the RSI using these values. You could not use the RSI instruction as that looks at the last 14 bars values.

  Using arrays is only necessary if you are not storing a new value at every bar.

   

   

  1 user thanked author for this post.

  Mattias

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  03/10/2021 at 2:08 PM #163719
  robertogozzi

  No, because an array is made of several elements, which one should be used?

  You can use only, say, element 12 indirectly:

  Element = $myArray[12] myRSI = RSI[14](Element)

  1 2	Element = $myArray[12] myRSI   = RSI[14](Element)

   

  1 user thanked author for this post.

  Mattias

  

  robertogozzi

  Moderator

  Topics: 69

  Replies: 17260

  Been thanked: 4118 times
  03/10/2021 at 5:00 PM #163738
  Vonasi

  Here is an example of how we can store values in an array and then do calculations on the last ‘p’ array locations. This example calculates the RSI of the last 14 high fractals. Unfortunately it is just a simple average based RSI and not smoothed as I believe the normal RSI is (I couldn’t figure out the how to do that calculation in the time I had!)

  p = 14 if high < high[1] and high[1] > high[2] then a = a + 1 $myarray[a] = high[1] endif up = 0 down = 0 if a >=14 then for b = lastset($myarray) downto lastset($myarray)-13 if $myarray[b] > $myarray[b-1] then up = up + ($myarray[b] - $myarray[b-1]) endif if $myarray[b] < $myarray[b-1] then down = down + ($myarray[b-1] - $myarray[b]) endif next upavg = up/p downavg = down/p rs = upavg/downavg myrsi = 100-(100/(1+rs)) endif return myrsi

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

  p = 14   if high < high[1] and high[1] > high[2] then a = a + 1 $myarray[a] = high[1] endif   up = 0 down = 0 if a >=14 then for b = lastset($myarray) downto lastset($myarray)-13 if $myarray[b] > $myarray[b-1] then up = up + ($myarray[b] - $myarray[b-1]) endif   if $myarray[b] < $myarray[b-1] then down = down + ($myarray[b-1] - $myarray[b]) endif next   upavg = up/p downavg = down/p rs = upavg/downavg myrsi = 100-(100/(1+rs)) endif   return myrsi

   

  1 user thanked author for this post.

  Mattias

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  03/10/2021 at 11:21 PM #163756
  Mattias

  Thanks a lot, both of you!

  Vonasi, it worked. I had the array code covered (I did an average from it, I believe I learned from your code somewhere from the forum as well), but it saved a lot of time with the RSI-calculation.

  So thanks again.

  

  Mattias

  Participant

  Topics: 9

  Replies: 76

  Been thanked: 19 times
  03/10/2021 at 11:48 PM #163758
  Vonasi

  Thanks for the thanks – but it is not an RSI calculation that returns the same result as the built in platform RSI. I tested it using every close value (which would be the same as the built in platform RSI uses) and the result was different because I believe the normal RSI either uses smoothing by doing the following:

  The second, and subsequent, calculations are based on the prior averages and the current gain loss:

  • Average Gain = [(previous Average Gain) x 13 + current Gain] / 14.
  • Average Loss = [(previous Average Loss) x 13 + current Loss] / 14.

   

  Borrowed from:

  https://school.stockcharts.com/doku.php?id=technical_indicators:relative_strength_index_rsi

   

  I’ve got rather a lot on at the moment (like trying to sort out a Greek biometric residency permit due to bloody Brexit) which is taking up a lot of time right now otherwise I would have persisted with the calculations – sorry but hopefully I put you on the right path.

   

  1 user thanked author for this post.

  Mattias

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
• Author
  Posts