[R] State-space model with long tails

Hi All: I have a dataset that when estimated as a dynamic state-space model, appears that the trend term would be better modeled with long-tails, such as a t-distribution, rather than Gaussian. I have looked at the manuals for the packages dlm, KFAS and sspir, and if I can estimate such a mode

Re: [R] State space model

I found an old thread on R-Sig-Finance with the same problem and a possible solution https://stat.ethz.ch/pipermail/r-sig-finance/2007q2/001362.html I've used with success a few times but it seems a bit slow. If anyone has a better way of modelling state-dependent volatility using one of the avai

[R] State space model

Hi, I'm trying to estimate the parameters of a state space model of the following form measurement eq: z_t = a + b*y_t + eps_t transition eq y_t+h = (I -exp(-hL))theta + exp(-hL)y_t+ eta_{t+h}. The problem is that the distribution of the innovations of the transition equation depend on the pr

Re: [R] state space model for poisson distribution

Have you looked at the 'sspir' package? A good introduction to the package can be found in C. Dethlefsen and S. Lundbye-Christensen (2006) "Formulating state space models in r with focus on longitudinal regression models", Journal of Statistical Software, 16(1) [http://www.jstatsoft.org/

[R] state space model for poisson distribution

Hi Rers, I have a poission time series model with 5 parameters. I just wanted to remove two of the lag on response in the model and put it as a system model. I am not sure about the codes to combine these two on R. If anybody has any R example (code), please post it. My original model:

Re: [R] State-space model estimation with EM

Package "dlm" has a function for maximum likelihood estimation of parameters in general (linear Normal) state space model. The function, dlmMLE, computes the likelihood based on singular value decomposition and appears to be fairly robust. No EM algorithm, though. Giovanni > Date: Sun, 11 No

[R] State-space model estimation with EM

Hi - I follow some references and now implement my own state-space model estimation. I have a question. In case, my equations are like this: y(t) = Ax(t)+Bu(t)+eps(t) # observation eq x(t) = Cx(t-1)+Du(t)+eta(t) # state eq Using EM, after backward recursion, you will use the smoothed state estim