brquasiFit: Fitting function for ['glm()'] for reduced-bias estimation...

In ikosmidis/brquasi: Bias reduction in quasi likelihood estimation

Fitting function for ['glm()'] for reduced-bias estimation and inference using quasi likelihoods

Description

'brquasiFit' is a fitting method for ['glm()'] for mean-bias reduction in quasi-likelihood estimation, using the implicit and explicit reduced-bias M-estimators in Kosmidis & Lunardon (2020).

Usage

brquasiFit(
  x,
  y,
  weights = rep(1, nobs),
  start = NULL,
  etastart = NULL,
  mustart = NULL,
  offset = rep(0, nobs),
  family = gaussian(),
  control = list(),
  intercept = TRUE,
  singular.ok = TRUE
)

brquasi_fit(
  x,
  y,
  weights = rep(1, nobs),
  start = NULL,
  etastart = NULL,
  mustart = NULL,
  offset = rep(0, nobs),
  family = gaussian(),
  control = list(),
  intercept = TRUE,
  singular.ok = TRUE
)

Arguments

x	'x' is a design matrix of dimension 'n * p'.
y	'y' is a vector of observations of length 'n'.
weights	an optional vector of ‘prior weights’ to be used in the fitting process. Should be NULL or a numeric vector.
start	starting values for the parameters in the linear predictor. If 'NULL' (default) then the maximum likelihood estimates are calculated and used as starting values.
etastart	applied only when start is not 'NULL'. Starting values for the linear predictor to be passed to 'glm.fit' when computing starting values using maximum likelihood.
mustart	applied only when start is not 'NULL'. Starting values for the vector of means to be passed to 'glm.fit' when computing starting values using maximum likelihood.
offset	this can be used to specify an a priori known component to be included in the linear predictor during fitting. This should be NULL or a numeric vector of length equal to the number of cases. One or more offset terms can be included in the formula instead or as well, and if more than one is specified their sum is used. See model.offset.
family	any one of the ['quasi()'] families.
control	a list of parameters controlling the fitting process. See ['brquasiControl()'] for details.
intercept	logical. Should an intercept be included in the null model?
singular.ok	logical. If 'FALSE', a singular model is an error.

Details

Estimation is performed using a quasi Newton-Raphson iteration (similar to the quasi Fisher-scoring iteration described, for example, in 'vignette("iteration", "brglm2")', which, in the case of mean-bias reduction, resembles an iterative correction of the asymptotic bias of the Fisher scoring iterates.

Author(s)

Ioannis Kosmidis [aut, cre] [e-mail]

See Also

['glm.fit()'] ['glm()'] ['quasi()']

ikosmidis/brquasi documentation built on Jan. 27, 2023, 9:04 p.m.