rmeasure_spec: The measurement-model simulator

In marvellous122/R-to-JS: Statistical Inference for Partially Observed Markov Processes

Description

Specification of rmeasure

Details

The measurement model is the link between the data and the unobserved state process. It can be specified either by using one or both of the rmeasure and dmeasure arguments.

Suppose you have a procedure to simulate observations given the value of the latent state variables. Then you can furnish

  rmeasure = f

to pomp algorithms, where f is a C snippet or R function that implements your procedure.

Using a C snippet is much preferred, due to its much greater computational efficiency. See Csnippet for general rules on writing C snippets.

In writing an rmeasure C snippet, bear in mind that:

1. The goal of such a snippet is to fill the observables with random values drawn from the measurement model distribution. Accordingly, each observable should be assigned a new value.

2. In addition to the states, parameters, covariates (if any), and observables, the variable t, containing the time of the observation, will be defined in the context in which the snippet is executed.

The demos and the tutorials on the package website give examples as well.

It is also possible, though far less efficient, to specify rmeasure using an R function. In this case, specify the measurement model simulator by furnishing

  rmeasure = f

to pomp, where f is an R function with prototype

  f(x, t, params, \dots)

It can also take any additional arguments if these are passed along with it in the call to pomp. When f is called,

• x will be a named numeric vector of length nvar, the number of state variables.

• t will be a scalar quantity, the time at which the measurement is made.

• params will be a named numeric vector of length npar, the number of parameters.

f must return a named numeric vector of length nobs, the number of observable variables.

Default behavior

The default rmeasure is undefined. It will yield missing values (NA).

See Also

Other information on model implementation: Csnippet, accumulators, covariate_table, distributions, dmeasure_spec, dprocess_spec, parameter_trans, pomp2-package, prior_spec, rinit_spec, rprocess_spec, skeleton_spec, transformations, userdata

marvellous122/R-to-JS documentation built on May 20, 2019, 9:55 a.m.