hurdleModel: Run binomial logistic models to predict species' occurrences...

In nicholasjclark/BBS.occurrences: BBS community composition and co-occurrence analysis

Description

Run binomial logistic models to predict species' occurrences and gaussian CRFs to predict species' local abundances.

Usage

hurdleModel(flyway, n_reps, n_cores)

Arguments

flyway	A character string representing the name of the specified flyway to model
n_reps	Positive integer representing the number of times to repeat 10-fold cv.glmnet regularized regressions (default is 10)
n_cores	Positive integer stating the number of processing cores to split the job across. Default is parallel::detect_cores() - 1

Details

This function runs the hurdle model by first estimating predictors of species' occurrence probability using lassoBinomial and then fitting a conditional random field to species' scaled abundances using lassoGaussian. Each of these models is run n_reps times to allow for uncertainty in coefficient estimates. Predictions are then used to estimate model fits and predict community network B'os and species' eigencentrality in each site. All models are run for sites within a unique bioregion, ensuring that network metrics reflect variation among sites with similar climate and species compositions.

nicholasjclark/BBS.occurrences documentation built on July 19, 2020, 8:31 p.m.