Fast Kalman Filter

This is a fast and flexible implementation of the Kalman filter and smoother, which can deal with NAs. It is entirely written in C and relies fully on linear algebra subroutines contained in BLAS and LAPACK. Due to the speed of the filter, the fitting of high-dimensional linear state space models to large datasets becomes possible. This package also contains a plot function for the visualization of the state vector and graphical diagnostics of the residuals.

The Fast Kalman Filer Toolbox

The Fast Kalman Fileter Toolbox for R allows the implimentation of multivariate time series filtering with external time varying inputs using the linear Kalman fileter in R.

The canonical link to the package on CRAN which includes examples and documentation is https://CRAN.R-project.org/package=FKF.

Development can be tracked via the github repository which contains source code for the package starting with version 0.1.5.

The latest developmental release can be installed with

devtools::install_github("waternumbers/FKF")

Prebuild packages (usually the latest development version) are available from the r-universe.