RMark (version 3.0.0)

popan.derived: Computes some derived abundance estimates for POPAN models

Description

Computes estimates, standard errors, confidence intervals and var-cov matrix for population size of each group at each occasion and the sum across groups by occasion for POPAN models. If a marklist is provided the estimates are model averaged.

Usage

popan.derived(x,model,revised=TRUE,normal=TRUE,N=TRUE,NGross=TRUE,drop=FALSE)
 
popan.Nt(Phi,pent,Ns,vc,time.intervals) 
 
popan.NGross(Phi,pent,Ns,vc,time.intervals)

Value

popan.derived returns a list with the following elements depending on the values of N and NGross:

 N -
dataframe of estimates by group and occasion and se, lcl,ucl and
group/occasion data N.vcv - variance-covariance matrix of abundance
estimates in N Nbyocc - dataframe of estimates by occasion (summed across
groups) and se, lcl,ucl and occasion data Nbyocc.vcv - variance-covariance
matrix of abundance estimates in Nbyocc NGross - dataframe of gross
abundance estimates by group and se, lcl,and ucl NGross.vcv -
variance-covariance matrix of NGross abundance estimates 

popan.Nt returns a list with the following elements:

 N
- dataframe of estimates by group and occasion and se, lcl,ucl and
group/occasion data N.vcv - variance-covariance matrix of abundance
estimates in N 

popan.NGross returns a list with the following elements:

 NGross - vector of gross abundance estimates by group vcv -
variance-covariance matrix of abundance estimates in NGross 

Arguments

x

processed data list resulting from process.data

model

a single mark POPAN model or a marklist of POPAN models

revised

if TRUE, uses revised version of model averaged standard error eq 6.12; otherwise uses eq 4.9 of Burnham and Anderson (2002)

normal

if TRUE, uses confidence interval based on normal distribution; otherwise, uses log-normal

N

if TRUE, will return abundance estimates by group and occasion and total by occasion

NGross

if TRUE, will return gross abundance estimate per group

drop

if TRUE, models with any non-positive variance for betas are dropped

Phi

interval-specific survival estimates for each group

pent

occasion-specific prob of entry estimates (first computed by subtraction) for each group

Ns

group specific super-population estimate

vc

variance-covariance matrix of the real parameters

time.intervals

vector of time interval values

Author

Jeff Laake

Details

popan.derived computes all of the real parameters using covariate.predictions and handles all of the computation using popan.Nt. Description for functions popan.Nt and popan.NGross are given here for completeness but it is not intended that they be called directly.

If a model is a marklist of models, the values returned by popan.derived are model averaged using model weights in the model.table; otherwise, it returns the values for the specified model.

References

BURNHAM, K. P., AND D. R. ANDERSON. 2002. Model selection and multimodel inference. A practical information-theoretic approach. Springer, New York.

Examples

# \donttest{
# This example is excluded from testing to reduce package check time
# Example
data(dipper)
dipper.processed=process.data(dipper,model="POPAN",groups="sex")
run.dipper.popan=function()
{
dipper.ddl=make.design.data(dipper.processed)
Phidot=list(formula=~1)
Phitime=list(formula=~time)
pdot=list(formula=~1)
ptime=list(formula=~time)
pentsex.time=list(formula=~time)
Nsex=list(formula=~sex)
#
# Run assortment of models
#
dipper.phisex.time.psex.time.pentsex.time=mark(dipper.processed,
     dipper.ddl,model.parameters=list(Phi=Phidot,p=ptime,
     pent=pentsex.time,N=Nsex),invisible=FALSE,adjust=FALSE,delete=TRUE)
dipper.psex.time.pentsex.time=mark(dipper.processed,dipper.ddl,
     model.parameters=list(Phi=Phitime,p=pdot,
     pent=pentsex.time,N=Nsex),invisible=FALSE,adjust=FALSE,delete=TRUE)
#
# Return model table and list of models
#
return(collect.models() )
}
dipper.popan.results=run.dipper.popan()
popan.derived(dipper.processed,dipper.popan.results)
# }