Statistical Inference for Partially Observed Markov Processes
Approximate Bayesian computation
accumulator variables
Coerce to data frame
as.pomp
Tools for reproducible computations
Basic POMP model components.
Useful probes for partially-observed Markov processes
Beta-binomial distribution
Nicholson's blowflies.
The Liu and West Bayesian particle filter
B-spline bases
Historical childhood disease incidence data
Extract, set, or alter coefficients
Concatenate
Concatenate
Conditional log likelihood
Continue an iterative calculation
Covariates
Estimate a covariance matrix from algorithm traces
C snippets
Model of cholera transmission for historic Bengal.
Design matrices for pomp calculations
dinit specification
dinit workhorse
dmeasure specification
dmeasure workhorse
dprior workhorse
dprocess specification
dprocess workhorse
Ebola outbreak, West Africa, 2014-2016
Effective sample size
Elementary computations on POMP models.
emeasure specification
emeasure workhorse
Parameter estimation algorithms for POMP models.
Eulermultinomial and gamma-whitenoise distributions
Filtering mean
Filtering trajectories
flow workhorse
Forecast mean
Gompertz model with log-normal observations.
Hitching C snippets and R functions to pomp_fun objects
Ensemble Kalman filters
Kalman filter
listie
Loading and unloading shared-object libraries
Log likelihood
The log-mean-exp trick
Lookup table
Monte Carlo adjusted profile
Melt
Iterated filtering: maximum likelihood by iterated, perturbed Bayes ma...
Nonlinear forecasting
Objective functions
obs
Two-dimensional discrete-time Ornstein-Uhlenbeck process
parameter transformations
Create a matrix of parameters
partrans workhorse
Parus major population dynamics
Particle filter
pomp plotting facilities
The particle Markov chain Metropolis-Hastings algorithm
The basic pomp class
pre-built pomp examples
The "pomp_fun" class
Inference for partially observed Markov processes
Constructor of the basic pomp object
Prediction mean
Prediction variance
Print methods
prior specification
Probe matching
Probes (AKA summary statistics)
MCMC proposal distributions
pStop, pWarn, pMess
Resample
Ricker model with Poisson observations.
rinit specification
rinit workhorse
rmeasure specification
rmeasure workhorse
rprior workhorse
rprocess specification
rprocess workhorse
rw_sd
Two-dimensional random-walk process
Simulated annealing with box constraints.
Saved states
Show methods
Simulations of a partially-observed Markov process
Compartmental epidemiological models
skeleton specification
skeleton workhorse
Spectrum matching
Power spectrum
Spy
Latent states
Summary methods
Methods to extract and manipulate the obseration times
The zero time
Traces
Trajectory matching
Trajectory of a deterministic model
Transformations
Undefined
Facilities for making additional information to basic components
Verhulst-Pearl model
vmeasure specification
vmeasure workhorse
Window
Workhorse functions for the pomp algorithms.
Weighted particle filter
Weighted quantile function
Tools for data analysis with partially observed Markov process (POMP) models (also known as stochastic dynamical systems, hidden Markov models, and nonlinear, non-Gaussian, state-space models). The package provides facilities for implementing POMP models, simulating them, and fitting them to time series data by a variety of frequentist and Bayesian methods. It is also a versatile platform for implementation of inference methods for general POMP models.