integrated_density: create integrated density objects

In jdyen/greta.integrated: Integrated Population Models in greta

Description

An integrated_density object specifies the form of density dependence and associated prior distributions for an integrated population analysis

Usage

ricker(parameters = uniform(0, 1))

beverton(parameters = uniform(0, 1))

element_wise(parameters = uniform(0, 1), masks = list())

no_density(parameters = uniform(0, 1))

is.integrated_density(object)

## S3 method for class 'integrated_density'
print(x, ...)

## S3 method for class 'integrated_density'
summary(object, ...)

## S3 method for class 'integrated_density'
plot(x, ...)

Arguments

object	an integrated_density object
x	an integrated_density object
...	additional arguments to print, summary, and plot methods (currently ignored)
lambda	parameter of Ricker and Beverton-Holt models of density dependence

Details

Prior distributions can be specified as single-dimensional greta distribution, e.g., normal(0, 1). Link functions and transformations can be specified directly in-line, e.g., ilogit(normal(0, 1)) specifies normal priors with a mean of zero and a standard deviation of one, transformed with an inverse-logit link.

Value

An object of class integrated_density, which can be used to create and integrated_process object

Examples

## Not run: 

library(integrated)

# a really basic age-structured model with five age classes
process <- leslie(5, density = ricker(lambda = uniform(0, 1)))

# setting custom priors
process <- leslie(5, density = beverton(lambda = uniform(0, 1)),
                  priors = list(survival = ilogit(normal(0, 1)),
                                fecundity = exp(normal(0, 1))))

## End(Not run)

jdyen/greta.integrated documentation built on May 3, 2019, 8:04 p.m.