Drawing from the smoothing distribution
# S4 method for pfilterd_pomp
filter.traj(object, vars, ...)# S4 method for pfilterList
filter.traj(object, vars, ...)
# S4 method for pmcmcd_pomp
filter.traj(object, vars, ...)
# S4 method for pmcmcList
filter.traj(object, vars, ...)
result of a filtering computation
optional character; names of variables
ignored
The smoothing distribution is the distribution of $$X(t_k) | Y(t_1)=y^*_1, \dots, Y(t_n)=y^*_n,$$ where \(X(t_k)\) is the latent state process and \(Y(t_k)\) is the observable process at time \(t_k\), and \(n\) is the number of observations.
To draw samples from this distribution, one can run a number of independent particle filter (pfilter) operations, sampling the full trajectory of one randomly-drawn particle from each one.
One should view these as weighted samples from the smoothing distribution, where the weights are the likelihoods returned by each of the pfilter computations.
One accomplishes this by setting filter.traj = TRUE in each pfilter computation and extracting the trajectory using the filter.traj command.
In particle MCMC (pmcmc), the tracking of an individual trajectory is performed automatically.
More on particle-filter based methods in pomp:
bsmc2(),
cond.logLik(),
eff.sample.size(),
filter.mean(),
kalman,
mif2(),
pfilter(),
pmcmc(),
pred.mean(),
pred.var(),
saved.states(),
wpfilter()