An S3-class that defines the named list returned by pf_filter() and pf_smoother_two_filter().
pf_particles is a label used to denote outputs from selected functions in patter. The structure of this class is not strictly defined and primarily exists to streamline documentation. At the time of writing, pf_particles objects may comprise the following elements:
xinit---A data.table of initial particle samples;
states---A data.table of simulated states;
diagnostics---A data.table of diagnostic statistics;
convergence---A logical variable that defines whether or not the algorithm converged;
xinitxinit is a data.table that defines initial particles (from sim_states_init()). This includes one row for each particle and one column for each state dimension (e.g., map_value, x, y).
statesstates is a data.table that defines simulated particle states, with the following columns:
path_id---An integer vector that defines the particle index;
timestep---An integer vector that defines the time step;
timestamp---A POSIXct vector of time stamps;
Additional columns with the values of each state dimension;
Particles are equally weighted, as the .n_record particles recorded at each time step are selected by resampling (see pf_filter()).
diagnosticsdiagnostics is a data.table that stores diagnostic statistics for each time step. This includes timestep, timestamp and the following columns:
ess---A numeric vector that defines the effective sample size;
maxlp---A numeric vector that defines the maximum log posterior;
convergenceconvergence is a logical variable that defines whether or not the algorithm converged (that is, reached the end of the time series).
Particle filters and smoothers sample states (particles) that represent the possible locations of an individual through time, accounting for all data and the individual's movement.
To simulate artificial datasets, see sim_*() functions (especially sim_path_walk(), sim_array() and sim_observations()).
To assemble real-world datasets for the filter, see assemble_*() functions.
pf_filter() runs the filter:
To run particle smoothing, use pf_smoother_two_filter().
To map emergent patterns of space use, use a map_*() function (such as map_pou(), map_dens() and map_hr()).