# munit_measure: munit_measure In spatPomp: Inference for Spatiotemporal Partially Observed Markov Processes

## Description

`munit_measure` returns a moment-matched parameter set given an empirically calculated measurement variance and latent states. This is used in `girf()` and `igirf()` when they are run with `kind='moment'`.

## Usage

 ```1 2``` ```## S4 method for signature 'spatPomp' munit_measure(object, x, vc, unit, time, params, Np = 1) ```

## Arguments

 `object` An object of class `spatPomp` `x` A state vector for all units `vc` The empirically calculated variance used to perform moment-matching `unit` The unit for which to obtain a moment-matched parameter set `time` The time for which to obtain a moment-matched parameter set `params` parameters to use to obtain a moment-matched parameter set `Np` Number of particle replicates for which to get parameter sets

## Value

An array with dimensions `dim(array.params)[1]` by `dim(x)[2]` by `length(unit)` by`length(time)` representing the moment-matched parameter set(s) corresponding to the variance of the measurements, `vc`, and the states, `x`.

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14``` ```b <- bm(U=3) s <- states(b)[,1,drop=FALSE] rownames(s) -> rn dim(s) <- c(3,1,1) dimnames(s) <- list(variable=rn, rep=NULL) p <- coef(b); names(p) -> rnp dim(p) <- c(length(p),1); dimnames(p) <- list(param=rnp) o <- obs(b)[,1,drop=FALSE] array.params <- array(p, dim = c(length(p), length(unit_names(b)), 1, 1), dimnames = list(params = rownames(p))) vc <- c(4, 9, 16); dim(vc) <- c(length(vc), 1, 1) munit_measure(b, x=s, vc=vc, Np=1, unit = 1, time=1, params=array.params) ```

spatPomp documentation built on Sept. 5, 2021, 5:35 p.m.