evinb {evinf}

Running an extreme value inflated negative binomial model with bootstrapping

Description

Running an extreme value inflated negative binomial model with bootstrapping

Usage

evinb(
  formula_nb,
  formula_evi = NULL,
  formula_pareto = NULL,
  data,
  bootstrap = TRUE,
  n_bootstraps = 100,
  multicore = FALSE,
  ncores = NULL,
  block = NULL,
  boot_seed = NULL,
  max.diff.par = 0.01,
  max.no.em.steps = 500,
  max.no.em.steps.warmup = 5,
  c.lim = c(50, 1000),
  max.upd.par.pl.multinomial = 0.5,
  max.upd.par.nb = 0.5,
  max.upd.par.pl = 0.5,
  no.m.bfgs.steps.multinomial = 3,
  no.m.bfgs.steps.nb = 3,
  no.m.bfgs.steps.pl = 3,
  pdf.pl.type = "approx",
  eta.int = c(-1, 1),
  init.Beta.multinom.PL = NULL,
  init.Beta.NB = NULL,
  init.Beta.PL = NULL,
  init.Alpha.NB = 0.01,
  init.C = 200,
  verbose = FALSE
)

Arguments

formula_nb

Formula for the negative binomial (count) component of the model

formula_evi

Formula for the extreme-value inflation component of the model. If NULL taken as the same formula as nb

formula_pareto

Formula for the pareto (extreme value) component of the model. If NULL taken as the same formula as nb

data

Data to run the model on

bootstrap

Should bootstrapping be performed. Needed to obtain standard errors and p-values

n_bootstraps

Number of bootstraps to run. For use of bootstrapped p-values, at least 1,000 bootstraps are recommended. For approximate p-values, a lower number can be sufficient

multicore

Should multiple cores be used?

ncores

Number of cores if multicore is used. Default (NULL) is one less than the available number of cores

block

Optional string indicating a case-identifier variable when using block bootstrapping

boot_seed

Optional bootstrap seed to ensure reproducible results.

max.diff.par

Tolerance for EM algorithm. Will be considered to have converged if the maximum absolute difference in the parameter estimates are lower than this value

max.no.em.steps

Maximum number of EM steps to run. Will be considered to not have converged if this number is reached and convergence is not reached

max.no.em.steps.warmup

Number of EM steps in the warmup rounds

c.lim

Integer range defining the possible values of C

max.upd.par.pl.multinomial

Maximum parameter change step size in the extreme value inflation component

max.upd.par.nb

Maximum parameter change step size in the count component

max.upd.par.pl

Maximum parameter change step size in the pareto component

no.m.bfgs.steps.multinomial

Number of BFGS steps for the multinomial model

no.m.bfgs.steps.nb

Number of BFGS steps for the negative binomial model

no.m.bfgs.steps.pl

Number of BFGS steps for the pareto model

pdf.pl.type

Probability density function type for the pareto component. Either 'approx' or 'exact'. 'approx' is adviced in most cases

eta.int

Initial values for eta

init.Beta.multinom.PL

Initial values for beta parameters in the extreme value inflation component. Vector of same length as number of parameters in the extreme value inflation component or NULL (which gives starting values of 0)

init.Beta.NB

Initial values for beta parameters in the count component. Vector of same length as number of parameters in the count component or NULL (which gives starting values of 0)

init.Beta.PL

Initial values for beta parameters in the pareto component. Vector of same length as number of parameters in the pareto component or NULL (which gives starting values of 0)

init.Alpha.NB

Initial value of Alpha NB, integer or NULL (giving a starting value of 0)

init.C

Initial value of C. Integer which should be within the C_lim range.

verbose

Should progress be printed for the first run of evinb

Value

An object of class 'evinb'

Examples

data(genevzinb2)
model <- evinb(y~x1+x2+x3,data=genevzinb2, n_bootstraps = 10)