Different RSI Charts / any idea to rebuild?

• 02/24/2024 at 8:51 AM #228674
  joe8711

  Dear PRT-code-community,

  First of all, thanks to all, who offer their know-how inside here. Since a few weeks, I’m reading some very interesting topics here.
  But now, I have a “special” question and hopefully someone has an idea.
  MANY THANKS!

  I use PRT also with different indicators.
  I also have access to an RSI indicator, which coding is blocked and not open to see.
  Now, I tried to rebuild this with several settings in PRT.
  The result goes in the right direction, but some peaks are not on the same level.
  RSI1) => the basic RSI from PRT with a period of 14 days.
  RSI2) => the “blocked” RSI, where I not know some details
  RSI3) => RSI code from here, which looks very similar to RSI1) (please see code bellow).
  ***I still tried different periods, but the result is not really rebuildable.***

  Thanks in advance for your ideas and help!

  Best,
  Joe8711

  RSI from PRT-code

  // RSI indicator // // Formula: // // (https://www.prorealcode.com/topic/indicatore-rsi/#post-51250) // (https://www.investopedia.com/terms/r/rsi.asp) // p = 14 //periods Ob = 70 //OverBought Os = 100 - Ob //OverSold // Bullish = MAX(0, CLOSE - CLOSE[1]) Bearish = MAX(0, CLOSE[1] - CLOSE) // mmBullish = WILDERAVERAGE[p](Bullish) mmBearish = WILDERAVERAGE[p](Bearish) // RS = mmBullish / mmBearish myRSI = 100 - (100 / (1 + RS)) // RETURN MyRSI AS "Rsi",Os AS "Os",Ob AS "Ob"

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

  // RSI indicator // // Formula: // //    (https://www.prorealcode.com/topic/indicatore-rsi/#post-51250) //    (https://www.investopedia.com/terms/r/rsi.asp) // p         = 14                          //periods Ob        = 70                          //OverBought Os        = 100 - Ob                    //OverSold // Bullish   = MAX(0, CLOSE - CLOSE[1]) Bearish   = MAX(0, CLOSE[1] - CLOSE) // mmBullish = WILDERAVERAGE[p](Bullish) mmBearish = WILDERAVERAGE[p](Bearish) // RS        = mmBullish / mmBearish myRSI     = 100 - (100 / (1 + RS)) // RETURN MyRSI AS "Rsi",Os AS "Os",Ob AS "Ob"

   

  

  joe8711

  Participant

  Topics: 1

  Replies: 1

  Been thanked: 1 time
  02/24/2024 at 10:55 AM #228677
  JS

  Hi Joe,

  In your screenshot, RSI1 and RSI3 are the same…

  The “blocked” RSI is more responsive compared to the other RSI and for this purpose a “normal” average was used instead of the “WilderAverage”…

  Average RSI

  REM Computes the daily variations UP = MAX(0, close - close[1]) DOWN = MAX(0, close[1] - close) REM Computes the moving average of gains on positive days REM and losses on negative days upMA = Average[p](UP) downMA = Average[p](DOWN) REM Now we can compute the RS RS = upMA / downMA REM And finally the RSI myRSI = 100 - 100 / (1 + RS) DrawHLine(70) DrawHLine(30) RETURN myRSI AS "Relative Strength Index"

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

  REM Computes the daily variations   UP = MAX(0, close - close[1]) DOWN = MAX(0, close[1] - close)   REM Computes the moving average of gains on positive days REM and losses on negative days   upMA = Average[p](UP) downMA = Average[p](DOWN)   REM Now we can compute the RS   RS = upMA / downMA   REM And finally the RSI   myRSI = 100 - 100 / (1 + RS)   DrawHLine(70) DrawHLine(30)   RETURN myRSI AS "Relative Strength Index"

  1 user thanked author for this post.

  joe8711

  

  JS

  Participant

  Topics: 26

  Replies: 1701

  Been thanked: 768 times
  02/24/2024 at 11:09 AM #228679
  robertogozzi

  The code you posted is the correct standard formula, as used by PRT.

  I couldn’t spot any difference between the two indicators.

   

  

  robertogozzi

  Moderator

  Topics: 69

  Replies: 17131

  Been thanked: 4026 times
  02/24/2024 at 11:22 AM #228682
  joe8711

  Thanks to both of you!

  The reply from JS is exactly, what I’m looking for.
  Only “p = 14” was missing, but thats it.

  Many thanks, good job!

  1 user thanked author for this post.

  JS

  

  joe8711

  Participant

  Topics: 1

  Replies: 1

  Been thanked: 1 time
• Author
  Posts