w2n: Scaling function: working to natural parameters

In momentuHMM: Maximum Likelihood Analysis of Animal Movement Behavior Using Multivariate Hidden Markov Models

Scaling function: working to natural parameters

Description

Scales each parameter from the set of real numbers, back to its natural interval. Used during the optimization of the log-likelihood.

Usage

w2n(
  wpar,
  bounds,
  parSize,
  nbStates,
  nbCovs,
  estAngleMean,
  circularAngleMean,
  consensus,
  stationary,
  fullDM,
  DMind,
  nbObs,
  dist,
  Bndind,
  nc,
  meanind,
  covsDelta,
  workBounds,
  covsPi
)

Arguments

wpar	Vector of working parameters.
bounds	Named list of 2-column matrices specifying bounds on the natural (i.e, real) scale of the probability distribution parameters for each data stream.
parSize	Named list indicating the number of natural parameters of the data stream probability distributions
nbStates	The number of states of the HMM.
nbCovs	The number of beta covariates.
estAngleMean	Named list indicating whether or not to estimate the angle mean for data streams with angular distributions ('vm' and 'wrpcauchy').
circularAngleMean	Named list indicating whether to use circular-linear or circular-circular regression on the mean of circular distributions ('vm' and 'wrpcauchy') for turning angles. See fitHMM.
consensus	Named list indicating whether to use the circular-circular regression consensus model
stationary	FALSE if there are time-varying covariates in formula or any covariates in formulaDelta. If TRUE, the initial distribution is considered equal to the stationary distribution. Default: FALSE.
fullDM	Named list containing the full (i.e. not shorthand) design matrix for each data stream.
DMind	Named list indicating whether fullDM includes individual- and/or temporal-covariates for each data stream specifies (-1,1) bounds for the concentration parameters instead of the default [0,1) bounds.
nbObs	Number of observations in the data.
dist	Named list indicating the probability distributions of the data streams.
Bndind	Named list indicating whether DM is NULL with default parameter bounds for each data stream.
nc	indicator for zeros in fullDM
meanind	index for circular-circular regression mean angles with at least one non-zero entry in fullDM
covsDelta	data frame containing the delta model covariates
workBounds	named list of 2-column matrices specifying bounds on the working scale of the probability distribution, transition probability, and initial distribution parameters
covsPi	data frame containing the pi model covariates

Value

A list of:

...	Matrices containing the natural parameters for each data stream (e.g., 'step', 'angle', etc.)
beta	Matrix of regression coefficients of the transition probabilities
delta	Initial distribution

Examples

## Not run: 
m<-example$m
nbStates <- 2
nbCovs <- 2
parSize <- list(step=2,angle=2)
par <- list(step=c(t(m$mle$step)),angle=c(t(m$mle$angle)))
bounds <- m$conditions$bounds
beta <- matrix(rnorm(6),ncol=2,nrow=3)
delta <- c(0.6,0.4)

#working parameters
wpar <- momentuHMM:::n2w(par,bounds,list(beta=beta),log(delta[-1]/delta[1]),nbStates,
m$conditions$estAngleMean,NULL,m$conditions$Bndind,
m$conditions$dist)

#natural parameter
p <-   momentuHMM:::w2n(wpar,bounds,parSize,nbStates,nbCovs,m$conditions$estAngleMean,
m$conditions$circularAngleMean,lapply(m$conditions$dist,function(x) x=="vmConsensus"),
m$conditions$stationary,m$conditions$fullDM,
m$conditions$DMind,1,m$conditions$dist,m$conditions$Bndind,
matrix(1,nrow=length(unique(m$data$ID)),ncol=1),covsDelta=m$covsDelta,
workBounds=m$conditions$workBounds)

## End(Not run)

momentuHMM documentation built on Oct. 19, 2022, 1:07 a.m.