pfilter: Particle filter

Description

A plain vanilla sequential Monte Carlo (particle filter) algorithm. Resampling is performed at each observation.

Usage

# S4 method for data.frame
pfilter(
  data,
  Np,
  params,
  rinit,
  rprocess,
  dmeasure,
  pred.mean = FALSE,
  pred.var = FALSE,
  filter.mean = FALSE,
  filter.traj = FALSE,
  save.states = FALSE,
  ...,
  verbose = getOption("verbose", FALSE)
)
# S4 method for pomp
pfilter(
  data,
  Np,
  pred.mean = FALSE,
  pred.var = FALSE,
  filter.mean = FALSE,
  filter.traj = FALSE,
  save.states = FALSE,
  ...,
  verbose = getOption("verbose", FALSE)
)
# S4 method for pfilterd_pomp
pfilter(data, Np, ..., verbose = getOption("verbose", FALSE))
# S4 method for 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.

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 Np either as a vector of positive integers of length

length(time(object,t0=TRUE))

or as a function taking a positive integer argument. In the latter case, Np(k) must be a single positive integer, representing the number of particles to be used at the k-th timestep: Np(0) is the number of particles to use going from timezero(object) to time(object)[1], Np(1), from timezero(object) to time(object)[1], and so on, while when T=length(time(object)), Np(T) is the number of particles to sample at the end of the time-series.

params

optional; named numeric vector of parameters. This will be coerced internally to storage mode double.

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 rinit=NULL sets the initial-state simulator to its default. For more information, see ?rinit_spec.

rprocess

simulator of the latent state process, specified using one of the rprocess plugins. Setting rprocess=NULL removes the latent-state simulator. For more information, see ?rprocess_spec for the documentation on these plugins.

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 dmeasure=NULL removes the measurement density evaluator. For more information, see ?dmeasure_spec.

pred.mean

logical; if TRUE, the prediction means are calculated for the state variables and parameters.

pred.var

logical; if TRUE, the prediction variances are calculated for the state variables and parameters.

filter.mean

logical; if TRUE, the filtering means are calculated for the state variables and parameters.

filter.traj

logical; if TRUE, a filtered trajectory is returned for the state variables and parameters. See filter.traj for more information.

save.states

logical. If save.states=TRUE, the state-vector for each particle at each time is saved.

...

additional arguments supply new or modify existing model characteristics or components. See pomp for a full list of recognized arguments.

When named arguments not recognized by pomp are provided, these are made available to all basic components via the so-called userdata facility. This allows the user to pass information to the basic components outside of the usual routes of covariates (covar) and model parameters (params). See ?userdata for information on how to use this facility.

verbose

logical; if TRUE, diagnostic messages will be printed to the console.

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 list of 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 (described here) to cause the C snippets to be written to a file of your choice, thus avoiding the use of temporary files altogether.

References

M.S. Arulampalam, S. Maskell, N. Gordon, & T. Clapp. A tutorial on particle filters for online nonlinear, non-Gaussian Bayesian tracking. IEEE Transactions on Signal Processing 50, 174--188, 2002.

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.

Examples

Run this code

# NOT RUN {
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=TRUE)
fm <- filter.mean(pf)    	## extract the filtering means
ft <- filter.traj(pf)    	## one draw from the smoothing distribution
ss <- saved.states(pf)          ## the latent-state portion of each particle
# }

Run the code above in your browser using DataLab