Kalman Filter Program
(KALMAN)

The Kalman Filter is a recursive algorithm for updating linear projections of the system comprised by of an Observation Equation and a State Equation.

Observation Equation:

State Equation:

The user will supply a vector of observations, coefficients of variation on observation error and scale vector (qt).  The model will estimate two parameters: the state variable at time zero (b0) and the standard error on the state variable (sw).

The model will return a vector of Filtered and Smoothed observations.

There is a technical manual available with details of equations used in the model.  

Meinhold, R and Meinhold, N. 1983. Understanding the Kalman Filter. The American Statistician, Vol. 37, No. 2 (May, 1983), pp. 123-127

Back to Top 

Kalman Version 2.3 (July 2009)  Includes the following changes

• Calculation engine version 1.1 which has been modified to eliminate message "Attempting to take the derivative of sqrt(prevariable x) at x=0" and subsequent "Run Failed" message in the GUI.

• GUI updates to eliminate a problem with the scroll bars on the Data Observations page and to correct a "Run Failed" message in certain situations when working with a new case.

• New 'Combined' plot showing both the Filter and Smooth series on one chart.

Kalman Version 2.1 (January 2009) was upgraded from 2.0 to support Windows Vista.  

Back to Top 

Older Version Changes