fxi.secrgam: Probability Density of Home Range Centre

In david-borchers/secrgam: Spatially Explicit Capture-Recapture with Generalised Additive Density Models

Description

Calculates the probability density function for the estimated location of one or more range centres, i.e. f(X | Ω). Acts as a wrapper function for fxi.secr.

Usage

fxi.secrgam(fit, i = 1, col = 1:length(i), X = NULL, plt = TRUE,
  add = FALSE, sessnum = 1, border = 0, normal = TRUE, ...)

Arguments

fit	a fitted secrgam model (returned from secrgam.fit)
i	integer or character vector of individuals for which to plot contours
col	integer or character vector of colour for individuals' contours
X	2-column matrix of x- and y- coordinates (e.g. the original mask)
plt	logical to do contour plot
add	logical to add to existing plot
sessnum	session number if fit$capthist spans multiple sessions
border	width of blank margin around the outermost detectors
normal	logical; should values of pdf be normalised?
...	additional arguments passed to contour.

Details

fxi.secr returns a vector of probability densities assuming a uniform density distribution,

f(x_i | ω_i) = P(ω_i | x_i) / ∑ P(ω_i | x_i)

fxi.secrgam returns a vector of probability densities for any density distribution,

f(x_i | ω_i) = P(ω_i | x_i) D(x_i) / ∑ [ P(ω_i | x_i) D(x_i) ]

Needs to be tested for cases where density model has fixed parameters. Also assumes that log link is used for density.

Examples

data(Boland.fits1)

# capture history data
plot(fit1.a3$capthist)

# look at data for animal 20
animal = 20
k = apply(fit1.a3$capthist[animal,,], 2, sum) > 0
points(traps(fit1.a3$capthist)[k,], col = 2, pch = 19)
fxi.secrgam(fit1.a3, i = animal, add = TRUE)

david-borchers/secrgam documentation built on May 14, 2019, 9:30 p.m.