sspir                 package:sspir                 R Documentation

_S_t_a_t_e _S_p_a_c_e _M_o_d_e_l_s _i_n _R

_D_e_s_c_r_i_p_t_i_o_n:

     The main contribution of this package is to give a formula
     language for specifying dynamic generalized linear models. That
     is, an extension of glm formulae by marking terms with 'tvar' to
     specify that their coefficients are time-varying. The package also
     provides (extended) Kalman filter and Kalman smoother for models
     within Gaussian, Poisson and binomial families. To get started,
     try 'demo(gas)', 'demo(vandrivers)' and 'demo(mumps)'.

_W_h_a_t _s_s_p_i_r _d_o_e_s _n_o_t _i_n_c_l_u_d_e:

   _O_p_t_i_m_i_z_a_t_i_o_n To keep full generality, the Kalman filter and smoother
        use the 'SS' functions 'Fmat', 'Gmat', 'Vmat', 'Wmat'. The
        special cases where these matrices are time-invariant, the
        algorithms can be considerably speeded up by implementing these
        special cases eg. in C. For simple models, see 'StructTS'.

   _D_i_f_f_u_s_e _i_n_i_t_i_a_l_i_z_a_t_i_o_n We use m_0 and C_0 as initialization of the
        state process. This may cause numeric problems which may be
        solved using diffuse initialization (see Durbin and Koopman
        (2000)).

   _A_R_I_M_A _m_o_d_e_l_s These are not directly supported, but since they can be
        expressed as state space models, it is possible to specify them
        as 'SS' objects.

   _I_m_p_o_r_t_a_n_c_e _s_a_m_p_l_i_n_g We use iterated extended Kalman smoothing and
        use the likelihood from the approximating Gaussian state space
        model. This may be improved using importance sampling.

   _M_u_l_t_i_v_a_r_i_a_t_e _o_b_s_e_r_v_a_t_i_o_n_s We have plans including methods for
        combining 'ssm' objects for different time-series so that they
        can possibly share components in the latent process.

   _M_u_l_t_i-_p_r_o_c_e_s_s _m_o_d_e_l_s We have plans for including methods for
        combining 'ssm' objects defining different models for the same
        time-series. Providing prior probabilities for these models, it
        is possible to calculate posterior probabilites for the models,
        thereby discriminating between the models.

   _M_a_r_k_o_v _c_h_a_i_n _M_o_n_t_e _C_a_r_l_o A state space model may be part of a
        hierarchical model giving priors on hyper-parameters. Inference
        may be done using MCMC methods and for the state space model,
        the Forwards filtering, Backwards sampling method may be used.
        We plan to include at least some support for this.

   _M_i_s_s_i_n_g _v_a_l_u_e_s Missing values are not allowed in the covariates.

   _F_a_m_i_l_y Currently, only three combinations of distribution/link
        functions are supported: Gaussian/identity, Poisson/log,
        Binomial/logit. To add new variations, edit the function
        'getFamily'

_O_t_h_e_r _s_o_f_t_w_a_r_e:

   _O_x/_S_s_f_P_a_c_k SsfPack is a package for Ox by SJ Koopman, N Shephard and
        JA Doornik. Multivariate Gaussian state space models with some
        support for non-Gaussian models. <URL:
        http://www.ssfpack.com/>. Closely related to S+FinMetrics, a
        package for S-Plus, developed by Insightful.  <URL:
        http://www.insightful.com/products/finmetrics/default.asp>.

   _S_t_r_u_c_t_T_S StructTS by BD Ripley is a part of R. Univariate Gaussian
        state space models of the class structural time series models. 

   _K_E_E Kalman Estimating Equations is a package for S-Plus by SJ
        Knudsen. Poisson-Tweedie log-linear State-Space model. <URL:
        http://www.statdem.sdu.dk/~sjk/kee/index.html>.

   _I_E_K_S IEKS library is a package for S-Plus and R by B Klein.
        Implementation of the iterated extended Kalman filter and
        smoother to discrete data from an exponential family. The
        latent process is assumed to be linear and Gaussian. <URL:
        http://genetics.agrsci.dk/~bmk/>. 

   _M_a_t_l_a_b Kalman filter toolbox for Matlab by K Murphy. <URL:
        http://www.ai.mit.edu/~murphyk/Software/Kalman/kalman.html>. 

   _B_a_t_s BATS software by A Pole, M West and J Harrison. <URL:
        http://www.isds.duke.edu/~mw/bats.html>. 

   _d_s_e Dynamic Systems Estimation library (dse1 and dse2) are packages
        for  R by P Gilbert. Multivariate Gaussian state space models
        with focus on ARMA models. See <URL:
        http://www.bank-banque-canada.ca/pgilbert>.  

   _C_T_S_M Continuous Time Stochastic Modelling (CTSM) is a stand alone
        program by NR Kristensen, LE Christiansen and H Madsen.
        Stochastic differential equations, linear and non-linear
        Gaussian state space models. <URL:
        http://www.imm.dtu.dk/ctsm/>.

     Probably this list is incomplete. Feel free to contribute with
     more links to packages.

_A_u_t_h_o_r(_s):

     Claus Dethlefsen and Sren Lundbye-Christensen.

_S_e_e _A_l_s_o:

     'ssm'

