Learn R Programming

unmarked (version 1.4.3)

occuPEN: Fit the MacKenzie et al. (2002) Occupancy Model with the penalized likelihood methods of Hutchinson et al. (2015)

Description

This function fits the occupancy model of MacKenzie et al (2002) with the penalized methods of Hutchinson et al (2015).

Usage

occuPEN(formula, data, knownOcc=numeric(0), starts, method="BFGS",
    engine=c("C", "R"), lambda=0, pen.type = c("Bayes","Ridge","MPLE"), ...)

Value

unmarkedFitOccuPEN object describing the model fit.

Arguments

formula

Double right-hand side formula describing covariates of detection and occupancy in that order.

data

An unmarkedFrameOccu object

knownOcc

Vector of sites that are known to be occupied. These should be supplied as row numbers of the y matrix, eg, c(3,8) if sites 3 and 8 were known to be occupied a priori.

starts

Vector of parameter starting values.

method

Optimization method used by optim.

engine

Either "C" or "R" to use fast C++ code or native R code during the optimization.

lambda

Penalty weight parameter.

pen.type

Which form of penalty to use.

...

Additional arguments to optim, such as lower and upper bounds

Author

Rebecca A. Hutchinson

Details

See unmarkedFrame and unmarkedFrameOccu for a description of how to supply data to the data argument.

occuPEN fits the standard occupancy model based on zero-inflated binomial models (MacKenzie et al. 2006, Royle and Dorazio 2008) using the penalized likelihood methods described in Hutchinson et al. (2015). See occu for model details. occuPEN returns parameter estimates that maximize a penalized likelihood in which the penalty is specified by the pen.type argument. The penalty function is weighted by lambda.

The MPLE method includes an equation for computing lambda (Moreno & Lele, 2010). If the value supplied does not equal match the one computed with this equation, the supplied value is used anyway (with a warning).

References

Hutchinson, R. A., J. V. Valente, S. C. Emerson, M. G. Betts, and T. G. Dietterich. 2015. Penalized Likelihood Methods Improve Parameter Estimates in Occupancy Models. Methods in Ecology and Evolution. DOI: 10.1111/2041-210X.12368

MacKenzie, D. I., J. D. Nichols, G. B. Lachman, S. Droege, J. Andrew Royle, and C. A. Langtimm. 2002. Estimating Site Occupancy Rates When Detection Probabilities Are Less Than One. Ecology 83: 2248-2255.

MacKenzie, D. I. et al. 2006. Occupancy Estimation and Modeling. Amsterdam: Academic Press.

Moreno, M. and S. R. Lele. 2010. Improved estimation of site occupancy using penalized likelihood. Ecology 91: 341-346.

Royle, J. A. and R. Dorazio. 2008. Hierarchical Modeling and Inference in Ecology. Academic Press.

See Also

unmarked, unmarkedFrameOccu, occu, computeMPLElambda, occuPEN_CV, nonparboot

Examples

Run this code

# Simulate occupancy data
set.seed(344)
nSites <- 100
nReps <- 2
covariates <- data.frame(veght=rnorm(nSites),
    habitat=factor(c(rep('A', nSites/2), rep('B', nSites/2))))

psipars <- c(-1, 1, -1)
ppars <- c(1, -1, 0)
X <- model.matrix(~veght+habitat, covariates) # design matrix
psi <- plogis(X %*% psipars)
p <- plogis(X %*% ppars)

y <- matrix(NA, nSites, nReps)
z <- rbinom(nSites, 1, psi)       # true occupancy state
for(i in 1:nSites) {
    y[i,] <- rbinom(nReps, 1, z[i]*p[i])
    }

# Organize data and look at it
umf <- unmarkedFrameOccu(y = y, siteCovs = covariates)
obsCovs(umf) <- covariates
head(umf)
summary(umf)


# Fit some models
fmMLE <- occu(~veght+habitat ~veght+habitat, umf)
fm1pen <- occuPEN(~veght+habitat ~veght+habitat, umf,lambda=0.33,pen.type="Ridge")
fm2pen <- occuPEN(~veght+habitat ~veght+habitat, umf,lambda=1,pen.type="Bayes")

# MPLE:
fm3pen <- occuPEN(~veght+habitat ~veght+habitat, umf,lambda=0.5,pen.type="MPLE")
MPLElambda = computeMPLElambda(~veght+habitat ~veght+habitat, umf) 
fm4pen <- occuPEN(~veght+habitat ~veght+habitat, umf,lambda=MPLElambda,pen.type="MPLE")

# nonparametric bootstrap for uncertainty analysis:
fm1pen <- nonparboot(fm1pen,B=20) # should use more samples
vcov(fm1pen,method="nonparboot")




Run the code above in your browser using DataLab