I was wondering if anyone has ported the examples from Durbin & Koopman "Time Series Analysis by State Space Methods" to R?

You can find RATS code for the examples online and obviosly SsfPack/Ox but no signs of an R companion for this book...


There is a package in CRAN (KFAS) which implements a good portion of the algorithms described in Durbin & Koopman, including "exact" non-informative distributions for the state vector, or parts of it.

Although it is not paticularly tied to Durbin & Koopman's book, you might also be interested in package dlm and the companion book Dynamic Linear Models with R.

  • $\begingroup$ I've run into the same problem as the OP, and also found that dlm and the Dynamic Linear Models with R book are very useful. $\endgroup$ – Wayne Jan 23 '11 at 19:21
  • 2
    $\begingroup$ I would also add that I bought "An Introduction to State Space Time Series Analysis (Practical Econometrics)", which was a very gentle introduction to the topic, and it too, used SsfPack/Ox. Makes it much less useful, in one sense, though it does have the benefit of once you implement it in dlm (in R) yourself, you can check your answer against an independent implementation. $\endgroup$ – Wayne Jan 24 '11 at 18:00
  • $\begingroup$ dlm package and DLMs with R book were very useful for me, too. $\endgroup$ – egbutter Nov 30 '12 at 15:31

A walk through the Commandeur and Koopman's book with R examples is available here. Also check the official book's page where links to new software and tutorials are posted.

An entire JSS volume was dedicated to state space modelling in different software. See particularly the introductory paper and the R paper. Kalman Filtering in R has a good overview of available R packages for state space modeling.

Sections "Dynamic Regression Models" and "Multivariate Time Series Models" in TimeSeries CRAN View contain a few more references.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.