model.diversity {DiversityOccupancy} | R Documentation |
This function takes a diversityoccu object and heuristically searches for the glm that best explains the alpha diversity of the modelled species.
model.diversity(DivOcc, method = "h", delta = 2, squared = FALSE)
DivOcc |
is an object returned by the divesityoccu function of this package |
method |
The method to be used to explore the candidate set of models. If "h" an exhaustive screening is undertaken. If "g" the genetic algorithm is employed (recommended for large candidate sets). If "l", a very fast exhaustive branch-and-bound algorithm is used. Package leaps must then be loaded, and this can only be applied to linear models with covariates and no interactions. |
delta |
The number of models that will be returned will be the ones that have a maximum AICc difference with the top model equal to delta. |
squared, |
if FALSE (Default), only GLMs with linear components will be evaluated; If TRUE, GLMs with both linear and quadratic components will be evaluated. WARNING if squared is TRUE, the number of parameters duplicates and the models grow exponentially, this may result in to many variables for a CPU to compute. |
This function fits every first order glm possible and ranks them by AICc.
An object with the best fitted model, the coefficients of that model, a table with the top 5 fitted models ranked by AICc and the data used for the model
Derek Corcoran <derek.corcoran.barrios@gmail.com>
## Not run:
#Load the data
data("IslandBirds")
data("Daily_Cov")
data("siteCov")
#Model the abundance for 5 bat species and calculate alpha diversity from that
BirdDiversity <-diversityoccu(pres = IslandBirds, sitecov = siteCov,
obscov = Daily_Cov,spp = 5, form = ~ Day + Wind + Time + Rain +
Noise ~ Elev + AgroFo + SecVec + Wetland + Upland)
#Select the best model that explains diversity using genetic algorithms
set.seed(123)
glm.Birdiversity <- model.diversity(BirdDiversity, method = "g")
#see the best models
glm.Birdiversity$Best.model
#plot the response of diversity to individual variables
plot(glm.Birdiversity, elev)
#To add the quadratic components of models
glm.birdiversity <- model.diversity(BirdDiversity , method = "g", squared = TRUE)
responseplot.diver(glm.birdiversity, Elev)
## End(Not run)