makePar: 'makePar' gets parameters for subsequent filter/smoother

In HMMoce: Improved Analysis of Marine Animal Movement Data Using Hidden Markov Models

Description

Function builds movement kernels for 2 different behavior states and calculates, if desired, switching probability using an expectation maximization routine

Usage

makePar(migr.spd, grid, L.arr, p.guess = c(0.7, 0.8), calcP = FALSE)

Arguments

migr.spd	is numeric input to calc.param
grid	is a grid output by resample.grid that corresponds to the extent and resolution of L.arr (below).
L.arr	is the likelihood array used for state switch probability calculation (see expmax). This is typically the L.mle array returned from make.L because it's typically more coarse (and thus faster) than the higher-resolution L array.
p.guess	is vector of length 2 indicating probability of staying in states 1 and 2, respectively
calcP	is logical indicating whether to use expmax to calculate state-switching probabilities

Value

list of parameters including movement kernels (K1, K2) and switch probability (P.final)

Examples

## Not run: 
par0 <- makePar(migr.spd=2, grid=g.mle, L.arr=L.mle, p.guess=c(.9,.9), calcP=T)

## End(Not run)
      

HMMoce documentation built on Nov. 17, 2017, 5:57 a.m.