remMNS: remMNS

In dslramsey/eradicate: Functions to support the "eradication tools" shiny app

remMNS

Description

remMNS fits multinomial removal models (e.g. Haines 2019, Dorazio et al 2005) to 'stacked' data from from M sites over T primary periods with each primary consisting of J secondary periods. Currently supported models include the Poisson, Negative binomial and zero-inflated Poisson (ZIP).

Usage

remMNS(lamformula, detformula, data, starts, method="BFGS", se=TRUE, ...)

Arguments

lamformula	formula for the latent abundance component. For stacked data representing multiple primary periods, the reserved terms .season or .trend can be added to the formula to model abundance for each primary period as a factor (.season) or numeric trend (.trend)
detformula	formula for the removal detection component. Only site-level covariates are allowed for the removal detection component. This differs from the similar model in unmarked.
data	A eFrameMNS object containing the response (counts) for each site, indexed by sessions as well as site-level covariates. see eFrameMNS for how to format the required data.
mixture	model for the latent abundance, either Poisson (P), negative binomial (NB) or zero-inflated Poisson (ZIP).
starts	Initial values for parameters
method	Optimisation method
se	flag to return the standard error (hessian).

Value

a efit model object.

Examples

 rem<- san_nic_open$rem
 emf <- eFrameMNS(y=rem)
 mod <- remMNS(~.season, ~1, data=emf)
 Nhat<- calcN(mod)

dslramsey/eradicate documentation built on March 16, 2024, 1:40 p.m.