point.sim.pt: Point Transect Method : Simulation

Description Usage Arguments Details Value See Also Examples

Description

This function simulates sample objects of class 'sample.pt' and estimates abundance and related parameters for each simulated sample object.

Usage

1
2
point.sim.lt(pop.spec, survey.spec, design.spec, B = 999, plot = FALSE, title = FALSE, conditional = TRUE, model = "half.normal", seed = NULL, show =  
FALSE, ...) 

Arguments

pop.spec

population specification; either an object of class 'population' or 'pars.population'

survey.spec

survey specification; an object of class 'pars.survey.pt'

design.spec

design specification; either an object of class 'design.pt' or 'pars.design.pt'

B

number of simulations required

plot

argument for point.est.pt - if TRUE, a density histogram of observed radial distances with the estimated probability density function superimposed, is plotted for each simulation

title

Flag telling if you want the title "Radial distance distribution and fitted detection function"

conditional

argument for point.est.pt - if FALSE, the full likelihood (Equation (7.32) of Borchers et al. 2002) is maximized; if TRUE the conditional likelihood (Equation (7.33) of Borchers et al. 2002) is maximized and abundance is estimated using Equation (7.34) of Borchers et al. 2002.

model

argument for point.est.pt - ="half.normal" (no other detection function models have implemented yet)

seed

Number passed to set.seed() to initialise random number generator

show

if TRUE displays the histograms of observations, and the fitted distribution function for each simulated survey as it is run

...

extra plot arguments

Details

This function simulates sample objects of class 'sample.pt' by simulating from the observation model (using survey.spec) and if pop.spec is of class 'pars.population' from the state model (using pop.spec to generate new populations on each simulation) and if design.spec is of class 'pars.design.pt' from the design (using design.spec to generate new design realizations on each simulation.)

Value

An object of class point.sim.lt with the following elements: A results matrix, each row of which contains the following values:

Nhat.grp

MLE of group abundance

Nhat.ind

MLE of individual abundance (= Nhat.grp * Es)

Es

Estimate of mean group size (simple mean of observed group sizes)

esa

MLE effective survey area: 2piint_0^w x p(x),dx, where w is the trunction distance, and integration is from 0 to w.

nbar

"encounter rate" - the number of groups detected per point

true

The true (simulated) values of group abundance, animal abundance and mean group size

conditional

Equal to the argument 'conditional' passed to the function

random.pop

TRUE if population is randomised

random.design

TRUE if design is randomised

parents

Details of WiSP objects passed to function

created

Creation date and time

seed

Equal to the argument 'seed' passed to the function

See Also

setpars.population, setpars.design.pt setpars.survey.pt, point.est.pt

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
pt.reg <- generate.region(x.length = 100, y.width = 50)
pt.dens <- generate.density(pt.reg)

#heterogeneous population
pt.poppars<-setpars.population(density.pop = pt.dens, number.groups = 1000, size.method = "poisson",
                               size.min = 1, size.max = 30, size.mean = 10, exposure.method = "beta",
                               exposure.min = 0, exposure.max = 1, exposure.mean = 0.4,
                               exposure.shape = 0.5, type.values=c("Male","Female"),
                               type.prob=c(0.48,0.52))
pt.pop<-generate.population(pt.poppars)

pt.despars<-setpars.design.pt(pt.reg, n.transects=8, n.units=32, visual.range=3.5)
pt.des<-generate.design.pt(pt.despars)

pt.survpars<-setpars.survey.pt(pt.pop, pt.des, disthalf.min=2, disthalf.max=4)
pt.samp<-generate.sample.pt(pt.survpars)

# simulate (design and population randomisation):
pt.sim<-point.sim.pt(pop.spec=pt.poppars, design.spec=pt.despars, survey.spec=pt.survpars, B=9, seed=NULL, plot=TRUE)
summary(pt.sim)
plot(pt.sim)
plot(pt.sim, type="hist")
plot(pt.sim, type="box")

# simulate (design randomisation only):
pt.sim<-point.sim.pt(pop.spec=pt.pop, design.spec=pt.despars, survey.spec=pt.survpars, B=99, seed=NULL, plot=FALSE)
summary(pt.sim)
plot(pt.sim)

dill/wisp documentation built on May 15, 2019, 8:31 a.m.