estpsi {dwp}

Estimate Probability Carcass lands in Searched Area

Description

Estimated probability that carcass lands in searched area. This is an intermediate step in estimating dwp but is also interesting in its own right. The estimation involves integrating the modeled carcass distribution (model) over the search plots at the turbines. Data for the search plots is stored in the generic argument, x, which can take any of a number of different forms, as described in the Arguments section (below).

Usage

estpsi(x, ...)

## S3 method for class 'rings'
estpsi(x, model, extent = "full", nsim = 1000, zrad = 200, ...)

## S3 method for class 'ringscc'
estpsi(
  x,
  model,
  modnames = NULL,
  extent = "full",
  nsim = 1000,
  zrad = 200,
  ...
)

## S3 method for class 'xyLayout'
estpsi(x, model, extent = "full", nsim = 1000, zrad = 200, ...)

## S3 method for class 'rpA'
estpsi(x, model, extent = "full", nsim = 1000, zrad = 200, ...)

## S3 method for class 'rdat'
estpsi(x, model, extent = "full", nsim = 1000, zrad = 200, ...)

## S3 method for class 'data.frame'
estpsi(x, model, extent = "full", nsim = 1000, zrad = 200, ...)

Arguments

x

data

rings

a formatted site map created from raw data via function prepRing (or as a component of a list returned by addCarcass).

ringscc

a list of rings objects, one for each carcass class; created from raw data via function prepRing (or as a component of a list returned by addCarcass).

xyLayout

formatted site map data derived from (x, y) coordinates covering every square meter of searched areas at each turbine; derived from the function initLayout, when called with xy data.

rpA

(intended as an internal function that would rarely be called directly by users) a list of data frames (one for each turbine) giving the fraction of area searched (pinc at each distance (r). rpA data are embedded in rings objects that are created from site "maps" via prepRing.

rdat

(intended as an internal function that would rarely be called directly by users) list of data frames giving the area searched ("exposure"), in a 1 meter ring with outer radius "r" and the number of carcasses found "ncarc" in each ring, with search class scVar optional. There is also a summary data frame $rdat[["total"]] that sums the exposures and carcass counts for all turbines across the site. The $rdat[["total"]] is the data frame used in fitting the GLMs. rdat objects are components of the return value of prepRing

data.frame

(intended as an internal function that would rarely be called directly by users) a data frame giving the fraction of area searched (pinc at each distance (r).

...

ignored

model

A fitted dd model or an array of estimated parameters (ddSim object); or, if x is a ringscc object, a list of dd models (one for each carcass class), or a ddArraycc accompanied by a vector of model names to use (one for each carcass class).

extent

calculate dwp within searched radius only ("win") or for full complement of carcasses ("full"), including those that fall outside the search radius.

nsim

number of parametric bootstrap iterations for accounting for uncertainty in the estimator. Default is nsim = 1000. Use nsim = 0 for the estimate of psi based on the MLE of the given model without accounting for uncertainty.

zrad

radius

modnames

if x is a ringscc object, a vector of names of model to use for each carcass class; otherwise, modnames is ignored.

Value

A psiHat object, which is either 1) an array giving the expected fraction of carcasses lying in the searched area at each turbine with nsim rows and one column for each turbine + one row for the total; or 2) a list of such arrays, one for each carcass class if x is a ringscc object. The uncertainty in the expected fractions is characterized by simulation and reflected in the variation in psi values within each column.