There is interrelationship between process noise and measurement noise. The averaging constant K is given as:
K = (Estim. Error + Process Noise)/ [(Estim. Error + Process Noise) + Measurement Noise Variance] When measurement noise is much smaller than the observation noise, K= 1 and there is no averaging. When measurement noise is much greater, K is very small and averaging is very significant. Of course, one more cog in the wheel is Estimation Error, which is recomputed at each time step, making Kalman adaptive, unlike EMA, - where K is constant. Estimation Error is measure of how well the algorithm suppresses noise. On a different topic... In KalmanTest: addParam(FAST_ERROR, 0, 100, 1, 0); If the last parameter is indeed the value used for back test, it is probably best to make it very small but not zero. I think setting the value of short-term noise to zero can do strange things to an algorithm designed to suppress that noise. ________________________________ From: nonlinear5 <[email protected]> To: JBookTrader <[email protected]> Sent: Tue, November 30, 2010 8:49:17 AM Subject: [JBookTrader] Re: Status of Kalman filter? I noticed that there is also kalman.setProcess_noise_cov(), in addition to the kalman.setMeasurement_noise_cov() which I am currently using. I replaced setMeasurement_noise_cov with setProcess_noise_cov(), and now the Tension indicator much more closely resembles the original Tension where EMA was used as a filter. -- You received this message because you are subscribed to the Google Groups "JBookTrader" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/jbooktrader?hl=en. -- You received this message because you are subscribed to the Google Groups "JBookTrader" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/jbooktrader?hl=en.
