pfilter {pomp} | R Documentation |
Particle filter
Description
A plain vanilla sequential Monte Carlo (particle filter) algorithm. Resampling is performed at each observation.
Usage
## S4 method for signature 'data.frame'
pfilter(
data,
Np,
params,
rinit,
rprocess,
dmeasure,
pred.mean = FALSE,
pred.var = FALSE,
filter.mean = FALSE,
filter.traj = FALSE,
save.states = c("no", "weighted", "unweighted", "FALSE", "TRUE"),
...,
verbose = getOption("verbose", FALSE)
)
## S4 method for signature 'pomp'
pfilter(
data,
Np,
pred.mean = FALSE,
pred.var = FALSE,
filter.mean = FALSE,
filter.traj = FALSE,
save.states = c("no", "weighted", "unweighted", "FALSE", "TRUE"),
...,
verbose = getOption("verbose", FALSE)
)
## S4 method for signature 'pfilterd_pomp'
pfilter(data, Np, ..., verbose = getOption("verbose", FALSE))
## S4 method for signature 'objfun'
pfilter(data, ...)
Arguments
data |
either a data frame holding the time series data,
or an object of class ‘pomp’,
i.e., the output of another pomp calculation.
Internally, |
Np |
the number of particles to use.
This may be specified as a single positive integer, in which case the same number of particles will be used at each timestep.
Alternatively, if one wishes the number of particles to vary across timesteps, one may specify length(time(object,t0=TRUE)) or as a function taking a positive integer argument.
In the latter case, |
params |
optional; named numeric vector of parameters.
This will be coerced internally to storage mode |
rinit |
simulator of the initial-state distribution.
This can be furnished either as a C snippet, an R function, or the name of a pre-compiled native routine available in a dynamically loaded library.
Setting |
rprocess |
simulator of the latent state process, specified using one of the rprocess plugins.
Setting |
dmeasure |
evaluator of the measurement model density, specified either as a C snippet, an R function, or the name of a pre-compiled native routine available in a dynamically loaded library.
Setting |
pred.mean |
logical; if |
pred.var |
logical; if |
filter.mean |
logical; if |
filter.traj |
logical; if |
save.states |
character;
If |
... |
additional arguments are passed to |
verbose |
logical; if |
Value
An object of class ‘pfilterd_pomp’, which extends class ‘pomp’. Information can be extracted from this object using the methods documented below.
Methods
logLik
the estimated log likelihood
cond_logLik
the estimated conditional log likelihood
eff_sample_size
-
the (time-dependent) estimated effective sample size
pred_mean
,pred_var
the mean and variance of the approximate prediction distribution
filter_mean
the mean of the filtering distribution
filter_traj
-
retrieve one particle trajectory. Useful for building up the smoothing distribution.
saved_states
retrieve saved states
as.data.frame
coerce to a data frame
plot
diagnostic plots
Note for Windows users
Some Windows users report problems when using C snippets in parallel computations.
These appear to arise when the temporary files created during the C snippet compilation process are not handled properly by the operating system.
To circumvent this problem, use the cdir
and cfile
options to cause the C snippets to be written to a file of your choice, thus avoiding the use of temporary files altogether.
Author(s)
Aaron A. King
References
M.S. Arulampalam, S. Maskell, N. Gordon, and T. Clapp. A tutorial on particle filters for online nonlinear, non-Gaussian Bayesian tracking. IEEE Transactions on Signal Processing 50, 174–188, 2002. doi:10.1109/78.978374.
A. Bhadra and E.L. Ionides. Adaptive particle allocation in iterated sequential Monte Carlo via approximating meta-models. Statistics and Computing 26, 393–407, 2016. doi:10.1007/s11222-014-9513-x.
See Also
More on pomp elementary algorithms:
elementary_algorithms
,
kalman
,
pomp-package
,
probe()
,
simulate()
,
spect()
,
trajectory()
,
wpfilter()
More on sequential Monte Carlo methods:
bsmc2()
,
cond_logLik()
,
eff_sample_size()
,
filter_mean()
,
filter_traj()
,
kalman
,
mif2()
,
pmcmc()
,
pred_mean()
,
pred_var()
,
saved_states()
,
wpfilter()
More on full-information (i.e., likelihood-based) methods:
bsmc2()
,
mif2()
,
pmcmc()
,
wpfilter()
Examples
pf <- pfilter(gompertz(),Np=1000) ## use 1000 particles
plot(pf)
logLik(pf)
cond_logLik(pf) ## conditional log-likelihoods
eff_sample_size(pf) ## effective sample size
logLik(pfilter(pf)) ## run it again with 1000 particles
## run it again with 2000 particles
pf <- pfilter(pf,Np=2000,filter.mean=TRUE,filter.traj=TRUE,save.states="weighted")
fm <- filter_mean(pf) ## extract the filtering means
ft <- filter_traj(pf) ## one draw from the smoothing distribution
ss <- saved_states(pf,format="d") ## the latent-state portion of each particle
as(pf,"data.frame") |> head()