This script is a simplified version of John Ehlers's adaption of Dr. Kalman's optimum estimator as applied to price action (More can be found on this here: dimensionetrading.com/Pattern_e_indicatori/Ehlers _Optimal Tracking Filters_.pdf). Here I have adapted two of these optimum estimators to work together to provide crossover signals. The user can choose the input of this filter in the 'input source'. The 'Ratio of Uncertainties' controls how adaptive the moving averages are, increasing this number will increase adaptivity and vice versa for decreasing. The 'Kalman Gain' allows the user to choose how much error to let into the calculation. The smaller this number is the quicker the moving average will approach price action.
In practice this indicator is much smoother than most other moving averages and has significantly less whiplash while still getting very early entries. If anyone wants to adapt this script for their own uses please feel free. Message me what you make with it, I am very curious what this can do when in the right hands!
my man, you always do great work. You are an asset to the Tradingview community. Sincerely, thank you!
jaggedsoft
⋅
Thanks for sharing, using different gain levels as a crossover is a novel idea. The kalman filter is amazing, it helped get the Apollo astronauts safely to the moon and back, an event that helped shape our modern world. It's reliable enough for lives to depend on it and since it can track noisy instruments with zero lag, it should do alright with tracking volatile asset prices as well.
ICEKI
⋅
@jaggedsoft, Appreciated your sharing mcbw!
John Ehlers's indicators is one of my favorite to work with. It's is great to seeing you combined with Kalman filter into it. Thank You so much. Also thank you jaggedsoft for the useful information shared too =D
mcbw_
⋅
Hi @ICEKI! It's always a pleasure to see your comments on my work! I hope that this indicator is of use to you
ICEKI
⋅
@mcbw_, You're very welcome; I love all your work; keep it up bro <3
mcbw_
⋅
Thanks @jaggedsoft! It is interesting to see how different gains can alter this as an indicator. Using this on a higher timeframe in combination with the QQE seems like the start of a decent system.