Conversion of indicator: Kalman Filter from Tradingview

• 07/06/2022 at 6:46 AM #196839
  Jiacky

  Indicator Kalman Filter. It allows efficiently smoothing the noise, extracting the main trend from it.

  Kalman Filter is a variety of recursive filters. To assess the system status as of the current operation tact, it needs a status assessment (as the system status assessment and the error assessment of defining this status) at the preceding operation tact, and measuring at the current tact. This property distinguishes it from packet filters requiring to know at the current operation tact the history of measurements and/or assessments.

  More about Kalman Filter.

  The indicator has three input parameters:

  • K – Kalman factor;
  • Sharpness – factor for calculating the error minimization;
  • Applied price – price used for calculations.

  Calculations:

  Kalman[i] = Error + Velocity[i]

  1	Kalman[i] = Error + Velocity[i]

  where:

  Error = Kalman[i-1] + Distance * ShK Velocity[i] = Velocity[i-1] + Distance * K / 100 Distance = Price[i] - Kalman[i-1] ShK = sqrt(Sharpness * K / 100) Code:

  1 2 3 4 5 6

  Error = Kalman[i-1] + Distance * ShK Velocity[i] = Velocity[i-1] + Distance * K / 100 Distance = Price[i] - Kalman[i-1] ShK = sqrt(Sharpness * K / 100)   Code:

  //@version=5
  indicator(‘Kalman Filter [Loxx]’, overlay = true, shorttitle=”KF [Loxx]”, timeframe=””, timeframe_gaps=true)

  src = input(ohlc4)
  Sharpness = input.float(1.0)
  K = input.float(1.0)

  greencolor = color.lime
  redcolor = color.red

  velocity = 0.0
  kfilt = 0.0

  Distance = src – nz(kfilt[1], src)
  Error = nz(kfilt[1], src) + Distance * math.sqrt(Sharpness*K/ 100)
  velocity := nz(velocity[1], 0) + Distance*K / 100
  kfilt := Error + velocity

  plot(kfilt, color=velocity > 0 ? greencolor : redcolor, linewidth = 2)

  

  Jiacky

  Participant

  Topics: 5

  Replies: 12

  Been thanked: 1 time
  07/06/2022 at 6:48 AM #196841
  Jiacky

  Screenshot:

  

  

  Jiacky

  Participant

  Topics: 5

  Replies: 12

  Been thanked: 1 time
  07/06/2022 at 6:49 AM #196844
  Jiacky

  Code is attached. Thank you

  

  Jiacky

  Participant

  Topics: 5

  Replies: 12

  Been thanked: 1 time
  07/06/2022 at 8:52 AM #196854
  Nicolas

  You can find 3 versions of the Kalman Filter in the code attached to this topic (huge indicator with 67 different type of moving averages):

  https://www.prorealcode.com/topic/indicateur-afr-average-filter-regression/#post-36039

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
  07/06/2022 at 6:12 PM #196913
  Jiacky

  Thank you Nicolas.

  

  Jiacky

  Participant

  Topics: 5

  Replies: 12

  Been thanked: 1 time
• Author
  Posts